BR A IN RE S E A RCH 1 1 84 ( 20 0 7 ) 6 5 –7 1

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Dendritic spine plasticity: Looking beyond development

Kimberly J. Harms, Anna Dunaevsky ⁎

Department of Neuroscience, Brown University, Providence, RI, USA

A R T I C LE I N FO AB S T R A C T

Article history: Most excitatory synapses in the CNS form on dendritic spines, tiny protrusions from the
Accepted 24 February 2006 dendrites of excitatory neurons. As such, spines are likely loci of synaptic plasticity. Spines
Available online 5 April 2006 are dynamic structures, but the functional consequences of dynamic changes in these
structures in the mature brain are unclear. Changes in spine density, morphology, and
Keywords: motility have been shown to occur with paradigms that induce synaptic plasticity, as well as
Dendritic spines altered sensory experience and neuronal activity. These changes potentially lead to an
Synapse alteration in synaptic connectivity and strength between neuronal partners, affecting the
Plasticity efficacy of synaptic communication. Here, we review the formation and modification of
excitatory synapses on dendritic spines as it relates to plasticity in the central nervous
system after the initial phase of synaptogenesis. We will also discuss some of the molecular
links that have been implicated in both synaptic plasticity and the regulation of spine
morphology.
© 2006 Published by Elsevier B.V.

1. Introduction clear. Here, we address the formation and modification of

excitatory synapses on dendritic spines as it relates to
The central nervous system is remarkably flexible. Plasticity in plasticity in the central nervous system later in development
cortical function is evident not only during development but after the initial phase of synaptogenesis. We will also discuss
in the adult during learning and memory. However, is this some of the molecular links that have been implicated in both
plasticity due to modification of existing synapses or forma- synaptic plasticity and the regulation of spine morphology.
tion of new synapses? Although synapse number gradually
declines with age, new synapse formation does occur
throughout adulthood. Most excitatory synapses in the central 2. Spines are altered by sensory experience
nervous system form on dendritic spines (Fig. 1), tiny and learning
protrusions from the dendrites of excitatory neurons (Harris,
1999; Nimchinsky et al., 2002) and as such are likely loci of Sensory experience and learning have been hypothesized to
synaptic plasticity. Spines are dynamic structures; individual influence the number and structure of synapses in the CNS,
spines appear and disappear over time, as well as change reflecting a change in synaptic connectivity. Studies examin-
shape (Dailey and Smith, 1996; Fischer et al., 1998; Dunaevsky ing the effects sensory experience on synaptic structural
et al., 1999, 2001; Harris, 1999). There is evidence that spines plasticity have demonstrated changes in the number and
and filopodia are involved in synaptogenesis during develop- morphology of dendritic spines. Increasing sensory experi-
ment (Fiala et al., 1998; Ziv and Smith, 1996), but the functional ence by raising rats in an enriched environment (Diamond et
consequences of these structures in the mature brain are less al., 1975; Leggio et al., 2005) or deflecting a whisker of a freely

⁎ Corresponding author. Fax: +1 401 863 1074.

E-mail address: [email protected] (A. Dunaevsky).

0006-8993/$ – see front matter © 2006 Published by Elsevier B.V.

doi:10.1016/j.brainres.2006.02.094
66 BR A IN RE S EA RCH 1 1 84 ( 20 0 7 ) 6 5 –71

