University of Southern California Dissertations and Theses

Synaptic second wind: phosphene fading of the Argus II and glutamate secretion in the inner retina

(USC Thesis Other)

Synaptic second wind: phosphene fading of the Argus II and glutamate secretion in the inner retina

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

SYNAPTIC SECOND WIND

PHOSPHENE FADING OF THE ARGUS II
& GLUTAMATE SECRETION OF THE INNER RETINA

by

David A. Sutter

A Thesis Presented to the
FACULTY OF THE USC VITERBI SCHOOL OF ENGINEERING
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOMEDICAL ENGINEERING)

May 2020

Copyright 2020 David A. Sutter
ii

Epigraph

Of what avail are forty freedoms without a blank spot on the map?

- Aldo Leopold

iii
Acknowledgements

With this achievement, I have but merely left the starting line of a lifelong pursuit to more
deeply appreciate our world. This is not without the tremendous support and guidance of mentors,
family, and friends. I would like to thank my M.S. advisor and friend, Dr. Robert H. Chow who
was the source of my alacrity to puzzle through the biological world. His lessons and enthusiasm
will stay with me for the rest of my career. To my Thesis Defense Committee, Dr. Robert Chow,
Dr. Mark Humayun, Dr. Gerald Loeb, Dr. Christina Zavaleta, and Dr. Qifa Zhou, thank you for
challenging my work and asking difficult questions.
Thank you to the friends I have made along the way, Johnny Castillo, Edder Lopez,
Victoria Wolsey, Nicko Guzman, and Balint Der, who daily, encouraged my many questions and
ideas. To my friends who share my passion for the ocean, its power, ecosystem, and wonder, cheers
to many more adventures. Finally, I would like to send a hug and a kiss to my mother, father, and
brother, to whom I dedicate my work.

iv
Table of Contents

Epigraph ii

Acknowledgements iii

List of Figures x

Abstract xi

Chapter 1: Introduction 1
1.1 Motivation and Hypothesis 1
1.2 The Retina 4
1.2.1 Rod and Cone Photoreceptors 5
1.2.2 Bipolar Cells 6
1.2.3 Retinal Ganglion Cells 7
1.3 Glutamate Receptors 9
1.4 Degenerative Diseases of the Retina 10
1.5 Argus II Retinal Prosthetic 11
1.6 Calcium Channel Inactivation 14
1.7 Desensitization 15
1.8 Amacrine Cell Inhibition 16
1.9 Vesicle Depletion 16

Chapter 2: Experimental Setup & General Methods 18
2.1 Glutamate Sensor and Injection Surgery 18
2.2 System and Stimulation 19

Chapter 3: Results 21
3.1 Glutamate Secretion Kinetics May Be Due To Underlying Vesicle Pool Kinetics 21
3.2 There Are Two Distinct Phases to Glutamate Release 23
3.3 Recovery Kinetics Differentiate Above and Below Current Threshold for Second Phase 25
3.4 Recovery Kinetics On The Time Course of Minutes 26

Chapter 4: Conclusions 28
4.1 Discussion 28
4.2 Future Experiments 30

Bibliography 32

v
List of Figures

1.0: Schematic of Hypothesis. Bipolar-RGC synapse illustrating Ca
2+
channels, ribbon synapses,
synaptic vesicles, glutamate receptors and amacrine cells. . . . . . . . . . . . . . . . . . . . . . . . 2

1.1: Physiology of the retina. Illustration of the eye, macula (pink shaded region), layers of the retina,
and retinal pigmented epithelium [13] . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.2: Patient Reported Phosphene Fading. Patient reported light intensity in response to a 10s duration
20Hz stimulus for nine Argus II subjects. Each plot represents an average of five trials for each patient.
Green dots represent verbal estimations made at these time points [3]. . . . . . . . . . . . . . . . . . 12

1.3: Direct vs. Indirect Inactivation. (A) Action potential recording from RGCs resulting from indirect
activation with 8Hz stimulation versus (B) direct activation at 225Hz stimulation [47]. . . . . . . . . 13

1.4a,b: Calcium Channel Inactivation. (A) Inward Ca2+ current in rod-bipolar cells is tenacious and
non-inactivating. (B) Inward Ca2+ current in cone-bipolar cells rapidly inactivates [50]. . . . . . . . 14

1.5: Glutamate Receptor Desensitization. Inward Ca2+ current isolated pharmacologically. The
scaled control shows rapid inactivation in the time course of 1 sec. Desensitization of the glutamate
receptor is blocked by CTZ resulting in prolonged inactivation of the inward current on the scale of 3
seconds [2]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1.6a,b: Synaptic Vesicle Depletion Model. Modeling of (A) the readily releasable pool with respect to
internal Ca2+ concentration representing number of vesicles with capacitance measurements, and (B)
vesicle secretion with respect to internal Ca2+ concentration, representing exocytosis dC/dt [57]. . . . 17

2.1a,b: AAV iGluSnFR expression in mouse retina cross-section. Retinal layers are labeled. Neuron
nuclei are labeled with DAPI in blue, and glutamate labeling in green is predominantly confined to the
inner plexiform layer. (a) No AAV virus. Green labeling comes from autoflorescence. (b) AAV virus
transfection of retina. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1a,b,c: Synaptic depression, rebound, and two phases to glutamate release. The blue spikes
represent glutamate florescent responses to repeated stimulation by brief pulse trains applied via a multi-
electrode array on a whole mount retina. (a) Following a strong initial glutamate response, subsequent
spikes are reduced in amplitude, with a “rebound” in subsequent signals. (b) Same stimulation protocol
as A, but there is a second phase to each glutamate spike that is more obvious (zoomed inset). (c)
Following progressively shorter intervals between stimulation, the signal augments and then depletes.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2a,b,c: The First and Second Phase of Glutamate Release Arise From Independent Processes. (a)
MEA stimulation at progressively increasing current amplitudes. (b) Heat map analysis of first and
second phases where dark blue is no glutamate and dark red shows high glutamate secretion. The blue
ring distinguishes the electrode borders. (c) Glutamate secretion in response to one 20Hz pulse train with
and without CPA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.3a,b,c: Degree of synaptic depression and rebound depend on whether second phase glutamate
secretion is stimulated. (a) 10 pulse train delivered above the current threshold for second phase. (b) 10
pulse train delivered below the current threshold for second phase. (c) A summary of n=3 unique
wholemount retina with protocols from a and b. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.4: Recovery from synaptic depression is enhanced when second phase secretion is elicited. Orange
lines show under threshold stimulation and blue lines show over stimulation for n=7. . . . . . . . . 26
vi
Abstract

The epiretinal prosthetic has shown immense promise in the restoration of sight for those with Retinitis
Pigmentosa. Yet, like many neural prosthetics, its non-native interaction with the underlying neural tissue prompts
many innately biological questions regarding the neural circuitry’s own capabilities and plasticity. The clinical
manifestation of ‘percept fading’ provided impetus for the following study, and the underlying biological mechanisms
of fading in the retina was explored. A virally transduced glutamate florescent sensor was used to explore the effects
of repeated high frequency stimulation by a multielectrode array on the dynamics of glutamate secretion at the retinal
bipolar cell - ganglion cell synapse. It was observed that depending on the strength of the repeated stimuli, we see
strong depletion or less strong depletion accompanied by a rebound in secretion. Surprisingly, the stronger depletion
does not occur with stronger stimulation, rather with weaker stimulation. Using glutamate as a proxy, it is observed
that stronger stimulation, and not weaker stimulation, results in activation of calcium release from intracellular stores.
I hypothesize that prolonged low-level calcium entry via a calcium-release-activated current greatly facilitates
replenishment of the Readily releasable pool without the concomitant rapid secretion that would otherwise be
stimulated at higher calcium concentrations. It is this mechanism of efficient refilling the RRP enables the ribbon
synapses to maintain high vesicle secretion rates in the retina. Understanding the mechanisms underlying vesicle pool
depletion and replenishment offers insights into potential strategies for alleviating the problem of percept fading in
patients with the surgically implanted Argus II epiretinal microchip.

