Neuroscience Research 75 (2013) 83–88
Contents lists available at SciVerse ScienceDirect
Neuroscience Research
journal homepage: www.elsevier.com/locate/neures
Review article
The axon as a unique computational unit in neurons
Takuya Sasaki∗
Division of Cerebral Structure, National Institute for Physiological Sciences, Okazaki 444-8787, Japan
a r t i c l e i n f o
Article history:
Received 1 November 2012
Received in revised form
11 December 2012
Accepted 17 December 2012
Available online 5 January 2013
Keywords:
Axon
Action potential
Synaptic transmission
Depolarization
Calcium
Oscillation
a b s t r a c t
In the mammalian cortex, axons are highly ramiﬁed and link an enormous number of neurons over large
distances. The conventional view assumes that action potentials (APs) are initiated at the axon initial
segment in an all-or-none fashion and are then self-propagated orthodromically along axon collaterals
without distortion of the AP waveform. By contrast, recent experimental results suggest that the axonal
AP waveform can be modiﬁed depending on the activation states of the ion channels and receptors on
axonal cell membranes. This AP modulation can regulate neurotransmission to postsynaptic neurons.
In addition, the latest studies have provided evidence that cortical axons can integrate somatic burst
ﬁrings and promote activity-dependent ectopic AP generation, which may underlie the oscillogenesis
of fast rhythmic network activity. These seminal observations indicate that axons can perform diverse
functional operations that extend beyond the prevailing model of axon physiology.
© 2013 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2. Techniques for direct recording from axons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3. Analog modulation of axonal action potential by somatic depolarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4. The modulation of action potential propagation during axonal conduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5. AP initiation at distal axons independent from the soma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6. Possible roles of ectopic APs at distal axons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
7. Concluding remarks and future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
1. Introduction
At the cellular level, the neuronal signaling chain can be
divided into several functional components: the dendritic integration
of synaptic inputs, AP generation, AP propagation, and
synaptic transmission. In the context of information ﬂow, it has
been assumed that AP initiation and propagation are regenerative
events that never vary in intensity or duration. From this perspective,
axons have been regarded as simple cables that faithfully
transmit electrical impulses. With the development of direct patchclamp
techniques for small axonal structures, the results of recent
Abbreviations: AP, action potential; eAP, ectopic action potential; AMPA, 2amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic
acid.
∗ Tel.: +81 564 59 5279; fax: +81 564 69 5276.
E-mail address: t.sasaki.0224@gmail.com
neurophysiological studies have begun to challenge this traditional
model (Debanne, 2004; Bucher and Goaillard, 2011). Because the
generation and propagation of APs are insured by the balance of
Na+ inﬂux and K+ efﬂux through voltage-gated ion channels in
axons (Hodgkin and Huxley, 1952), the axonal AP waveform can
be modulated by the activation state of these channels and by spatial
patterns of axonal arborization (Alle and Geiger, 2006; Shu
et al., 2006, 2007; Kole et al., 2007; Sasaki et al., 2011, 2012a).
Local shaping of the axonal AP waveform subsequently controls
the Ca2+ signals at presynaptic terminals and potentiates neurotransmitter
release to postsynaptic cells. The most recent studies
have demonstrated that APs can be ectopically initiated even from
the distal axonal region, which is far from the axon initial segment,
by integrating repetitive ﬁrings at the soma (Shefﬁeld et al., 2011;
Dugladze et al., 2012). These results suggest that axonal ﬁbers can
transmit not only digital-like (all-or-none) but also analog-like signals,
enabling axons to serve as a unique computational unit to
0168-0102/$ – see front matter © 2013 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved.
http://dx.doi.org/10.1016/j.neures.2012.12.004
84 T. Sasaki / Neuroscience Research 75 (2013) 83–88
maintain and process a variety of gradient information in neural
circuits (Alle and Geiger, 2008). Given that a cortical pyramidal cell
locally projects to several hundred postsynaptic targets (Ishizuka
et al., 1990; Li et al., 1994; Major et al., 1994), these unique properties
of axons may have a signiﬁcant impact on network dynamics.
In this article, I review the recent advances in the understanding of
axonal physiology, focusing primarily on hippocampal and neocortical
neurons.
2. Techniques for direct recording from axons
Traditionally, direct electrophysiological recording from thin
axonal ﬁbers has been considered almost impossible due to the relative
complexity and small diameters of these structures. Several
pioneering studies, however, have overcome this technical issue.
In 1994, pioneering works have ﬁrst demonstrated whole-cell
patch-clamp recording techniques from subcellular compartments
including dendrites and axons (up to 30 ␮m from the axon hillock)
in cortical pyramidal cells (Stuart and Sakmann, 1994) and cerebellar
Purkinje cells (Stuart and Hausser, 1994). Owing to the
development of these techniques, axonal initiation and propagation
of APs could be clearly demonstrated in various brain regions
such as substantia nigra (Hausser et al., 1995). While dendritic
recordings have subsequently been utilized to study integration
and computation within dendrites, few studies have performed
axonal recordings (Colbert and Johnston, 1996; Schmitz et al.,
2001), presumably due to technical limitations. For this reason,
axonal recordings have been limited to giant axonal structures
(3–5 ␮m), such as the mossy ﬁber boutons of dentate gyrus granule
cells (Fig. 1A; Geiger and Jonas, 2000; Bischofberger et al., 2006)
Fig. 1. Patch-clamp techniques for recording electrical signals from axons. (A)
Whole-cell recording from a mossy ﬁber bouton of a dentate gyrus granule cell.
