Learning to hear: plasticity of auditory cortical processing

Johannes C Dahmen and Andrew J King

Sensory experience and auditory cortex plasticity are intimately the most well known model [1] – have provided us with
related. This relationship is most striking during infancy when valuable insights into how early sensory experience shapes
changes in sensory input can have profound effects on the the maturation of cortical circuits. One reason for the visual
functional organization of the developing cortex. But a system’s popularity in plasticity research lies in the relative
considerable degree of plasticity is retained throughout life, as ease with which aspects of an animal’s visual environment
demonstrated by the cortical reorganization that follows can be controlled. By contrast, it is hard to prevent the
damage to the sensory periphery or by the more controversial acoustic input arising from self-vocalizations, the animal’s
changes in response properties that are thought to accompany movement or handling of the animal and makes the kind of
perceptual learning. Recent studies in the auditory system have deprivation experiments that have proven such a success
revealed the remarkably adaptive nature of sensory processing in visual neurobiology more difficult to carry out in the
and provided important insights into the way in which cortical auditory domain.
circuits are shaped by experience and learning.

Addresses
On the other hand, it is often easier to control stimulus
Department of Physiology, Anatomy and Genetics, Sherrington input statistics in the auditory system and some of the
Building, University of Oxford, Parks Road, Oxford OX1 3PT, UK clearest effects of perceptual learning on neuronal
response properties have been described in the auditory
Corresponding author: King, Andrew J. (andrew.king@dpag.ox.ac.uk)
cortex. Moreover, various approaches for distorting an
animal’s early acoustic environment in a controlled
Current Opinion in Neurobiology 2007, 17:456–464 fashion have now been developed, which have yielded
important insights into the role of experience in cortical
This review comes from a themed issue on development. Finally, the success of cochlear implants
Sensory systems
Edited by Peter Mombaerts and Tony Zador
has provided further impetus for studying how abnormal
sensory experience influences the structure and function
Available online 21st August 2007 of the auditory system, as well as a model for determining
0959-4388/$ – see front matter how the artificial input provided by sensory prostheses is
# 2007 Elsevier Ltd. All rights reserved. processed and interpreted within the brain. In this article,
we review recent findings in each of these areas.
DOI 10.1016/j.conb.2007.07.004
Plasticity of cortical processing during
development
Introduction Most studies of visual cortex plasticity have focused on
Sensory experience shapes the structure and function of the role of experience in the development of response
neural systems over various timescales and throughout an properties, such as ocular dominance and orientation
organism’s lifetime in a way that, presumably, is adaptive selectivity, which emerge at the level of the cortex [1].
for the organism. Although most pronounced during By contrast, the impact of environmental manipulations
infancy when the neural architecture is first established on the developing auditory cortex has usually been
and then refined to optimize the processing of sensory measured in terms of the selectivity of the neurons to
information, plasticity of sensory systems continues into sound frequency (Figure 1), which has its origin in the
adulthood with neural representations remaining capable tuning of the hair cells in the cochlea. This partly reflects
of undergoing substantial reorganization in response to our current understanding of auditory cortical processing
the altered inputs associated with peripheral injury or – it is still largely unclear what additional response
perceptual learning. Despite several decades of research characteristics arise within the cortex – but also the results
in this field, our understanding of the nature and beha- of other studies demonstrating plasticity in the cortical
vioural significance of the physiological changes brought representation of sound frequency in adult animals [2,3].
about by manipulating the sensory environment at differ-
ent stages of life remains far from complete. In infant rats, electrophysiological recordings have shown
that tone-evoked cortical responses first occur during
Traditionally, vision has been the system of choice for the second week after birth and are initially both more
exploring the effects of sensory experience on cortical broadly tuned and cover a smaller range of best frequen-
plasticity, particularly during development. A number of cies – the sound frequencies to which the neurons are
elegant experiments designed to control the early visual most sensitive – than those found in adult animals [4]. An
environment – with ocular dominance plasticity providing adult-like tonotopic representation is present in the

Current Opinion in Neurobiology 2007, 17:456–464 www.sciencedirect.com

 Learning to hear: plasticity of auditory cortical processing Dahmen and King 457

Figure 1

Reorganization of the cortical frequency map. (a) Schematic of a normal frequency map in a patch of mature A1. Neurons selective for high
frequencies can be found in the upper right hand corner while neurons tuned to progressively lower frequencies follow in the opposite direction.
The area occupied by each frequency band is roughly equal. (b) Schematic of a frequency map in which the region occupied by neurons
selective for frequencies of about 15 kHz is expanded. Frequencies just below and just above 15 kHz are underrepresented. Such distortions
in the tonotopic organization of the cortex have been observed following a number of experimental manipulations, including continuous
pure-tone stimulation during infancy, pairing of tones with cholinergic basal forebrain stimulation in adults and frequency discrimination training.