Chapter 1

Introduction

1.1 Motivation and Hypothesis

Retinitis Pigmentosa patients who have been treated with the surgically implanted Argus
II epiretinal microchip describe visual percept fading after the onset of maintained high-frequency
multielectrode stimulation [1, 2]. Previous work shows that epiretinal electrodes activate neurons
of the inner retina – either the bipolar cells (BCs) or retinal ganglion cells (RGCs) or both. The
system output is alteration in the frequency of RGC action potential firing, which is transmitted in
the visual pathway to elicit percepts. Previous work has shown apparent synaptic depression at the
BC-RGC synapse. Understanding the mechanisms that underlie percept fading may yield insights
into how to reduce percept fading.
Previous work from the Chow Lab and others has shown that, depending on stimulus
duration, extracellular current stimulation results in direct RGC stimulation (current stimuli 0.1-
0.4ms) or indirect RGC stimulation (current stimuli 14-18ms), in which cells of the inner retina
are stimulated first and, in turn, synaptically activate RGCs [3,4]. Clinical and in vivo data has
shown that while direct stimulation of RGCs has superior temporal resolution, it frequently causes
activation of passing axons resulting in ‘streaking’. While the temporal resolution of indirect
stimulation is poor, the spatial resolution is superior and requires less stimulus energy [5].
However, following repeated high-frequency stimulation, the response to indirect activation leads
to progressively smaller signals, measured postsynaptically. Multiple mechanisms that may
2
underlie percept fading are suggested, based on this understanding. Repeated high frequency
stimulation could elicit progressive reduction in signaling at the BCs due to:

I. Use-dependent calcium channel inactivation in BCs
II. Synaptic depression at the BC to RGC synapse
III. Desensitization of the postsynaptic glutamate receptors at the RGCs
IV. Inhibitory input from amacrine cells

Figure 1.0: Schematic of Hypothesis. Bipolar-RGC synapse illustrating Ca
2+
channels,
ribbon synapses, synaptic vesicles, glutamate receptors and amacrine cells

In the following work, I use the recently developed genetically encoded fluorescent
glutamate sensor iGluSnFR [6] to focus on glutamate secretion dynamics at BC-RGC synapses
being repeatedly stimulated. Use of iGluSnFR enables direct monitoring of BC glutamate release
and removes the effects of downstream postsynaptic mechanisms (kinetics of glutamate receptor
activation and desensitization, as well as the complex input-output function of RGC action
potential firing) that normally confound using electrophysiological monitoring of postsynaptic
RGC responses as means to monitor presynaptic BC output.
The BC-RGC synapse is a ribbon synapse. Synaptic architecture is similar among cells
with ribbon synapses, including cochlear hair cells, photoreceptors, Calyx of Held neurons,
3
vestibular organ receptors, and pinealocytes. What each of these cells shares is that each must
sustain high vesicle secretion rates with little fatigue. High secretion would result in rapid depletion
of readily releasable vesicles, unless there is equally rapid vesicle replenishment.
As will be shown, my results suggest that the kinetics of depletion and of resupply of the
readily releasable pool (RRP) of vesicles are highly complex. It is known that both depletion and
resupply of vesicles depend significantly on underlying calcium concentration dynamics. Basal
cytoplasmic calcium is normally in the 100 nM range. Calcium elevation to the micromolar range
triggers rapid vesicle secretion/exocytosis [7,8,9], while smaller elevations to the hundreds of
nanomolar range facilitates vesicle replenishment [8,9,10,11,12]. Depending on the strength of the
repeated stimuli, I observe strong depletion or less strong depletion accompanied by a rebound in
secretion. Surprisingly, the stronger depletion does not occur with stronger stimulation, rather with
weaker stimulation. As I detail, stronger stimulation, and not weaker stimulation, may result in
activation of calcium release from intracellular stores. I hypothesize prolonged low-level calcium
entry via a calcium-release-activated current (Icrac, SOC, or ORAI-STIM system) greatly
facilitates replenishment of the RRP without the concomitant rapid secretion that would otherwise
be stimulated at higher calcium concentrations. This mechanism of efficient refilling the RRP
enables the ribbon synapses to maintain high vesicle secretion rates.
Understanding the mechanisms underlying vesicle pool depletion and replenishment offers
insights into potential strategies for alleviating the problem of percept fading in patients with the
surgically implanted Argus II epiretinal microchip. Future work will thoroughly examine the role
of ORAI-STIM in vesicle pool refilling, as well as, examine other potential pre- and postsynapatic
mechanisms contributing to apparent synaptic depression at the BC-RGC synapse.
4
1.2 The Retina

The retina is a translucent, yet light-sensitive tissue lining the interior of the eye. Incoming
light is filtered and focused by the cornea and lens, creating an incident beam to strike the retina.
As one of the five sensory input channels informing our perception of the world, the retina is
responsible for efficiently and continuously delivering organized visual data to upstream nuclei in
the brain. To that end, it is composed of different cellular layers, each with unique roles. Light
information is first converted to biological signals at the posterior cell layer of the retina, made up
of photoreceptors. Following phototransduction, changes in synaptic release cause neural signals
to back-propagate through the anterior cell layers of the retina, from the photoreceptor layer,
through outer nuclear layer (ONL), outer plexiform layer (OPL), inner nuclear layer (INL), inner
plexiform layer (IPL), retinal ganglion cell layer (RGCL), and finally to axon layer culminating in
the optic nerve.

Figure 1.1: Physiology of the retina. Illustration of the eye, macula (pink shaded
region), layers of the retina, and retinal pigmented epithelium [13].

