Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Signal processing by the mammalian retina near absolute visual threshold
(USC Thesis Other)
Signal processing by the mammalian retina near absolute visual threshold
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
SIGNAL PROCESSING BY THE MAMMALIAN RETINA
NEAR ABSOLUTE VISUAL THRESHOLD
by
Haruhisa Okawa
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(NEUROSCIENCE)
August 2009
Copyright 2009 Haruhisa Okawa
ii
Acknowledgments
I would like to thank my thesis advisor, Dr. Alapakkam Sampath. Since I was the
first student in the lab and it was only two of us during the first one year, we had to start
everything from scratch, including building experimental setups and organizing the lab.
This precious hands-on experience under his guidance taught me a lot about how to
design experiments. Once the setup was ready, all of my projects gradually took off. Our
constant discussions regarding the results of every experiment trained me to think
critically and to design additional experiments, which kept me on the right track for
successfully finishing those projects. I also want to thank my committee members, Dr.
Jeannie Chen, Dr. Judith Hirsch, Dr. Li Zhang and Dr. David Hinton.
Outside the lab, Kanako Haikata supported me with her constant love and
patience for the last two years. Her mental support helped me a lot to go through the
busiest times before my graduation and focus on my research. My family, particularly my
mom and dad, always advised me to do what I believe and encouraged me to stick to
what I want to do. I thank Yoshio Takashima, who is the only Japanese Neuroscience
graduate student besides me who did not give up finishing Ph.D. We helped each other
through a variety of activities we shared. I also thank other Japanese friends in Zilkha
Neuroscience building, Yukihiro Koike and Hirohisa Shiraishi, and my labmates, Joshua
Miyagishima, Cyrus Arman, Dr. Alan Horsagar, Alice Cho and Dr. Johan Pahlberg. I
thank all of my friends whom I could not name here. Without your friendships, I would
not have been able to settle down here and have fruitful academic and private life.
iii
Table of Contents
Acknowledgements
List of Figures
Abstract
Chapter 1 - Introduction
1.1 Maximizing the presynaptic voltage change
1.1.1 Phototransduction
1.1.2 Gap junctional coupling in the outer retina
1.2 Presynaptic optimization for signal transmission
1.2.1 Voltage sensitivity of Ca
2+
channels
1.2.2 Properties of glutamate release
1.3 Postsynaptic thresholding and the elimination of noise
Chapter 2 - Genetic dissection of the relationship between optimal processing of
sensory signals and behavioral sensitivity
2.1 Introduction
2.2 Materials & Methods
2.2.1 GCAPs
-/-
mice
2.2.2 Preparation
2.2.3 Electrophysiological recording and analysis
2.2.4 Mouse behavioral assay
2.2.5 Calibration of light intensity
2.3 Results
2.3.1 Higher signal-to-noise ratio in GCAPs
-/-
rods
2.3.2 Elevated behavioral threshold in GCAPs
-/-
mice
2.3.3 Altered response properties of GCAPs
-/-
rod bipolar cells
2.3.4 Increased noise in GCAPs
-/-
rod bipolar cells
2.3.5 Reduced SNR in GCAPs
-/-
rod bipolar cells
2.3.6 Nonlinear threshold is poorly positioned in GCAPs
-/-
rod bipolar
cells
2.3.7 Increased postsynaptic saturation increase SNR in GCAPs
-/-
rod
bipolar cells
2.4 Discussion
2.4.1 Influence of rod and rod bipolar SNR on behavior
2.4.2 Can the nonlinear threshold adapt to accommodate changing SNR?
Chapter 3 - Cooperative control of sensitivity by two splice variants of Gα
o
in
retinal On-bipolar cells
3.1 Introduction
3.2 Materials & Methods
3.2.1 Mice and preparation
ii
v
vi
1
2
3
5
6
8
10
12
19
19
21
21
22
22
24
26
26
26
29
31
32
34
37
40
40
41
43
46
46
48
48
iv
3.2.2 Electrophysiology and light stimulation
3.2.3 Western blotting
3.3 Results
3.3.1 The default state of rod bipolar transduction channels is closed
3.3.2 Gα
o1
-/-
rod bipolar cells have weak On-response mediated by Gα
o2
3.3.3 The sensitivity of Gα
o2
-mediated On-response
3.3.4 mGluR6 controls both Gα
o1
and Gα
o2
3.3.5 Reduced gain and sensitivity in Gα
o2
-/-
rod bipolar cells
3.3.6 Reducing the total amount of Gα
o
by half does not alter the
property of rod bipolar response
3.4 Discussion
3.4.1 The default state of the transduction channel in the G-protein
cascade is closed
3.4.2 Gα
o1
and Gα
o2
work in cooperation to set the sensitivity in rod
bipolar cells
3.4.3 Nonlinear thresholding of rod signals is mediated downstream of
Gα
o
activation
3.4.4 A model for the mGluR6 signaling pathway
Chapter 4 - Conclusions
References
49
50
50
50
53
55
58
58
59
61
62
63
64
65
68
74
v
List of Figures
Figure 1.1: Propagation of the dim flash response from rods to rod bipolar cells
Figure 1.2: Structure and signal transfer at the rod-to-rod bipolar synapse
Figure 1.3: Convergence at the rod-to-rod bipolar synapse
Figure 2.1: Light-evoked responses of WT and GCAPs
-/-
rod photoreceptors
Figure 2.2: Behavioral estimate of absolute visual threshold in WT and GCAPs
-/-
mice
Figure 2.3: Flash response families and dark noise in WT and GCAPs
-/-
rod
bipolar cells
Figure 2.4: Estimation of SNR in WT and GCAPs
-/-
rod bipolar cells
Figure 2.5: Model for the nonlinear threshold at the rod-to-rod bipolar synapse
Figure 2.6: Comparison of SNRs of rods and rod bipolar cells
Figure 3.1: The default state of transduction channels in rod bipolar cells is
closed
Figure 3.2: Rod bipolar response is partially mediated by Gα
o2
Figure 3.3: The property of the Gα
o2
-mediated response measured in
Gα
o1
-/-
Cx36
-/-
AIIACs
Figure 3.4: Gα
o2
-/-
rod bipolar cells exhibited reduced gain and light sensitivity
Figure 3.5: Reduced Gα
o
expression in Gα
o
+/-
retina does not alter rod bipolar
response
Figure 3.6: A revised model of the signal transduction in rod bipolar cells
3
7
14
27
30
33
36
38
42
52
54
56
59
60
66
vi
Abstract
When few photons are available, such as on a moonless night, our vision is limited by the
probability of photon absorption in rod photoreceptors, and the reliable transmission of
light-evoked signals through the retina. Dark-adapted rod photoreceptors are capable of
reliably signaling single photon absorptions. However, these single photon responses are
challenged by neural noise arising from rod photoreceptors themselves, and the
subsequent retinal circuitry. Rod bipolar cells, the second-order neurons specialized for
low light level vision, pool 20 – 100 rod inputs. This convergence imposes on rod bipolar
cells the difficult task of identifying sparse single photon responses among the majority
that generate noise. It has been shown that saturation within the postsynaptic signaling
cascade in the rod bipolar cells can effectively eliminate signals from rods whose
amplitudes are statistically dominated by noise near absolute visual threshold. My first
study emphasizes the delicate match between the signal-to-noise ratio of the rod
photoresponse and the properties of the signaling cascade in the postsynaptic rod bipolar
cells. In particular, I provide evidence that the improved signal-to-noise ratio in rod
photoreceptors does not guarantee improved behavioral performance if the rod bipolar
signaling cascade is not well-tuned to the rod signal-to-noise ratio. In my second study, I
show that two splice variants of the heterotrimeric G-protein alpha subunit, G
o
α, are
involved in the signaling cascade of On-bipolar cells. In rod (On) bipolar cells these
subunits may together subserve the fine-tuning of the stoichiometry in the signaling
cascade to improve the detection of light near absolute threshold. Collectively, my work
demonstrates how the signaling cascade in the rod bipolar cells can be fine-tuned to the
vii
rod inputs to improve vision near absolute threshold, and how such adjustments are
limited by biological constraints.