Presynaptic boutons are directly accessible to the patch electrode due to their large
diameters (3–5 ␮m). (B) Whole-cell recording from an axonal bleb of a cortical pyramidal
neuron. Axonal blebs are enlarged structures that form at the cut ends of
axons after slicing and (C) Cell-attached recording from an intact unmyelinated
axon (∼1 ␮m) of a cortical pyramidal cell. The patch pipettes are coated with a ﬂuorophore,
which allows visual targeting of the ﬂuorescently labeled axon using online
confocal manipulation. Although this method is only applicable to cell-attached
recordings of extracellular AP waveforms, it is relatively easy to perform in comparison
with the above two techniques.
and the giant synaptic expansions of the Calyx of Held (Forsythe,
1994; Awatramani et al., 2005). In 2006, Shu et al. reported a technological
breakthrough in recording signals from axonal blebs up
to 400 ␮m from the axon hillock in cortical pyramidal cells. After
most of pyramidal cell axons are cut during slicing procedure, they
form patchable swellings (3–6 ␮m blebs) at the ends of the axons on
the surface of slices (Fig. 1B). The axonal bleb recording technique
has been subsequently utilized to study axon physiology in cortical
pyramidal cells (Shu et al., 2006; Kole et al., 2007; Sasaki et al.,
2011). Recently, we have developed a new patch-clamp technique
for the analysis of intact unmyelinated axons (∼1 ␮m) that uses a
glass pipette coated with Alexa Fluor-conjugated albumin (Fig. 1C)
(Ishikawa et al., 2010; Sasaki et al., 2012c). The advantage of this
technique is that ﬂuorescently labeled axon tracts and patch-clamp
pipettes can be simultaneously identiﬁed in a single ﬂuorescent
image without switching back and forth between ﬂuorescence and
transmitted images. Although this simpliﬁed technique is restricted
to cell-attached recordings, similar to previous studies (Clark et al.,
2005; Monsivais et al., 2005; Perkins, 2006; Atherton et al., 2008;
Palmer et al., 2010; Rudolph et al., 2011; Dugladze et al., 2012), it
can be used to measure APs as extracellular unit-like waveforms,
which are sufﬁcient to estimate the precise location of AP initiation
site (Clark et al., 2005; Khaliq and Raman, 2006; Palmer et al.,
2010) or relative changes in the duration of intracellular APs (Sasaki
et al., 2012a). These technical advancements in axonal recording
techniques will facilitate the further exploration of the density,
properties, and distribution of receptors and ion channels on
axons.
3. Analog modulation of axonal action potential by somatic
depolarization
At the resting potential, the duration of somatic APs is constant
(Fig. 2A) (Sasaki et al., 2012b), consistent with the general view of
AP physiology. However, a different scenario is observed when the
soma is sufﬁciently depolarized. Because neurons contain a variety
of voltage-gated Na+ and K+ channels that are required for AP
generation, the activation and/or inactivation of these channels can
shape AP waveforms. This effect can be easily conﬁrmed by a simple
experiment with a somatic whole-cell recording. Subthreshold
depolarization just prior to AP initiation at the soma signiﬁcantly
prolongs the subsequent AP waveform (Fig. 2B). The magnitude of
the increase in AP duration depends on the depolarization level
prior to the AP. The major mechanism underlying the AP broadening
is the inactivation of Kv1 channels rather than changes in
Na+-channel opening (Shu et al., 2007). It has been estimated that
a 10 mV depolarization relative to the resting membrane potential
would be expected to inactivate ∼10% of the Kv1 channels
(Kole et al., 2007). Simultaneous axonal and somatic patch-clamp
recordings have revealed that subthreshold voltage ﬂuctuations at
the somatodendritic axis propagate a signiﬁcant distance along the
axon in granule cells in the dentate gyrus (Alle and Geiger, 2006)
and in pyramidal cells in the prefrontal cortex (Shu et al., 2006). The
depolarization propagated along axons can induce AP broadening
at axonal regions as well as in the soma, enabling the propagation
of broadened APs along axonal ﬁbers. Because the range of
the somatic inﬂuence is spatially limited by the axonal path length,
with a length constant ( ) ranging from 400 to 600 ␮m (Alle and
Geiger, 2006; Shu et al., 2006; Kole et al., 2007), depolarizationbroadened
APs return to a normal width at some distance from
soma (Fig. 2C). In addition to the distance restriction, we recently
demonstrated that the AP broadening effect decays more steeply at
branch points than at axonal shaft segments (Fig. 2D) (Sasaki et al.,
2012a), indicating that branch points act as a ﬁlter to modulate
the strength of the propagated impulses. This idea is supported by
T. Sasaki / Neuroscience Research 75 (2013) 83–88 85
Fig. 2. Axonal topology interacts with the somatic modulation of axonal AP broadening
and synaptic outputs. (A) The duration of somatic APs initiated from the
resting potential is constant from trial to trial. APs were evoked by a somatic current
injection every 10 s. (B) Somatic AP width is modulated by depolarization
prior to AP initiation at the soma. Voltage traces are color-coded by the amplitudes
of the rectangular currents preceding the AP generation. (C and D) Changes in
depolarization-induced axonal AP width plotted against the distance from the soma
(C) and against the number of branch points (D) that APs traverse. Each curve represents
the best ﬁt of the exponential decay function. (E) The somatic depolarization
of presynaptic neurons facilitates neurotransmission to close (within 100 ␮m, top)
but not to distant (more than 300 ␮m, bottom) postsynaptic cells. The traces represent
presynaptic APs (left) and postsynaptic currents (right) when the presynaptic
CA3 neuron is at the resting (black) or 20 mV depolarized (red) potential and (F)
Summary of somatic depolarization-induced changes in synaptic efﬁcacy between
close and distant synaptically coupled neurons.
morphological evidence that sibling axon branches are narrower
than parent axons and by the cable theory, which deﬁnes as
proportional to the square root of the ﬁber diameter (Dayan and
Abbott, 2001).