Light transduction occurs at the photoreceptor cell layer through rods and cones. The
anterior portion of the eye, including zones 2 and 3 are densely populated with rods (100 million),
while the macula, in zone 1, is dominated by cones (3 million) [13]. Rods are slender and
5
specialized for low light discrimination, even responsive to a single photon. It takes less than 10
photons for our brain to register a ‘flash of light’ [13, 14]. Interestingly, the photopigments
associated with color vision and cones have a more prehistoric origin than rods, yet trichromatic
vision with three distinct cone photopigments – short (blue), medium (green), and long (red) –
rather than two (dichromatic), seems to be a more recent, higher order development of primates
[15]. Each photopigment is delegated photons within a different wavelength range to encompass
the visible spectrum (350-800nm).
Rod and cone photoreceptors have unconventional synaptic release. In darkness,
photoreceptors experience an inward sodium current supported by high cystolic cGMP, keeping
cGMP gated Na+ channels open and the membrane potential at -40mV [14, 16]. Glutamate is
released continuously in response to this depolarized state, but serves as both inhibitory and
excitatory presynaptic input. The photosensitive visual pigment is called rhodopsin, upon which
is the protein opsin and its prosthetic group, 11-cis retinal. 11-cis retinal is a polyene; a molecule
with six alternating pi and sigma bonds making it a potent light absorber [17, 18]. Upon incident
photon, this electron dense network isomerizes into 11-trans retinal, causing 5 Angstroms of
movement within the time scale of picoseconds. The G-protein coupled receptor, opsin, is now
activated and triggers the phosphorylation of the G-protein transducin. Transducin amplifies
cGMP phosphodiesterase which rapidly hydrolyzes cGMP to GMP. Decreased cytosolic cGMP
closes cGMP-gated Na+ channels which results in the hyperpolarization of the membrane and
decreased glutamate secretion [19].
The ONL is composed of bipolar and horizontal cells, each with complex responses to
glutamate. Horizontal cells have a negative feedback relationship with cones. When a cone is
hyperpolarized in response to light it leads to the hyperpolarization of a horizontal cell in a sign-
6
conserving manner [14]. Moreover, this hyperpolarization spreads via gap junctions to nearby
horizontal cells [14]. It has not yet been demonstrated what type of receptors in a horizontal cell
respond to the fall in glutamate concentration. The negative feedback mechanism of a horizontal
cell is such that they are presynaptic to cones in a sign-inverting way. Recent research suggests
that horizontal cells modify the voltage sensitivity of Ca2+ channels in cone synapses to mediate
synaptic vesicle secretion. The negative feedback effect has been described as color opponency
[20]. A feedforward inhibitory network also exists as horizontal cells synapse onto bipolar cells,
thought to partially contribute of the center-surround response [21].
Bipolar cells (BC) mediate neural signals from rod and cone photoreceptors to ganglion cells
via glutamate (Glu). The postsynaptic processes of all bipolar cells contain glutamate receptors,
yet some are ionotropic and others are metabotropic. In response to glutamate secretion from
cones, bipolar cells with metabotropic glutamate receptors close ionic channels and are therefore
hyperpolarized [14]. These are termed ON-bipolar cells because in response to light they
depolarize. Alternatively, there are other bipolar cells with ionotropic receptors which open
channels in response to cone released glutamate [14]. These are termed OFF-bipolar cells because
they depolarize without light input. Morphology differentiates these cell types as well. The axon
of ON-bipolar cells arborizes the inner half of the inner synaptic layer (ISL) while the OFF-bipolar
cell arborizes the outer half [14].
The terminals of bipolar cells contain synaptic ribbons in a dyad configuration. In other words,
there are almost always two postsynaptic processes at the ribbon synapses, one belonging to an
amacrine cell and the other to a ganglion cell [14]. While it was first thought that amacrine cells
“lacked axons” (Greek a = lacking + makros = long, + inos = fiber), it has since been discovered
that many amacrine cells have axons, yet they coalesce separately from the optic nerve [14]. As
7
the most diverse population of retinal cells (30-40 types), the functions of only one, AII amacrine
cells, are understood well. Amacrine cells known to contribute integration and interposition relays
upon temporal features of visual information presented to ganglion cells [22]. Moreover, they are
thought to be critical in ganglion cell properties like directional selectivity and local edge detection
[23]. These complex, yet delicate preprocessing features are only possible through the secretion of
inhibitory neurotransmitters, GABA or glycine, presynaptic to ganglion cells, as well as upstream,
presynaptic to the bipolar synapse.
Neural input is consolidated in the retinal ganglion cell layer before consigned to the higher
visual processing centers of the brain. Unlike the dendrites of bipolar, amacrine, and horizontal
cells, the dendrites of ganglion cells are strictly postsynaptic [14]. Unlike axons of upstream
neurons, ganglion cells summate graded potentials, and upon meeting a voltage threshold at the
axon hillock, an action potential is delivered through an unmyelinated axon. Mature ganglion cells
flaunt unique voltage gated potassium channels known as Kv3 channels, which are designed for
high frequency repetitive firing and fast repolarization [24]. Much like the bipolar cell, RGCs have
a center-surround receptive field and have an ON or OFF input distinction. Ganglion cell type
morphology is diverse, with three general classes, W, X and Y types [25]. W-type ganglion cells
tend to have small somas, make up 40% of the total population, are excited by rod, and detect all
direction of light movement [25]. X-type ganglion cells have medium sized soma diameter, make
up 55% of the population, have cone input [25]. Y-type ganglion cells have giant cell bodies and
make up 5% of the population, they respond to rapid eye movement or rapid changes in light
intensity. Based on function and axonal projection, RGC population can be divided into five
classes – Midget cell, parasol cell, bistratified cells, photosensitive ganglion cells, and those that
project to the superior colliculus [25]. P-type RGCs project to the parvocellular layers of the lateral
8
geniculate nucleus (LGN). In a total RGC population of 1-2 million, 80% are P-type midget cells
with a characteristic small dendritic arbor and slow conduction velocity. P-type cells respond to
changes in color, but weakly to changes in contrast [17]. M-type retinal ganglion cells make up
10% of the RGC population, project to the magnocellular layers of the LGN, and have large cell
bodies and wide dendritic arbors. Due to their large arborization, they receive many inputs and
compensate with fast conduction velocity. Unlike P-type, M-type cells respond to low contrast
stimuli but are indifferent to changes in color [17]. K-type RGCs have bi-stratified dendritic arbors
and large receptive fields with no surround inhibition. They are always ON to the blue cone and
OFF to red and green cones. These bistratified ganglion cells are now the focus of frequency
dependent activation for use in progressive retinal prosthetics [17, 26, 27].
A step back to observe general retinal organization reveals that information consolidates as it
travels anteriorly. Thus, 110 million photoreceptors synapse to 36 million bipolar cells which
synapse to 1-2 million ganglion cells [14]. In the retinal periphery, a set of 15-30 rods may synapse
to one bipolar cell, creating a wide receptive field center. At the fovea, there is a one-to-one ratio
of cones to bipolar cells and a detailed receptive field [14]. The lateral topography of the retina is
just as diverse and clinically relevant as its layers. Non-uniform anatomy, complex feedback
systems, combined with the ever-changing plasticity of neural tissue excites me. As a biomedical
engineer keen to one day remedy degenerated retinas I see this tissue as wondrous and humbling.

9
1.3 Glutamate Receptors

Glutamate is the primary excitatory neurotransmitter in vertebrate retina, used in
feedforward and some feedback systems throughout its layers. Glutamate receptors (GluR)
regulate glutamate mediated ion entry leading to postsynaptic excitation. Metabotropic and
ionotropic glutamate receptors can co-occupy postsynaptic membranes but as mentioned
previously, their distinction in the retina is most well-known for differentiating ON and OFF
pathways, respectively. Ionotropic glutamate receptors (iGluR) include subtypes, α-amino-3-
hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPA), N-methyl-D-aspartate receptor
(NMDA), and Kainate Receptors (KA), and each of these contain identifiable protein subunits. It
is important to review this anatomy because the kinetics of activation and desensitization of GluR
observed in the following experiments.
It has been shown in primates that photoreceptors possess metabotropic glutamate
receptors and rod axon terminals have KA subunits GluR6/7 [28, 29]. Primate horizontal cells
express iGluR 2,3,4 and KA subunits GluR6/7 on their primary dendrites, colocalized to
photoreceptor ribbon synapses. Interestingly, AMPA-R in horizontal cells desensitize quickly
<2ms, and the desensitization can be blocked by AMPA antagonist, cyclothiazide (CTZ),
prolonging the EPSC decay time, suggesting that HC desensitization may function to extend the
dynamic range of HC light response [30, 31, 32]. The ON and OFF pathways of bipolar cells are
differentiated by mGluR (selectively agonized by L-AP4), versus iGluR AMPA and KA receptors.
The bipolar synaptic terminal is riddled with complexity, as it secretes glutamate upon ganglion
and amacrine cells, while receiving GABA inhibitory input from amacrine cells. Moreover,
presynaptic mGluR activity has been shown to modulate synaptic release and ganglion cell EPSC
kinetics [30].
10
Metabotropic glutamate receptors have been divided into three categories based on
sequencing and pharmacological interaction, such groupings also seem to have functional
differences [33]. The postsynaptic terminal is dominated by Group I receptors while Group II and
III are largely presynaptic. Group II receptors are sequestered to periterminal regions of the axon,
such that they might be activated only by prolonged and robust synaptic stimulation creating a
glutamate ‘spillover’ [34]. In contrast, Group III receptors are presynaptic auto-receptors, localized
near vesicle release sites, with rapid feedback suppressing glutamate release by bipolar cells [35].
In rats and primates, ganglion cell dendrites have been shown to express all iGluR subunits –
GluR1-4 of AMPA receptors, NR1 and NR2A-C of NMDA receptors and GluR6/7 of KA
receptors [36].

1.4 Degenerative Diseases of the Retina

Age related macular degeneration and retinitis pigmentosa are of the two most prevalent
retinopathies, causing blindness due to degeneration of the photoreceptors [37]. Various genetic
and lifestyle factors influence the onset and extent of photoreceptor damage, yet despite
degeneration of the outer retina, the inner retina largely maintains a healthy profile.
Age-related macular degeneration (AMD) is the leading cause of visual impairment in
patients over 60 years [37]. A distinctive feature of AMD is central vision loss and macular
accumulation of drusen, caused by atrophy of the retinal pigment epithelium (RPE) [38]. Clinical
histology studies have shown that more than 80% of rods and cones are destroyed while the
circuitry of the inner retina is largely unimpaired [39]. Moreover, inner retinal health is directly
correlated with healthy retinal pigment epithelium (RPE), such that 70% of ganglion cell nuclei
remained unperturbed when RPE is normal [40].
11
Unlike AMD, patients with retinitis pigmentosa (RP) present with peripheral vision loss
and genetic perturbation, yet share the same significant loss of photoreceptors with a relatively
intact inner retina. In every 4000 live births 1 patient will acquire a genetic defect that leads to RP
[41, 42]. RP is unique from AMD in its impact through the inner retina. While there is robust
bipolar and ganglion cell health when compared to the decimated photoreceptor population, patient
morphology shows inconsistent preservation of the inner retinal layers. Moreover, the synaptic
machinery seems to rewire desperately, such that degenerated rods, amacrine, and horizontal cells
sprout neurites that extend throughout all three layers, convoluting all normal feedback systems.
In the late stages of RP, cells even begin to migrate between layers [40, 43].

