Brain 24/02/2007 17:45:00
Crandall et al 2007,x
Humans and songbirds learn vocalisation during a sensorimotor
sensitive period- “babbling phase”.
Prolonged bursting is key aspect of sensory critical period.
The „High Vocal Center‟ of songbirds is analogous to the cortex in
humans, and codes aspects of song structure including syllables, motifs
and higher-order patterns.
Zebra finches – rapid bursts of activity in premotor area HVC before
and after learned vocalisations. Show inverse correlation with song
maturation. Same neurons also respond to auditory stimuli. In other
words, prolonged motor discharge and sensory input coincide in these
neurons. This would provide the cellular basis for activity-dependent
mechanisms of sensorimotor shaping.
Wang 2004x
Cheng and Mersenich 2003
Continuous noise disrupted functional organisation of the rat auditory
cortex.
Normally the auditory cortex A1 is organised according to a
„tonographic map‟ of sound frequency, although in juveniles the neurons
tend to have a broader tuning frequency and the map is less organised.
Reorganisation has been shown to follow deafness or training, but is more
susceptible to effects of the inputs in infancy, leading to the postulate of a
„critical period‟.
Difficult to remove all input from auditory system, so used white
noise.
White noise led to „immature‟ cortex, pulses of one frequency led to
overreprentation of that in tonographic map.
Adult rats reared with white noise and then exposed to highly
structured noise showed changes like those in young rats.
Showed „critical period‟ not fixed as previously thought, but not
known if this is different. In normal development, critical period linked to
levels of NMDA receptors and BDNF – unknown whether early
disruptionaltered chemical signalling mechanisms or another mechanism
of cortical plasticity is responsible.
In humans- hearing begins at about 20 weeks gestation, so this
could have implications for pregnant mothers in noisy environments. Also
implications for deaf children given cochlear implants at later stage.
Berardi 2000x
Neurotrophins, NMDA receptors and GABAergic inhibition all play roles
in determining critical periods for experience-dependant plasticity.
Manipulation of levels of BDNF alters the timing of this critical period in
the mouse visual system, establishing a causal relation between
neurotrophins and the critical period.
Moore et al 1995x
Myelination of the human auditory pathway from the proximal end of the
cochlear nerve to the brainstem (inferior colliculus) occurs between the
26th and 29th week of gestation, but is not complete until the age of 1yr.
Coincides with acoustimotor reflexes and brainstem auditory evoked
responses, both of which require rapid, synchronised conduction of
impulses in the audit. Nerve and brainstem
Suports idea that 26-29 wks is a critical period.
Dmitrieva and Gottlieb 1994x
Ducklings- even a short period of auditory deprivation affected the
development of auditory sensitivity, as studied by brainstem auditory-
evoked potentials.
Thresholds and latencies of brainstem responses to auditory stimuli did
not show their normal decline in these ducks.
There was a species-specific critical period where this effect was most
pronounced.
Hung 1989x
Chinese children- recorded auditory brainstem evoked potentials.
Auditory clicks.
Peripheral transmission reached adult level at 3mths, brainstem
transmission matched adult at 1yr.
At 6mths waveforms matched adult.
Critical period for maturation of peripheral and central auditory
pathways?
Sadato et al 2004x
Middle superior temporal sulcus responds selectively to human voices.
Found that in deaf signers, makes shift from auditory to visual input in
age-dependant manner- less activation in late-deaf signers.
Chang et al 2005x
Investigated the role of inhibitory patterns in the development of
responses to auditory inputs differing in frequency and temporal
characteristics.
Initially both excitatory and inhibitory responses in A1 are tuned to a
broad frequency range. Temporal resolution is poor due to broad
inhibitory effects. With increasing age, the excitatory and inhibitory
effects become tuned to narrower frequencies. Tuning of inhibitory effects
lagged behind excitatory tuning, and corresponded with the critical
period.
Discrimination of rapid sounds is important to the processing of speech
sounds. Wide-ranging inhibitory effects prevent repeated firing in infant
rats.