(LTP). LTP induces changes in spine number and morphology

in regions important in learning and memory, such as the
hippocampus and cortex. Induction of LTP followed by EM
analysis revealed increases in spine size and number in the
hippocampus and spine number in the neocortex (Ivanco et
al., 2000). In studies where stimulated synapses were
identified ultrastructurally by the presence of a calcium
Fig. 1 – Dendritic spines in the motor cortex. DiI-labeled precipitate, LTP leads to an increase in number of multiple
dendrite of a pyramidal neuron in a slice from the forelimb synapse boutons (MSBs), single presynaptic boutons con-
motor cortex of an adult rat. Notice the variety of spine tacting multiple spines from the same dendrite. This result
shapes, including stubby spines, long and thin spines, and was interpreted as evidence for the formation of new spines,
mushroom spines with a clear spine head (arrows). Scale bar: possibly by splitting of existing spines (Toni et al., 1999).
5 μm. Harris and colleagues also reported an increased number of
MSBs following LTP, although careful EM analysis suggested
that new spines emerged from the dendrite rather than
moving animal leads to an increase in the number, size, and formed by splitting of existing spines (Fiala et al., 2002). The
density of synaptic inputs onto dendritic spines (Knott et al., formation of MSBs with LTP nicely parallels their formation
2002). Conversely, decrement of sensory experience through with learning paradigms discussed above. These experi-
dark rearing decreases the density of spines in the visual ments provided correlative evidence that changes in den-
cortex. Dark rearing also produces shorter spines with larger dritic spine morphology occur with stimulation protocols
spine heads in layer 3 of the visual cortex, which is partially that lead to synaptic enhancement, and that similar
reversible by subsequent light exposure (Wallace and Bear, mechanisms function during learning and sensory enhance-
2004). ment of synaptic connectivity.
Changes in synapse number have been reported with
various learning paradigms. Increases in the number of
synaptic contacts per neuron with motor skill training have 4. Live imaging of activity-induced changes in
been reported in the rat motor cortex and the cerebellum spines
(Kleim et al., 1996, 1998). Training on spatial learning tasks has
also been shown to increase spine density in the rat Although many studies have indicated that LTP and LTD can
hippocampus (Eyre et al., 2003; Moser et al., 1994; Moser et induce changes in spine number and morphology, not all
al., 1997). Associative memory formation due to eye-blink static studies have agreed (Chang and Greenough, 1984; Sorra
conditioning increases spine density in CA1 pyramidals of the and Harris, 1998; Geinisman et al., 2000; Harris et al., 2003). The
hippocampus (Leuner et al., 2003). In addition, studies have development of 2-photon microscopy and the use of vital dyes
shown that the number of multisynaptic boutons increase in as well as GFP technology have allowed imaging of dendritic
the hippocampus following trace eye-blink conditioning spines in living tissue (Chen et al., 2000) and given researchers
(Geinisman et al., 2001), onto parallel fiber presynaptic the opportunity to follow basal and stimulated spine dynam-
varicosities by separate spines emerging from a given Purkinje ics in living vertebrate organisms. Researches can now
cell dendrite with motor learning (Federmeier et al., 2002), and examine spines in real-time and follow short- and long-term
in the neocortex following experience in the enriched sequential changes, rather than rely on comparing numbers
environment (Jones et al., 1997). Changes in spine morphology from static glimpses or changes across separate populations.
have also been reported with learning (Geinisman et al., 2000). Approaches using live 2-photon imaging of spines in slices
Thus, the number and quality of excitatory synaptic connec- have further demonstrated that spines can dynamically
tions are variable as an animal is influenced by and interacts respond to stimulation that leads to increases or decreases
with its environment. These changes in synapse number and in synaptic strength. Engert and Bonhoeffer showed that
shape potentially lead to an alteration in synaptic connectivity induction of LTP in a small dendritic region of CA1 pyramidal
and strength between neuronal partners, affecting the efficacy neurons in hippocampal slices correlated with the emersion of
of synaptic communication. new spines in the stimulated region, while spine disappear-
ance occurred randomly throughout the slice. Furthermore,
spine formation only occurred when persistent, but not
3. Spines are altered by synaptic activity transient, changes in synaptic strength were induced (Engert
and Bonhoeffer, 1999). Emergence of new dendritic protru-
The strength of excitatory synapses can be modified in an sions has also been induced by synaptic activation with local
experience-dependent manner, and some forms of learning tetanic stimulation (Maletic-Savatic et al., 1999). Not only does
have been shown to enhance synaptic efficacy in the spine formation occur with LTP, but synaptic strength has also
hippocampus (Moser et al., 1993; Sacchetti et al., 2001), the been correlated with spine size (Matsuzaki et al., 2001, 2004;
amygdala (Rogan et al., 1997; McKernan and Shinnick- Murthy et al., 2001), and increases in spine size have been
Gallagher, 1997), and the neocortex (Rioult-Pedotti et al., detected after induction of LTP (Matsuzaki et al., 2004; Lang et
1998; Roman et al., 1999; Saar et al., 1999). Structural changes al., 2004). Spines that are initially small are more likely to
linked to synaptic strength can be addressed experimentally undergo a persistent expansion with stimulation than initially
with stimulation patterns that induce long-term potentiation large spines, implying that small spines, likely with a small
 BR A IN RE S E A RCH 1 1 84 ( 20 0 7 ) 6 5 –7 1 67