1
Chapter 1
Introduction
Across mammalian species, the retinal circuitry underlying visual processing in
dim light is well conserved. Individual rod photoreceptors, themselves capable of
reliably signaling the absorption of individual photons (Baylor et al., 1979; Baylor et al.,
1984), are pooled in a specialized circuitry referred to as the Rod Bipolar Pathway
(Dacheux and Raviola, 1986; Smith et al., 1986). Following a series of convergent
connections in this pathway, the ganglion cells, which are the output cells of the retina,
send signals from thousands of pooled rods (Walraven et al., 1990) to higher visual
centers. It has been appreciated for more than a half-century that absolute behavioral
threshold for light detection requires only a few photon absorptions in this pool (Barlow,
1956; Hecht et al., 1942; van Rossum and Smith, 1998; reviewed by Field et al., 2005).
Several factors influence the transmission of the rod photoresponse through the retina
including the fidelity of the rod photoresponse itself, the great degree of convergence of
rod signals, and the number of stages of processing. These factors collectively place
fundamental limits on the performance of rod vision, and thus must each be optimized for
single-photon transmission near absolute threshold. Of particular interest is the very first
synapse of the Rod Bipolar Pathway, which pools the responses of 20-100 rods and has
the unenviable task of discriminating the small graded potential change in the few rods
absorbing photons from the remainder of the rods generating electrical noise.
2
The process of rod-to-rod bipolar cell signal transmission has strong implications
for setting the absolute threshold for seeing. Indeed recent experiments on transgenic
mice with altered rod photoresponses, due to the lack of the rod photoreceptor protein
recoverin, indicate that deficits in absolute visual threshold can be attributable to the
properties of rod-to-rod bipolar signal transfer (Sampath et al., 2005). Furthermore,
mutations in pre- and post-synaptic proteins that alter the function of this synapse are
known to produce visual disorders like Congenital Stationary Night Blindness (Boycott et
al., 2001; Hemara-Wahanui et al., 2005; Mansergh et al., 2005; O'Conner et al., 2006;
Rao et al., 1994). The goal of this review is to highlight the process by which the rod
photoresponse traverses the rod-to-rod bipolar synapse to ensure its reliable transmission
through the retina. We emphasize the mechanisms that maximize the voltage change
sensed by the rod synaptic terminal (called a spherule) following photon absorption, that
convert this voltage change to a reduction in glutamate release, and that remove rod noise
postsynaptically. These mechanisms collectively improve the fidelity of the rod single-
photon response and allow the retina to transmit these signals to higher visual centers
where they contribute to visually guided behavior.
1.1 Maximizing the presynaptic voltage change
Near absolute threshold perhaps the most important goal for the rod photoreceptor
is to signal the largest possible voltage change per absorbed photon. Two fundamental
processes are involved in signaling this voltage change to the rod spherule, where
glutamate release is controlled (Figure 1.1). First, the rod phototransduction cascade
must amplify the signal generated by a single rhodopsin molecule to yield the largest
possible change in the outer segment current with respect to the underlying noise.
3
Fig. 1.1. Propagation of the dim flash response from rods to rod bipolar cells.
(A) As many as 20-100 rod photoreceptors (R) synapse onto a single rod bipolar cell (RB).
Signals generated in a single rod may also be spread to neighboring rods and cones (C)
through gap junctions. (B) Suction electrodes measure the light-induced reduction in rod
photocurrent generated by the phototransduction cascade following a brief flash. For dim
flashes the reduction in current hyperpolarizes the rod photovoltage proportionally to the
change in photocurrent, thereby causing a reduction in synaptic glutamate release at the rod
spherule. The reduction in glutamate release is sensed postsynaptically on the rod bipolar
cell by a metabotropic glutamate receptor, mGluR6, and results in a depolarizing response.
Note the voltage gain between the rod photovoltage and the rod bipolar cell voltage
(Ashmore and Falk, 1980; Capovilla et al., 1987). Dashed lines emphasize the speeding of
the rod bipolar response compared to the rod photocurrent. Data presented is schematic.
Second, the voltage change produced by the closure of the outer segment current needs to
be transmitted to the rod spherule with the least amount of loss.
1.1.1 Phototransduction
The phototransduction cascade in vertebrate rods is arguably the best-elucidated
G-protein coupled signal transduction mechanism, and the subject of many review
articles (Burns and Baylor, 2001; Fain et al., 2001; Lamb and Pugh, 2006). The
activation of single rhodopsin in the outer segment by light is known to trigger a cascade
4
of events that ultimately results in the closing of cGMP-gated channels that are normally
open in darkness. Through a series of amplifying steps this cascade leads to the
degradation of more than 10
5
cGMP per photon, generating a robust rod photoresponse
(Yee and Liebman, 1978).
A great deal is known about the biophysical mechanisms that underlie the rod
photoresponse, specifically the forms of the intrinsic noise that limit the fidelity of rod
signals in the retina. Light-detection near threshold ultimately requires that light-evoked
responses can be distinguished from two types of noise intrinsic to the rod photoreceptor:
discrete noise and continuous noise (Baylor et al., 1980). Discrete noise is produced by
the thermal activation of the visual pigment and has an identical form to the rod single-
photon response (Yau et al., 1979). The identity of discrete noise to the rod single-
photon response prevents the implementation of any mechanism to eliminate this noise.