One consequence of activity-dependent axonal AP broadening
is an intra-axonal elevation of the Ca2+ concentration, which favors
glutamate release at presynaptic boutons. Optical imaging has conﬁrmed
that somatic depolarization prior to AP initiation produces
larger AP-induced Ca2+ transients in presynaptic boutons than
those evoked at the resting potential, which has been demonstrated
in hippocampal pyramidal cells (Sasaki et al., 2012a,b) and cerebellar
interneurons (Christie and Jahr, 2008; Christie et al., 2011),
but not at hippocampal mossy ﬁber synapses (Alle and Geiger,
2006; Scott et al., 2008). Two potential mechanisms underlying
the AP-induced Ca2+ dynamics can be proposed: (i) subthreshold
depolarization spreads into axon arbor and facilitates Ca2+ inﬂux at
presynaptic terminals through opening of voltage-dependent calcium
channels, albeit at low probability, which is accumulated on
the subsequent AP-induced Ca2+ inﬂux (Christie et al., 2011), and
(ii) broadened APs enhance Ca2+ signaling at synaptic terminals
by increasing the duration of depolarization. The AP-induced Ca2+
increase then facilitates neurotransmitter release from presynaptic
terminals. As observed for the depolarization-induced AP broadening
effect (Fig. 2C), the depolarization-induced synaptic facilitation
also depends on the distance from the soma; somatic depolarization
in presynaptic cells strengthens synaptic transmission to close
postsynaptic cells but not to distant cells (Fig. 2E). These ﬁndings
suggest that nearby cortical neurons can communicate with each
other in a mixed digital- and analog-like manner through AP conduction
along axons.
4. The modulation of action potential propagation during
axonal conduction
The AP waveform can be transiently modulated even at the middle
of an axon while the AP travels down the axon. Debanne et al.
(1997), for instance, reported that failure of AP propagation can
be induced as a result of the activation of A-type K+ channels on
axonal membranes when neurons are brieﬂy hyperpolarized prior
to the AP generation. In cerebellar Purkinje cells, simple spikes and
spikelets of complex spikes fail to propagate at somatic ﬁring rates
of ∼200 Hz (Khaliq and Raman, 2005). In a majority of cortical pyramidal
neurons, excitatory synapses are formed at focal swellings
(axonal varicosities) with intervals of 4–6 ␮m in an en passant manner;
these swellings are separated from other varicosities by short
axonal shaft segments (Shepherd et al., 2002). It has been reported
that presynaptic membranes and axonal shafts express a wide variety
of receptors and voltage-dependent ion channels (Engelman
and MacDermott, 2004). We recently demonstrated that AP waveforms
can also be broadened during axonal conduction (Sasaki
et al., 2011). The local application of glutamate to the axonal shaft
evokes a broadening of the axonal APs recorded at a point downstream
from the drug application site (Fig. 3A). This effect can be
mimicked by a local puff of high (20 mM) K+ and 4-aminopyridine
(4-AP), an A-type K+ channel blocker. Taken together, these experiments
demonstrate that local depolarization at an axonal segment
results in AP broadening originating from the focal site through the
transient inactivation of IA conductances (Fig. 3B). The broadened
APs elicit larger Ca2+ increases in presynaptic boutons and thereby
facilitate synaptic transmission to postsynaptic neurons.
In pyramidal cells, most axonal shafts and presynaptic varicosities
are in close contact with astrocytes, the largest class of glial
cells in the brain. We found that the activation of these periaxonal
astrocytes by Ca2+ uncaging triggers axonal AP broadening that is
dependenton the activation of AMPAreceptors (Fig. 3A). This observation
suggests that one of the physiological factors that evokes AP
broadening might be an extrinsic signal arising from astrocytes,
as outlined in Fig. 3B. Similar to the signaling pathway associated
with somatic depolarization-induced AP broadening (Fig. 2), the
broadened APs facilitate neurotransmitter release at downstream
synapses. Another functional implication from this ﬁnding is that
astrocyte-mediated signaling can inﬂuence synaptic outputs at distal
regions, well beyond the physical territory of astrocytes (which
have a diameter of approximately 50 ␮m), through the regulation
of axon propagation.
Several questions related to this signaling ﬂow remain to be
resolved. The most crucial issue is how AMPA receptor activation
leads to axonal AP broadening. In cortical pyramidal neurons, it has
been debated whether AMPA receptors are expressed on axonal cell
membranes (Martin et al., 1993, 1998; Engelman and MacDermott,
2004; Schicker et al., 2008). Two possible pathways could account
for local AP broadening: (1) the direct activation of AMPA receptors
expressed on axonal membranes could evoke depolarization
86 T. Sasaki / Neuroscience Research 75 (2013) 83–88
Fig. 3. APs are subject to waveform modulation as they travel down axons. (A) Dual
somatic and axonal patch-clamp recordings obtained from a hippocampal pyramidal
cell. The top panel shows a schematic representation of the recording conﬁguration.
Drug application or Ca2+
uncaging of astrocytes on the axon path results in a substantial
increase in the duration of the APs that are propagated down the axon and
(B) A schematic illustration showing how astrocytes regulate neurotransmission at
distant synapses through axonal modulation. Local axonal depolarization elicited by
periaxonal astrocytes induces AP broadening along axons, which enhances postsynaptic
currents. Thus, astrocytes serve as extrinsic instructors of neurotransmission
by shaping axonal AP waveforms.
at the axonal segment; or (2) extracellular glutamate could activate
AMPA receptors on nearby dendrites, which would in turn
increase the extracellular concentration of K+ to indirectly evoke
axonal AP broadening. The discrimination of these two mechanisms
will require further investigation. Furthermore, glutamate release
from astrocytes as a gliotransmitter under physiological conditions
is still a matter of debate. Several recent reports have disputed
the physiological relevance of gliotransmission (Fiacco et al., 2007,
2009; Agulhon et al., 2008, 2010; Petravicz et al., 2008; Hamilton
and Attwell, 2010). However, the Ca2+ uncaging method used in
our study might be too artiﬁcial and nonspeciﬁc to truly represent
spontaneously generated astrocyte activity. This point also requires
future experimental evaluation to better interpret the signiﬁcance
of the astrocyte-axon interaction.