The different effects of excitatory and inhibitory transmission were
investigated using a GABA antagonist and glutamate. Inhibitory receptive
fields were narrowed but not eliminated by the GABA antagonist, and this
effect was most pronounced in infant rats. Glutamate had no effect on the
excitatory receptive fields. Forward masking by asynchronous two-tonal
stimuli was not affected by GABA antagonist, and is thought to be
mediated by the thalmocortical synapse.
Rats reared with white noise failed to show this tuning of excitatory and
inhibitory responses. „Rescue‟ rats showed similar receptive fields to
normal adult rats after rearing in a normal environment, showing that the
sensitive period can be extended in these circumstances.
- local inhibitory circuits contribute to cortical spectral selectivity
- GABA makes intracortical contribution to simultaneous 2 tone inhibition.
-exposure to patterned sound input critical to A1 response maturation
- subcortical factors contribute to cortical response patterns- these
mature early
- anatomical and synaptic properties change over development, eg GABA
neurons in auditory cortex peak in late development, and large number of
projections throughout the auditory cortex from pyramidal neurons
probably contributes to the broad inhibitory effects.
These excitatory and inhibitory effects provide a mechanism for sorting
temporally synchronous inputs.
Lack of input keeps inhibitory network in prolonged state of immaturity,
and patterned stimuli + functional cortical inhibition contribute to normal
cortical development.
Nakahura et al 2004x
How are the cortical representations of dynamic sound shaped by the
spectro-temporal patterns of auditory input during development?
Two sequences of three tones (low/high)- region of poor response
separated them in A1. Third tone in sequence also showed decreased
reponse. Double peak of receptive fields also found- never seen in
controls.
Forward/reverse sequence- no response to second tone in reversed
condition shows temporal sequence important.
Effects endured into adulthood.
Relevant to phoneme-specific responses to native language?
Fishman et al 2001
Monkeys- auditory stream segregation.
At low presentation rates, both tones evoked responses resulting in a
corresponding pattern of activity.
At faster rates, response to the tone which was not the best frequency of
the neuronal field was suppressed, this effect increased with both speed
of presentation, and the difference in the sounds.
Suggests cortical origin of auditory stream segregation, including
masking effects which can be observed in humans.
Kohler et al 2002x
Monkey ventral premotor cortex- neurons which respond both when it
performs an action and when it hears the related sound.
Monkey homolog of Broca‟s Area: origins of language?
These neurons did not respond to non-action related sounds.
A large subset of these neurons responded to both sound and vision
alone as well as both. Neurons were found which were selective in their
responses eg to peanut being broken, paper ripping.
Multimodal mirror neurons had been described before, but this was the
first time they had been observed to code for actions when they could
only be heard, rather than for spatial stimuli.
They also fire whilst the monkey is performing the action: execution,
planning, goals?
Gestural communication?
Left-right asymmetry in ventral premotor cortex of great apes, together
with access to auditory input could mean this is where Broca‟s area
evolved from.
Raggio and Schreiner 1999x
Evaluated effects of auditory deprivation on spatial distribution of
response thresholds to electrical stimulation of the cochlea (cats).
Dorsal and ventral A1 are normally separated by a rostrocaudal band of
neurons with low thresholds and narrow frequency tuning. Electrical
cochlear stimulation reversed this threshold pattern.
Alternating patches of high and low threshold found along the
cochleotopic axis (dorsoventral) suggest functional modules in A1. This
was present even after many years of deafness, even when cochleotopic
mapping was absent. May provide framework for different auditory
processing streams. The spatial uniformity of these bands was however
compromised in long term deaf cats. Basic functional circuits may remain
operational? Implication for prognoses?
Loss of cochleotopic mapping may limit information flow to speech and
language cortex in humans, despite perfect channel separation
peripherally (cochlear implant).
Unexpected differences between acute and short term groups suggested
that adjustment of excitatory and inhibitory processes took place. In
particular, spatial extent of activation increased and suggested reduction
in inhibition.
Kral et al 2000x
Congenitally deaf cats: functional deficits of auditory cortex when
stimulated with cochlear implants.