In fact the rate of occurrence of discrete noise events has historically thought to be what
fundamentally limits absolute threshold, as it generate a ‘dark light’ that the frequency of
light-evoked events need to exceed for detection (Barlow, 1956; but see also Field et al.,
2005). To some degree nature has mitigated this form of noise through the evolution of
rhodopsin molecules with great thermal stability (Baylor et al., 1984).
While discrete noise events need to be considered in downstream retinal neurons
pooling thousands of rods, their contribution at an individual rod-to-rod bipolar synapse
is relatively small. Near absolute threshold an individual synapse will rarely ‘see’ a
discrete noise event. If discrete noise events occur at a rate of 0.006 s
-1
rod
-1
( Baylor et
al., 1984), and the integration time of the rod response is 0.2 s (Walraven et al., 1990),
then an individual rod-to-rod bipolar synapse will ‘see’ a discrete event on average once
5
every ~ 833 integration times. The constraints for the detection of a photon instead rely
on discriminating a light-evoked event from continuous noise resulting from thermal
activation of the rod phosphodiesterase (Rieke and Baylor, 1996). At the rod-to-rod
bipolar synapse, a linear convergence of rod inputs would sum this noise and occlude
single-photon responses that are sparse in the array of rods at absolute visual threshold.
Instead, as described below, nonlinear signal transmission at this synapse can help to
differentiate signals from rods absorbing a photon from those generating continuous
noise (Field and Rieke, 2002; Sampath and Rieke, 2004).
1.1.2 Gap-junctional coupling in the outer retina
The light-evoked closure of outer segment cGMP-gated channels by the
phototransduction mechanism is processed by the rod’s input impedence and inner
segment conductances to yield a hyperpolarizing voltage change (~ 1 mV per photon;
Schneeweis and Schnapf, 1995) that is passively transmitted to the rod spherule (Figure
1.1). Near absolute threshold the form of this single-photon voltage change appears to
mirror the outer segment current (Baylor and Nunn, 1986), but at higher light levels the
voltage and current begin to deviate as the voltage is shaped by inner segment
hyperpolarization-activated and Ca
2+
-activated conductances (Bader et al., 1982; Fain et
al., 1978).
The coupling of rod photoreceptors to both rods and cones in the mammalian
retina can provide one source of voltage loss to the rod spherule (Figure 1.1A). Thus, at
the rod spherule the light-induced hyperpolarization and reduction in glutamate release
per photon will become progressively diminished the more extensively the rod is
electrically-coupled to its neighbors. It has been suggested that near absolute threshold
6
the rods are poorly coupled to their neighbors to allow the spherule to sense the largest
voltage change in the rods absorbing photons (Smith et al., 1986). Recent recordings
from primate rods and cones indeed show that in the dark-adapted retina many but not all
rods are dye-coupled to neighboring rods and cones, and this coupling is suggested to
impair absolute visual threshold (Hornstein et al., 2005). However, the functional
consequences of dye coupling remain to be seen, and highlights an important issue. The
fact that rods are coupled at all to neighboring photoreceptors is perplexing if the main
goal of rod vision is the discrimination of individual photons. The observed coupling
would suggest the retina is simultaneously balancing single-photon detection near
absolute threshold (where coupling will degrade the signal) with light detection at higher
levels where almost every rod receives a photon (where it is postulated that coupling
improves rod signal-to-noise by averaging across neighboring cells; Smith et al., 1986)).
1.2 Presynaptic optimization for signal transmission
The transmission of the rod photoresponse to rod bipolar cells is controlled at a
specialized synapse at the rod spherule. As shown in Figure 1.2, the bipolar cell dendrite
(as well as horizontal cell dendrites) project into an invaginated structure that is opposite
the release site for glutamate. In darkness a consequence of the open cGMP-gated
channels in the rod outer segment is a relatively depolarized resting membrane potential
and a steady influx of Ca
2+
at the synapse. This steady Ca
2+
influx results in continuous
glutamate release from a specialized synaptic ribbon (Sjostrand, 1953). The released
glutamate is sensed postsynaptically by a metabotropic glutamate receptor, mGluR6, and
leads to the closure of cation channels (Figure 1.2B). The light-evoked hyperpolarization
produced by the closing of outer segment cGMP-gated channels is transmitted passively
7
Fig. 1.2. Structure and signal transfer at the rod-to-rod bipolar synapse.
(A) The rod spherule is a specialized invaginating structure where the dendrites of
horizontal (H) and rod bipolar cells (RB) are apposed to a glutamate release site controlled
by a ribbon. Adapted from Rao-Mirotznik et al. (1998). Ca
2+
channels (Ca
V
1.4) are
located near the active zone (AZ) and allow the continuous release of glutamatergic
vesicles in darkness. In ‘On’ rod bipolar cells glutamate is sensed by mGluR6 receptors
located near the mouth of the invagination. Inset: Release of glutamate is dependent on
Ca
2+
influx through Ca
V
1.4 Ca
2+
channels whose voltage sensitivity can be optimized by
the binding of internal factors, like CABP4. CABP4 binding to Ca
V
1.4 pushes the
activation curve to more negative voltages into a regime where Ca
2+
concentration
changes in the synapse are more pronounced at physiological voltages. Adapted from
Haeseleer et al. (2004). (B) The signaling cascade in rod bipolar cell dendrites is poorly
understood. mGluR6 activation leads to the activation of G
o
α, which though unknown
mechanisms leads to the closure of a non-selective cation channel whose identity is also
unknown. The light-evoked reduction in glutamate release relieves activity in this
cascade and opens cation channels leading to depolarization.
to the rod spherule where it reduces Ca
2+
influx and transiently reduces glutamate release,
thereby reducing postsynaptic G-protein coupled signaling and opening cation channels
and depolarizing the rod bipolar cell. In the transformation of the rod photovoltage to a
8
bipolar cell voltage, the time course of signaling also becomes speeded (see Figure 1.1),
which may allow the rod system to encode accurately the arrival time of photons (Bialek
and Owen, 1990; Rieke and Baylor, 1998).
Effective signal transmission across this synapse requires that the rod
photovoltage is sufficient to alter the open probability of Ca
2+
channels and change the
internal Ca
2+
concentration, and for the change in internal Ca
2+
concentration to reduce
significantly the level of glutamate release. The presynaptic properties of signal
transmission also have the important role of setting the dark release rate of glutamate,
which may play an instructive role in nonlinear signal transmission at this synapse (see
below; (Sampath and Rieke, 2004)).