5. AP initiation at distal axons independent from the soma
In general, APs propagate orthodromically from the axon initial
segment toward axon terminals. Although unidirectional propagation
is ensured by the absolute refractory period, the nature
of the axons allows APs to propagate equally in both directions.
This phenomenon can be observed in a simple electrophysiological
experiment. The stimulation of distal axonal ﬁbers triggers ectopic
APs, also termed as antidromic APs, originating from the local
site, which subsequently conduct toward both the somatodendritic
compartment (antidromic direction) and the axon terminal
(orthodromic direction). These peculiar APs are distinguishable
at the soma by their stereotypical kinetics; the voltage trajectory
of ectopic APs increases sharply from the resting potential
in the absence of depolarization prior to the AP. The existence
of ectopic APs was initially conﬁrmed under pathological circumstances
(Pinault, 1995), but more recently, several studies have
demonstrated that ectopic APs can occur during active states even
in normal cortical networks (Papatheodoropoulos, 2008; Bahner
et al., 2011; Shefﬁeld et al., 2011). In a subset of hippocampal and
neocortical interneurons, repetitive spikes at the soma are integrated
at the distal axonal region and eventually trigger persistent
autonomous ﬁring originating from the axonal segment (Shefﬁeld
et al., 2011). One of the mechanisms involved in ectopic APs seems
to be the opening of gap junctions, rather than somatic depolarization
or synaptic transmission, but this point is still unresolved
due to the poor speciﬁcity of gap junction blockers (Rouach et al.,
2003; Chepkova et al., 2008; Tovar et al., 2009; Ye et al., 2009;
Behrens et al., 2011). The axonal APs occasionally fail to invade
the soma, instead giving rise to prepotentials, termed spikelets, at
the somatodendritic compartment (Shefﬁeld et al., 2011). A recent
study by Dugladze et al. (2012) provided an important clue about
the control of the antidromic propagation of ectopic APs. During
in vitro gamma oscillation (30–80 Hz) evoked by kainic acid application,
ectopic APs are initiated in the distal part of the axon in
hippocampal pyramidal cells. Curiously, the somatic invasion by
antidromic APs is blocked by the activation of a nearby axo-axonic
cell, a type of interneuron that exerts a powerful inhibitory effect
on the axon initial segment. These results demonstrate that the
axon initial segment has a gating property that determines whether
ectopic APs can reach the somatic region and that axo-axonic cells
can separate axons from somatodendritic activity during fast network
oscillations.
6. Possible roles of ectopic APs at distal axons
Although the fundamental signiﬁcance of ectopic APs has not
been fully elucidated, their functional implications are profound.
Given that an ectopic AP occurs at the middle of an axonal segment,
this type of AP can trigger synaptic transmission to all of
the upstream and downstream postsynaptic targets. Therefore, the
impact of an ectopic AP might be almost equivalent to that of a
somatic AP. Because the frequency of ectopic APs at axons is considerably
higher than that of somatic APs during fast network activity
(Bahner et al., 2011; Dugladze et al., 2012), the vast majority of
ectopic APs may provide a powerful means of information transfer
from the hippocampus to other brain regions, such as the neocortex.
Another remarkable implication is that ectopic APs are likely to
be associated with fast network oscillations such as gamma oscillations
and ultrafast ripple oscillations (80–200 Hz), which have
been thought to be crucial in memory consolidation (Girardeau
et al., 2009; Ego-Stengel and Wilson, 2010) and in transfer from
the hippocampus to the neocortex (Buzsaki, 1989; Buzsaki et al.,
1992; Eichenbaum, 2000; Diekelmann and Born, 2010; Carr et al.,
2011). Previous studies have demonstrated that such fast rhythmic
activity is abolished by gap junction blockers, demonstrating
that gap junction-mediated electrical couplings are essential for
fast network oscillations (Draguhn et al., 1998; LeBeau et al., 2003).
Notably, in vivo whole-cell recordings from hippocampal pyramidal
cells recently revealed that bursts of spikelets and a barrage of
postsynaptic currents frequently occur in awake animals (Harvey
et al., 2009; Maier et al., 2011; Chorev and Brecht, 2012), and both
of these phenomena are well phase-locked to the high-frequency
oscillation in local ﬁeld recordings. In combination with the fact
that somatic spikelets represent ectopic APs initiated at axonal
compartments through electrical coupling, as mentioned above
(Bahner et al., 2011; Shefﬁeld et al., 2011; Dugladze et al., 2012),
it might be possible that ectopic APs are involved in the initiation
and/or maintenance of fast network activity. Consistent with
this idea, anatomical and electrophysiological evidence has implied
T. Sasaki / Neuroscience Research 75 (2013) 83–88 87
the presence of gap junctions between axons at a site 50–120 ␮m
from the soma (Gladwell and Jefferys, 2001; Schmitz et al., 2001).
However, the density of axonal gap junctions seems extremely low
(approximately 1–2 per neuron) (Traub et al., 1999), and understanding
the exact nature of this phenomenon will require further
evaluation. In line with this hypothesis, an in vitro study using hippocampal
slices has shown that high-frequency oscillations can
be generated even when the axonal plexus in the striatum oriens
is isolated from the pyramidal cell bodies (Traub et al., 2003a).