1.5 Argus II Retinal Prosthetic

Treatments for AMD and RP preceding the retinal prosthetic were limited to preventing
progression of the retinopathy rather than restoring lost vision. Anti-VGEF drugs and laser are
used to treat neovascularization in AMD. Acetazolamid and Vitamin A palmitate are given to
relieve macular edema in RP. The Argus II is an epiretinal prosthetic, which was developed to
treat vision loss instead of disease progression. At present, it is only FDA approved in the treatment
of RP [44].
Fundamentally, the Argus II is an array of electrodes placed adjacent to RGCs to stimulate
residual healthy cells of inner retina. The Argus II has 60 platinum electrodes with diameters of
225um. The visual world is captured by a camera which is processed by a video processing unit
(VPU). The pixilation is consolidated to a 6x10-pixel plane and transferred as radio frequency
12
signals. These signals are converted to electrical pulses upon the platinum electrode array whose
amplitude corresponds to the brightness of the external stimulus [45].
The Argus II has been successful in returning functional sight to patients with RP. Yet,
there are many interesting and improvable observations about the electrode interface. In particular,
it seems that the target of stimulation is inconsistent. Sometimes, passing axons are stimulated
causing the patient to see streaks. Other times, RGCs are stimulated indirectly by upstream neurons
like bipolar cells [45].
When cells of the inner retina are stimulated, the consequent neural signal is convoluted
by the properties of the synaptic pathway. This includes filtering, inactivation, inhibition, or
desensitizing of the neural signal. One clinical manifestation of this is called ‘percept fading’
where patients’ visual field fades before their eyes.

Figure 1.2: Patient Reported Phosphene Fading. Patient reported light intensity
in response to a 10s duration 20Hz stimulus for nine Argus II subjects. Each plot
represents an average of five trials for each patient. Green dots represent verbal
estimations made at these time points [3].

In an Argus II clinical study, perceived light intensity was measured in response to a 10
second stimulus at 20Hz in nine patients [46]. Data from this experiment is shown in Figure 1.2.
13
Noticeably, the perceived light intensity was inconsistent, owing to a trend of percept fading. This
same trend of desensitization to stimulation has been reported in vivo when studying mouse and
rabbit retina in response to indirect activation of RGCs [47]. This type of desensitization can be
observed in rabbit RGCs recordings when stimulated at 8Hz.

Figure 1.3: Direct vs. Indirect Inactivation. (A) Action potential recording from RGCs
resulting from indirect activation with 8Hz stimulation versus (B) direct activation at 225Hz
stimulation [47].

Figure 1.3a shows that an initial pulse produces a burst of 10 action potentials from RGCs. In
consequent pulses the cell responds with fewer action potentials. In contrast, Figure 1.3b shows
direct activation of RGCs and action potential response. Unlike indirect stimulation, the action
potential response to direct stimulation of RGCs does not waver even at a frequency of 225Hz.
This difference in stimulus tracking suggests that the mechanisms of ‘desensitization’ is occurring
at the BC-RGC synapse. The precise mechanisms of this ‘desensitization’ is still unknown
however here I suggest multiple mechanisms and explore them. Repeated high frequency
stimulation could elicit progressive reduction in signaling at the BCs due to:

I. Use-dependent calcium channel inactivation in BC’s.

II. Synaptic depression at the BC to RGC synapse.

III. Desensitization of the postsynaptic glutamate receptors at the RGC’s.

IV. Inhibitory input from amacrine cells.

14
1.6 Calcium Channel Inactivation

Synaptic vesicle secretion in retinal bipolar cells (RBCs) is primarily mediated by Ca2+
entry via L-type Ca
v
1.3 Ca
2+
channels [48]. Cav1.3 channels serve diverse tissues from auditory
hair cells to pancreatic cells and myocytes and in some cases, are known to inactivate, halting
synaptic release. Fitting its promiscuity, inactivation of CaV1.3 by intracellular Ca2+ [Ca2+
dependent inactivation (CDI)] can be variable, tuned tissue to tissue. For instance, strong CDI is
present in sinoatrial CaV1.3 [48]. Oppositely, in auditory hair cells, Ca2+ currents by CaV1.3 are
much more tenacious and CDI is much weaker or absent [49].
Whole cell recordings at the synaptic terminal of rod bipolar cells (RBCs) reveal
tenaciously stable Ca2+ current, as shown in Figure 2.1a below [50]. This suggests that RBC
CaV1.3 channels do not inactivate. However, it has recently been shown that cone bipolar cells
(CBCs) have an uncharacteristic rapid and pronounced inactivating Ca2+ current shown in Figure
2.1b [50, 51]. The hypothesis from these experiments is that L-type Ca
2+
channels exist on both
RBC and CBCs but those upon CBCs inactivate. Similar experiments in in vitro mouse bipolar
cells have shown T-type Ca2+ channels to also inactivate [50]. Therefore, it may be possible that
a difference in CDI exists between cone and rod bipolar cells responsible for the observed
‘desensitization’. A reliable way to test this hypothesis is to repeat the experiments done by Protti
et. al., 1998 using the photoreceptor peel method.

Figure 1.4a,b: Calcium Channel Inactivation. (A) Inward Ca2+ current in rod-
bipolar cells is tenacious and non-inactivating. (B) Inward Ca2+ current in cone-
bipolar cells rapidly inactivates [50]
15
1.7 Desensitization

Desensitization describes a process where ligand gated channels undergo a use dependent
decrease in synaptic strength. In the case of the observed ‘desensitization’ at the BC-RGC synapse,
it may be possible that postsynaptic glutamate receptors are desensitizing, mitigating the effect of
presynaptic release and attenuating the RGC response. It is known that AMPA and Kainate
glutamate receptors on retinal ganglion cell dendrites desensitize [52]. The exact role of their
desensitization is unknown but it seems to shape synaptic responses in the CNS and contribute to
the temporal sensitivity of the inner retina [53]. In experiments by Lukasiewicz and Valentino at
Univesity of Washington in St. Louis, it was shown that normal AMPA and KA iGluR desensitize
to half max in T
1/2
= 45ms [53]. When 10uM of the positive allosteric modulator, Cyclothiazide
(CTZ) is used, the desensitization rate slows 10-fold to T
1/2
= 423ms shown in Figure 2.2 below
[53]. The mechanism of desensitization is still unknown, however recent cryo-EM studies suggest
the role of transmembrane AMPA regulatory proteins (TARPs) that modify gating complexes [54].

Figure 1.5: Glutamate Receptor Desensitization. Inward Ca2+ current
isolated pharmacologically. The scaled control shows rapid inactivation in the
time course of 1 sec. Desensitization of the glutamate receptor is blocked by
CTZ resulting in prolonged inactivation of the inward current on the scale of 3
seconds [2].
16
1.8 Amacrine Cell Inhibition

Amacrine cells are known to release GABA upon bipolar cells attenuating the
glutamatergic feedforward signal upon RGCs. For this reason, it must be considered that amacrine
cell inhibition may be a part of the ‘desensitization’ at the BC-RGC synapse. Previous studies have
pharmacologically blocked amacrine cell input upon bipolar cells and shown that desensitization
persisted [55]. While this is surprising, it does not rule out amacrine cell contribution and a more
rigid experimentation is required.

1.9 Vesicle Depletion

In many systems, repeated high frequency stimulation is known to deplete the presynaptic
readily releasable pool (RRP) and cause synaptic depression. This vesicle depletion model for
synaptic depression is based upon synaptic vesicle translocation and priming through different
‘pools’ before exocytosis [56, 57, 58, 59]. Electron microscopy reveals hundreds of vesicles in a
synaptic terminal, and they follow an order of secretion. A few vesicles are docked at the
membrane, primed to fuse upon Ca2+ entry and SNARE complex activation. These vesicles make
up the readily releasable pool (RRP) while distant vesicles make up the reserve pool (RP) [56]. As
stimulation persists, the readily releasable pool is depleted despite replenishment from the reserve
pool. This inevitably leads to a decrease in vesicle exocytosis following stimulation resulting in
depression. The size of the RRP, transition rates between pools, and other recycling mechanisms
play large roles in the kinetics of depletion based depression.
17

Figure 1.6a,b: Synaptic Vesicle Depletion Model. Modeling of (A) the readily
releasable pool with respect to internal Ca2+ concentration representing number of
vesicles with capacitance measurements, and (B) vesicle secretion with respect to
internal Ca2+ concentration, representing exocytosis via dC/dt [57].

Vesicle exocytosis as well as priming the readily releasable pool are both Ca2+ dependent.
Previous modeling in neuroendocrine cells by Dr. Robert Chow suggests that the RRP is most
filled at 500nM internal Ca2+ while secretion most readily occurs above 1uM internal Ca2+ as
shown in Figures 2.3ab [57]. The models presented above have been fitted and validated by
capacitance change measurements by Augustine and Neher [57].