primary auditory cortex (A1) by about 4 weeks after birth in adulthood, the concept of a sensitive period has had to be
[4,5], with sideband inhibition maturing 2–3 weeks later revised [1,10]. Nevertheless, the neural circuitry that
[5]. This postnatal development of the auditory cortex is emerges during this time seems to constrain the plasticity
now known to be more heavily dependent upon the that is possible in later life. Until recently, it was unclear
acoustic environment experienced during early life than when artificial sounds have to be provided in order to alter
was previously thought to be the case. Thus, rat pups the functional organization of auditory cortex. Now de
exposed to pulsed single-tone stimuli develop an over- Villers-Sidani et al. [11] have shown that rat pups must
representation of that sound frequency at the expense of be exposed to repeated single-tone pulses within a narrow
other frequencies [4]. Similar experiments in mice, in three-day window during the second postnatal week to
which cortical activity was measured using transcranial expand the region of A1 tuned to that frequency. Although
flavoprotein fluorescence imaging, did not show this reor- surprisingly short compared to the critical period described
ganization, although larger responses were observed at the previously for the rat’s visual cortex [12], this window
exposure frequency [6]. Rearing rats in continuous noise, in coincides with the maturation of various A1 response
an attempt to remove patterned acoustic inputs, appears to properties that have been described using simple tonal
maintain the cortex in its immature state, with subsequent stimuli [11]. As in the visual system [13], however, elim-
exposure to either normal conditions or pulsed tone stimuli ination of patterned sensory input (by noise rearing) can
leading to a rapid reorganization of A1 frequency selectiv- extend the sensitive period for auditory development [5,7].
ity [5,7]. Whilst providing a dramatic demonstration of the
importance of experience in molding emerging neural For educational purposes, the importance of knowing
representations, it is unlikely that this plasticity arises about the timing of developmental sensitive periods is
solely within the cortex. Indeed, it has previously been obvious. It has long been recognized that language can be
shown that raising animals in abnormal acoustic environ- acquired more readily in infancy than in later life, which
ments can alter both frequency tuning [8] and other could be because early experience leads to neural changes
response properties [9] at subcortical levels too. that optimize perception in the native language and
constrain the subsequent learning of a new language
One of the most important questions in the study of sensory [14]. Musicians can also provide useful insights into
systems plasticity is the extent to which the influence of experience-dependent plasticity in humans [15], with a
experience is restricted to a sensitive or critical period of recent study showing that a sensitive period exists for the
development. Because extensive plasticity is also possible effects of musical training on motor performance [16].

www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:456–464

458 Sensory systems

Plasticity in later life approach was taken by Cheung et al. [24], who showed
The formation of new synapses and elimination of old ones that changes in self-generated vocalizations can alter the
is likely to be essential for producing a persistent physical way in which marmoset A1 neurons respond to these
remodelling of neural circuits in the face of changes in the sounds. They first surgically modified the vocal tracts of
sensory environment [17,18]. This turnover of synapses is the monkeys, which resulted in a change in the spectral
likely to be high in the infant, possibly preconditioning the structure of the ‘twitter’ call used in social communication.
developing neocortex to adapt rapidly to its sensory Recordings made several months later from A1 showed that
environment. The relatively lower synaptic turnover in responses to both altered and native calls were weaker and
the adult brain may limit its capacity to undergo plastic temporally less precise than those in control animals.
change [19,20]. In the adult animal, more stability may be Abnormal responses were also observed to frequency-
desirable and even necessary to achieve the efficiency and modulated sweeps resembling those found in the twitter
reliability of a mature neural system. However, if sensory calls, whereas the cortical representation of pure tones was
experience is altered sufficiently, occurs in association with unchanged. The plasticity observed in this study therefore
the release of neuromodulatory reinforcement signals or appears to be specific to the processing of complex sounds.
possesses behavioural relevance, cortical representations
can undergo considerable reorganization in adulthood. Perceptual learning and cortical plasticity
It has been known for a long time that perceptual abilities
Lesion-induced plasticity provides an example of the in a variety of tasks can be improved by training. The time
cortical changes arising from altered sensory inputs course and specificity of learning point to plasticity
[3,21]. By damaging hair cells in specific regions of the throughout life at different levels of cortical processing
cochlea, exposure to high-intensity sounds or other forms and the challenge today lies in identifying the changes in
of acoustic trauma can cause frequency-specific hearing neuronal response properties that are causally related to
loss. As a consequence, the area of A1 in which the lesioned the perceptual improvements [25]. Evidence for learning-
frequency range would normally be represented becomes related physiological changes in adult auditory cortex
occupied by an expanded representation of neighbouring have been accumulating since Recanzone et al. [26] first
sound frequencies. These changes can be largely pre- reported that training monkeys on a frequency discrimi-
vented, however, if the decrease in input due to the hearing nation task led to an expanded representation of the
loss is offset by appropriate acoustic stimulation. Noreña trained frequencies in A1. Enlargements in the repres-
and Eggermont [22] found that placing cats that had entations of the trained frequency range have also been
previously been exposed to intense noise in an enriched reported more recently in other species [27,28]. In the
acoustic environment for several weeks reduced the hear- study by Polley et al. [28], rats were trained on either a
ing loss and stopped the associated cortical tonotopic map frequency or intensity recognition task. Both groups
reorganization that would normally result. In a subsequent exhibited an improvement in performance with training
study, normally hearing juvenile cats were exposed to the and changes were observed in A1 and in a secondary
same enriched acoustic environment [23]. Instead of auditory area that differed according to the task used.
leading to an enlarged cortical representation of these Thus, while reorganization of the tonotopic map accom-
sounds, as might be expected from the effects of tone panied learning in the frequency domain, the proportion
rearing during the sensitive period, Noreña et al. [23] of neurons tuned to the trained intensity range increased
found that the frequencies which featured in the enriched in animals that had undergone learning in the intensity
acoustic environment were actually under-represented at recognition task. Importantly, the same set of stimuli was
the expense of neighbouring frequencies, which is remi- used in both tasks, indicating that, in contrast to the
niscent of the cortical changes following frequency-specific receptive field changes described during infancy, the
hearing loss. What brings about those changes, which were observed cortical plasticity could not be attributed simply
not accompanied by any hearing loss, is largely unclear, to exposure to a particular stimulus configuration.
although the authors propose that forward suppression of
responses to the rapid tone-pips comprising the enriched Devoting more neurons to processing those aspects of a
acoustic environment could have led to the cortical neurons sensory stimulus that have particular behavioural relevance
becoming insensitive to those frequencies. In addition to may seem like a plausible way of improving perceptual
elucidating the mechanisms underlying the tonotopic reor- acuity within that dimension. However, such redistribution
ganization of A1, a major question left unanswered by this might not be a necessary condition for perceptual learning.
study and by the majority of those in which animals have In cats, improvements in sound frequency discrimination
been raised in abnormal acoustic conditions is what the have been reported without an accompanying increase in
perceptual consequences of these changes are. the representation of the trained frequency range [29,30],
although other changes in cortical response properties were
Most experiments investigating the effects of altered observed. Recent studies in humans have shown that rapid
auditory input on neural plasticity employ an external improvements in temporal discrimination [31] and speech
sound source that can be controlled easily. A different segregation [32] are accompanied by enhanced responses