Correspondingly, in silico network models also predict that the electrical
coupling of pyramidal cell axons contributes to the formation
of high-frequency oscillations (Traub et al., 1999, 2003b, 2012;
Traub and Bibbig, 2000; Bahner et al., 2011). These observations
are plausible in terms of the kinetics of intercellular interactions
because general chemical synapses alone might be too slow to
account for such high-frequency (80–200 Hz) oscillations. Thus,
electrical couplings might be more effective mechanisms for synchronizing
the activity in a bundle of axons at such a ﬁne temporal
scale. Through these unique intercellular communications, ectopic
axonal APs might play a central role in the generation and maintenance
of cortical oscillations, independent of somatodendritic
activity.
7. Concluding remarks and future perspectives
Thanks to the recent achievement of direct axonal recording in
the mammalian cortex, a growing number of studies have revealed
fascinating roles for axons that challenge the conventional all-ornone
view of APs. In combination with non-linear somatodendritic
integration, axonal computation may contribute to enriching the
information processing capabilities of single neurons. This paper
introduced two unique properties of cortical axons; modulation of
AP waveform and ectopic AP initiation. Although it has been still
unknown whether these properties are linked with each other, it
is possible to consider their relationships. For example, ectopic APs
initiated in the axon that propagate orthodromic and antidromic
directions could be differentially affected by somatic depolarization
due to the location dependency of depolarization-induced
broadening of axonal APs. Other possible event is that broadened
APs might be more effective to trigger ectopic APs from
axonal segments compared with APs induced from resting potential.
Future studies are needed to clarify these phenomena. In
addition, several details concerning the subcellular mechanisms of
these phenomena also remain unresolved; for example, the distribution
of ion channel clusters on complex axonal trees, the factors
driving signaling pathways that modulate axonal conduction, the
physiological conditions required for ectopic AP generation, and
the frequency of conduction failures are all unknown (Dyball et al.,
1988; Baccus, 1998; Soleng et al., 2003). At the neural circuit level,
it remains unknown how many synapses are scaled by transient
changes in analog-like AP waveforms and how effectively axonal
AP dynamics serve to upgrade neural population coding. Answering
these questions will require substantial advances in recording
techniques for axons, including quantitative immunoﬂuorescence
methods (Kole et al., 2008; Kuba et al., 2010) and optical imaging
techniques with higher temporal and spatial resolution using
genetically encoded Ca2+ indicators (Dreosti et al., 2009; Tian et al.,
2009; Zhao et al., 2011; Ohkura et al., 2012), Na+ indicators (Kole
et al., 2008), and voltage-sensitive dyes (Palmer et al., 2010; Popovic
et al., 2011). Optogenetic tools will enable to stimulate local axonal
sites and examine how the evoked APs spread throughout axonal
branches. In addition, more realistic computational models will
also be indispensable to quantitatively predict the potential consequences
of single-axonal computation for neural circuit functions
with respect to the storage of working memories and information
processing capability. Unveiling novel axonal function using these
innovative approaches will facilitate our understanding of the limits
of single-neuron computation.
Acknowledgments
I thank Dr. Yuji Ikegaya for valuable comments on an earlier version
of this manuscript. This review was written to acknowledge the
receipt of the Japan Neuroscience Society Young Investigator Award
in 2012. This work was supported by a Postdoctoral Fellowship
from the Japan Society for the Promotion of Science.
References
Agulhon, C., Fiacco, T.A., McCarthy, K.D., 2010. Hippocampal short- and longterm
plasticity are not modulated by astrocyte Ca2+
signaling. Science 327,
1250–1254.
Agulhon, C., Petravicz, J., McMullen, A.B., Sweger, E.J., Minton, S.K., Taves, S.R., Casper,
K.B., Fiacco, T.A., McCarthy, K.D., 2008. What is the role of astrocyte calcium in
neurophysiology? Neuron 59, 932–946.
Alle, H., Geiger, J.R., 2006. Combined analog and action potential coding in hippocampal
mossy ﬁbers. Science 311, 1290–1293.
Alle, H., Geiger, J.R., 2008. Analog signalling in mammalian cortical axons. Curr. Opin.
Neurobiol. 18, 314–320.
Atherton, J.F., Wokosin, D.L., Ramanathan, S., Bevan, M.D., 2008. Autonomous initiation
and propagation of action potentials in neurons of the subthalamic nucleus.
J. Physiol. 586, 5679–5700.
Awatramani, G.B., Price, G.D., Trussell, L.O., 2005. Modulation of transmitter release
by presynaptic resting potential and background calcium levels. Neuron 48,
109–121.
Baccus, S.A., 1998. Synaptic facilitation by reﬂected action potentials: enhancement
of transmission when nerve impulses reverse direction at axon branch points.
Proc. Natl. Acad. Sci. U.S.A. 95, 8345–8350.
Bahner, F., Weiss, E.K., Birke, G., Maier, N., Schmitz, D., Rudolph, U., Frotscher, M.,
Traub, R.D., Both, M., Draguhn, A., 2011. Cellular correlate of assembly formation
in oscillating hippocampal networks in vitro. Proc. Natl. Acad. Sci. U.S.A. 108,
E607–E616.
Behrens, C.J., Ul Haq, R., Liotta, A., Anderson, M.L., Heinemann, U., 2011. Nonspeciﬁc
effects of the gap junction blocker meﬂoquine on fast hippocampal network
oscillations in the adult rat in vitro. Neuroscience 192, 11–19.
Bischofberger, J., Engel, D., Li, L., Geiger, J.R., Jonas, P., 2006. Patch-clamp recording
from mossy ﬁber terminals in hippocampal slices. Nat. Protoc. 1, 2075–2081.