number of AMPA receptors, are preferential sites of LTP (Trachtenberg et al., 2002). Other studies (Zuo et al., 2005b)
induction (Matsuzaki et al., 2004). Conversely, induction of found that long-term (2–4 weeks) sensory deprivation,
LTD in the hippocampus by low-frequency stimulation leads through trimming all whiskers on one side of the face, reduced
to spine shrinkage (Zhou et al., 2004) and retraction (Zhou et the rate of spine elimination in barrel cortex layer 5 pyramidal
al., 2004; Nagerl et al., 2004). These changes in spine number neuron apical tufts in young adolescent mice (4–6 weeks) but
and morphology are bidirectionally modifiable and reversible. had no effect on the rate of spine addition. When whiskers
Thus, changes in spine morphology may be linked to changes were allowed to re-grow, the restored input accelerated spine
in synaptic function. elimination, which quickly compensated for the period of rate
reduction. A lesser decrement in spine elimination was seen
with extended whisker trimming in the adult, although
5. In vivo imaging of spines whisker restoration did not lead to a recovery of the
elimination. Additionally, even with extended (1–6 months
In addition to imaging dendritic spines briefly in artificial slice of age) whisker trimming, the spine elimination rate does
preparations, studies for the first time have followed individ- eventually fall to match the rate of spine addition (Zuo et al.,
ual spines and looked at spine turnover in the intact adult 2005b). These results are in contrast to other work in which
cortex (Grutzendler et al., 2002; Trachtenberg et al., 2002). Two rats are sensory deprived through dark rearing until 30 days of
groups in particular have looked at spine turnover over many age, when they were found to have 15% fewer spines in the
weeks in developing and adult mouse cortex, reporting 96% of visual cortex than those reared under normal conditions
spines to be stable over 1 month in the adult primary visual (Wallace and Bear, 2004). The difference may lie in the point at
cortex (Grutzendler et al., 2002) and 50% of spines to be stable which sensory deprivation began: before eye opening so that
over 1 month in the adult somatosensory cortex (Trachtenberg there was never any visual input or after 1 month of normal
et al., 2002). These results appear contradictory, and it is sensory development. The stability of spines in specific areas
unclear how stable spines are in the adult cortex. Much of this may be influenced by development of cortical plasticity in the
confusion arises from the variability in imaging techniques region. In general, dendritic spine stability increases with age,
and parameters used, regions of cortex imaged, animal strain but the absolute rates of spine elimination and addition,
and age, and classification of spine groups, although attempts which determine relative spine stability, may differ in
to reconcile some of these differences still do not totally agree different areas and layers of the cortex, corresponding to the
on the level of spine stability in the adult cortex (Holtmaat et plasticity of those layers. For example, in layer 2/3 of the visual
al., 2005; Zuo et al., 2005a). cortex, spines were found to have a higher density and a lower
Although the above studies disagree on the overall stability percentage of transient spines than in layer 5 (Holtmaat et al.,
of dendritic spines, they do have some common conclusions. 2005), and spines in visual cortex are more stable than those in
Foremost, these groups demonstrated that there are at least somatosensory cortex (Trachtenberg et al., 2002; Grutzendler
some spines which are quite stable and can persist for weeks et al., 2002; Holtmaat et al., 2005). Thus, sensory changes at
or even months in the adult animal. At the same time, there different developmental stages could have varying effects.
are spines which are much more short lived and are added and
eliminated within hours or days (Holtmaat et al., 2005; Zuo et
al., 2005a; Trachtenberg et al., 2002; Grutzendler et al., 2002). 6. Spine motility and sensory experience
The rate of spine addition is fairly stable throughout develop-
ment. In contrast, the spine elimination rate is initially high Live imaging studies of dendritic spines have shown that they
early in development, quickly drops down, and then gradually are dynamic structures, undergoing rapid changes in their
declines until it approximates the addition rate, leading to shape and size, termed spine motility. Motility of spines and
fairly stable spine densities in the adult while spine turnover is filopodia has been proposed to be important in synaptogen-
still ongoing at a low level (Trachtenberg et al., 2002; esis (Ziv and Smith, 1996; Fiala et al., 1998; Petrak et al., 2005),
Grutzendler et al., 2002; Zuo et al., 2005b; Holtmaat et al., and spine motility does decrease as development proceeds
2005). Large spines, such as thick mushroom-shaped spines, (Dunaevsky et al., 1999; Lendvai et al., 2000; Majewska and Sur,
tend to be more stable than thin spines or filopodia (Zuo et al., 2003; Oray et al., in press). In addition, sensory deprivation
2005a; Holtmaat et al., 2005; Trachtenberg et al., 2002). Thus, it studies suggest that spine motility is involved in experience-
seems that there are at least two groups of dendritic spines in dependent plasticity, suggesting that motility may play a role
the cortex: those that tend to be stable and those that are in new synapse formation (Lendvai et al., 2000). In the barrel
short-lived, which can be loosely correlated with their cortex, deprivation via 1–3 days of whisker trimming reduces
morphology. Furthermore, of all the spines formed, only a the motility of spines and filopodia in slices, although only
small fraction may be stabilized. However, it remains to be during a restricted developmental window. In addition, the
seen what determines which spines are stabilized and which tuning of sensory maps was perturbed by this deprivation,
are eliminated, whether small spines are already tagged for indicating that sensory deprivation, by decreasing spine
removal, or are simply more susceptible to it. motility, may have disrupted experience-dependent forma-
Sensory experience can alter spine stability in vivo. tion and rearrangement of synaptic connections (Lendvai et
Trachtenberg and colleagues found that 4 days of deprivation al., 2000). In contrast to monocular deprivation, binocular
in a subset of whiskers by trimming in a chessboard pattern deprivation from before eye opening increases spine motility
decreased the proportion of stable spines in the somatosen- in slices from the primary visual cortex during the peak of the
sory cortex, without changing the overall spine density critical period. The authors in this study proposed that
68 BR A IN RE S EA RCH 1 1 84 ( 20 0 7 ) 6 5 –71