1.2.1 Voltage sensitivity of Ca
2+
channels
In the face of continuous depolarization, as experienced by vertebrate rods and
cones in darkness (V
rod
~ -40 mV; Schneeweis and Schnapf, 1995), the photoreceptor
synapse must be sensitive to changes in membrane voltage under circumstances where
many types of voltage-activated Ca
2+
channels are in the inactivated state (Hille, 2001).
Two fundamental properties of Ca
2+
channels must hold to account for the strong
dependence of glutamate release on Ca
2+
concentration (Attwell et al., 1987; Belgum and
Copenhagen, 1988; Thoreson et al., 2004). First, the Ca
2+
channels must support a stable
Ca
2+
flux that can maintain the dark release rate of glutamate. Second, the voltage
sensitivity of the channels must be tuned to the range of voltages experienced by the rod
in darkness. It is known that the Ca
2+
channels expressed at the rod synapse are Ca
v
1.4
L-type channels (Morgans, 2001), whose properties appear to be suited ideally for rod-to-
rod bipolar cell transmission.
9
A striking feature of Ca
v
1.4 channels, even compared to other L-type channels in
its family, is the lack of Ca
2+
-dependent inactivation and the relatively slow voltage-
dependent inactivation. Either form of inactivation would gradually reduce the dark Ca
2+
flux and thus glutamate release. Indeed recent work has identified a specialized domain
in C-terminus of Ca
v
1.4 that prevents Ca
2+
-dependent inactivation (Wahl-Schott et al.,
2006). Furthermore, voltage-dependent inactivation in these channels appears slow, with
a time constant on the order of seconds and dependent on the combination of a and b
subunits expressed (Koschak et al., 2003). What is intriguing about these findings is that
they suggest that the Ca
2+
change in the synapse should track the hyperpolarizing rod
photovoltage (see also below), thereby raising questions about the mechanism that
transforms the slow rod light response into the fast bipolar cell light response (see also
Figure 1.1). If the Ca
2+
change at the synapse isn’t accelerated with respect to the rod
photovoltage and the bipolar cell tracks faithfully the change in synaptic glutamate
(Shiells and Falk, 1994), then fast glutamate changes in the synaptic cleft must result
from fast reuptake (Hasegawa et al., 2006) or an additional presynaptic mechanism(s).
For instance, this may result from Ca
2+
induced changes in the efficacy of release (Hsu et
al., 1996). One potential mechanism subserving this speeding may thus be a rapidly-
shifting Ca
2+
dependence of glutamate release to lower Ca
2+
concentrations, allowing
glutamate release to reach dark levels before the Ca
2+
concentration recovers.
A requirement for signal transmission near absolute visual threshold is that Ca
v
1.4
must be sensitive to small voltage changes. Indeed an amplification of voltage, or
voltage gain, in rod to bipolar cell signal transfer increases the detectability of the light
response above synaptic and cellular noise (Ashmore and Falk, 1980; Capovilla et al.,
10
1987). Thus in an ideal situation the rod voltage should be positioned at or near the
steepest point in the relationship between voltage and channel opening (Attwell et al.,
1987; Belgum and Copenhagen, 1988; Thoreson et al., 2004). As alluded to previously,
mutations that alter the voltage sensitivity are known to impair low light level vision
(Hemara-Wahanui et al., 2005; Zeitz et al., 2006). More recently proteins like CABP4,
Calcium Binding Protein-4, have proven essential in optimizing the position of this
voltage dependence by shifting it to more negative membrane potentials where small
changes in voltage will alter more significantly the Ca
2+
current (Haeseleer et al., 2004).
It is possible that CABP4, as well as perhaps other unidentified proteins, give the synapse
the ability to alter the voltage sensitivity of Ca
V
1.4 as the rod voltage changes with light
history. This control of the voltage sensitivity of the Ca
2+
channel allows the synapse to
signal robustly changes in the rod’s voltage and to set the release rate of glutamate over a
wide range of light intensities (see below). Lastly, the low open probability and low
unitary conductance of Ca
v
1.4 channels are ideal for generating voltage-dependent
changes in Ca
2+
concentration that are essential for the low noise transmission of small
graded signals (Doering et al., 2005).
1.2.2 Properties of glutamate release
Unlike conventional synapses where action potentials create transient spots of
high Ca
2+
concentration near Ca
2+
channels, rod synapses must sustain a high rate of
vesicle release under the low Ca
2+
concentrations normally present in the spherule under
physiological conditions (V
rod
~ -40 mV; 0.5 – 3µM [Ca
2+
]; Rieke and Schwartz, 1996).
Indeed rod terminals carry a vesicle pool sensitive to Ca
2+
concentration as low as 1 µM
which likely provides sustained vesicle release by sensing averaged Ca
2+
concentrations
11
instead of high Ca
2+
concentrations near channels (Thoreson, 2004). The resulting vesicle
release is asynchronous and thus will reduce fluctuations in synaptic glutamate (Rieke
and Schwartz, 1996). The properties of Ca
v
1.4 described previously assist in creating the
uniform and low noise Ca
2+
distribution, and thus are suitable for asynchronous release.
Within the narrow range of the physiological Ca
2+
concentration, the cooperativity of
Ca
2+
ions on the vesicle release rate becomes negligible and the exocytosis rate appears to
correlate linearly with the Ca
2+
concentration (Thoreson et al., 2004). Collectively with
other evidence that shows linearity in processes involved in the transmission of rod dim
flash responses (Field and Rieke, 2002; Rieke and Schwartz, 1996; Schneeweis and
Schnapf, 1995; Thoreson et al., 2004), the properties of rod exocytosis allow the graded
release of glutamate that faithfully represents the amplitude of rod single-photon
response. This linearity enables rod bipolar cells to interpret glutamate concentrations as
scaled rod voltages and may form the basis for an instructive signal that sets their
threshold (see below).
The unique invaginating structure of the rod synapse is also desirable for low
noise transmission of small graded signals (Figure 1.2). Release sites at rod terminals are
spread longitudinally along the bottom of the ribbon while mGluR6 receptors are located
on the shaft of the rod bipolar dendrites near the mouth of the invagination (Vardi et al.,
2000). The diffusion of glutamate molecules through this distance reduces the glutamate
concentration at the mGluR6 receptors, thereby placing it in a range where high-affinity
mGluR6 receptors are most sensitive in darkness (Hasegawa et al., 2006; Rao-Mirotznik
et al., 1995; Rao-Mirotznik et al., 1998). This mGluR6 location also reduces variability in
distance from discrete release sites, thus suppressing potential noise. Based on the
12
structure of the spherule and a diffusion model, the dark release rate that realizes this
glutamate concentration at mGluR6 and that does not generate frequent false-positive
events was estimated as ~ 100 vesicles/s (Rao-Mirotznik et al., 1998; Schein and Ahmad,
2005; van Rossum and Smith, 1998). Additionally, it has been proposed that vesicle
release might be regular rather than stochastic because Poisson fluctuation in the vesicle
release would overwhelm rod continuous noise, making the release rate resulting from
single-photon absorptions indistinguishable from the dark release rate (Schein and
Ahmad, 2005). Such regularity in the vesicle release may be partly achieved by imposing
a refractory period after vesicle release at individual release sites (Zenisek et al., 2000).