18
Chapter 2

Experimental Setup & General Methods

2.1 Glutamate Sensor and Injection Surgery

A virally transduced florescent sensor was used to visualize glutamate secretion from
retinal bipolar neurons. 2.5 to 3 month-old wild-type mice (C57BL6/J) were anesthetized with a
Ketamine/xylazine solution 80-100 mg/kg, 5-10mg/kg via intraperitoneal injection. As previously
described in publications from our laboratory using genetically encoded calcium indicators [60],
these mice received an intravitreal injection of an AAV vector encoding the iGluSnFR fluorescent
glutamate sensor [61]. The vector was packaged into viruses at the Florida viral vector core of
Professor Loren Looger. The eyes were allowed 2-3 weeks for protein expression [62]. The AAV
was not specifically targeted to a particular cell type, and presumably the iGluSnFR was expressed
both on BCs and RGCs; however, only the BCs secrete glutamate, and therefore iGluSnFR should
report glutamate secreted at the BC-RGC synapse. This was confirmed by previous studies and
my own immunohistochemistry as shown below in Figure 2.1 [61].

19

Figure 2.1a,b: AAV iGluSnFR expression in mouse retina cross-section. Retinal layers
are labeled. Neuron nuclei are labeled with DAPI in blue, and glutamate labeling in green
is predominantly confined to the inner plexiform layer. (a) No AAV virus. Green labeling
comes from autoflorescence. (b) AAV virus transfection of retina.

I euthanized the mice and removed the injected eye. The retina was dissected out and placed
flat (after several cuts to relieve curvature) in AMES buffer on a multielectrode array (MEA
fabricated in house by Dr. YC Chang; interface built by Multichannel Systems, a Division of
Harvard Bioscience, Inc.).

2.2 System and Stimulation

As described previously [60, 62], the MEA was mounted on an inverted microscope stage,
and the microscope was focused on the inner plexiform layer, where the BC-RGC synapses are
located. A computer-controlled stimulus generator (STG-2008, Multi Channel Systems,
Reutlingen, Germany) generated voltage stimuli that were converted to current using a custom
system. Each individual stimulus consisted of a 20-Hz burst of 10, biphasic, anodic-first, 1ms
duration, current pulses, and for most experiments these bursts were applied multiple times,
separated by 5 sec to up to 5 min. The amplitude of each stimulus was typically set to 1.5 times
20
the threshold that elicited a glutamate response (fluorescence intensity change, see below) from
bipolar cells in a given region of interest.
Simultaneously, I illuminated the wholemount retina at 488 nm and imaged fluorescence
at 565 nm at a frame rate of 10 Hz. Fluorescence intensity changes of iGluSnFR were analyzed
using custom software written in Matlab.

21
Chapter 3

Results

3.1 Glutamate Secretion Kinetics May Be Due To Underlying Vesicle
Pool Kinetics

Figure 3.1a,b,c: Synaptic depression, rebound, and two phases to glutamate release. The
blue spikes represent glutamate florescent responses to repeated stimulation by brief pulse trains
applied via a multi-electrode array on a whole mount retina. (a) Following a strong initial
glutamate response, subsequent spikes are reduced in amplitude, with a “rebound” in subsequent
signals. (b) Same stimulation protocol as A, but there is a second phase to each glutamate spike
that is more obvious (zoomed inset). (c) Following progressively shorter intervals between
stimulation, the signal augments and then depletes.

22
Figure 3.1 shows my first observations of glutamate secretion in response to repeated
extracellular stimulation by MEA. As shown in Figures 3.1a and 3.1b, repeated MEA stimulation
(as described in Methods) every 5 seconds to wholemount retina expressing iGluSnFR elicited
clearly discernible fluorescence responses, represented in the figures as transient upward spikes.
Note that the size of the second spike is markedly reduced compared to the first, likely reflecting
synaptic depression – that is, depletion of readily releasable vesicles. After initial synaptic
depression, I commonly observed a partial “rebound” in the subsequent spike amplitudes. In some
experiments (e.g. Fig. 1b inset), the response to a single burst of stimuli consisted of two easily
resolved phases – a large fast transient followed by a smaller and slower transient.
In Figure 3.1C, I applied repeated stimuli of the same strength, but separated by
progressively shorter intervals (30 s, 20 s, 10 s, 5 s, 3 s, 2 s). Note that the second glutamate spike
was larger than the first – partial rebound – then the spikes became progressively smaller, probably
due to a shrinking RRP of glutamate vesicles, due to progressively less time allowed for vesicle
resupply to occur. In summary, these kinetics could be explained by vesicle pool augmentation
(overfilling), followed by depletion.

23

3.2 There Are Two Distinct Phases to Glutamate Release

Figure 3.2a,b,c: The First and Second Phase of Glutamate Release Arise From Independent Processes. (a)
MEA stimulation at progressively increasing current amplitudes. (b) Heat map analysis of first and second phases
where dark blue is no glutamate and dark red shows high glutamate secretion. The blue ring distinguishes the
electrode borders. (c) Glutamate secretion in response to one 20Hz pulse train with and without CPA.

The first and second phases of glutamate release likely arise from independent processes,
based on three findings:
(1) the second phase is observed only when stimulation strength exceeds a defined current
threshold. In Figure 3.2A, repeated stimuli of progressively larger strength are applied at regular
intervals of 20 s. The glutamate signals increase as the stimulation amplitude progressively
increases. The second phase is absent for the first three applications, and emerges only with the
stronger fourth stimulation at a current amplitude at approximately 1.3 times the minimum current
threshold, and is abundantly obvious during the fifth stimulation. The second phase starts after the
burst stimulus has ended.
24
(2) the second phase seems to have a more focal response than the first (Figure 3.2B).
MATLAB image analysis of the glutamate response shows that the second phase corresponds with
a subset of neurons more centered above the stimulating electrode. Neurons closer to the
stimulating electrode experience higher current densities and may, therefore, be more strongly
stimulated.
(3) the second phase can be blocked with the addition of cyclopiazonic acid (CPA), without
disrupting the first phase (Figure 3.2C). CPA is known to block the SERCA Ca2+ ATPase pump
in the Endoplasmic Reticulum (ER). This suggests that the second phase may represent glutamate
secretion in response to intracellular Ca2+ release from the ER, perhaps triggered by Ca2+-induced
Ca2+ release (CICR) when basal [Ca2+] exceeds a certain threshold.

25
3.3 Recovery Kinetics Differentiate Above and Below Current
Threshold for Second Phase

Figure 3.3a,b,c: Degree of synaptic depression and rebound depend on whether second phase glutamate
secretion is stimulated. (a) 10 pulse train delivered above the current threshold for second phase. (b) 10 pulse train
delivered below the current threshold for second phase. (c) A summary of n=3 unique wholemount retina with
protocols from a and b.

As shown in Figure 3.3A, repeating stimulations every 5 seconds at an intensity below the
threshold to activate second-phase glutamate secretion generally elicits strong synaptic depression
and very little rebound, whereas repeating stimulations above the threshold (Figure 3.3B) show
less depletion and significant rebound. Figure 3.3C shows a composite of multiple cells stimulated
either over (OT, blue symbols) or below the threshold (UT, orange symbols) for activating second-
phase glutamate secretion. The apparently counterintuitive observation that a stronger stimulus
results in less synaptic depression and greater rebound might be explained if second-phase
secretion activates processes that enhance RRP refilling.
To further characterize the kinetics of the rebound phenomenon illustrated in Figure 3.1B
and 3.3, I monitored the size of the glutamate fluorescence signal at a range of intervals after an
26
initial depleting stimulus. Each experimental trial began with a pair of stimuli at a stimulus
intensity sufficient to activate second phase response and separated by a 5-second interval. As
expected, the second response was invariably smaller than the first, due to synaptic depression,
then after a delay I applied another stimulus, to monitor the recovery from depression. I repeated
the experiment for a number of trials, randomly varying the duration (including delays of 1 min, 2
min, 3 min, 4 min, 5 min) of the delays (Figure 3.4 orange). After the depression of the second
response, the signal had regained its original amplitude after 1 minute. By 2 min, the glutamate
signal appears to have peaked, and the signal either plateaus or is trending back down to its original
size by 5 min.

3.4 Glutamate Secretion Kinetics May Be Due To Underlying Vesicle
Pool Kinetics

Figure 3.4: Recovery from synaptic depression is enhanced when second phase secretion is elicited. Orange lines
show under threshold stimulation and blue lines show over stimulation for n=7.

27
The same stimulus protocol was administered using a current stimulation amplitude that
was below threshold for activating the second phase (Figure 3.4 blue). Surprisingly, not only was
there more profound synaptic depression when stimulating below second phase threshold, but
recovery was slower, as well. The greatest difference in time courses was observed in the first
minute following the depleting paired stimuli. Less depletion and faster recovery despite more
intense stimulation is counterintuitive, and suggests that some process associated with CICR plays
a role in promoting vesicle recycling.