Bucher, D., Goaillard, J.M., 2011. Beyond faithful conduction: short-term dynamics,
neuromodulation, and long-term regulation of spike propagation in the axon.
Prog. Neurobiol. 94, 307–346.
Buzsaki, G., 1989. Two-stage model of memory trace formation: a role for noisy brain
states. Neuroscience 31, 551–570.
Buzsaki, G., Horvath, Z., Urioste, R., Hetke, J., Wise, K., 1992. High-frequency network
oscillation in the hippocampus. Science 256, 1025–1027.
Carr, M.F., Jadhav, S.P., Frank, L.M., 2011. Hippocampal replay in the awake state: a
potential substrate for memory consolidation and retrieval. Nat. Neurosci. 14,
147–153.
Chepkova, A.N., Sergeeva, O.A., Haas, H.L., 2008. Carbenoxolone impairs LTP and
blocks NMDA receptors in murine hippocampus. Neuropharmacology 55,
139–147.
Chorev, E., Brecht, M., 2012. In vivo dual intra- and extracellular recordings suggest
bidirectional coupling between CA1 pyramidal neurons. J. Neurophysiol. 108,
1584–1593.
Christie, J.M., Chiu, D.N., Jahr, C.E., 2011. Ca2+
-dependent enhancement of release by
subthreshold somatic depolarization. Nat. Neurosci. 14, 62–68.
Christie, J.M., Jahr, C.E., 2008. Dendritic NMDA receptors activate axonal calcium
channels. Neuron 60, 298–307.
Clark, B.A., Monsivais, P., Branco, T., London, M., Hausser, M., 2005. The site of action
potential initiation in cerebellar Purkinje neurons. Nat. Neurosci. 8, 137–139.
Colbert, C.M., Johnston, D., 1996. Axonal action-potential initiation and Na+
channel
densities in the soma and axon initial segment of subicular pyramidal neurons.
J. Neurosci. 16, 6676–6686.
Dayan, P., Abbott, L.F., 2001. Theoretical Neuroscience: Computational and Mathematical
Modeling of Neural Systems. Massachusetts Institute of Technology
Press, Cambridge, Mass.
Debanne, D., 2004. Information processing in the axon. Nat. Rev. Neurosci. 5,
304–316.
Debanne, D., Guerineau, N.C., Gahwiler, B.H., Thompson, S.M., 1997. Action-potential
propagation gated by an axonal I(A)-like K+
conductance in hippocampus.
Nature 389, 286–289.
Diekelmann, S., Born, J., 2010. The memory function of sleep. Nat. Rev. Neurosci. 11,
114–126.
Draguhn, A., Traub, R.D., Schmitz, D., Jefferys, J.G., 1998. Electrical coupling underlies
high-frequency oscillations in the hippocampus in vitro. Nature 394, 189–192.
Dreosti, E., Odermatt, B., Dorostkar, M.M., Lagnado, L., 2009. A genetically encoded
reporter of synaptic activity in vivo. Nat. Methods 6, 883–889.
88 T. Sasaki / Neuroscience Research 75 (2013) 83–88
Dugladze, T., Schmitz, D., Whittington, M.A., Vida, I., Gloveli, T., 2012. Segregation
of axonal and somatic activity during fast network oscillations. Science 336,
1458–1461.
Dyball, R.E., Grossmann, R., Leng, G., Shibuki, K., 1988. Spike propagation and conduction
failure in the rat neural lobe. J. Physiol. 401, 241–256.
Ego-Stengel, V., Wilson, M.A., 2010. Disruption of ripple-associated hippocampal
activity during rest impairs spatial learning in the rat. Hippocampus 20, 1–10.
Eichenbaum, H., 2000. A cortical-hippocampal system for declarative memory. Nat.
Rev. Neurosci. 1, 41–50.
Engelman, H.S., MacDermott, A.B., 2004. Presynaptic ionotropic receptors and control
of transmitter release. Nat. Rev. Neurosci. 5, 135–145.
Fiacco, T.A., Agulhon, C., McCarthy, K.D., 2009. Sorting out astrocyte physiology from
pharmacology. Annu. Rev. Pharmacol. Toxicol. 49, 151–174.
Fiacco, T.A., Agulhon, C., Taves, S.R., Petravicz, J., Casper, K.B., Dong, X., Chen, J.,
McCarthy, K.D., 2007. Selective stimulation of astrocyte calcium in situ does
not affect neuronal excitatory synaptic activity. Neuron 54, 611–626.
Forsythe, I.D., 1994. Direct patch recording from identiﬁed presynaptic terminals
mediating glutamatergic EPSCs in the rat CNS, in vitro. J. Physiol. 479 (Pt 3),
381–387.
Geiger, J.R., Jonas, P., 2000. Dynamic control of presynaptic Ca2+
inﬂow by fastinactivating
K+
channels in hippocampal mossy ﬁber boutons. Neuron 28,
927–939.
Girardeau, G., Benchenane, K., Wiener, S.I., Buzsaki, G., Zugaro, M.B., 2009. Selective
suppression of hippocampal ripples impairs spatial memory. Nat. Neurosci. 12,
1222–1223.
Gladwell, S.J., Jefferys, J.G., 2001. Second messenger modulation of electrotonic
coupling between region CA3 pyramidal cell axons in the rat hippocampus.
Neurosci. Lett. 300, 1–4.
Hamilton, N.B., Attwell, D., 2010. Do astrocytes really exocytose neurotransmitters?
Nat. Rev. Neurosci. 11, 227–238.
Harvey, C.D., Collman, F., Dombeck, D.A., Tank, D.W., 2009. Intracellular dynamics
of hippocampal place cells during virtual navigation. Nature 461, 941–946.
Hausser, M., Stuart, G., Racca, C., Sakmann, B., 1995. Axonal initiation and active
dendritic propagation of action potentials in substantia nigra neurons. Neuron
15, 637–647.
Hodgkin, A.L., Huxley, A.F., 1952. A quantitative description of membrane current
and its application to conduction and excitation in nerve. J. Physiol. 117,
500–544.
Ishikawa, D., Takahashi, N., Sasaki, T., Usami, A., Matsuki, N., Ikegaya, Y., 2010. Fluorescent
pipettes for optically targeted patch-clamp recordings. Neural Netw.
23, 669–672.
Ishizuka, N., Weber, J., Amaral, D.G., 1990. Organization of intrahippocampal projections
originating from CA3 pyramidal cells in the rat. J. Comp. Neurol. 295,
580–623.
Khaliq, Z.M., Raman, I.M., 2005. Axonal propagation of simple and complex spikes
in cerebellar Purkinje neurons. J. Neurosci. 25, 454–463.
Khaliq, Z.M., Raman, I.M., 2006. Relative contributions of axonal and somatic Na
channels to action potential initiation in cerebellar Purkinje neurons. J. Neurosci.
26, 1935–1944.
Kole, M.H., Ilschner, S.U., Kampa, B.M., Williams, S.R., Ruben, P.C., Stuart, G.J., 2008.
Action potential generation requires a high sodium channel density in the axon
initial segment. Nat. Neurosci. 11, 178–186.
Kole, M.H., Letzkus, J.J., Stuart, G.J., 2007. Axon initial segment Kv1 channels control
axonal action potential waveform and synaptic efﬁcacy. Neuron 55, 633–647.
Kuba, H., Oichi, Y., Ohmori, H., 2010. Presynaptic activity regulates Na(+) channel
distribution at the axon initial segment. Nature 465, 1075–1078.
LeBeau, F.E., Traub, R.D., Monyer, H., Whittington, M.A., Buhl, E.H., 2003. The role of
electrical signaling via gap junctions in the generation of fast network oscillations.
Brain Res. Bull. 62, 3–13.
Li, X.G., Somogyi, P., Ylinen, A., Buzsaki, G., 1994. The hippocampal CA3 network: an
in vivo intracellular labeling study. J. Comp. Neurol. 339, 181–208.
Maier, N., Tejero-Cantero, A., Dorrn, A.L., Winterer, J., Beed, P.S., Morris, G., Kempter,
R., Poulet, J.F., Leibold, C., Schmitz, D., 2011. Coherent phasic excitation during
hippocampal ripples. Neuron 72, 137–152.
Major, G., Larkman, A.U., Jonas, P., Sakmann, B., Jack, J.J., 1994. Detailed passive cable
models of whole-cell recorded CA3 pyramidal neurons in rat hippocampal slices.
J. Neurosci. 14, 4613–4638.
Martin, L.J., Blackstone, C.D., Levey, A.I., Huganir, R.L., Price, D.L., 1993. AMPA glutamate
receptor subunits are differentially distributed in rat brain. Neuroscience
53, 327–358.
Martin, L.J., Furuta, A., Blackstone, C.D., 1998. AMPA receptor protein in developing
rat brain: glutamate receptor-1 expression and localization change at regional,
cellular, and subcellular levels with maturation. Neuroscience 83, 917–928.
Monsivais, P., Clark, B.A., Roth, A., Hausser, M., 2005. Determinants of action potential
propagation in cerebellar Purkinje cell axons. J. Neurosci. 25, 464–472.
Ohkura, M., Sasaki, T., Kobayashi, C., Ikegaya, Y., Nakai, J., 2012. An improved genetically
encoded red ﬂuorescent ca(2+)
indicator for detecting optically evoked
action potentials. PLoS One 7, e39933.
Palmer, L.M., Clark, B.A., Grundemann, J., Roth, A., Stuart, G.J., Hausser, M., 2010.
Initiation of simple and complex spikes in cerebellar Purkinje cells. J. Physiol.
588, 1709–1717.
Papatheodoropoulos, C., 2008. A possible role of ectopic action potentials in the
in vitro hippocampal sharp wave-ripple complexes. Neuroscience 157, 495–501.
Perkins, K.L., 2006. Cell-attached voltage-clamp and current-clamp recording and
stimulation techniques in brain slices. J. Neurosci. Methods 154, 1–18.
Petravicz, J., Fiacco, T.A., McCarthy, K.D., 2008. Loss of IP3 receptor-dependent Ca2+
increases in hippocampal astrocytes does not affect baseline CA1 pyramidal
neuron synaptic activity. J. Neurosci. 28, 4967–4973.
Pinault, D., 1995. Backpropagation of action potentials generated at ectopic axonal
loci: hypothesis that axon terminals integrate local environmental signals. Brain
Res. Brain Res. Rev. 21, 42–92.
Popovic, M.A., Foust, A.J., McCormick, D.A., Zecevic, D., 2011. The spatio-temporal
characteristics of action potential initiation in layer 5 pyramidal neurons: a
voltage imaging study. J. Physiol. 589, 4167–4187.
Rouach, N., Segal, M., Koulakoff, A., Giaume, C., Avignone, E., 2003. Carbenoxolone
blockade of neuronal network activity in culture is not mediated by an action
on gap junctions. J. Physiol. 553, 729–745.
Rudolph, S., Overstreet-Wadiche, L., Wadiche, J.I., 2011. Desynchronization of multivesicular
release enhances Purkinje cell output. Neuron 70, 991–1004.
Sasaki, T., Matsuki, N., Ikegaya, Y., 2011. Action-potential modulation during axonal
conduction. Science 331, 599–601.