1.3 Postsynaptic thresholding and the elimination of noise
Of the three known pathways for rod signals to reach ganglion cells in the
mammalian retina (Rod Bipolar, Rod-Cone, and Rod-Off Pathways; reviewed by
Bloomfield and Dacheux, 2001), only the Rod Bipolar Pathway pools enough rods to
account for the high sensitivity of rod vision near absolute visual threshold. It has been
recognized for more than 20 years that to account for this high sensitivity, where one can
detect few photoisomerizations in thousands of pooled rods (c.f. Barlow et al., 1971), rod
outputs cannot be pooled linearly. Early measurements of dark noise from primate rods
indicate a noise variance that would swamp out a single-photon response in a rod bipolar
cells pooling 20-100 rods if the rod output were simply summed (Baylor et al., 1984; see
also Figure 1.3). A threshold-like mechanism at the synapse between rods and rod
bipolar cells has been suggested as a way of eliminating noise from rods, and has been
studied analytically (Clark and van Rossum, 2006; Field and Rieke, 2002; van Rossum
and Smith, 1998).
13
As mentioned above, a main source noise that must be considered at an individual
rod-to-rod bipolar synapse is the continuous noise generated in the phototransduction
cascade by the spontaneous activation of cGMP phosphodiesterase (Rieke and Baylor,
1996). If a threshold is going to be effective in distinguishing single-photon events from
the continuous noise, it must be precisely positioned. First, the amplitude of the
threshold must be high enough to exclude as much of the continuous noise as possible.
Second, the amplitude of the threshold must not be too high to exclude single-photon
events. Such positioning becomes problematic when the amplitude distribution of the rod
continuous noise overlaps significantly with the amplitude distribution of single-photon
responses, requiring a tradeoff between these two parameters.
Field and Rieke (2002) approached this issue in the mouse retina by measuring
the distributions of noise amplitude and single-photon response amplitude in rods, and
determining how these shape rod-to-rod bipolar signal transmission. They found that the
responses of rod bipolar cells depended supralinearly on flash strength compared to the
linear relation in rods. By combining their data on rod bipolar cell responses with a
model where rod signals are passed through a sharp threshold and summed (see also van
Rossum and Smith, 1998), Field and Rieke (2002) determined that to explain the data the
threshold would need to be positioned at 1.3 times the average amplitude of the rod
single-photon response. This is close to the optimal position they determined for the
ideal separation of the rod single-photon response from the underlying rod (continuous)
noise. In other words, the threshold eliminates rod single-photon responses with
amplitudes less than 1.3 times average, while those larger than 1.3 times average are
retained. A striking feature of this conclusion is that the threshold eliminates almost 75%
14
Fig. 1.3. Convergence at the rod-to-rod bipolar synapse.
(A) A rod bipolar cell pools inputs from many rods, but near absolute visual threshold only
one rod may absorb a photon (red) while the remaining rods are generating electrical noise
(blue). A nonlinear threshold (dashed) may improve photon detection at this synapse by
retaining responses in rods absorbing a photon and discarding responses of the remaining
rods. It should be noted that optimal position of the threshold might be expected to increase
given the gap junctional coupling of rods (see text). (B) Linear versus nonlinear signal
processing can improve the fidelity of rod signals. If rod outputs from (A) are simply summed
the resulting trace is noisy, but when summed after applying a threshold for each rod in (A)
the response is more detectable. Adapted from Field and Rieke (2002).
of the rod single-photon responses. While the elimination of such a large fraction of the
rod single-photon responses seems high for a system trying to maximize visual
sensitivity, such a nonlinear operation improves the signal-to-noise ratio of rod signals by
~350-fold over their simple linear pooling. Some debate exists about the exact position
of this nonlinearity depending on the model (0.85 times amplitude; Taylor and Smith,
2004), but nonetheless a postsynaptic threshold that eliminates a significant fraction of
the rod single-photon responses appears to optimize the transfer of rod signals
downstream to the retina. One might expect that the position of such a postsynaptic
threshold should vary by species depending on the signal-to-noise ratio of their rod
15
single-photon responses (see below). Along these lines, one may also expect that near
absolute threshold the extent rods are electrically-coupled to their neighbors (see Figure
1.3) will raise the position of the threshold, since coupling will increase noise variance in
the spherule of the rod absorbing a photon and thus degrade the signal-to-noise ratio of
the single-photon response (Hornstein et al., 2005). Such coupling may explain why the
position of the threshold measured by Field and Rieke (2002) is higher than the optimal
position determined for uncoupled rods.
Sampath and Rieke (2004) further investigated the physiological mechanism
underlying this nonlinear threshold. Previous studies have shown linearity in the
mechanisms that convert the rod voltage to changes in glutamate release rate in the rod
spherule (Field and Rieke, 2002; Rieke and Schwartz, 1996; Schneeweis and Schnapf,
1995; Thoreson et al., 2004). However, given the high rate of release in darkness of
glutamate (Trifonov, 1968) a reasonable hypothesis is that saturation at the synapse might
be responsible for suppressing small single-photon responses that don’t reduce the
synaptic glutamate enough (van Rossum and Smith, 1998). Sampath and Rieke (2004)
were able to demonstrate that the extent of nonlinearity of rod bipolar cell responses
depends on the magnitude of G-protein coupled signaling. By utilizing high affinity
agonists and antagonists to activate or inactivate selectively a fraction of the postsynaptic
glutamate receptor, mGluR6, they found that they could exacerbate or ameliorate the
extent of nonlinearity. In fact for many cells where they applied the mGluR6 antagonist
LY341495, they found that the synapse can be restored to linearity. This result indicates
that a majority of nonlinearity can be explained by saturation within the G-protein
16
cascade that controls the rod bipolar response, and not by saturation at the level of
mGluR6 receptors.
Taken collectively, the information presented above allows us to describe the role
of nonlinear signaling at the rod-to-rod bipolar synapse and how it is controlled. Near
absolute threshold the ability of the retina to distinguish single-photon absorptions from
noise will ultimately be dependent on individual rods. In the mouse retina, the signal-to-
noise ratio of the rod single-photon response is ~ 3, with distributions of noise amplitude
and single-photon responses amplitude overlapping significantly (Field and Rieke, 2002).