28
Chapter 4

4.1 Discussion

I have demonstrated that the genetically encoded glutamate fluorescent indicator iGluSnFR
reports dynamic signals at the BC-RGC synapse. Glutamate secretion at synapses occurs when the
cytoplasmic calcium concentration (normally in the range of ~100 nM at rest) is elevated to the
micromolar range typically by calcium entry through voltage-gated calcium channels, thereby
triggering secretion from an exhaustible finite pool of readily releasable vesicles. Following
depletion of the finite pool of readily releasable vesicles, pool replenishment occurs by processes
stimulated by low-level elevation of calcium concentration in the range of a few hundred
nanomolar. Normally, following action potential stimulation, calcium channel closing and
cessation of calcium entry leads the micromolar level calcium concentrations to collapse rapidly,
and the residual calcium stays in the range of a few hundred nanomolar only transiently, limiting
the duration of optimal RRP refilling.
I observed that weak current injection stimulation elicited a single spike of glutamate
secretion, most likely due to calcium entry through activated voltage-gated calcium channels,
whereas stronger stimuli elicit two phases of glutamate release, the second of which occurs
significantly after the current stimulation has ended and which is blocked by inhibitors of calcium
release from intracellular stores. I hypothesize the second phase represents glutamate secretion
triggered by calcium-induced calcium release (CICR), activated by entry of calcium in the first
phase of calcium entry through voltage-gated calcium channels.
Repeated stimuli of identical strength every 5 seconds results in distinct patterns of
glutamate secretion, depending on whether second-phase glutamate secretion was triggered. In the
29
absence of second-phase secretion, there is a reduction in spike amplitudes initially (synaptic
depression), until a reduced steady-state plateau of spike amplitude is reached. In contrast, when
the stimulus is strong enough to elicit second-phase secretion, I observed initial depression of the
spike amplitude (but not as marked as in the absence of second phase secretion), followed by a
partial rebound in amplitude. Experiments designed to monitor rebound showed that the process
that underlies rebound may last for minutes.
I hypothesize that second-phase glutamate secretion is elicited by calcium-induced calcium
release from intracellular stores, which activates Icrac/ORAI-STIM to refill the calcium stores.
Entry of calcium is via low conductance ORAI channels that do not cause the elevation of
cytoplasmic calcium to levels as high as calcium entry through voltage-gated calcium channels
[63]. Activation of Icrac/ORAI has been reported to lead to maintained and low-level (in the range
of a few hundreds nanomolar) elevation of cytoplasmic calcium, which is ideal to stimulate vesicle
pool replenishment with minimal vesicle exocytosis. Other mechanisms that may also help
maintain low-level calcium elevation for replenishment of the readily releasable pool of vesicles.
Other mechanisms also exist that may cause prolonged low-level calcium elevation, such
as the reversal of the Na+/Ca2+ exchanger. Previous studies suggest that this mechanism also may
facilitate replenishment of the readily releasable pool of vesicles [64]. I have not ruled out a role
for the Na+/Ca2+ exchanger in assisting in maintaining low-level cytoplasmic calcium elevation.
Possible shared mechanism in pancreatic ß cells may underlie glucose intolerance in mice
lacking mechanism for calcium release from intracellular stores In a previous study, it was found
that transgenic knockout of IP3 receptors in pancreatic beta cells predisposes the mice to glucose
intolerance [65]. My results here on the role of calcium-release-activated current in replenishing
the readily releasable pool of vesicles suggest that glucose intolerance may arise in the IP3R
30
knockout animals because of loss of one mechanism of calcium release from intracellular stores
and, therefore, of activation of Icrac/ORAI-STIM.
The possible role of Icrac/ORAI-STIM in vesicle pool replenishment has been proposed
for another ribbon synapse, the auditory hair cell [66]. The common finding for a role for
Icrac/ORAI-STIM in optimal RRP refilling in both retinal bipolar and auditory hair cells supports
the idea that this is a shared mechanism is critical in enabling ribbon synapses to maintain high
vesicle secretion rates with little fatigue.

4.2 Future Experiments

A tantalizing hypothesis must be tested, that Icrac/ORAI-STIM plays a key role in
replenishing the readily releasable pool of vesicles. I will use known inhibitors or stimulators of
IP3R, ryanodine receptors, and Icrac/ORAI/STIM to study effects on glutamate secretion. I will
use a light-activated STIM to stimulate ORAI independent of calcium store depletion, which will
enable me to test whether elevating basal calcium, without prior calcium store depletion enhance
vesicle pool refilling.
As I have demonstrated, glutamate secretion at the BC-RGC synapse appears to show
complex kinetics of both depression and rebound. I will also perform critical control experiments
to rule out a contribution by the iGluSnFR sensor itself. Another important control will be to test
whether the late phase glutamate secretion is due not to calcium-induced calcium release but to
activation of a polysynaptic or autaptic pathway that leads to delayed glutamate secretion.
Ion channels may contribute to or masquerade as synaptic depression or facilitation. I will
use patch clamp to tease out contributions of ion channels. To rule out a major role of presynaptic
calcium channel inactivation in what I have interpreted as synaptic depression, I will perform patch
31
clamp recording of bipolar neurons and examine whether calcium channel inactivation could lead
to a progressive reduction in BC glutamate secretion over time. I will also examine whether
calcium channel facilitation might contribute to the apparent rebound secretion I observed using
iGluSnFR.
To evaluate the possible contribution of glutamate receptor desensitization at RGCs on
percept fading, I will simultaneously monitor iGluSnFR fluorescence as a direct measure of
glutamate secretion and the postsynaptic glutamate-activated currents at RGCs voltage-clamped
at, say, -70 mV (to prevent action potential firing). In principle, glutamate-activated currents
should track with glutamate secretion, unless there is significant desensitization.
Maintained low-level calcium elevation could be mediated not only by the mechanism I
have focused on here -- Icrac/ORAI-STIM -- but also by the reversed Na-Ca exchanger. I will
evaluate the contribution of the Na-Ca exchanger by using specific pharmacologic blockers of the
Na-Ca exchanger. The significance of these experiments is that they will give insights into
potential strategies to remedy the percept fading reported by retinitis pigmentosa patients treated
with the Argus II epiretinal chip implant.