Transient expression of acetylcholinesterase in auditory cortex during
critical time period : Ach is involved in learning effects in auditory cortex
and may play role in formation of thalmocortical projections.
Different layers of auditory cortex normally activated in well-defined
sequence. This was not found. Small currents in layer IV (input layer with
most thalmocortical synapses), but lacking fine structure, and in
infragranular layers, then in supragranular layers. The amplitude of
responses was also reduced. Activation of infragranular layers, caused by
supragranular activation, at longer latencies is substantially reduced. This
is the output layer of the cortex, projecting both to other cortical areas
and the thalamus. Thalamo-cortical loops are thus probably not
functional. Projections to the inferior colliculus are probably also affected.
In normal cats, all layers are activated in sequence, activation of layer IV
shows fine structure, and lastly the supragranular and infragranular layers
are activated in an alternating pattern. This may be crucial for complex
analysis in normals.
This shows substantial deficits in cortical function in response to
stimulation of the auditory nerve.
Cell morphology has also been shown to be different: synapses appear
immature, with reduction in branching of end bulbs and presynaptic
vesicles. There is reduction in dendrites of pyramidal cells.
Unknown whether caused by failure of maturation or degeneration of
cortico-cortico or cortico-thalamic connections.
Kitzes and Semple 1985x
Gerbil inferior colliculus: responses after ablation of contralateral cochlea
indicate that responses to ipsilateral ear result in part from interactions
with projections from contralateral ear.
Kilman et al 2002x
Neonatal rats: 2d of deprivation of activation of auditory cortex resulted
in reduction of staining for GABA receptors, and in number of synaptic
sites staining for GABA.
Showed that, like excitatory transmitters, inhibitory transmitter activity
can be scaled to match activity. Normal part of homeostatic mechanism
preventing either silence or epileptiform activity.
Kapfer et al 2002x
Medial superior olivary neurons encode microsecond differences in arrival
time of low-frequency sound at two ears.
Experience-dependant loss of synapses. This could minimise summation
of inputs, to refine temporal window. In the hippocampus, feed-forward
GABAergic inhibition has been shown to curtail this window, but the
timescale is much larger.
Propose that a mechanism may remove synapses activated too early or
too late.
Hirano et al 2000x
Auditory association area A2 in prelingually deaf.
Cochlear implants after age of 8.
Prelingually deaf showed high blood flow in A2 at rest (PET), indicating a
lack of synaptic revision (loss due to maturation). Suggests there was
little functional differentiation in processing for these subjects. There was
also no increase in blood flow when listening, unlike postlingually
deaf/hearing. Activation was observed during lipreading for prelingually
deaf lipreaders, and during listening for one who never learned to lipread
and had made progress with speech since the implant.
Suggests functional differentiation of A2 should differ according to what
modality used during critical periods of speech acquisition.
Francis and Manis 2000x
Cochlear ablation impairs afferent innervation of ventral cochlear
nucleus.
Electrical responses, cell and synaptic morphology of brainstem are
altered. Electrical properties of VCN neurons also altered.
Implications for cochlear implants?
Kotak et al 2005x
Hearing loss (gerbils) increades excitability in A1 and causes greater
thalmocortical excitation. This is as a result of attempt to maintain
optimum cortical excitability.
Includes increased susceptibility to NMDA receptor antagonist.
Semple and Scott 2003
Primate auditory cortex shows core (including A1) with sharp frequency
tuning and tonotopic organisation, surrounded by less tonotopic areas,
and then non-tonotopic areas.
Increased stimulus complexity (humans) associated with increased
activation throughout core and surrounding areas, but tonal stimuli just
activate the core (transverse temporal gyrus).
Core region shows intense staining in layer IV for parvalbumin (thalamic
projection neurons).
Thalmocortical auditory projections provide basis for hierarchical
transformation, as signal processed sequentially through core and
surrounding regions. Parallel processing in multiple core fields. Then
distributed to multimodal areas in temporal, parietal and frontal lobes.
2 major streams? What and where pathways.
Where- A1 to frontal eye field and parietal (spatial)
What- from anterior core and belt to temporal lobe.