However, in other mammalian species with a higher signal-to-noise ratio (~ 5 in primates
and ~ 4 in guinea pigs; Field and Rieke, 2002) signal transmission should require less
nonlinearity. Under these circumstances the rod single-photon response is more
distinguishable from the underlying (continuous) noise, and thus the threshold can be
positioned lower with respect to the average single-photon response amplitude to provide
optimal separation between signal and noise. We predict a strong negative correlation
between the rod signal-to-noise ratio (or any other measure of the fidelity of rod signals;
Clark and van Rossum, 2006) and extent of nonlinearity across all mammalian species.
The origin of nonlinear signaling within the G-protein cascade, rather than at the
level of the mGluR6 receptors (Sampath and Rieke, 2004), provides a site for potential
modulation. It would seem that the extent of nonlinearity could be set by the relative
expression of the G-protein, G
o
, or other components of the signaling cascade that remain
largely unknown. However if the expression of some component of the G-protein
cascade is setting the extent of nonlinear signaling at the rod-to-rod bipolar synapse,
especially given the near-optimal position of the threshold in mice (Field and Rieke,
17
2002), how its concentration is set becomes an interesting problem. In particular the
expression level of these signaling elements would either need to be encoded genetically,
guided by an instructive signal, or some combination of both. Near absolute threshold,
where a single rod in 10,000 may be absorbing a photon, the benefit of nonlinear signal
transmission is significant, providing ~ 350-fold improvement of the signal-to-noise ratio
over the linear combination of rod signals (Field and Rieke, 2002). However, this benefit
diminishes as the light level increases. Perhaps the extent of nonlinearity is genetically
encoded for optimal separation of signal and noise near absolute threshold, but even at
higher light levels some modulation of nonlinearity may modestly improve the rod
signal-to-noise ratio. Such modulation may result from an instructive signal, which in
this case could be the dark release rate and fluctuations in synaptic cleft glutamate
concentration as these define the functional properties required for the separation of light-
evoked signals from noise.
A problem with the idea of an instructive signal controlling the extent of
nonlinearity is that from the perspective of a single rod-to-rod bipolar cell synapse, or a
single rod bipolar cell, this signal may not occur at a high enough frequency to guide this
mechanism. Under these circumstances the instructive signal needs to originate at a level
of the retina that integrates over many rods (i.e. AII amacrine cell). Alternatively, at light
levels above 1 photon absorbed in 100 rods per rod integration time (i.e. where a rod
bipolar cell ‘sees’ at least 1 absorbed photon in its receptive field per integration time) the
rod bipolar cell itself could act as the integrator. These instructive signals could guide the
expression of G
o
or downstream elements either at the level of nuclear expression, or by
local protein synthesis in the dendrites for faster changes (Kang and Schuman, 1996).
18
Early psychophysical studies of detection threshold and recordings from ganglion
cells have emphasized the notion that the visual system can detect few photon
absorptions in large pools of rod photoreceptors. Such sensitivity can only be borne out
of early biophysical mechanisms that preserve information about single-photon
absorptions in individual rods, and allow those signals to be propagated efficiently
through the retina. In this review we emphasize that precise tuning of the rod-to-rod
bipolar cell synapse is mediated by both pre- and post-synaptic mechanisms. Presynaptic
mechanisms must amplify small graded signals generated by a single rhodopsin
molecule, transmit the voltage change with the least loss to the synapse, and reduce
significantly the synaptic Ca
2+
concentration and glutamate release. Furthermore the dark
release rate and fluctuations in synaptic cleft glutamate concentration appears to serve as
the instructive signal for the postsynaptic nonlinear removal of noise. Thus coordinated
pre- and post-synaptic mechanisms at the rod-to-rod bipolar cell synapse are crucial for
the transmission of the rod single-photon response and our exquisite visual sensitivity.
19
Chapter 2
Genetic dissection of the relationship between
optimal processing of sensory signals and
behavioral sensitivity
2.1 Introduction
The signal and noise of sensory receptors set the fundamental limit to sensation,
but subsequent neural processing must optimally process these signals for sensory
systems to reach this limit. The exquisite sensitivity of sensory receptors has often been
highlighted (Baylor et al., 1979; Bhandawat et al., 2005; Hudspeth and Corey, 1977) and
has been proposed as limiting behavioral performance (Barlow, 1956; see also Bialek,
1987). However, the necessity of precisely tuning post-receptor processing to match the
signal-to-noise properties of the sensory receptors, and the behavioral consequences of
failing to do so, has received significantly less attention. Furthermore, it remains unclear
whether the biophysical mechanisms that underlie this neural processing can be tuned to
alterations in the properties of the sensory receptors, and what impact this has finally on
threshold for behavioral detection.
Near absolute visual threshold humans are capable of identifying few photon
absorptions among thousands of rod photoreceptors (Hecht et al., 1942; Teich et al.,
1982; van der Velden, 1946). Noise in rods has been considered limiting for detection
due to the statistical variability in the absorption of single photons, and intrinsic dark
noise in the phototransduction cascade. The detection of light thus involves
20
discriminating single photon responses in a handful of rods from the substantial dark
noise generated by the vast majority (reviewed by Field et al., 2005). Given the sparse
photon absorptions near absolute visual threshold in the array of rod photoreceptors, the
nonlinear summation of rod signals has been proposed to be critical for eliminating noise
in rods not absorbing photons, thereby increasing the signal-to-noise ratio (SNR) of the
single photon response in downstream cells (Field and Rieke, 2002; van Rossum and
Smith, 1998). An optimally positioned nonlinear threshold (Field and Rieke, 2002)
produced by postsynaptic saturation at the rod-to-rod bipolar cell synapse (Sampath and
Rieke, 2004) appears to underlie mechanistically the elimination of rod noise. However,
several outstanding questions remain about how this optimal processing is tuned and its
implications for behavioral detection: (1) Can the position of the nonlinear threshold, or
extent of postsynaptic saturation, adjust to separate optimally the rod signal and noise as
their distributions change? and (2) What are the behavioral consequences if the
downstream processing of rod signals are not set optimally to the single photon response
SNR?
Here we show that increasing the SNR of rod photoreceptors does not necessarily
result in improved behavioral performance. Rather that downstream processing of the rod
responses must also be altered to process efficiently the rod signals and reject rod noise.
We studied the properties of the single photon response and noise in rods and rod bipolar
cells of GCAPs
-/-
mice, which display a higher SNR of the rod single photon response.
Despite this higher SNR, we found surprisingly that behavioral performance in GCAPs
-/-
mice deteriorated. A poorly positioned nonlinear threshold at the synapse between rods
and rod bipolar cells could mechanistically explain the elevated behavioral threshold. In
21
particular, the extent of postsynaptic saturation did not adjust to the altered distributions
of signal and noise in GCAPs
-/-
rods, and therefore did not separate effectively the two.