32
Bibliography

1. L. Yue, J. D. Weiland, B. Roska, and M. S. Humayun, “Retinal stimulation strategies to restore vision:
Fundamentals and systems,” Progress in Retinal and Eye Research, vol. 53, pp. 21–47, Jul. 2016, doi:
10.1016/j.preteyeres.2016.05.002.
2. A. Pérez Fornos et al., “Temporal Properties of Visual Perception on Electrical Stimulation of the Retina,”
Invest. Ophthalmol. Vis. Sci., vol. 53, no. 6, p. 2720, May 2012, doi: 10.1167/iovs.11-9344.
3. D. K. Freeman, J. F. Rizzo, and S. I. Fried, “Encoding visual information in retinal ganglion cells with
prosthetic stimulation,” J. Neural Eng., vol. 8, no. 3, p. 035005, Jun. 2011, doi: 10.1088/1741-
2560/8/3/035005.
4. Weitz AC, Nanduri D, Behrend MR, Gonzalez-Calle A, Greenberg RJ, Humayun MS, Chow RH, Weiland
JD. Improving the spatial resolution of epiretinal implants by increasing stimulus pulse duration. Sci.
Transl. Med. 2015; 7, pp. 318ra203.
5. D. K. Freeman, J. F. Rizzo, and S. I. Fried, “Encoding visual information in retinal ganglion cells with
prosthetic stimulation,” J. Neural Eng., vol. 8, no. 3, p. 035005, Jun. 2011, doi: 10.1088/1741-
2560/8/3/035005.
6. Marvin JS, Borghuis BG, Tian L, Cichon J, Harnett MT, Akerboom J, Gordus A, Renninger SL, Chen TW,
Bargmann CI, et al. An optimized fluorescent probe for visualizing glutamate neurotransmission. Nature
Methods. 2013; 10, pp 162-170.
7. Zucker RS. Exocytosis: A molecular and physiological perspective. Neuron. 1996; 17, pp. 1049–1055.
8. Heinemann C, von Rüden, L, Chow RH, Neher E. Two-step model of secretion control in neuroendocrine
cells. Pflugers Arch.. 1993; 424, 105–112.
9. Heinemann C, Chow RH, Neher E, Zucker RS. Kinetics of the secretory response in bovine chromaffin
cells following flash photolysis of caged Ca2+ Biophysical Journal. 1994; 67, pp. 2546–2557.
10. Zucker RS, Regehr WG. Short-term Synaptic Plasticity. Annual Review of Physiology. 2002; 64, pp. 355-
405.
11. Neher E, Zucker RS. Multiple calcium-dependent processes related to secretion in bovine chromaffin cells.
Neuron. 1993; 10, pp. 21–30.
12. von Ruden L, Neher E. A Ca2+-dependent early step in the release of catecholamines from adrenal
chromaffin cells. Science. 1993; 262, pp. 1061–1065.
13. L. Yue, J. D. Weiland, B. Roska, and M. S. Humayun, “Retinal stimulation strategies to restore vision:
Fundamentals and systems,” Progress in Retinal and Eye Research, vol. 53, pp. 21–47, Jul. 2016, doi:
10.1016/j.preteyeres.2016.05.002.
14. Rodieck, R. W. (1998). The first steps in seeing. Sinauer Associates.
15. B. M. Carlson, “Special Senses—Vision and Hearing,” in The Human Body, Elsevier, 2019, pp. 177–207.
16. J. Troscianko, J. Wilson-Aggarwal, D. Griffiths, C. N. Spottiswoode, and M. Stevens, “Relative
advantages of dichromatic and trichromatic color vision in camouflage breaking,” Behavioral Ecology,
vol. 28, no. 2, pp. 556–564, Apr. 2017, doi: 10.1093/beheco/arw185.
17. Kandel ER, Schwartz JH, Jessell TM, Siegelbaum SA, Hudspeth AJ. Principles of neural science. 5th
edition. Columbus, Ohio: McGraw-Hill Education; 2012.
18. Ij. M. Kramer, “Sensory Signal Processing; Visual Transduction and Olfaction,” in Signal Transduction,
Elsevier, 2016, pp. 329–379.
19. Pugh E. N. Jr. and Lamb T. D. (1993). Amplification and kinetics of the activation steps in
phototransduction. Biochim. Biophys. Acta 1141, 111-149. 10.1016/0005-2728(93)90038-H.
20. P. Ahnelt and H. Kolb, “Horizontal cells and cone photoreceptors in human retina: A Golgi-electron
microscopic study of spectral connectivity,” J. Comp. Neurol., vol. 343, no. 3, pp. 406–427, May 1994, doi:
10.1002/cne.903430306.
21. X. L. Yang and S. M. Wu, “Feedforward lateral inhibition in retinal bipolar cells: input-output relation of
the horizontal cell-depolarizing bipolar cell synapse.,” Proceedings of the National Academy of Sciences,
vol. 88, no. 8, pp. 3310–3313, Apr. 1991, doi: 10.1073/pnas.88.8.3310.
33
22. HARTVEIT, E. 1999. Reciprocal synaptic interactions between rod bipolar cells and amacrine cells in the
rat retina. J Neurophysiol, 81, 2923-36.
23. MASLAND, R. H. 2012. The tasks of amacrine cells. Vis Neurosci, 29, 3-9.
24. J. Henne, “Voltage-gated potassium channels in retinal ganglion cells of trout: A combined biophysical,
pharmacological, and single-cell RT-PCR approach,” p. 9.
25. R. L. Rockhill, F. J. Daly, M. A. MacNeil, S. P. Brown, and R. H. Masland, “The Diversity of Ganglion
Cells in a Mammalian Retina,” J. Neurosci., vol. 22, no. 9, pp. 3831–3843, May 2002, doi:
10.1523/JNEUROSCI.22-09-03831.2002.
26. D. K. Freeman, J. F. Rizzo, and S. I. Fried, “Encoding visual information in retinal ganglion cells with
prosthetic stimulation,” J. Neural Eng., vol. 8, no. 3, p. 035005, Jun. 2011, doi: 10.1088/1741-
2560/8/3/035005
27. J. D. Weiland and M. S. Humayun, “Retinal Prosthesis,” IEEE Trans. Biomed. Eng., vol. 61, no. 5, pp.
1412–1424, May 2014, doi: 10.1109/TBME.2014.2314733.
28. Hare, W.A., Owen, W.G., 1992. Effects of 2-amino-4-phosphonobutyric acid on cells in the distal layers
of the tiger salamander’s retina. J. Physiol. 445, 741–757.
29. Harvey, D.M., Calkins, D.J., 2002. Localization of kainate receptors to the presynaptic active zone of the
rod photoreceptor in primate retina. Vis. Neurosci. 19, 681–692.
30. H. Hirasawa, R. A. Shiells, and M. Yamada, “Analysis of spontaneous EPSCs in retinal horizontal cells of
the carp,” Neuroscience Research, vol. 40, no. 1, pp. 75–86, May 2001, doi: 10.1016/S0168-
0102(01)00212-7.
31. X. Yang, “Characterization of receptors for glutamate and GABA in retinal neurons,” Progress in
Neurobiology, vol. 73, no. 2, pp. 127–150, Jun. 2004, doi: 10.1016/j.pneurobio.2004.04.002.
32. P. Lukasiewicz, J. Lawrence, and T. Valentino, “Desensitizing glutamate receptors shape excitatory
synaptic inputs to tiger salamander retinal ganglion cells,” J. Neurosci., vol. 15, no. 9, pp. 6189–6199, Sep.
1995, doi: 10.1523/JNEUROSCI.15-09-06189.1995.
33. J.-P. Pin and R. Duvoisin, “The metabotropic glutamate receptors: Structure and functions,”
Neuropharmacology, vol. 34, no. 1, pp. 1–26, Jan. 1995, doi: 10.1016/0028-3908(94)00129-G.
34. Asztely, E, Erdemli, G. and Kullmann, D.M. (1997) Extra- synaptic glutamate spillover in the
hippocampus: dependence on temperature and the role of active glutamate uptake. Neuron, 18: 281-293.
35. M. H. Higgs, C. Romano, and P. D. Lukasiewicz, “Presynaptic effects of group III metabotropic glutamate
receptors on excitatory synaptic transmission in the retina,” Neuroscience, vol. 115, no. 1, pp. 163–172,
Nov. 2002, doi: 10.1016/S0306-4522(02)00381-0.
36. Cohen, E., & Miller, R. (1994). The role of NMDA and non-NMDA excitatory amino acid receptors in the
functional organization of primate retinal ganglion cells. Visual Neuroscience, 11(2), 317-332.
doi:10.1017/S0952523800001668
37. Schuman, S. G., Koreishi, A. F., Farsiu, S., Jung, S. H., Izatt, J. A., and Toth, C. A. Photoreceptor layer
thinning over drusen in eyes with age-related macular degeneration imaged in vivo with spectral-domain
optical coherence tomography. Ophthalmology , 116(3):488–496 e2, 2009.
38. L. Yue, J. D. Weiland, B. Roska, and M. S. Humayun, “Retinal stimulation strategies to restore vision:
Fundamentals and systems,” Progress in Retinal and Eye Research, vol. 53, pp. 21–47, Jul. 2016, doi:
10.1016/j.preteyeres.2016.05.002.
39. Medeiros, N. E. and Curcio, C. A. Preservation of ganglion cell layer neurons in age-related macular
degeneration. Investigative ophthalmology & visual science , 42(3):795–803, 2001.
40. Kim, S. Y., Sadda, S., Humayun, M. S., E.de Juan, Jr., Melia, B. M., and Green, W. R. Morphometric
analysis of the macula in eyes with geographic atrophy due to age-related macular degeneration. Retina ,
22(4):464–70, 2002
41. Bunker, C. H., Berson, E. L., Bromley, W. C., Hayes, R. P., and Roderick, T. H. Prevalence of retinitis
pigmentosa in Maine. American journal of ophthalmology , 97(3):357–65, 1984.