Sasaki, T., Matsuki, N., Ikegaya, Y., 2012a. Effects of axonal topology on the somatic
modulation of synaptic outputs. J. Neurosci. 32, 2868–2876.
Sasaki, T., Matsuki, N., Ikegaya, Y., 2012b. Heterogeneity and independency of unitary
synaptic outputs from hippocampal CA3 pyramidal cells. J. Physiol. 590,
4869–4880.
Sasaki, T., Matsuki, N., Ikegaya, Y., 2012c. Targeted axon-attached recording with
ﬂuorescent patch-clamp pipettes in brain slices. Nat. Protoc. 7, 1228–1234.
Schicker, K.W., Dorostkar, M.M., Boehm, S., 2008. Modulation of transmitter release
via presynaptic ligand-gated ion channels. Curr. Mol. Pharmacol. 1, 106–129.
Schmitz, D., Schuchmann, S., Fisahn, A., Draguhn, A., Buhl, E.H., Petrasch-Parwez, E.,
Dermietzel, R., Heinemann, U., Traub, R.D., 2001. Axo-axonal coupling. A novel
mechanism for ultrafast neuronal communication. Neuron 31, 831–840.
Scott, R., Ruiz, A., Henneberger, C., Kullmann, D.M., Rusakov, D.A., 2008. Analog modulation
of mossy ﬁber transmission is uncoupled from changes in presynaptic
Ca2+
. J. Neurosci. 28, 7765–7773.
Shefﬁeld, M.E., Best, T.K., Mensh, B.D., Kath, W.L., Spruston, N., 2011. Slow integration
leads to persistent action potential ﬁring in distal axons of coupled interneurons.
Nat. Neurosci. 14, 200–207.
Shepherd, G.M., Raastad, M., Andersen, P., 2002. General and variable features of varicosity
spacing along unmyelinated axons in the hippocampus and cerebellum.
Proc. Natl. Acad. Sci. U.S.A. 99, 6340–6345.
Shu, Y., Hasenstaub, A., Duque, A., Yu, Y., McCormick, D.A., 2006. Modulation of intracortical
synaptic potentials by presynaptic somatic membrane potential. Nature
441, 761–765.
Shu, Y., Yu, Y., Yang, J., McCormick, D.A., 2007. Selective control of cortical axonal
spikes by a slowly inactivating K+
current. Proc. Natl. Acad. Sci. U.S.A. 104,
11453–11458.
Soleng, A.F., Chiu, K., Raastad, M., 2003. Unmyelinated axons in the rat hippocampus
hyperpolarize and activate an H current when spike frequency exceeds 1 Hz. J.
Physiol. 552, 459–470.
Stuart, G., Hausser, M., 1994. Initiation and spread of sodium action potentials in
cerebellar Purkinje cells. Neuron 13, 703–712.
Stuart, G.J., Sakmann, B., 1994. Active propagation of somatic action potentials into
neocortical pyramidal cell dendrites. Nature 367, 69–72.
Tian, L., Hires, S.A., Mao, T., Huber, D., Chiappe, M.E., Chalasani, S.H., Petreanu, L.,
Akerboom, J., McKinney, S.A., Schreiter, E.R., Bargmann, C.I., Jayaraman, V., Svoboda,
K., Looger, L.L., 2009. Imaging neural activity in worms, ﬂies and mice with
improved GCaMP calcium indicators. Nat. Methods 6, 875–881.
Tovar, K.R., Maher, B.J., Westbrook, G.L., 2009. Direct actions of carbenoxolone on
synaptic transmission and neuronal membrane properties. J. Neurophysiol. 102,
974–978.
Traub, R.D., Bibbig, A., 2000. A model of high-frequency ripples in the hippocampus
based on synaptic coupling plus axon–axon gap junctions between pyramidal
neurons. J. Neurosci. 20, 2086–2093.
Traub, R.D., Cunningham, M.O., Gloveli, T., LeBeau, F.E., Bibbig, A., Buhl, E.H., Whittington,
M.A., 2003a. GABA-enhanced collective behavior in neuronal axons
underlies persistent gamma-frequency oscillations. Proc. Natl. Acad. Sci. U.S.A.
100, 11047–11052.
Traub, R.D., Pais, I., Bibbig, A., LeBeau, F.E., Buhl, E.H., Hormuzdi, S.G., Monyer, H.,
Whittington, M.A., 2003b. Contrasting roles of axonal (pyramidal cell) and dendritic
(interneuron) electrical coupling in the generation of neuronal network
oscillations. Proc. Natl. Acad. Sci. U.S.A. 100, 1370–1374.
Traub, R.D., Schmitz, D., Jefferys, J.G., Draguhn, A., 1999. High-frequency population
oscillations are predicted to occur in hippocampal pyramidal neuronal networks
interconnected by axoaxonal gap junctions. Neuroscience 92, 407–426.
Traub, R.D., Schmitz, D., Maier, N., Whittington, M.A., Draguhn, A., 2012. Axonal
properties determine somatic ﬁring in a model of in vitro CA1 hippocampal
sharp wave/ripples and persistent gamma oscillations. Eur. J. Neurosci. 36,
2650–2660.
Ye, Z., Oberheim, N., Kettenmann, H., Ransom, B.R., 2009. Pharmacological crossinhibition
of connexin hemichannels and swelling activated anion channels. Glia
57, 258–269.
Zhao, Y., Araki, S., Wu, J., Teramoto, T., Chang, Y.F., Nakano, M., Abdelfattah, A.S.,
Fujiwara, M., Ishihara, T., Nagai, T., Campbell, R.E., 2011. An expanded palette of
genetically encoded Ca(2) indicators. Science 333, 1888–1891.