These results demonstrate that optimal processing of sensory signals is critical to capture
fully the signals present in the sensory receptor cells near absolute threshold.
2.2 Materials & Methods
2.2.1 GCAPs
-/-
mice
GCAPs
-/-
mice, generated from the targeted deletion of Guanylyl Cyclase
Activating Protein 1 and 2, have previously been used to study the properties of Ca
2+
feedback on the synthesis of cGMP in the vertebrate rod phototransduction cascade, and
their photocurrents have been characterized extensively (Burns et al., 2002; Mendez et
al., 2001). Briefly, that the lack of feedback on cGMP synthesis resulted in larger and
slower single photon responses than WT, and also resulted in an increase in rod
continuous noise. Despite these alterations in response properties, the dark current of
GCAPs
-/-
rods was statistically indistinguishable from WT (Mendez et al., 2001), also
consistent with recordings presented here; the dark current in WT rods was 14.2 + 0.4 pA
(mean + SEM, n = 22), and GCAPs
-/-
rods was 15.3 + 0.4 pA (mean + SEM, n = 23), p-
value = 0.07. Furthermore, measurements of the resting membrane potential were made
in current clamp from WT and GCAPs
-/-
rods in dark-adapted retinal slices (see Okawa et
al., 2008); the membrane potential of WT rods was -42.0 + 1.8 mV (mean + SEM, n=9),
and in GCAPs
-/-
rods was -43.8 + 2.0 mV (mean + SEM, n=8), p-value = 0.5. Thus the
targeted deletion of GCAPs alters the properties of the light response without changing
the rod photoreceptor dark current or resting membrane potential, which controls
glutamate release at the rod spherule (Trifonov, 1968). To minimize the strain
22
differences in the physiological or behavioral experiments, we used mice bred for > 5
generations into a C57Bl/6 background (WT). Finally, GCAPs are expressed only in the
rod and cone photoreceptors of the retina, and their targeted deletion has little effect on
the expression of other genes or retinal morphology (Mendez et al., 2001).
2.2.2 Preparation
Procedures used for the handling of retinal tissue, and electrophysiological
recordings from rod photoreceptors and rod bipolar cells were similar to those in
Sampath et al. (2005). Mice were dark-adapted overnight and sacrificed according to
protocols approved by the Institutional Animal Care and Use Committee of the
University of Southern California. Briefly, eyes were enucleated under infrared
illumination, the lens and cornea removed, and the resulting eyecups were stored at 32
o
C
in Ames’ media equilibrated with 5% CO
2
/95% O
2
. All subsequent manipulations of the
retinal tissue were done under infrared illumination (> 950 nm) using infrared image
converting goggles. For electrophysiological recordings the retinas were prepared as
described below and superfused with Ames’ media heated to 35-37
o
C.
2.2.3 Electrophysiological Recording and Analysis
Light-evoked currents from rod photoreceptors were measured with suction
electrodes from finely chopped pieces of retinal tissue. Clusters of cells with the outer
segments protruding were targeted and individual rod outer segments were drawn gently
into a suction electrode containing Ames’ media buffered with 10 mM HEPES to pH 7.4.
Light-evoked currents were measured following 10 ms flashes from an LED (λ
max
~ 470
nm, FWHM ~ 30 nm) or 30 ms flashes from tungsten-halogen source passed through an
interference filter (λ
max
~ 500 nm, FWHM ~ 15 nm), low-pass filtered at 20 Hz with an 8-
23
pole Bessel filter, and digitized at 1 kHz. It should be noted that the external solution
used in these experiments, Ames media, has a different ionic composition that the Locke
solution used in previous studies (Baylor et al., 1984; Burns et al., 2002), which in turn
makes the rod photoresponse slower and more sensitive to light.
Light-evoked currents from rod bipolar cells were recorded with patch electrodes.
Whole-cell and perforated-patch recordings were made from rod bipolar cells in dark-
adapted retinal slices as described previously (Armstrong-Gold and Rieke, 2003;
Sampath et al., 2005). The internal solution for these experiments consisted of (in mM):
125 K-Aspartate, 10 KCl, 10 HEPES, 5 NMG-HEDTA, 0.5 CaCl
2
, 1 ATP-Mg, 0.2 GTP-
Mg; pH was adjusted to 7.2 with NMG-OH. Flash families were measured in response to
a 10 ms flash from a blue LED (λ
max
~ 470 nm, FWHM ~ 30 nm) whose strength varied
from generating a just-measurable response and increased by factors of 2. Membrane
currents were low-pass filtered at 300 Hz by an 8-pole Bessel filter, and digitized at 1
kHz. The signal-to-noise ratio (SNR) in both rods and rod bipolar cells was calculated
from amplitude histograms generated from light-evoked responses to a repeated dim
flash. For each rod photoreceptor response amplitudes of individual trials were
determined by the correlation of those trials with the average of all trials. The amplitude
histogram of the rod was fit by the equation:
!
P(A) ="A
exp(#n)n
n
n!
[2$(%
D
2
+n%
A
2
)]
#1/2
exp[#
(A#nA#D)
2
2(%
D
2
+n%
A
2
)
]
n=0
&
'
,where P(A) is the probability of obtaining responses in amplitude between A - ΔA/2 and
A + ΔA/2, σ
D
is the standard deviation of dark noise, σ
A
is the standard deviation of
single photon response amplitude with
!
A as its mean, and n is the number of
24
photoisomerizations produced by a flash with
!
n is its average (see Field and Rieke,
2002).
!
D is the mean amplitude of dark noise and was introduced only for GCAPs
-/-
rods
whose noise peaks in the histograms skewed slightly from zero. This equation assumes
that photon absorptions by rods obey Poisson statistics and that dark noise and variations
in single photon responses are independent and additive. For rod bipolar cells, whose
dim flash responses demonstrate a supralinear dependence on flash strength (Field and
Rieke, 2002), the histogram peak representing the dark noise and single photon response
amplitudes were fit with Gaussian distributions. These Gaussian distributions were then
summed to provide the overall fit to the histogram. The signal-to-noise ratio (SNR) for
both rods and rod bipolar cells was defined as distance between noise and single photon
response peaks divided by the square root of the sum of their variances, which for rods
becomes
!
(A"D)/(2#
D
2
+#
A
2
)
1/2
, and for rod bipolar cells becomes
!
(A"D)/(#
D
2
+#
A
2
)
1/2
(see Dunn et al., 2006). The bins of histograms were chosen for each rod in a way that
maximizes the SNR based on the number of trials, noting that histograms for rod bipolar
cells always contained fewer trials due to the limited stability of the perforated-patch
recordings (and GCAPs
-/-
rod bipolar cells had the fewest trials given the slower response
kinetics).