motility could be downregulated by synaptic activity, perhaps potential postsynaptic partners towards active presynaptic
serving to stabilize a nascent synaptic connection (Majewska structures and subsequent synapse formation.
and Sur, 2003). Thus, like spine turnover, sensory deprivation
can lead to different effects on spine motility depending on
the point at which sensory deprivation began: before initial 8. Molecular links of spine structure and
sensory input or after a period of normal sensory develop- synaptic plasticity
ment. These studies suggest that the timing of sensory
experience may help to shape synaptic connections. However, Although many molecules are implicated in the regulation of
other studies have shown that in slices from the cerebellum spine morphology and dynamics (see Tashiro and Yuste, 2003;
(Dunaevsky et al., 2001; Deng and Dunaevsky, 2005) and Lippman and Dunaevsky, 2005 for review), a few have been
hippocampus (Umeda et al., 2005), motile spines often have also shown to play key roles in synaptic plasticity and
and retain presynaptic contacts. Furthermore, imaging both learning. One of the most prevalent forms of LTP is dependent
the presynaptic varicosity and postsynaptic spine simulta- on activation of NMDA receptors and subsequent calcium
neously showed that both structures can undergo rapid influx (Bliss and Collingridge, 1993; Nicoll and Malenka, 1995).
morphological changes that are independently regulated but This activation leads to, among other changes, a redistribution
maintain an overlapping region (Umeda et al., 2005) (but see of AMPA receptors to the post-synaptic density (Malinow and
Deng and Dunaevsky, 2005). So does motility represent Malenka, 2002; Song and Huganir, 2002). Removal of AMPA
dynamic, directed synapse formation? Some studies suggest receptors from the postsynaptic membrane is likewise
that motility and the resulting changes in spine morphology thought to play a role in long-term depression (LTD) of
regulate diffusion of the spines biochemical components and synaptic responses (Beattie et al., 2000; Lu et al., 2001; Xiao et
the spine's ability to compartmentalize local calcium transi- al., 2004). Thus NMDA-regulated redistribution of AMPA
ents (Majewska et al., 2000). It is possible the motility performs receptors seems to be a key mechanism for modifying
multiple functions, both seeking out synaptic partners and synaptic strength. Calcium/calmodulin-kinase II (CaMKII), a
regulating transduction of signals that could then serve to serine/threonine kinase regulated by the binding of calcium-
stabilize the connection. calmodulin and autophosphorylation, is able to phosphory-
late a wide range of postsynaptic targets, including NMDA and
AMPA receptors (Soderling et al., 2001), and its function is
7. Spine motility and synaptic activity important for synaptic plasticity and learning and memory
(Giese et al., 1998; Frankland et al., 2001; Meyer and Shen,
Several lines of evidence suggest that synaptic activity 2000). CaMKII is cytoplasmic and partially actin bound in the
influences the motility of dendritic spines. Blocking evoked inactive state but rapidly translocates to the postsynaptic
presynaptic release of neurotransmitter by bath application of density upon synaptic activation, where it remains in the
tetrodotoxin leads to an increase in spine motility (Korkotian active state during stimulation for minutes afterward. This
and Segal, 2001). Conversely, activation of AMPA or NMDA translocation requires elevated calcium and NMDAR activa-
receptors decreases spine motility (Fischer et al., 2000; Oray et tion, suggesting that it may play a major role in activity-
al., in press). Further studies have shown that motility can be dependent synaptic plasticity (Shen and Meyer, 1999; Meyer
decreased with calcium influx via AMPAR or NMDAR activa- and Shen, 2000). Actin itself may also be important in LTP
tion or depolarization and activation of voltage-gated calcium induction, as well as maintenance of AMPARs at the synapse
channels (Oertner and Matus, 2005). Although spine motility is (Kim and Lisman, 1999; Krucker et al., 2000; Cox et al., 2003)
decreased by glutamate receptor activation, other studies and is regulated by activation of NMDA receptors and calcium
show that spines are also maintained by spontaneous influx (Star et al., 2002). These same signaling pathways are
miniature release of glutamate and AMPA receptor activation also important in the regulation of spine dynamics.
(McKinney et al., 1999). These results suggest that spines are Actin is abundant in spines and regulates rapid shifts in
stabilized by synaptic activity, supporting the notion that spine morphology (Fischer et al., 1998; Zito et al., 2004).
motility facilitates synapse formation. More recent studies Stimulation patterns that induce LTP also induce actin to
have shown that blockade of evoked activity with TTX for an shift from a soluble to a polymerized form, simultaneously
hour or more leads to an increase in filopodia, stubby spines, enlarging dendritic spine heads (Okamoto et al., 2004; Lin et
and spine head protrusions (Petrak et al., 2005; Richards et al., al., 2005). In contrast, low-frequency stimulation, used to
2005). This seems at odds with the evidence presented above induce LTD, depolymerizes actin and shrinks spine heads
that LTP increases spine density. However, Richards et al. go (Okamoto et al., 2004). Studies using 2-photon uncaging of
on to show that the new protrusions may extend towards sites glutamate upon individual spines demonstrated that stimu-
of glutamate release and form new synaptic connections lated spines undergo long-lasting increases in size (Matsuzaki
(Richards et al., 2005). In addition, Lohmann et al. demonstrate et al., 2004). The initial expansion is dependent on stimulation
in slice cultures that local calcium transients can bidirection- of NMDA receptors, calmodulin activity, and actin polymeri-
ally regulate the motility of filopodia; low levels of calcium zation. A persistent increase in the size of these spines is
increase filopodial growth until higher levels halt that growth blocked by inhibition of CaMKII and is accompanied by
and stabilize a potential new connection (Lohmann et al., increases in AMPAR-mediated currents (Matsuzaki et al.,
2005). It is likely that patterns of local activity dictate the level 2004). Inducing LTP or polymerizing actin by phalloidin
of calcium in the spine. Thus, the strength of synaptic activity treatment also leads to an increase in CaMKII in the spine
could dictate the direction of its regulation of motility, guiding and the bound fraction of CaMKII (Okamoto et al., 2004;
 BR A IN RE S E A RCH 1 1 84 ( 20 0 7 ) 6 5 –7 1 69