But- localisation and pattern recognition are not necessarily independent.
Cross talk between streams.
GABAergic inhibitory circuits in top and deep layers of A1
Subcortical contribution to functional cortical maps- complex excitation
and inhibition from thalamus.
Temporal and spectral variation influence different areas of auditory
cortex- multiple time scales may encode single stimulus.
Hormones 24/02/2007 17:45:00
Beech and Beauvois 2006x
The possible influence of prenatal androgens on the temporal
processing of sounds in the left hemisphere.
Measured lengths of ring fingers on L/R hands, indicating exposure to
testosterone in utero.
Found that people exposed to more androgens were impaired on
phonology, and reading, and that auditory saltation (ability to
differentiate rapid sounds) was also impaired.
Campos-Barros et al 2000x
Thyroid hormone signalling is essential for cochlear development and
onset of hearing (mouse)
In humans, deafness associated with congenital hypothyroidism.
Tightly constrained to a critical period, after which addition of hormone
will not rescue cochlea formation.
Language 24/02/2007 17:45:00
Buchwald 1994x
Lack of exposure to specific types of sensory pattern during critical
periods can result in lack of responsiveness to those stimuli in adults.
Japanese adults show both behaviourally and electrophysiologically a
failure to discriminate English sounds /r/ and /l/.
Language structure produces measurable effect on aspects of brain
development and function.
Boothroyd 1993
Monaural hearing aid: 80 and 40 % phoneme rcognition
After several months of binaural hearing aid: 75% in previously unaided
ear
Critical Period 24/02/2007 17:45:00
Seebach et al 1994
Neural net model of early speech acquisition (consonants) did not require
innate knowledge of sppech or linguistic structure.
Kuhl 2000
Skinner- language learning as result of reinforcement.
Chomsky- innate language faculty puts constraints on language. Fodor-
language module.
New approach- infants map statistical properties of language they are
exposed to.
Partitioning at phonetic level not unique to humans. Although adults
show language-specific recognition of phonemes, infants and monkeys do
not have this bias. They are able to detect different speech sounds, and
prosodic cues. This implies speech has exploited auditory features.
Evidence that babies are exploiting statistical cues in speech, rather than
having modules for different phonemes or innate knowledge of language
structure.
Hurford 1991
Critical period for language acquisition ends at puberty.
Computer simulation of evolution of linguistic ability showed critical
periods ending at around puberty.
Gilbert 1994x
Neuronal plasticity in some form into adulthood.
Implicit sensory learning may take place in primary sensory cortex.
Properties of neurons and functional architecture of cortex may be
modified by experience.
Early sensory pathways implicated by specificity of learning effects.
„Critical period‟ thought to exist for sensory systems since Hubel and
Wiesel‟s kitten experiments (vision).
However eg size of cortical representation of trained frequency (audio)
and shift of best frequency of neurons toward trained frequency
demonstrates some plasticity remains.
Change in connection strengths, synaptic proliferation and collateral
sprouting.
All cortical areas are able to undergo dynamic changes in functional
architecture and specificity.
Feral children 24/02/2007 17:45:00
McCrone 2003 x
“Throughout history, feral children have seemed able to answer one of the biggest
questions about the human brain—what makes it different from that of an animal? Is
human consciousness innate, or is it the product of language and socialisation?”
“in the end, their fascinating tales lead nowhere in particular.”
-critical periods for language and other cognitive abilities
-3 way interaction of experience and synaptic pruning, genetic factors and cultural
factors
-no single experiment can unravel these
Bengal 1920s Rev. Joseph Singh described 2 girls raised by wolves aged 3 and 5.
Elder child survived but after extensive training never mastered more than about 40
words, but no sentences or conversation.
Showed that the human mind could not be innate- but also was not recovered
despite returning to human society.
Fujinaja et al 1990 x
extreme social isolation and to complex deprivation
we have found that their physical and motor development or recovery has
proceeded smoothly, whereas their linguistic and cognitive development has
continued to show such weaknesses as defective functioning of internal speech
(Vygotsky, 1962) and poor ability to deal with abstract, linguistic subjects