Current Opinion in Neurobiology 2007, 17:456–464 www.sciencedirect.com

 Learning to hear: plasticity of auditory cortical processing Dahmen and King 459

in auditory cortex. But identifying the cellular basis for receptive field plasticity that would otherwise result from
these changes can usually be done only in animals and conditioning [47] or NB stimulation [48–50], while mice
future progress here will rely on recordings being made lacking the M1 muscarinic receptor subtype show weaker
while a task is being performed [33,34] (see review by NB-stimulation-induced changes [51]. These results
Fritz and colleagues this issue). suggest that neuromodulatory inputs enable auditory cor-
tex to acquire and store mnemonic traces about the import-
Most electrophysiological studies of cortical plasticity ance of specific acoustic stimuli. Indeed, NB stimulation
have focused on changes in neuronal tuning character- can rapidly produce associative memory for the paired tone
istics, as defined by variations in spike firing rate with the frequency [52] and the specificity of the memory is de-
stimulus parameter in question. That temporal response pendent upon the strength of NB stimulation [53].
properties are also malleable has been shown by training
rats to use the repetition rate of noise pulses as a cue to This pairing paradigm also evokes receptive field plasticity
locate a food source [35]. This led to A1 neurons, usually in the inferior colliculus, which is mediated by corticofugal
relatively poor at following stimuli with high temporal feedback [54,55] and in other cortical fields [56]. But
modulation rates, developing stronger responses to high- whereas repeatedly pairing NB stimulation with a high-
rate noise pulses and improved phase-locking to those frequency tone expanded the representation of that fre-
stimuli. In this case, the cortical responses encode rapidly quency without altering the width of the tuning curves of
varying acoustic events, but spike timing unrelated to the individual neurons in A1, neurons in the posterior auditory
temporal structure of the sound can also convey signifi- field became more selective for that frequency while
cant stimulus-related information [36]. By training ferrets responses to other frequencies were reduced [56]. Com-
to discriminate marmoset twitter calls from other natural bining sensory experience with NB stimulation may there-
sounds, Schnupp et al. [37] found no evidence for the fore help to shed light on both the role of neurons in
emergence of call selectivity in terms of the overall different cortical areas in experience-driven plasticity
response rates of A1 neurons, whereas the amount of and how their response properties change with learning.
information contained in their temporal firing patterns
was increased compared to naı̈ve animals (Figure 2). Adaptation to altered auditory inputs
The studies reviewed so far have investigated auditory
Neuromodulatory influences on cortical plasticity by manipulating the statistics of the input by
plasticity exposing subjects, either passively or in the context of
Considerable evidence has accumulated over the last 20 behavioural training, to particular sounds. Another app-
years that neuromodulators regulate cortical plasticity roach is to alter inputs by modifying the ears rather than
according to the behavioural state of the animal [2]. Atten- the acoustic environment. This has been done most
tion has focused primarily on acetylcholine released from frequently in the context of studies of auditory localiz-
the cholinergic nucleus basalis (NB), part of the basal ation, which have revealed considerable plasticity at both
forebrain, although other neuromodulators, including a perceptual level and in the underlying neural processes.
dopamine [38], noradrenaline [39] and serotonin [40] have
been implicated as well. One approach to investigating the As with other aspects of sensory processing, auditory
possible role of neuromodulatory inputs in cortical localization mechanisms are particularly sensitive to exp-
plasticity is to examine whether receptive field changes erience during infancy [10,57]. This is necessary because
are still possible following their removal. Indeed, the the physical cues – differences in the sound level or arrival
reorganization that takes place in somatosensory cortex time at the two ears plus the spectral modifications pro-
following digit amputation can be prevented by depletion duced by the external ears—which provide the basis for
of acetylcholine [41]. However, this seems not to be the spatial hearing change in value as an animal grows. The
case in the auditory cortex, as near complete elimination of plasticity in the system therefore allows accurate sound
cortical projections from the cholinergic basal forebrain localization to develop by calibrating the neural circuits in
does not affect the reorganization of the frequency map in the light of experience of the cues available to individual
A1 resulting from partial cochlear lesions [3]. listeners.

Whilst this result casts doubt on the exact role played by But once again, recent studies have shown that consider-
the cholinergic projection to the cortex in auditory repres- able plasticity is also present in the adult brain. Humans can
entational plasticity, it has been shown in numerous adapt to altered monaural spectral cues that result from
studies that pairing electrical stimulation of the NB with long-term reshaping of one of the external ears with a mold
sound presentation produces stimulus-specific changes in [58] (Figure 3a). Similarly, adult ferrets can relearn to
A1 response properties and map reorganization, which localize sound after manipulating binaural cues by occlud-
resemble the plasticity produced by classical conditioning ing one ear [59]. This rapid adaptation was observed,
or by long-term behavioural training [42–46]. Moreover, however, only if the ear-plugged animals received auditory
blockade of cortical acetylcholine receptors prevents the localization training, with more frequent training leading to

www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:456–464

460 Sensory systems

Figure 2

Plasticity of temporal pattern codes in auditory cortex. (a) Spectrograms of three marmoset twitters. (b,c) Raster plots illustrating responses to
the twitters of two neurons in ferret A1. The firing patterns of neuron 1 are characterized by high intra-call reliability, that is the neuron fires
action potentials at roughly the same point in time during each presentation of the same twitter. Furthermore, the spike pattern evoked by one
twitter call is clearly different from that produced in response to a different twitter. Neuron 1 therefore conveys information in its firing that can be
used to discriminate between twitters. The spike patterns of neuron 2, however, are much less reliable and distinctive and appear less informative
with respect to twitter identity. (d,e) Schnupp et al. [37] quantified the amount of information transmitted by the discharge patterns of A1 neurons
and showed that these codes have to be analysed at a resolution of 10–20 ms in order to extract maximum information. They also trained ferrets to
discriminate marmoset twitters from the vocalizations of other species and found that neurons in A1 of the trained ferrets (d) transmit more
information than neurons in A1 of naı̈ve ferrets that had never been exposed to the stimuli (e). (f) Difference between plots (d) and (e). Adapted
with permission from [37].

more rapid and complete adaptation (Figure 3b,c). The mammals as well as owls focusing on the midbrain
importance of behavioural context in promoting plasticity [57]. However, it has recently been shown that bilateral
has also been demonstrated by Bergan et al. [60], who inactivation of A1 impairs the capacity of adult
showed that hunting increased adaptive adjustments in ferrets to recover accurate auditory localization after
the auditory space map in the optic tectum of adult barn plugging one ear [61]. Moreover, no adaptation was
owls whose visual field had been displaced by prismatic seen in animals in which layer V neurons projecting
spectracles. from A1 to the inferior colliculus had been ablated
[61], further emphasizing the likely importance of
The role of auditory cortex in the plasticity of sound descending corticofugal pathways in learning-induced
localization is largely untested, with most studies in plasticity.

Current Opinion in Neurobiology 2007, 17:456–464 www.sciencedirect.com

 Learning to hear: plasticity of auditory cortical processing Dahmen and King 461

Figure 3

Plasticity of spatial hearing in adulthood. (a) Distribution of head movements (open squares and circles) made by a human listener to different
broadband noise locations, which are indicated by the gray dots and grid lines. Localization responses aligned well with the auditory targets in the
normal-hearing condition (left panel), but covered a much more restricted range of elevations when a rubber mold was inserted unilaterally into
the concha cavity of the right external ear (middle panel). Over time, however, localization accuracy recovers, indicating adaptation to the altered
spectral cues (right panel). Adapted with permission from [58]. (b) Stimulus-response plots showing the distribution of responses (ordinate) made
by a ferret as a function of stimulus location in the horizontal plane (abscissa). The size of the dots indicates, for a given speaker angle, the
proportion of responses made to different locations. Before occlusion of the left ear, the animal achieved 100% correct scores at all stimulus
directions (left panel), but performed poorly, particularly on the side of the earplug, when the left ear was occluded (middle panel). Further testing
with the earplug still in place, however, led to a recovery in localization accuracy (right panel). (c) Mean change in performance (averaged
across all speaker locations) over time in three groups of ferrets with unilateral earplugs. No change was found in trained ferrets (n = 3) that received
an earplug for 6 weeks, but were tested only at the start and end of this period (blue circles and regression line). Two other groups of animals
received an equivalent amount of training while the left ear was occluded. Although the earplug was in place for less time, a much faster rate of
improvement was observed in the animals that received daily training (n = 3; red diamonds and regression line) compared to those that were
tested every six days (n = 6; black squares and regression line). Adapted from [59].

Hearing loss and cortical plasticity different timescales and throughout an animal’s lifetime.
The finding that auditory localization abilities can be The finding that changes in the statistics of the acoustic
restored in individuals with altered cues to sound-source environment can profoundly alter the response properties
location has considerable implications for the recovery of of cortical neurons in the developing animal whilst having
function in the hearing impaired. However, early sensor- little or no effect in older animals attests to a particular
ineural hearing loss prevents the maturation of long-term sensitivity of the infant cortex that is also observed in
potentiation in layer V cortical neurons, presumably limit- other sensory systems. Nevertheless, it is now clear that
ing the capacity for subsequent plasticity [62]. The learning-induced plasticity, most likely enabled by a
absence of sound-evoked inputs results in auditory cortex combination of neuromodulatory inputs and top-down
being taken over by other sensory modalities [63,64], influences that signal the behavioural significance of
although the extent of this cross-modal reorganization the stimuli, is also part of the normal operation of the
is limited to secondary areas in humans with residual adult cortex.
hearing [65]. While a reallocation of neural resources
among the intact senses seems adaptive, this appears to Despite much recent progress, countless questions
compromise the capacity of deaf subjects to make use of remain unanswered. First, although the response charac-
restored auditory inputs provided via cochlear implants teristics of neurons in the cortex can be altered in various
[66,67], emphasizing the importance of early implantation ways by experience or during learning, it has not so far
for maximizing the benefits of electrical hearing. been shown that any of these changes are causally related
to changes in perceptual abilities. Addressing this funda-
Conclusions mental issue will also provide valuable insights into the
The studies reviewed in this article provide considerable way in which biologically important signals are encoded
evidence that sensory experience can modify the way in and stored within the brain. Second, it is unclear to what
which auditory cortex processes its inputs over a range of extent plasticity is an emergent feature of the cortex itself

www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:456–464

462 Sensory systems

or inherited from subcortical structures. Indeed, given Cholinergic projections from the basal forebrain have been implicated in
neural plasticity mechanisms in the auditory cortex. In this study, the
growing evidence for the importance of descending cor- cholinergic input from the basal forebrain to A1 in adult cats was
ticofugal pathways in plasticity, future work will need to eliminated by an immunolesion technique before making restricted
cochlear lesions. Reorganization of the tonotopic map was still observed,
focus much more on the relationship between the cortex suggesting that acetylcholine may not be necessary for lesion-induced
and other parts of the auditory pathway. Third, although cortical plasticity in the auditory system.
substantial progress has been made in understanding the 4. Zhang LI, Bao S, Merzenich MM: Persistent and specific
influence of neuromodulators, such as acetylcholine, on influences of early acoustic environments on primary auditory
cortex. Nat Neurosci 2001, 4:1123-1130.
cortical response properties, the behavioural significance
of these effects and their relationship to top-down factors, 5. Chang EF, Bao S, Imaizumi K, Schreiner CE, Merzenich MM:
Development of spectral and temporal response selectivity
such as attention, require further investigation. Finally, in the auditory cortex. Proc Natl Acad Sci U S A 2005,
the mechanisms underlying plasticity in the auditory 102:16460-16465.
cortex are largely unexplored. While some forms of 6. Takahashi K, Hishida R, Kubota Y, Kudoh M, Takahashi S,
plasticity, such as reorganization of the cortical frequency Shibuki K: Transcranial fluorescence imaging of auditory
cortical plasticity regulated by acoustic environments in mice.
map, most likely involve some form of structural remo- Eur J Neurosci 2006, 23:1365-1376.
delling, alterations in synaptic connection strength may 7. Chang EF, Merzenich MM: Environmental noise retards
be sufficient to explain other, more short-term, changes. auditory cortical development. Science 2003,
Evidence for such structural or synaptic correlates of 300:498-502.

experience-dependent cortical plasticity comes mainly 8. Sanes DH, Constantine-Paton M: The sharpening of frequency
tuning curves requires patterned activity during development
from the somatosensory and visual systems and, at pre- in the mouse, Mus musculus. J Neurosci 1985, 5:1152-1166.
sent, it can only be assumed that similar mechanisms
9. Seidl AH, Grothe B: Development of sound localization
underlie plasticity in the auditory cortex. mechanisms in the mongolian gerbil is shaped by early
acoustic experience. J Neurophysiol 2005, 94:1028-1036.
Note added in proof 10. Knudsen EI: Sensitive periods in the development of the brain
A recently published article [68] describes the perceptual and behavior. J Cogn Neurosci 2004, 16:1412-1425.
consequences of providing continuous single-frequency 11. de Villers-Sidani E, Chang EF, Bao S, Merzenich MM: Critical
tonal stimulation during infancy. As expected, the cortical  period window for spectral tuning defined in the primary
auditory cortex (A1) in the rat. J Neurosci 2007, 27:180-189.
frequency maps of the rats used in this experiment Infant rats were continually exposed in a test chamber to single-frequency
showed an over-representation of the frequency to which tone pips, a manipulation known to induce an expansion of the cortical
representation of the presented frequency. By varying the timing of the
they had been exposed and an under-representation of exposure period, the authors were able to show that the maturation of A1
adjacent ones. Contrary to expectations, however, the is particularly susceptible to manipulations of the spectral content of its
auditory input during a three-day-long window, which opens on postnatal
animals were found to be impaired in their ability to day 11.
discriminate the over-represented frequencies, but
12. Fagiolini M, Pizzorusso T, Berardi N, Domenici L, Maffei L:
exhibited improved discrimination performance for the Functional postnatal development of the rat primary visual
adjacent, under-represented frequencies. However, the cortex and the role of visual experience: dark rearing and
monocular deprivation. Vision Res 1994, 34:709-720.
steepest parts of the slopes of a large number of neuronal
tuning curves lay in the range of the under-represented 13. Mower GD: The effect of dark rearing on the time course of
the critical period in cat visual cortex. Dev Brain Res 1991,
frequencies, which might explain the improved acuity for 58:151-158.
these stimuli.
14. Kuhl PK, Conboy BT, Padden D, Nelson T, Pruitt J: Early speech
perception and later language development: implications for
Acknowledgements the ‘critical period’. Lang Learn Dev 2005, 1:237-264.
Our research is funded by the Wellcome Trust through a four-year
15. Schlaug G, Norton A, Overy K, Winner E: Effects of music training
studentship to J.C. Dahmen and a Principal Research Fellowship to
on the child’s brain and cognitive development. Ann N Y Acad
A.J. King. Sci 2005, 1060:219-230.
16. Watanabe D, Savion-Lemieux T, Penhune VB: The effect of early
References and recommended reading musical training on adult motor performance: evidence
Papers of particular interest, published within the annual period of for a sensitive period in motor learning. Exp Brain Res 2006,
review, have been highlighted as: 176:332-340.
 of special interest 17. Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR,
 of outstanding interest Welker E, Svoboda K: Long-term in vivo imaging of experience-
dependent synaptic plasticity in adult cortex. Nature 2002,
420:788-794.
1. Hofer SB, Mrsic-Flogel TD, Bonhoeffer T, Hubener M: Lifelong
learning: ocular dominance plasticity in mouse visual cortex. 18. Holtmaat A, Wilbrecht L, Knott GW, Welker E, Svoboda K:
Curr Opin Neurobiol 2006, 16:451-459. Experience-dependent and cell-type-specific spine growth in
the neocortex. Nature 2006, 441:979-983.
2. Weinberger NM: The nucleus basalis and memory codes:
auditory cortical plasticity and the induction of specific, 19. Grutzendler J, Kasthuri N, Gan WB: Long-term dendritic spine
associative behavioral memory. Neurobiol Learn Mem 2003, stability in the adult cortex. Nature 2002, 420:812-816.
80:268-284.
20. Holtmaat AJGD, Trachtenberg JT, Wilbrecht L, Shepherd GM,
3. Kamke MR, Brown M, Irvine DRF: Basal forebrain cholinergic Zhang X, Knott GW, Svoboda K: Transient and persistent
 input is not essential for lesion-induced plasticity in mature dendritic spines in the neocortex in vivo. Neuron 2005,
auditory cortex. Neuron 2005, 48:675-686. 45:279-291.

Current Opinion in Neurobiology 2007, 17:456–464 www.sciencedirect.com

 Learning to hear: plasticity of auditory cortical processing Dahmen and King 463

21. Robertson D, Irvine DRF: Plasticity of frequency organization in Spectro-temporal receptive fields in A1 were measured while ferrets
auditory cortex of guinea pigs with partial unilateral deafness. performed frequency discrimination or tone detection tasks. Both tasks
J Comp Neurol 1989, 282:456-471. facilitated activity at the target frequency. Only the discrimination task,
however, was associated with a reduction of activity at the reference
22. Noreña AJ, Eggermont JJ: Enriched acoustic environment after frequency. These task-dependent differences in receptive field plasticity
noise trauma reduces hearing loss and prevents cortical map presumably reflect differences in the meaning attributed to identical
reorganization. J Neurosci 2005, 25:699-705. stimuli according to the context in which they were presented.
23. Noreña AJ, Gourevitch B, Aizawa N, Eggermont JJ: Spectrally 34. Blake DT, Heiser MA, Caywood M, Merzenich MM: Experience-
 enhanced acoustic environment disrupts frequency  dependent adult cortical plasticity requires cognitive
representation in cat auditory cortex. Nat Neurosci 2006, association between sensation and reward. Neuron 2006,
9:932-939. 52:371-381.
The authors examined the effects on the frequency organization of A1 of By recording chronically from adult owl monkeys, the authors show that
passively exposing juvenile cats to an acoustic environment whose simply pairing auditory stimuli and rewards is insufficient to induce
spectral content was highly enriched between 5 and 20 kHz. Contrary plasticity in A1. Subsequent pairing of the targets and rewards in a
to expectations, this manipulation did not lead to a representational classical conditioning paradigm led to increased excitability of the cortex,
expansion of the stimulated spectrum, but to a dramatic loss of selectivity whereas learning an operant frequency discrimination task was accom-
for the frequencies that were present in the acoustic input and an panied by reduced excitability and improved frequency selectivity.
enlargement in the representation of neighboring frequencies. This result
highlights the complexity with which the spectrotemporal content of the 35. Bao S, Chang EF, Woods J, Merzenich MM: Temporal plasticity
sensory input and the physiological state and developmental stage of the in the primary auditory cortex induced by operant perceptual
animal interact to shape the functional organization of auditory cortex. learning. Nat Neurosci 2004, 7:974-981.

24. Cheung SW, Nagarajan SS, Schreiner CE, Bedenbaugh PH, 36. Nelken I, Chechik G, Mrsic-Flogel TD, King AJ, Schnupp JWH:
 Wong A: Plasticity in primary auditory cortex of monkeys Encoding stimulus information by spike numbers and mean
with altered vocal production. J Neurosci 2005, response time in primary auditory cortex. J Comput Neurosci
25:2490-2503. 2005, 19:199-221.
A large part of an animal’s acoustic input is comprised of self-generated
sounds, especially vocalizations. The auditory system has developed a 37. Schnupp JWH, Hall TM, Kokelaar RF, Ahmed B: Plasticity of
number of mechanisms that help to reduce the impact of those sounds on  temporal pattern codes for vocalization stimuli in primary
the perception of external stimuli. Nevertheless, the authors show that a auditory cortex. J Neurosci 2006, 26:4785-4795.
surgical manipulation that irreversibly changes the spectral content of Neurons in marmoset primary auditory cortex often respond more
marmoset communication calls alters the way in which cortical neurons in vigorously to twitter communication calls than to the time-reversed
these animals respond to vocalizations and other complex sounds, versions of those calls. This is not the case, however, when cortical
without affecting the processing of pure tones. responses to these sounds are measured in other species, implying that
‘call specific’ neurons may exist in marmoset cortex. In this study, ferrets
25. Ghose GM: Learning in mammalian sensory cortex. were trained to recognize marmoset twitters. Subsequent recordings
Curr Opin Neurobiol 2004, 14:513-518. showed no evidence for the emergence of twitter-specific neurons.
Nevertheless, the twitters were represented by temporal variations in
26. Recanzone GH, Schreiner CE, Merzenich MM: Plasticity in the the firing of the neurons and the information available in these temporal
frequency representation of primary auditory cortex following codes was substantially higher in the trained animals than in controls,
discrimination training in adult owl monkeys. J Neurosci 1993, attesting to the importance of action potential timing in the encoding of
13:87-103. vocalizations.
27. Rutkowski RG, Weinberger NM: Encoding of learned 38. Bao S, Chan VT, Merzenich MM: Cortical remodelling induced
importance of sound by magnitude of representational area in by activity of ventral tegmental dopamine neurons.
primary auditory cortex. Proc Natl Acad Sci U S A 2005, Nature 2001, 412:79-83.
102:13664-13669.
39. Manunta Y, Edeline JM: Noradrenergic induction of selective
28. Polley DB, Steinberg EE, Merzenich MM: Perceptual learning plasticity in the frequency tuning of auditory cortex neurons.
 directs auditory cortical map reorganization through top- J Neurophysiol 2004, 92:1445-1463.
down influences. J Neurosci 2006, 26:4970-4982.
Adult rats were presented with a set of tones of variable frequency and 40. Ji W, Suga N: Serotonergic modulation of plasticity of the
intensity and trained to recognize either a specific target frequency or auditory cortex elicited by fear conditioning. J Neurosci 2007,
target intensity. After several weeks of training, rats in the frequency- 27:4910-4918.
recognition group exhibited an increased number of neurons in A1 that 41. Juliano SL, Ma W, Eslin D: Cholinergic depletion prevents
were selective for frequencies close to that of the target tone. By contrast, expansion of topographic maps in somatosensory cortex.
animals in the intensity-recognition group were characterized by a higher Proc Natl Acad Sci U S A 1991, 88:780-784.
number of neurons with nonmonotonic response-level functions that
peaked at values close to the target intensity. Similar but less pronounced 42. Bakin JS, Weinberger NM: Induction of a physiological memory
changes were also observed in a secondary auditory field. These results in the cerebral cortex by stimulation of the nucleus basalis.
suggest that top-down, task-dependent factors, presumably arising from Proc Natl Acad Sci USA 1996, 93:11219-11224.
other cortical areas, determine the nature of the changes that underlie
learning-induced improvements in perceptual abilities. 43. Kilgard MP, Pandya PK, Vazquez J, Gehi A, Schreiner CE,
Merzenich MM: Sensory input directs spatial and temporal
29. Brown M, Irvine DRF, Park VN: Perceptual learning on an plasticity in primary auditory cortex. J Neurophysiol 2001,
auditory frequency discrimination task by cats: association 86:326-338.
with changes in primary auditory cortex. Cereb Cortex 2004,
14:952-965. 44. Mercado E, Bao S, Orduña I, Gluck MA, Merzenich MM: Basal
forebrain stimulation changes cortical sensitivities to complex
30. Witte RS, Kipke DR: Enhanced contrast sensitivity in auditory sound. Neuroreport 2001, 12:2283-2287.
cortex as cats learn to discriminate sound frequencies.
Brain Res Cogn Brain Res 2005, 23:171-184. 45. Moucha R, Pandya PK, Engineer ND, Rathbun DL, Kilgard MP:
Background sounds contribute to spectrotemporal
31. van Wassenhove V, Nagarajan SS: Auditory cortical plasticity in plasticity in primary auditory cortex. Exp Brain Res 2005,
learning to discriminate modulation rate. J Neurosci 2007, 162:417-427.
27:2663-2672.
46. Pandya PK, Moucha R, Engineer ND, Rathbun DL, Vazquez J,
32. Alain C, Snyder JS, He Y, Reinke KS: Changes in auditory cortex Kilgard MP: Asynchronous inputs alter excitability, spike
parallel rapid perceptual learning. Cereb Cortex 2007, timing, and topography in primary auditory cortex.
17:1074-1084. Hear Res 2005, 203:10-20.
33. Fritz JB, Elhilali M, Shamma SA: Differential dynamic plasticity 47. Ji W, Gao E, Suga N: Effects of acetylcholine and atropine on
 of A1 receptive fields during multiple spectral tasks. plasticity of central auditory neurons caused by conditioning
J Neurosci 2005, 25:7623-7635. in bats. J Neurophysiol 2001, 86:211-225.

www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:456–464

464 Sensory systems

48. Miasnikov AA, McLin D 3rd, Weinberger NM: Muscarinic the auditory system can reinterpret the localization cues associated with
dependence of nucleus basalis induced conditioned receptive different directions in space.
field plasticity. Neuroreport 2001, 12:1537-1542.
59. Kacelnik O, Nodal FR, Parsons CH, King AJ: Training-induced
49. Yan J, Zhang Y: Sound-guided shaping of the receptive field in  plasticity of auditory localization in adult mammals. PLoS Biol
the mouse auditory cortex by basal forebrain activation. 2006, 4:627-638.
Eur J Neurosci 2005, 21:563-576. Localization of sounds in the horizontal plane is impaired if binaural cues
are distorted by occluding one ear. However, stimulus-specific training
50. Chen G, Yan J: Cholinergic modulation incorporated with a allows substantial adaptation to take place, which can occur in the
tone presentation induces frequency-specific threshold absence of vision or error feedback. The basis for this plasticity appears
decreases in the auditory cortex of the mouse. Eur J Neurosci to involve a re-weighting of different localization cues, such that the
2007, 25:1793-1803. animals learn to ignore the altered cues and to rely instead on other
auditory cues, especially the spectral information provided by the unoc-
51. Zhang Y, Hamilton SE, Nathanson NM, Yan J: Decreased
cluded ear, that are less affected by the earplug.
input-specific plasticity of the auditory cortex in mice lacking
M1 muscarinic acetylcholine receptors. Cereb Cortex 2006, 60. Bergan JF, Ro P, Ro D, Knudsen EI: Hunting increases adaptive
16:1258-1265.  auditory map plasticity in adult barn owls. J Neurosci 2005,
25:9816-9820.
52. Miasnikov AA, Chen JC, Weinberger NM: Rapid induction of
Prism rearing has proved to be a powerful approach for exploring the
specific associative behavioral memory by stimulation of the
adaptive plasticity of the barn owl’s auditory localization system. Adjust-
nucleus basalis in the rat. Neurobiol Learn Mem 2006, 86:47-65.
ments in the auditory space map in the optic tectum to match the optically
53. Weinberger NM, Miasnikov AA, Chen JC: The level of cholinergic displaced visual map take place during development, whereas plasticity
nucleus basalis activation controls the specificity of auditory in adults is thought to be much more limited. This study shows, however,
associative memory. Neurobiol Learn Mem 2006, 86:270-285. that much greater map plasticity is possible in adult birds if they are
allowed to hunt their prey rather than being provided with dead food and
54. Ma X, Suga N: Augmentation of plasticity of the central that the magnitude of the shifts in auditory spatial tuning correlates with
auditory system by the basal forebrain and/or somatosensory improvements in the accuracy with which the owls strike their prey.
cortex. J Neurophysiol 2003, 89:90-103.
61. King AJ, Bajo VM, Bizley JK, Campbell RAA, Nodal FR, Schulz AL,
55. Zhang Y, Hakes JJ, Bonfield SP, Yan J: Corticofugal feedback Schnupp JWH: Physiological and behavioral studies of spatial
 for auditory midbrain plasticity elicited by tones and electrical coding in the auditory cortex. Hear Res 2007, 229:106-115.
stimulation of basal forebrain in mice. Eur J Neurosci 2005,
22:871-879. 62. Kotak VC, Breithaupt AD, Sanes DH: Developmental hearing
Numerous studies have shown that pairing tones with electrical stimula- loss eliminates long-term potentiation in the auditory cortex.
tion of the nucleus basalis can alter the response properties of neurons in Proc Natl Acad Sci U S A 2007, 104:3550-3555.
the auditory cortex of adult animals. The authors show that this procedure
63. Finney EM, Fine I, Dobkins KR: Visual stimuli activate auditory
can also shift the frequency tuning of neurons in the inferior colliculus, but
cortex in the deaf. Nat Neurosci 2001, 4:1171-1173.
not if atropine, a muscarinic receptor antagonist, is applied to the cortex,
showing that the descending corticocollicular projection is responsible 64. Hunt DL, Yamoah EN, Krubitzer L: Multisensory plasticity in
for the observed plasticity of the midbrain. congenitally deaf mice: how are cortical areas functionally
specified? Neuroscience 2006, 139:1507-1524.
56. Puckett AC, Pandya PK, Moucha R, Dai W, Kilgard MP: Plasticity
in the rat posterior auditory field following nucleus basalis 65. Lambertz N, Gizewski ER, de Greiff A, Forsting M: Cross-modal
stimulation. J Neurophysiol 2007, 98:253-265. plasticity in deaf subjects dependent on the extent of hearing
loss. Brain Res Cogn Brain Res 2005, 25:884-890.
57. King AJ, Parsons CH, Moore DR: Plasticity in the neural coding
of auditory space in the mammalian brain. Proc Natl Acad Sci U 66. Doucet ME, Bergeron F, Lassonde M, Ferron P, Lepore F: Cross-
S A 2000, 97:11821-11828. modal reorganization and speech perception in cochlear
implant users. Brain 2006, 129:3376-3383.
58. Van Wanrooij MM, Van Opstal AJ: Relearning sound localization
 with a new ear. J Neurosci 2005, 25:5413-5424. 67. Lee HJ, Giraud AL, Kang E, Oh SH, Kang H, Kim CS, Lee DS:
By modifying the sound spectrum in a manner that depends on the Cortical activity at rest predicts cochlear implantation
direction of the source, the external ears generate spatial cues that allow outcome. Cereb Cortex 2007, 17:909-917.
the vertical location of auditory stimuli to be determined. The authors
found that reshaping one of the external ears with a mold initially 68. Han YK, Köver H, Insanally MN, Semerdjian JH, Bao S: Early
impaired sound localization, particularly, but not exclusively, on the experience impairs perceptual discrimination. Nature Neurosci
ipsilateral side. Performance then improved over time, indicating that 2007. [Epub ahead of print].

Current Opinion in Neurobiology 2007, 17:456–464 www.sciencedirect.com