Otmakhov et al., 2004), which can further recruit CaMKII morphological correlates of synaptic activity and efficacy
through self-scaffolding while calcium is present (Hudmon et change in the rat hippocampal slice. Brain Res. 309, 35–46.
Chen, B.E., Lendvai, B., Nimchinsky, E.A., Burbach, B., Fox, K.,
al., 2005). Since stimulation of actin polymerization can lead to
Svoboda, K., 2000. Imaging high-resolution structure of
transient increases in spine size in the absence of CaMKII
GFP-expressing neurons in neocortex in vivo. Learn. Mem.
activity, it seems reasonable to hypothesize that actin initiates 7, 433–441.
spine changes, which are then stabilized by CaMKII. However, Coleman, P.D., Riesen, A.H., 1968. Environmental effects on
injection of phosphorylated CaMKII or CamKII activators into cortical dendritic fields: I. Rearing in the dark. J. Anat. 102,
hippocampal neurons potentiated eEPSCs, the growth of 363–374.
filopodia, and formation of new spines ( Jourdain et al., 2003). Cox, P.R., Fowler, V., Xu, B., Sweatt, J.D., Paylor, R., Zoghbi, H.Y.,
2003. Mice lacking Tropomodulin-2 show enhanced long-term
Since actin and CaMKIIbeta directly interact, activated CaMKII
potentiation, hyperactivity, and deficits in learning and
may be capable of initiating actin polymerization. Thus it memory. Mol. Cell. Neurosci. 23, 1–12.
seems that actin and CaMKII interact to promote both changes Dailey, M.E., Smith, S.J., 1996. The dynamics of dendritic structure
in spines, receptor recruitment, and synaptic strength, in developing hippocampal slices. J. Neurosci. 16, 2983–2994.
although other molecules are certainly involved. Deng, J., Dunaevsky, A., 2005. Dynamics of dendritic spines and
their afferent terminals: spines are more motile than
presynaptic boutons. Dev. Biol. 277, 366–377.
Desmond, N.L., Levy, W.B., 1986. Changes in the postsynaptic
9. Conclusions density with long-term potentiation in the dentate gyrus.
J. Comp. Neurol. 253, 476–482.
Diamond, M.C., Lindner, B., Johnson, R., Bennett, E.L., Rosenzweig,
There is a great deal of plasticity in the adult CNS. Much of this
M.R., 1975. Differences in occipital cortical synapses from
plasticity is localized at dendritic spines, the postsynaptic sites
environmentally enriched, impoverished, and standard colony
of excitatory connections. Changes in spine density, morphol- rats. J. Neurosci. Res. 1, 109–119.
ogy, and motility have been shown to occur with paradigms Dunaevsky, A., Tashiro, A., Majewska, A., Mason, C., Yuste, R.,
that induce synaptic plasticity, as well as altered sensory 1999. Developmental regulation of spine motility in the
experience and neuronal activity. Increasing synaptic strength mammalian central nervous system. Proc. Natl. Acad. Sci.
tends to produce new spines, and enlarge and stabilize existing U. S. A. 96, 13438–13443.
Dunaevsky, A., Blazeski, R., Yuste, R., Mason, C., 2001. Spine
spines, whereas synaptic depression causes spine shrinkage
motility with synaptic contact. Nat. Neurosci. 4, 685–686.
and retraction. These changes seem to be paralleled in living Engert, F., Bonhoeffer, T., 1999. Dendritic spine changes associated
tissue in vitro, and in intact adult animals, where spines fall with hippocampal long-term synaptic plasticity. Nature 399,
into two groups: those that are large and for the most part 66–70.
stable, and those that are more transient and smaller. In Eyre, M.D., Richter-Levin, G., Avital, A., Stewart, M.G., 2003.
addition, developmental decreases in spine turnover parallel Morphological changes in hippocampal dentate gyrus
synapses following spatial learning in rats are transient. Eur.
nicely the decreased capacity for plasticity and termination of
J. Neurosci. 17, 1973–1980.
critical periods. However, much is left to learn. Although spine
Federmeier, K.D., Kleim, J.A., Greenough, W.T., 2002.
motility seems to be involved in synaptogenesis early in Learning-induced multiple synapse formation in rat
development, does continued spine motility serve the same cerebellar cortex. Neurosci. Lett. 332, 180–184.
purpose? Motility can be correlated with alterations in network Fiala, J.C., Allwardt, B., Harris, K.M., 2002. Dendritic spines do not
organization and biochemical compartmentalization. But does split during hippocampal LTP or maturation. Nat. Neurosci. 5,
induction of synaptic plasticity itself cause changes in motility? 297–298.
Fiala, J.C., Feinberg, M., Popov, V., Harris, K.M., 1998.
Motility can occur, while a synaptic connection is maintained,
Synaptogenesis via dendritic filopodia in developing
but it is unclear if spine movement alone is substantially
hippocampal area CA1. J. Neurosci. 18, 8900–8911.
altering synaptic transmission, or if it is due to the shuffling of Fifkova, E., Anderson, C.L., 1981. Stimulation-induced changes in
key postsynaptic proteins such as actin, CamKII, and glutamate dimensions of stalks of dendritic spines in the dentate
receptors. While global changes in activity can alter spines, is molecular layer. Exp. Neurol. 74, 621–627.
there a specific pattern of network activity most efficacious for Fifkova, E., Van Harreveld, A., 1977. Long-lasting morphological
specific changes in spine properties such as morphology? And changes in dendritic spines of dentate granular cells following
stimulation of the entorhinal area. J. Neurocytol. 6, 211–230.
lastly, although there is extensive evidence that spines can be
Fischer, M., Kaech, S., Knutti, D., Matus, A., 1998. Rapid actin-based
involved in synaptic plasticity, it has yet to be shown that plasticity in dendritic spines. Neuron 20, 847–854.
spines are sites specifically acted upon during learning and Fischer, M., Kaech, S., Wagner, U., Brinkhaus, H., Matus, A., 2000.
memory formation. Glutamate receptors regulate actin-based plasticity in
dendritic spines. Nat. Neurosci. 3, 887–894.
Frankland, P.W., O'Brien, C., Ohno, M., Kirkwood, A., Silva, A.J.,
REFERENCES 2001. Alpha-CaMKII-dependent plasticity in the cortex is
required for permanent memory. Nature 411, 309–313.
Geinisman, Y., Berry, R.W., Disterhoft, J.F., Power, J.M., Van der
Beattie, E.C., Carroll, R.C., Yu, X., Morishita, W., Yasuda, H., von Zee, E.A., 2001. Associative learning elicits the formation of
Zastrow, M., Malenka, R.C., 2000. Regulation of AMPA receptor multiple-synapse boutons. J. Neurosci. 21, 5568–5573.
endocytosis by a signaling mechanism shared with LTD. Nat. Geinisman, Y., Disterhoft, J.F., Gundersen, H.J., McEchron, M.D.,
Neurosci. 3, 1291–1300. Persina, I.S., Power, J.M., van der Zee, E.A., West, M.J., 2000.
Bliss, T.V., Collingridge, G.L., 1993. A synaptic model of memory: Remodeling of hippocampal synapses after
long-term potentiation in the hippocampus. Nature 361, 31–39. hippocampus-dependent associative learning. J. Comp.
Chang, F.L., Greenough, W.T., 1984. Transient and enduring Neurol. 417, 49–59.
70 BR A IN RE S EA RCH 1 1 84 ( 20 0 7 ) 6 5 –71

Giese, K.P., Fedorov, N.B., Filipkowski, R.K., Silva, A.J., 1998. Lippman, J., Dunaevsky, A., 2005. Dendritic spine morphogenesis
Autophosphorylation at Thr286 of the alpha and plasticity. J. Neurobiol. 64, 47–57.
calcium-calmodulin kinase II in LTP and learning. Science 279, Lohmann, C., Finski, A., Bonhoeffer, T., 2005. Local calcium
870–873. transients regulate the spontaneous motility of dendritic
Grutzendler, J., Kasthuri, N., Gan, W.B., 2002. Long-term dendritic filopodia. Nat. Neurosci. 8, 305–312.
spine stability in the adult cortex. Nature 420, 812–816. Lu, W., Man, H., Ju, W., Trimble, W.S., MacDonald, J.F., Wang, Y.T.,
Harris, K.M., 1999. Structure, development, and plasticity of 2001. Activation of synaptic NMDA receptors induces
dendritic spines. Curr. Opin. Neurobiol. 9, 343–348. membrane insertion of new AMPA receptors and LTP in
Harris, K.M., Fiala, J.C., Ostroff, L., 2003. Structural changes at cultured hippocampal neurons. Neuron 29, 243–254.
dendritic spine synapses during long-term potentiation. Majewska, A., Sur, M., 2003. Motility of dendritic spines in visual
Philos. Trans. R. Soc. Lond., B Biol. Sci. 358, 745–748. cortex in vivo: changes during the critical period and effects of
Holtmaat, A.J., Trachtenberg, J.T., Wilbrecht, L., Shepherd, G.M., visual deprivation. Proc. Natl. Acad. Sci. U. S. A. 100,
Zhang, X., Knott, G.W., Svoboda, K., 2005. Transient and 16024–16029.
persistent dendritic spines in the neocortex in vivo. Neuron 45, Majewska, A., Tashiro, A., Yuste, R., 2000. Regulation of spine
279–291. calcium dynamics by rapid spine motility. J. Neurosci. 20,
Hudmon, A., Lebel, E., Roy, H., Sik, A., Schulman, H., Waxham, 8262–8268.
M.N., De Koninck, P., 2005. A mechanism for Maletic-Savatic, M., Malinow, R., Svoboda, K., 1999. Rapid dendritic
Ca2+/calmodulin-dependent protein kinase II clustering at morphogenesis in CA1 hippocampal dendrites induced by
synaptic and nonsynaptic sites based on self-association. synaptic activity. Science 283, 1923–1927.
J. Neurosci. 25, 6971–6983. Malinow, R., Malenka, R.C., 2002. AMPA receptor trafficking and
Ivanco, T.L., Racine, R.J., Kolb, B., 2000. Morphology of layer III synaptic plasticity. Annu. Rev. Neurosci. 25, 103–126.
pyramidal neurons is altered following induction of LTP in Matsuzaki, M., Ellis-Davies, G.C., Nemoto, T., Miyashita, Y., Iino,
sensorimotor cortex of the freely moving rat. Synapse 37, M., Kasai, H., 2001. Dendritic spine geometry is critical for
16–22. AMPA receptor expression in hippocampal CA1 pyramidal
Jones, T.A., Klintsova, A.Y., Kilman, V.L., Sirevaag, A.M., neurons. Nat. Neurosci. 4, 1086–1092.
Greenough, W.T., 1997. Induction of multiple synapses by Matsuzaki, M., Honkura, N., Ellis-Davies, G.C., Kasai, H., 2004.
experience in the visual cortex of adult rats. Neurobiol. Learn. Structural basis of long-term potentiation in single dendritic
Mem. 68, 13–20. spines. Nature 429, 761–766.
Jourdain, P., Fukunaga, K., Muller, D., 2003. McKernan, M.G., Shinnick-Gallagher, P., 1997. Fear conditioning
Calcium/calmodulin-dependent protein kinase II contributes induces a lasting potentiation of synaptic currents in vitro.
to activity-dependent filopodia growth and spine formation. Nature 390, 607–611.
J. Neurosci. 23, 10645–10649. McKinney, R.A., Capogna, M., Durr, R., Gahwiler, B.H., Thompson,
Kim, C.H., Lisman, J.E., 1999. A role of actin filament in synaptic S.M., 1999. Miniature synaptic events maintain dendritic
transmission and long-term potentiation. J. Neurosci. 19, spines via AMPA receptor activation. Nat. Neurosci. 2, 44–49.
4314–4324. Meyer, T., Shen, K., 2000. In and out of the postsynaptic region:
Kleim, J.A., Barbay, S., Nudo, R.J., 1998. Functional reorganization signalling proteins on the move. Trends Cell Biol. 10, 238–244.
of the rat motor cortex following motor skill learning. Moser, E., Moser, M.B., Andersen, P., 1993. Synaptic potentiation in
J. Neurophysiol. 80, 3321–3325. the rat dentate gyrus during exploratory learning. Neuroreport
Kleim, J.A., Lussnig, E., Schwarz, E.R., Comery, T.A., Greenough, 5, 317–320.
W.T., 1996. Synaptogenesis and Fos expression in the motor Moser, M.B., Trommald, M., Andersen, P., 1994. An increase in
cortex of the adult rat after motor skill learning. J. Neurosci. dendritic spine density on hippocampal CA1 pyramidal cells
16, 4529–4535. following spatial learning in adult rats suggests the formation
Knott, G.W., Quairiaux, C., Genoud, C., Welker, E., 2002. Formation of new synapses. Proc. Natl. Acad. Sci. U. S. A. 91, 12673–12675.
of dendritic spines with GABAergic synapses induced by Moser, M.B., Trommald, M., Egeland, T., Andersen, P., 1997. Spatial
whisker stimulation in adult mice. Neuron 34, 265–273. training in a complex environment and isolation alter the
Korkotian, E., Segal, M., 2001. Regulation of dendritic spine motility spine distribution differently in rat CA1 pyramidal cells.
in cultured hippocampal neurons. J. Neurosci. 21, 6115–6124. J. Comp. Neurol. 380, 373–381.
Krucker, T., Siggins, G.R., Halpain, S., 2000. Dynamic actin Murthy, V.N., Schikorski, T., Stevens, C.F., Zhu, Y., 2001. Inactivity
filaments are required for stable long-term potentiation (LTP) produces increases in neurotransmitter release and synapse
in area CA1 of the hippocampus. Proc. Natl. Acad. Sci. U. S. A. size. Neuron 32, 673–682.
97, 6856–6861. Nagerl, U.V., Eberhorn, N., Cambridge, S.B., Bonhoeffer, T., 2004.
Lang, C., Barco, A., Zablow, L., Kandel, E.R., Siegelbaum, S.A., Bidirectional activity-dependent morphological plasticity in
Zakharenko, S.S., 2004. Transient expansion of synaptically hippocampal neurons. Neuron 44, 759–767.
connected dendritic spines upon induction of hippocampal Nicoll, R.A., Malenka, R.C., 1995. Contrasting properties of two
long-term potentiation. Proc. Natl. Acad. Sci. U. S. A. 101, forms of long-term potentiation in the hippocampus. Nature
16665–16670. 377, 115–118.
Leggio, M.G., Mandolesi, L., Federico, F., Spirito, F., Ricci, B., Gelfo, Nimchinsky, E.A., Sabatini, B.L., Svoboda, K., 2002. Structure
F., Petrosini, L., 2005. Environmental enrichment promotes and function of dendritic spines. Annu. Rev. Physiol. 64,
improved spatial abilities and enhanced dendritic growth in 313–353.
the rat. Behav. Brain Res. 163, 78–90. Oertner, T.G., Matus, A., 2005. Calcium regulation of actin
Lendvai, B., Stern, E., Chen, B., Svoboda, K., 2000. dynamics in dendritic spines. Cell Calcium 37, 477–482.
Experience-dependent plasticity of dendritic spines in the Okamoto, K., Nagai, T., Miyawaki, A., Hayashi, Y., 2004. Rapid
developing rat barrel cortex in vivo. Nature 404, 876–881. and persistent modulation of actin dynamics regulates
Leuner, B., Falduto, J., Shors, T.J., 2003. Associative memory postsynaptic reorganization underlying bidirectional
formation increases the observation of dendritic spines in the plasticity. Nat. Neurosci. 7, 1104–1112.
hippocampus. J. Neurosci. 23, 659–665. Oray, S., Majewska, A., Sur, M., in press. Effects of synaptic activity
Lin, B., Kramar, E.A., Bi, X., Brucher, F.A., Gall, C.M., Lynch, G., 2005. on dendritic spine motility of developing cortical layer v
Theta stimulation polymerizes actin in dendritic spines of pyramidal neurons. Cereb. Cortex. (electronic publication
hippocampus. J. Neurosci. 25, 2062–2069. ahead of print, August 24, 2005).
 BR A IN RE S E A RCH 1 1 84 ( 20 0 7 ) 6 5 –7 1 71

Otmakhov, N., Tao-Cheng, J.H., Carpenter, S., Asrican, B., actin in dendritic spines and its regulation by activity. Nat.
Dosemeci, A., Reese, T.S., Lisman, J., 2004. Persistent Neurosci. 5, 239–246.
accumulation of calcium/calmodulin-dependent protein Tashiro, A., Yuste, R., 2003. Structure and molecular organization
kinase II in dendritic spines after induction of NMDA of dendritic spines. Histol. Histopathol. 18, 617–634.
receptor-dependent chemical long-term potentiation. Toni, N., Buchs, P.A., Nikonenko, I., Bron, C.R., Muller, D., 1999. LTP
J. Neurosci. 24, 9324–9331. promotes formation of multiple spine synapses between a
Petrak, L.J., Harris, K.M., Kirov, S.A., 2005. Synaptogenesis on single axon terminal and a dendrite. Nature 402, 421–425.
mature hippocampal dendrites occurs via filopodia and Trachtenberg, J.T., Chen, B.E., Knott, G.W., Feng, G., Sanes, J.R.,
immature spines during blocked synaptic transmission. Welker, E., Svoboda, K., 2002. Long-term in vivo imaging of
J. Comp. Neurol. 484, 183–190. experience-dependent synaptic plasticity in adult cortex.
Richards, D.A., Mateos, J.M., Hugel, S., de Paola, V., Caroni, P., Nature 420, 788–794.
Gahwiler, B.H., McKinney, R.A., 2005. Glutamate induces the Trommald, M., Hulleberg, G., Andersen, P., 1996. Long-term
rapid formation of spine head protrusions in hippocampal slice potentiation is associated with new excitatory spine synapses
cultures. Proc. Natl. Acad. Sci. U. S. A. 102, 6166–6171. on rat dentate granule cells. Learn. Mem. 3, 218–228.
Rioult-Pedotti, M.S., Friedman, D., Hess, G., Donoghue, J.P., 1998. Umeda, T., Ebihara, T., Okabe, S., 2005. Simultaneous
Strengthening of horizontal cortical connections following skill observation of stably associated presynaptic varicosities and
learning. Nat. Neurosci. 1, 230–234. postsynaptic spines: morphological alterations of CA3-CA1
Rogan, M.T., Staubli, U.V., LeDoux, J.E., 1997. Fear conditioning synapses in hippocampal slice cultures. Mol. Cell. Neurosci. 28,
induces associative long-term potentiation in the amygdala. 264–274.
Nature 390, 604–607. Valverde, F., 1967. Apical dendritic spines of the visual cortex and
Roman, F.S., Truchet, B., Marchetti, E., Chaillan, F.A., light deprivation in the mouse. Exp. Brain Res. 3, 337–352.
Soumireu-Mourat, B., 1999. Correlations between Van Harreveld, A., Fifkova, E., 1975. Swelling of dendritic spines in
electrophysiological observations of synaptic plasticity the fascia dentata after stimulation of the perforant fibers as a
modifications and behavioral performance in mammals. Prog. mechanism of post-tetanic potentiation. Exp. Neurol. 49,
Neurobiol. 58, 61–87. 736–749.
Saar, D., Grossman, Y., Barkai, E., 1999. Reduced synaptic Wallace, W., Bear, M.F., 2004. A morphological correlate of
facilitation between pyramidal neurons in the piriform cortex synaptic scaling in visual cortex. J. Neurosci. 24, 6928–6938.
after odor learning. J. Neurosci. 19, 8616–8622. Xiao, M.Y., Wasling, P., Hanse, E., Gustafsson, B., 2004. Creation of
Sacchetti, B., Lorenzini, C.A., Baldi, E., Bucherelli, C., Roberto, M., AMPA-silent synapses in the neonatal hippocampus. Nat.
Tassoni, G., Brunelli, M., 2001. Long-lasting hippocampal Neurosci. 7, 236–243.
potentiation and contextual memory consolidation. Eur. J. Zhou, Q., Homma, K.J., Poo, M.M., 2004. Shrinkage of dendritic
Neurosci. 13, 2291–2298. spines associated with long-term depression of hippocampal
Shen, K., Meyer, T., 1999. Dynamic control of CaMKII translocation synapses. Neuron 44, 749–757.
and localization in hippocampal neurons by NMDA receptor Zito, K., Knott, G., Shepherd, G.M., Shenolikar, S., Svoboda, K., 2004.
stimulation. Science 284, 162–166. Induction of spine growth and synapse formation by regulation
Soderling, T.R., Chang, B., Brickey, D., 2001. Cellular signaling of the spine actin cytoskeleton. Neuron 44, 321–334.
through multifunctional Ca2+/calmodulin-dependent protein Ziv, N.E., Smith, S.J., 1996. Evidence for a role of dendritic filopodia
kinase II. J. Biol. Chem. 276, 3719–3722. in synaptogenesis and spine formation. Neuron 17, 91–102.
Song, I., Huganir, R.L., 2002. Regulation of AMPA receptors during Zuo, Y., Lin, A., Chang, P., Gan, W.B., 2005a. Development of
synaptic plasticity. Trends Neurosci. 25, 578–588. long-term dendritic spine stability in diverse regions of
Sorra, K.E., Harris, K.M., 1998. Stability in synapse number and size cerebral cortex. Neuron 46, 181–189.
at 2 Hr after long-term potentiation in hippocampal area CA1. Zuo, Y., Yang, G., Kwon, E., Gan, W.B., 2005b. Long-term sensory
J. Neurosci. 18, 658–671. deprivation prevents dendritic spine loss in primary
Star, E.N., Kwiatkowski, D.J., Murthy, V.N., 2002. Rapid turnover of somatosensory cortex. Nature 436, 261–265.