42. Hartong, D. T., Berson, E. L., and Dryja, T. P. Retinitis pigmentosa. Lancet , 368(9549):1795–809, 2006.
34
43. Milam, A. H., Li, Z. Y., and Fariss, R. N. Histopathology of the human retina in retinitis pigmentosa.
Progress in Retinal and Eye Research , 17(2):175–205, 1998.
44. J. D. Weiland and M. S. Humayun, “Retinal Prosthesis,” IEEE Trans. Biomed. Eng., vol. 61, no. 5, pp.
1412–1424, May 2014, doi: 10.1109/TBME.2014.2314733.
45. L. Yue, J. D. Weiland, B. Roska, and M. S. Humayun, “Retinal stimulation strategies to restore vision:
Fundamentals and systems,” Progress in Retinal and Eye Research, vol. 53, pp. 21–47, Jul. 2016, doi:
10.1016/j.preteyeres.2016.05.002
46. A. Pérez Fornos et al., “Temporal Properties of Visual Perception on Electrical Stimulation of the Retina,”
Invest. Ophthalmol. Vis. Sci., vol. 53, no. 6, p. 2720, May 2012, doi: 10.1167/iovs.11-9344.
47. D. K. Freeman, J. F. Rizzo, and S. I. Fried, “Encoding visual information in retinal ganglion cells with
prosthetic stimulation,” J. Neural Eng., vol. 8, no. 3, p. 035005, Jun. 2011, doi: 10.1088/1741-
2560/8/3/035005.
48. L. Shi, J. Y.-A. Chang, F. Yu, M. L. Ko, and G. Y.-P. Ko, “The Contribution of L-Type Cav1.3 Channels
to Retinal Light Responses,” Front. Mol. Neurosci., vol. 10, p. 394, Dec. 2017, doi:
10.3389/fnmol.2017.00394.
49. P. S. Yang et al., “Switching of Ca2+-Dependent Inactivation of CaV1.3 Channels by Calcium Binding
Proteins of Auditory Hair Cells,” Journal of Neuroscience, vol. 26, no. 42, pp. 10677–10689, Oct. 2006,
doi: 10.1523/JNEUROSCI.3236-06.2006.
50. D. A. Protti and I. Llano, “Calcium Currents and Calcium Signaling in Rod Bipolar Cells of Rat Retinal
Slices,” J. Neurosci., vol. 18, no. 10, pp. 3715–3724, May 1998, doi: 10.1523/JNEUROSCI.18-10-
03715.1998.
51. H. von Gersdorff and G. Matthews, “Calcium-dependent inactivation of calcium current in synaptic
terminals of retinal bipolar neurons,” J. Neurosci., vol. 16, no. 1, pp. 115–122, Jan. 1996, doi:
10.1523/JNEUROSCI.16-01-00115.1996.
52. P. Lukasiewicz, J. Lawrence, and T. Valentino, “Desensitizing glutamate receptors shape excitatory
synaptic inputs to tiger salamander retinal ganglion cells,” J. Neurosci., vol. 15, no. 9, pp. 6189–6199, Sep.
1995, doi: 10.1523/JNEUROSCI.15-09-06189.1995
53. Kandel ER, Schwartz JH, Jessell TM, Siegelbaum SA, Hudspeth AJ. Principles of neural science. 5th
edition. Columbus, Ohio: McGraw-Hill Education; 2012.
54. S. Chen, Y. Zhao, Y. Wang, M. Shekhar, E. Tajkhorshid, and E. Gouaux, “Activation and Desensitization
Mechanism of AMPA Receptor-TARP Complex by Cryo-EM,” Cell, vol. 170, no. 6, pp. 1234-1246.e14,
Sep. 2017, doi: 10.1016/j.cell.2017.07.045.
55. D. K. Freeman, J. F. Rizzo, and S. I. Fried, “Encoding visual information in retinal ganglion cells with
prosthetic stimulation,” J. Neural Eng., vol. 8, no. 3, p. 035005, Jun. 2011, doi: 10.1088/1741-
2560/8/3/035005.
56. Kandel ER, Schwartz JH, Jessell TM, Siegelbaum SA, Hudspeth AJ. Principles of neural science. 5th
edition. Columbus, Ohio: McGraw-Hill Education; 2012.
57. Heinemann C, von Rüden, L, Chow RH, Neher E. Two-step model of secretion control in neuroendocrine
cells. Pflugers Arch.. 1993; 424, 105–112.
58. Zucker RS, Regehr WG. Short-term Synaptic Plasticity. Annual Review of Physiology. 2002; 64, pp. 355-
405.
59. Neher E, Zucker RS. Multiple calcium-dependent processes related to secretion in bovine chromaffin cells.
Neuron. 1993; 10, pp. 21–30.
60. Weitz AC, Behrend MR, Lee NS, Klein RL, Chiodo VA, Hauswirth WW, Humayun MS, Weiland JD,
Chow RH. Imaging the response of the retina to electrical stimulation with genetically encoded calcium
indicators. J. Neurophysiol. 2013; 109, pp. 1979–1988.
61. Marvin JS, Borghuis BG, Tian L, Cichon J, Harnett MT, Akerboom J, Gordus A, Renninger SL, Chen TW
Bargmann CI, et al. An optimized fluorescent probe for visualizing glutamate neurotransmission. Nature
Methods. 2013; 10, pp 162-170.

35
62. Chang YC, Walston ST, Chow RH and Weiland JD. In vivo characterization of genetic expression of virus
transduced calcium indicators in retinal ganglion cells using a low-cost funduscope. 2016 IEEE 38th Annual
Int. Conf, of the Engineering in Medicine and Biology Society (EMBC). 2016; pp. 1316–9.
63. Hoth M, Penner R, Depletion of intracellular calcium stores activates a calcium current in mast cells.
Nature. 1992; 355, pp. 353-356.
64. Minami A, Xia YF, Zucker RS. Increased Ca2+ influx through Na+/Ca2+ exchanger during long-term
facilitation at crayfish neuromuscular junctions. J Physiol. 2007; 585, pp. 413– 427.
65. Ye R, Ni M, Wang M, Luo S, Zhu G, Chow RH, Lee AS. Inositol 1,4,5-trisphosphate receptor 1 mutation
perturbs glucose homeostasis and enhances susceptibility to diet-induced diabetes. J. Endocrinology. 2011;
210, pp. 209-217.
66. Castellano-Munoz M, Schnee ME, Ricci AJ. Calcium-induced calcium release supports recruitment of
synaptic vesicles in auditory hair cells. Journal of Neurophysiology. 2016; 115:226–239.

Abstract (if available)

Abstract The epiretinal prosthetic has shown immense promise in the restoration of sight for those with Retinitis Pigmentosa. Yet, like many neural prosthetics, its non-native interaction with the underlying neural tissue prompts many innate biological questions regarding the neural circuitry’s own capabilities and plasticity. The clinical manifestation of ‘percept fading’ provided impetus for the following study, and the underlying biological mechanisms was investigated. A virally transduced glutamate florescent sensor was used to explore the effects of repeated high frequency stimulation by a multi-electrode array on the dynamics of glutamate secretion at the retinal bipolar cell - ganglion cell synapse. It was observed that depending on the strength of the repeated stimuli, there is a strong depletion or less strong depletion accompanied by a rebound in secretion. Surprisingly, the stronger depletion does not occur with stronger stimulation, rather with weaker stimulation. Using glutamate as a proxy, it is observed that stronger stimulation, and not weaker stimulation, results in activation of calcium release from intracellular stores. I hypothesize that prolonged low-level calcium entry via a calcium-release-activated current greatly facilitates replenishment of the Readily Releasable Pool without the concomitant rapid secretion that would otherwise be stimulated at higher calcium concentrations. It may be this mechanism of efficient refilling the RRP which enables the ribbon synapse to maintain high vesicle secretion rates in the retina. Understanding the mechanisms underlying vesicle pool depletion and replenishment offers insights into potential strategies for alleviating the problem of percept fading in patients with the surgically implanted Argus II epiretinal microchip.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Manipulation of RGCs response using different stimulation strategies for retinal prosthesis

PDF

Investigation of the electrode-tissue interface of retinal prostheses

PDF

The structural and functional configurations of glutamate, GABA, and catecholamine pre-synaptic terminals in the parvicellular neuroendocrine part of the paraventricular nucleus of the hypothalamus

Asset Metadata

Creator Sutter, David Aime (author)

Core Title Synaptic second wind: phosphene fading of the Argus II and glutamate secretion in the inner retina

School Viterbi School of Engineering

Degree Master of Science

Degree Program Biomedical Engineering

Publication Date 05/15/2020

Defense Date 03/25/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Argus II,calcium,calcium mediated calcium entry,calcium release activated current,depletion,facilitation,glutamate,multi electrode array,OAI-PMH Harvest,readily releasable pool,retina,retinal prosthetic,ribbon synapse

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chow, Robert H. (committee chair), Humayun, Mark S. (committee member), Loeb, Gerald E. (committee member), Zavaleta, Cristina (committee member), Zhou, Qifa (committee member)

Creator Email dsutter@usc.edu,littlebigsut@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-308644

Unique identifier UC11665301

Identifier etd-SutterDavi-8529.pdf (filename),usctheses-c89-308644 (legacy record id)

Legacy Identifier etd-SutterDavi-8529.pdf

Dmrecord 308644

Document Type Thesis

Rights Sutter, David Aime

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA