Lecture 1: Hair Cells
Sound is pressure waves of alternating compressions and rarefactions. It’s characterized by
intensity, frequency, and direction.
Intensity is in decibels Sound Pressure Level: dB SPL = 20 log [P/P ]0 This is relative to an
absolute reference. Huge range of sounds are encountered in the world. Human hearing
spans 20 – 10,000 Hz, with peak sensitivity at 2 kHz.
90dB can produce temporary threshold shift, in which sensitivity is reduced. With severe
sound exposure, cochlear hair cells can be killed. They don’t regenerate. Other hearing loss
may be iatrogenic (gentamicin). Most common hearing loss is presbycusis, or loss of high
frequency hearing with age.
The pinna collects sound, amplifying it. The ear canal resonates frequencies around 4 kHz
and enhances them. These structures help you locate the source of a sound based on
The middle ear bones muse overcome the acoustic impedance mismatch of air/water. Water
has much higher acoustic impedance, and normally sound going from air to water is mostly
reflected. The small area of the stapes footplate and lengths of middle ear bones (lever)
increase the pressure wave ~26 fold.
Tensor tympani and stapedius attenuate transmission through the middle ear when noise is
A loss of middle ear function is “conductive” hearing loss. It may be caused by otosclerosis,
in which bony growth impedes the ear ossicles. Sensorineural hearing loss may be due to
loss of hair cells. They can be differentiated by using bone conduction through the skull to
test for an inner ear response.
External hearing aids may help with both conductive and sensorineural hearing loss. For
some conductive hearing loss surgery helps. For severe sensorineural hearing loss, a
cochlear implant may be necessary.
Movement of stapes footplate at oval window vibrates the scala vestibuli. Scala vestibuli on
top, bordered below by Reissner’s membrane, below which is scala media (bounded below
by the apical surface of the hair cells). There are inner (1 row) and outer hair cells (3 rows).
At the bottom of the scala media is the cochlear partition (hair cells, basilar membrane and
tectorial membrane). Below this is the scala tympani.
Scala vestibule and tympani have perilymph (low K, high Na), scala media endolymph (high
K, low Na). Endolymph is secreted into the scala media by the stria vascularis on its outer
wall. Hair cells’ apical surfaces are bathed in endolymph, their basolateral surfaces in
perilymph. The apical surfaces have tight junctions to maintain concentration barriers and
the +80 mV endolymphatic potential compared to the perilymph. These two forces drive K
ions into hair cells at their apical surfaces.
Meniere’s disease occurs with disrupted endolymphatic fluid circulation and results in
sensorineural hearing loss. Vestibular dysfunctions also occurs.
Vertical movement of the basilar and tectorial membranes causes shear between them,
pushing stereocilia (actin filled microvilli) on hair cells from side to side. Hair cells have 1
true cilium called the kinocilium, but it is absent in the adult cochlea. A tip link runs from
the top of 1 stereocilium to the side of the adjacent taller one.
Deflection toward the taller stereocilia results in stretching of the tip link and opening of ion
channels at the tips of stereocilia. Deflection in the opposite direction closes channels (at rest 10% are open). The channels are cation-permeant, though the [K] differential and
endolymphatic potential lead to a flux of mostly potassium into the hair cell from its apex.
This depolarizes the cell. Sinusoidal deflections of the cochlea result in sinusoidal changes
in hair cell membrane potential.
During a sustained deflection of hair cells, the ion current decays in a form of adaptation. A
hypothesis is that motors move the attachment of the tip link to relieve the mechanical
Hair cell depolarization leads to opening of VG Ca channels that trigger fusion of vesicles
containing glutamate. These excite AMPARs on afferent fibers which carry signals to their
cell bodies in the spiral ganglion. VG K channels in the basolateral surface of hair cells
allow an outward flux of K, and return the cell to its resting potential (hyperpolarizing it
At rest, hair cells release glutamate and afferent neurons have spontaneous activity. So, you
can signal by increasing or decreasing the firing rate. An individual cochlear afferent only
synapses with a single inner hair cell. At the synapse, the hair cell has a synaptic ribbon to
facilitate vesicle release.
There are 3 outer hair cells for every inner hair cell, but 95% of afferent fibers go to inner
hair cells via type I fibers. The contralateral superior olivary complex gives off a crossed
olivo-cochlear bundle that synapses massively on outer hair cells. This pathway is
inhibitory, and when it inhibits the outer hair cells it causes a loss of sensitivity and
frequency selection in afferent signals from the cochlea. So inhibiting the OHCs alters the
response of the IHCs.
When depolarized, OHCs shorten and when hyperpolarized they lengthen. This
electromotility is thought to be carried out by the anion channel prestin, which may be
translocated with changes in membrane potential and altering the membrane area. This
contributes to the vibration of the cochlear partition and enhances the motion detected by
inner hair cells.
One form of evidence for this idea of active biological motors is that the ear can generate
sounds called otoacoustic emissions, which are echoes or distortion products occurring with
certain stimulation. Distortion product otoacoustic emissions (DPOAEs) reflect normal
activity in the OHCs of the cochlea and may be used to diagnose sensorineural hearing loss.
Efferent innervations of the cochlea suppresses cochlear sensitivity by inhibiting OHCs. The
olivo-cochlear efferents release ACh on the OHCs. The OHCs have a unique nicotinic
AChR. This receptor allow a Ca influx and depolarizes the OHC. But nearby Ca-gated K
channels immediately open and hyperpolarize the cell.
Hopefully this unique AChR will provide specific drug targets for the auditory system.
Lecture 2: Auditory I
If you try and mask one tone with another, the farther apart the frequencies are the louder the
masking tone has to be to have an effect (the effect being that you can’t hear the other tone).
The identity of a sound is largely conveyed by its frequency content. The frequencies present
in the different vowel sounds, for example, show amplification and attenuation of the sound
produced at the vocal folds at varying frequencies. These are seen in spectrum plots, which
show energy content vs. frequency. The high energy peaks in a spectrum plot are called formants. Vowels are defined by their
first two formants (the first two resonant peaks in their spectra). So, frequency plays a role in
determining pitch as well as quality of a sound.
The cochlea and the downstream auditory transduction are tonotopic, breaking down sounds
into their component frequencies. Outer hair cells in the cochlea function to amplify sound
and sharpen tuning to help you discriminate between frequencies. The OHCs show level
dependent amplification and amplify softer sounds more. You can also discriminate between
frequencies better at lower sound levels, since loud sounds elicit responses over broader
ranges of frequencies.
With OHCs destroyed, you require a much louder sound to move the basilar membrane.
There is little difference with very loud sounds though, because loud sounds aren’t amplified
much by OHCs. This feature of amplification of softer sounds provides “compression,” so
that the same range of response in the cochlea occurs in response to a larger range of sound
Each auditory nerve fiber is tuned to be most sensitive to a narrow range of frequencies. Its
“best frequency” is the one to which it is most sensitive. Different fibers are tuned to
A tuning curve is a representation (threshold sound level vs. frequency) of how much sound
is needed to induce a response in an auditory nerve fiber at different frequencies.
Hearing loss may be categorized as: Conductive (loss of middle ear transmission), Metabolic
(damage to stria vascularis or cells supporting hair cells), Sensorineural (damage to hair
cells), of Central (tumor or lesion in auditory nerve or CNS).
Inner hair cells transduce the vibrations into nervous signals in the auditory nerve. Damage
to IHCs reduces the sensitivity of auditory nerve fibers.
Outer hair cells increase sensitivity and sharpen frequency tuning of auditory nerve fibers.
Damage to OHCs results in reduced sensitivity (requiring louder noise to reach threshold)
and stimulation by a broader range of frequencies.
These types of sensorineural hearing impairment produce the following problems:
o Soft sounds can’t be heard, due to a shift in the sensitivity (threshold).
o Sounds become louder faster (loudness recruitment), and in some cases any audible
sound may be too loud. This is because the volume at which sound becomes
uncomfortable remains the same, but the threshold to hear softer sounds may be greatly
o You can’t separate sounds by frequency as well since there is a decreased frequency
resolution (widened tuning curves). This causes the sound to be muffled.
Hearing loss often occurs more at higher frequencies, with auditory nerve fibers located near
the base of the cochlea.
Genetic deafness may be syndromic with other associated non-cochlear symptoms, or they
may be non-syndromic in which case they only have cochlear phenotypes. One gene for
non-syndromic deafness is the gene for prestin, the OHC motility motor.
Another non-syndromic gene is DFNB1, which encodes for connexin 26 and 31, proteins
which form gap junctions. They are thought to play a role in potassium recycling from the
perilymph back into the endolymph (by going from the scala tympani to the stria vascularis).
The connexins aren’t present in hair cells themselves. They play a role is maintaining the
[K] gradient and the endolymphatic potential of +80 mV. Lecture 3: Auditory 2
Auditory nerve fibers project to the cochlear nucleus, where each fiber splits and goes to each
division of the cochlear nucleus (DCN, AVCN, PVCN). At each step up to the auditory
cortex, the tonotopic map is preserved. The cochlear nucleus is the last place in the auditory
system where info from each ear is segregated. The system has lots of commissures and
decussations that allow for communication between the two sides.
Most other sensory modalities just have a receptors, primary afferent neurons, a primary
nucleus, projections to the thalamus and then projections to the cortex. The auditory system
goes from the primary nucleus (cochlear nucleus) to the superior olive, and also to the
inferior colliculus before heading to the thalamus. These two extra centers accommodate the
complexity required for sound localization.
In addition to sound localization, we have a couple of “de-noising” processes in which we
filter our non-essential info. We direct our auditory attention to particular sounds even when
many are present (cocktail party effect). We also suppress our perception of echoes and pay
attention to mostly the first wavefront that reaches us (the precedence effect). For both of
these processes, sound localization is important. Hearing aids often impair these functions.
The cues for sound localization are the Interaural Time Difference (ITD) with the associated
interaural phase difference, and the Interaural Level Difference (ILD). ITD is due to the
pathlength differences to each ear for a sound coming from one side. These differences are
tiny, so require very high time resolution. ILD is due to the head casting an acoustic shadow
on the ear farther from the source.
The cochlear nucleus outputs are given off by bushy cells. Bushy cells from both the
contralateral and ipsilateral cochlear nucleui give off excitatory fibers to the Medial Superior
Olive (MSO) on each side. Each Lateral Superior Olive (LSO) gets excitatory input from the
ipsilateral cochlear nucleus bushy cells, whereas it gets inhibitory input via an MNTB
interneuron from the contralateral cochlear nucleus bushy cells.
Neurons in the auditory nerve and cochlear nucleus fire in phase with the stimulus waveform,
called phase locking.
The MSO is a coincidence detector. If a sound comes from the right, it takes a little longer to
get to the left side. But, from the left cochlea it has to travel farther to get to the right MSO.
Basically, the time difference it takes for the extra length the AP has to travel across the
midline off sets the extra time it takes sound to reach the more distant ear. So, the signals
from each side should arrive at the MSOs on the side the sound comes from at around the
Depending on the orientation of the head, you’ll get different amounts of time delay (though
the neural delay should stay the same). So depending on the extent of the difference between
the two signals, the MSO discharges at different rates. The MSO is a coincidence detector
that responds to how closely the two inputs occur in time. It converts a time difference into a
discharge rate. By comparing the firing rates of the right and left MSOs, you can get a
decent idea of where a sound is coming from.
Bushy cell (in cochlear nucleus) specializations: Endbulb synapses with the incoming
auditory nerve fibers are huge, so when it gets input signals, it sends APs reliably. They also
have a fast membrane time constant, so little temporal summation occurs. The LSO gets excitatory input ispilaterally and inhibitory input contralaterally. It responds to
differences in sound intensity between the ears by detecting the balance of
excitation/inhibition in each LSO.
This detection of ITD and ILD well localize the sound to a cone for which these values will
be the same. It doesn’t give you much info on front vs. back or above vs. below, so there
must be another cue.
Spectral cues produced by the external ear resolve this ambiguity. The pinna reflects the
sound and attenuates different frequencies based on where the sound came from along the
vertical axis. The reflections have different pathlengths from the directly incoming sound,
and the time delay produces interference that varies with frequency. So, the frequencies that
are attenuated are another cue for where sound comes from. Thus, it is difficult to localize
sound along a vertical axis if you only hear one tone at a time. You need lots of frequencies
to figure out which ones are attenuated.
Lecture 4: Aphasia
Aphasia is a language impairment due to neurological disease or damage.
Broca’s aphasia presents as non-fluent agrammatic speech. People have good
comprehension, but tend to use lots of nouns. Repetition is also impaired, as is writing.
They generally have impaired articulation, called apraxia of speech. Basically they have
language production problems, but good comprehension.
Broca’s aphasia results from problems with the LEFT posterior, inferior frontal cortex. It’s
called Broca’s area, or Brodman’s area 44,45. It is supplied by the superior division of the
left middle cerebral artery (MCA). Stroke there is a common cause of Broca’s aphasia. It’s
near the motor cortex, so you often also get hemiplegia (motor deficit) in the right arm.
Wernicke’s aphasia presents as fluent, paragrammatic speech with English jargon or
neologisms (made up language). They have a tough time understanding spoken words and
sentences. Like their speech, their attempts at repetition are spontaneous jargon. They often
don’t realize that they don’t make sense, and it can be frustrating. Basically, they have good
language production, but comprehension problems.
Wernicke’s aphasia results from problems with the LEFT posterior, superior temporal lobe.
This is Wernicke’s area or Brodman’s area 22. It is supplied by the inferior division of the
left middle cerebral artery (MCA).
Conduction aphasia presents as fluent, paraphasic speech (misusing or mispronouncing
words). They have decent comprehension, but very poor repetition sometimes due to
working memory deficits.
Conduction aphasia results from damage to the arcuate fasciculus, which connects Broca’s
and Wernicke’s areas.
Transcortical Motor Aphasia, like Broca’s, shows nonfluent, agrammatic speech. People
also have good comprehension, but it differs in that their repetition is much better than
spontaneous speech. This is due to problems with areas supplied by the left anterior cerebral
artery (ACA), or the watershed between the left ACA and left middle cerebral artery (MCA).
Transcoritcal Sensory Aphasia, like Wernicke’s, shows fluent, paragrammatic speech, as well
as poor comprehension. But, as with transcortical motor, their repetition is much better.
This is due to problems in the watershed between the left MCA and posterior cerebral artery
(PCA). Mixed transcortical aphasia is when patients are mostly mute or echolalic (all they can do is
repeat what you say back to them). They have poor comprehension, but repetition is better
than spontaneous speech. This is due to isolation of the speech area, which occurs when the
watersheds between the ACA and MCA, and also PCA and MCA are damaged.
Global aphasia is nonfluent with repetitive utterances. They have poor comprehension, poor
repetition of something you say, and poor spontaneous speech. It results from damage to the
whole MCA distribution (Broca’s and Wernicke’s).
Anomic aphasia is fluent, grammatical speech with intact comprehension and repetition. The
only problem is poor word retrieval. Written naming may be spared, so that they could oddly
write down the thing they can’t name, then read what they wrote. That’s kind of rare though.
Anomic aphasia can result from any left MCA lesion, particularly around the angular gyrus
or posterior, inferior temporal gyrus.
Anomic aphasia can show great improvement with restored blood flow.
Lectures 5 and 7: Photoreceptors and Retina
The retina is an outgrowth of the prosencephalon, beginning as primary optic vesicles,
forming optic cups, and ultimately having a pigment epithelium and neural retina separated
by a subretinal space. Important structures there include the macula lutea (yellow spot),
within which is the fovea (pit, area of highest acuity).
The retina is layered. In order, the layers are the pigment epithelium, outer segment of rods
and cones (pigment disks), inner segment of rods and cones (metabolic organelles), outer
nuclear layer (rods and cones), outer plexiform layers (photoreceptors synapsing with
bipolar/horizontal cells), inner nuclear layer (bipolar/horizontal/amacrine/interplexiform
cells), inner plexiform layer (bipolar cells synapsing with ganglion/amacrine cells), ganglion
cell layer, and a nerve fiber layer (ganglion cell axons).
Exceptions to the layering patterns include some ganglion cells in the inner nuclear layer and
some amacrine cells in the ganglion cell layer. Also, Mueller glial cells fill in the rest of the
retinal space, and they have “feet” that form an inner limiting membrane at the vitreous
surface of the retina and an outer limiting membrane between the inner segment and outer
The inner 2/3 of the retina gets blood supply from arteries through the optic nerve, the outer
2/3 gets blood from choroidal circulation filtered through the pigment epithelium
Horizontal cells contact photoreceptors, bipolar cells and each other. Amacrine cells contact
bipolar cells, ganglion cells and each other.
o Rods are dim light detectors and can detect as little as one photon. They have limited
adaptability, and are easily saturated. The outer segment is formed by a highly expanded
and convoluted cilium. Its invaginations are internalized disks of membrane within the
plasma membrane. This is necessary because the visual pigment is a transmembrane
protein. There is only one kind of rod, and these cells make up 95% of photoreceptors.
o Cones have a lower sensitivity to light, but adapt much better and are used at higher
levels of light. The outer segment of cones is made up of sacs that are continuous with
the surface membrane. There are three different kinds of cones, each with a different
pigment molecule. These cells are abundant in the fovea. o Both types of photoreceptors have inner segments with lots of mitochondria, as well as
synaptic terminals that release glutamate.
o You can lose all of your rods and still function, but not cones (ex: macular degeneration)
since they are central to high acuity vision.
The pigment epithelium absorbs the photons that pass through the photoreceptors layer,
preventing back scatter. It is absorbed by melanin granules. There is a constant turnover of
disks/sacs in the photoreceptors, and as they are shed the pigment epithelial cells scavenge
them. Failure of this process can lead to Spargaf’s disease, a form of macular degeneration.
Also, after each use photoreceptor pigments must be regenerated, and this is dependent upon
the pigment epithelium.
Rods use the pigment rhodopsin, which has a peak sensitivity in the blue-green part of the
spectrum. There are three cone pigments: blue (short wavelength) pigment, green (medium)
pigment, and red (long…though its peak sensitivity is in yellow) pigment. Each
photoreceptor cell only expresses one of these pigments.
Absorption spectra must be broad and overlap to allow for the different ratios of activity
among pigments that allow us to see in color. It all depends on the ratio of activity in
different pigments, since activity in any one alone is influenced by intensity. This system
can distinguish up to 3 types of monochromatic light at a time. Colorblind people may lose
one, two or all three types of the cone receptors. With two, you can still see some colors.
A visual pigment consists of a chromophore (always 11-cis-retinal) and a 7 TM domain
protein called an opsin. The opsin tunes the frequency sensitivity of the chromophore to
Light absorption causes a series of very rapid changes in the rhodopsin. One intermediate,
metarhodopsin II, remains present for minutes and is the form that triggers vision.
Eventually it’s decomposed to the opsin protein and all-trans retinaldehyde. This is the
bleached form of the chromophore, which must be carried to the pigment epithelium,
converted back to 11-cis-retinal, and returned to the photoreceptor to rejoin the opsin protein.
In the dark, cation channels permeable to Na are open and depolarize the cell enough to
allow for a continuous release of glutamate from its synaptic terminal. With light, the
channels close, the photoreceptor is hyperpolarized, this spreads decrementally and glutamate
release is reduced.
Mechanism of channel opening/closing: basically, cGMP opens Na/Ca channels and Ca
closes them (directly, through guanylyl cyclase inhibition, and through rhodopsin kinase
Photoreceptors are electrically coupled to one another, and they also synapse with bipolar
and horizontal cells. Their synaptic terminals are spherules in rods and pedicles in cones.
They often contact a triad of two horizontal cell processes and one bipolar cell process,
which invaginates into the photoreceptor. At this point the photoreceptor has a synaptic
ribbon of unknown function. Other types of bipolar cells (flat or midget), with midget ones
being in the fovea and only synapsing with 1 cone.
Horizontal cells are electrically coupled to each other. They also feedback onto
photoreceptors, as well as signal forward onto bipolar cells.
Bipolar cells also have ribbons in their synapses. They synapse in dyads of mixed
composition. Some are reciprocal synapses with bipolar cells signaling to an amacrine cell
and the amacrine cell signaling back to the same bipolar cell. Amacrine cells are not electrically coupled, but they similarly feedback onto bipolar cells and
feedforward onto ganglion cells.
Ganglion cells may be diffuse and get input from many bipolar cells. Or, they may be
midget ganglion cells and get info from 1 midget bipolar cell.
GABA (horizontal cells) and glutamate (photoreceptors) are the main NTs in the retina,
though many others are present.
About 1% of ganglion cells are light sensitive, and they play a role in circadian rhythms.
They project to the suprachiasmatic nucleus of the hypothalamus where 24 hr cycles are
regulated. They also control the pupillary reflex. These ganglion cells detect light on their
own in addition to getting photoreceptor/bipolar input. Even people with bad rods and cones
still have these cells functioning just fine.
A receptive field for a cell is the area of the retina that induces a response in that cell when
light is shined on it. The receptive field of a cell is larger than the area of one photoreceptor
because the photoreceptors are coupled to each other directly and associated indirectly via
horizontal cells. About a photoreceptor is influenced by direct coupling with about 20 other
Photoreceptors release glutamate onto horizontal cells in sign preserving synapses, and
horizontal cells contact a large number of photoreceptors. In the dark, and photoreceptor is
depolarized, releases glutamate, and depolarizes a horizontal cell.
Receptive fields display center-surround properties. In the middle of the receptive field, On
bipolar cells will fire when the photoreceptors above them are illuminated. Light
hyperpolarizes the photoreceptor, so it stops releasing glutamate. The glutamate acts on
mGluR 6 receptors that are unusual in that they hyperpolarize the bipolar cell when they bind
glutamate. So light takes the glutamate away, and the On bipolar cell depolarizes and fires.
Off bipolar cells will hyperpolarize when light is shined on the photoreceptors above them.
These have typical, excitatory glutamate receptors. So light hyperpolarizes photoreceptors,
decreases glutamate release, and this hyperpolarizes the bipolar cell.
In either case, the opposite surround effect is mediated by horizontal cells. Photoreceptors
release glutamate, which binds typical excitatory receptors on horizontal cells in the
“surround” regions. But, these project laterally and inhibit the photoreceptor-bipolar cell
synapse in the middle of the RF. So it acts to reverse whatever transmission occurs in the
middle of the RF.
The center and surround effects cancel out if the entire RF is either illuminated or dark.
Amacrine cells are the first cells in the visual pathway that fire action potentials. Most don’t
show center/surround patterning, rather they fire with changes between light and dark (when
you turn a light on or of