Statistical significance in all experiments was ascertained using a two-tailed
Student’s t-test, for which p-values are provided.
2.2.4 Mouse Behavioral Assay
To determine how alterations in the rod photoresponse and noise influence
behavioral threshold for vision, WT and GCAPs
-/-
mice were trained to find a black wall
in a white water maze (Sampath et al., 2005). Mice were trained in bright light to escape
25
from the maze at the black wall by crawling out of the cool (17-18
o
C) water onto a
platform. A closed-circuit infrared-sensitive camera recorded individual trials and the
time required to find the platform was determined post hoc from the video. Mice were
placed into the maze 3 to 5 times per day, and the position of the target was randomized
for each trial. To randomize further the time between trials, we tested the mice from first
to last, last to first, etc. For safety, mice were removed from the maze if the time to
escape exceeded 30 s, and those trials were recorded as 30 s. Once the average time to
find the platform remained constant for 3 consecutive days at a given light intensity, the
light intensity was lowered by 1 log unit, and by 0.3 log units as the light intensity came
closer to absolute threshold. This process was repeated until the average time to find the
platform was the same as in darkness (~ 20 s). In some trials we found the mice learned
more quickly if trials were spaced with shorter intervals. Under those circumstances we
divided the mice into groups of four and 3-5 trials were given consecutively to each
group.
The maze was illuminated by a dissecting microscope light source (Thorlabs,
Newton, NJ), from which a liquid light guide illuminated a lens tube containing an
interference filter (λ
max
~ 500 nm, FWHM ~ 40 nm), and absorptive neutral density filters
(Thorlabs). A trifurcating fiber (Newport) originating on the opposite end of the lens
tube was placed at 120
o
around the maze, and were positioned to project off a
photographic reflector above the maze to yield uniform illumination. The uniformity of
the illuminating light was tested daily, and was always within 10% of the mean light
level. Photon flux was measured using a calibrated photometer (United Detector
26
Technologies, San Diego, CA), and the resulting flux was calculated in equivalent
photons at the rod photoreceptor peak spectral sensitivity (~ 501 nm).
2.2.5 Calibration of Light Intensity
Light intensities were calculated for recordings from rod photoreceptors and rod
bipolar cells, as well as behavioral experiments, as an effective photon flux at the peak
wavelength of spectral sensitivity for mouse rhodopsin (λ
max
~ 501 nm) to facilitate
comparison. In each instance the effective number of activated R per rod (R*/rod) was
determined by convolving the power-scaled output of the light source with the
normalized spectral sensitivity curve for mouse rhodopsin, correcting for added neutral
density in the light path. The R*/rod can then be calculated based on the effective
collecting area in each setup. In rod recordings light was focused on the preparation by a
10X 0.25NA (Nikon) water-immersion condenser objective, and the collecting area was
estimated based on the scaling between the time-dependant variance and mean response
to a dim flash as 0.5 µm
2
(Field and Rieke, 2002). In rod bipolar cell recordings light
was focused on the slice preparation by a 20X 0.75NA (Nikon) condenser objective and
the collecting area based on the scaling of the time dependent variance to the mean
response to a dim flash was estimated to be 0.18 µm
2
(Cao et al., 2008). Light delivery in
the behavioral setup is described above and an effective collecting area for rods at the
level of the cornea was 0.2 µm
2
(Lyubarsky and Pugh, 1996).
2.3 Results
2.3.1 Higher signal-to-noise ratio in GCAPs
-/-
rods
To investigate the influence of rod photoreceptor single photon response signal-
to-noise ratio (SNR) on retinal signaling and behavioral threshold we measured the
27
properties of the light-evoked response in rods of wild-type (WT) and GCAPs knockout
(GCAPs
-/-
) mice (see Experimental Procedures). Figure 2.1A demonstrates flash families
measured from WT and GCAPs
-/-
rods, and the derived rod single photon responses
(Figure 2.1B) show a similar initial rate of rise but continue in GCAPs
-/-
rods to a longer
15 10 5 0
30 20 10 0
Time (s)
100
50
0
4 2 0
40
20
0
10 5 0
Amplitude (pA)
n = 1598
n = 292
WT
GCAPs
-/-
A.
D. C.
B.
5%
5 pA
5 pA
2 pA
2 pA
0.5 s 0.5 s
0.5 s
WT GCAPs
-/-
Fig. 2.1. Light-evoked responses of WT and GCAPs
-/-
rod photoreceptors.
(A) Response families to 30 ms flashes (arrow) WT and GCAPs
-/-
rods. Flash strengths
for the WT rod yielded 0.43, 1.4, 4.1, 19, 58, and 180 Rh*, and for the GCAPs
-/-
rod
yielded 0.34, 0.76, 1.9, 4.0, 8.3, and 15 Rh*. (B) Derived single photon responses from
WT and GCAPs
-/-
rods plotted as a fractional suppression of the dark current versus time.
The WT response was derived from 2137 trials across 9 cells. The GCAPs
-/-
response was
derived from 4280 trials across 23 cells. The effective number of cells was estimated
from the total number of trials divided by the number of trials on the cell with the most,
and thus can be non-integer. The time-to-peak of the single photon response was 208 +
11 ms (n = 4.1) in WT rods and 503 + 39 ms (n = 18.4) in GCAPs
-/-
rods (mean + SEM).
The integration time of the averaged response was 540 ms for WT rods and 900 ms for
GCAPs
-/-
rods. (C) Individual trials to flashes of fixed strength in WT and GCAPs
-/-
rods.
In each epoch a flash was delivered (up triangle) at 0.2 s and on average generated 0.9
Rh* in WT rods, and 1.7 Rh* in GCAPs
-/-
rods. The average single photon response was
shown as the right-most epoch (grey) for comparison. Scale bars on the left are 2 pA.
(D) The SNR for WT and GCAPs
-/-
rods were determined from the fits of histograms fit to
the data in (C) (see Experimental Procedures for details). Note the difference in the
abscissa for WT and GCAPs-/- rods.
28
time-to-peak. To determine how these alterations in the rod photocurrent and noise
impact signal transfer to downstream retinal cells, we measured the SNR of the single
photon response for WT and GCAPs
-/-
rods (Dunn et al., 2006; see Experimental
Procedures). First, in WT rods an amplitude histogram was generated from suction
electrode recording by delivering repeatedly a flash of light that on average generated 0.9
R*, as shown in Figure 2.1C. Based on the correlation between each individual trial and
the average single photon response (Figure 2.1B), response amplitudes were computed
(Figure 2.1D). The histogram of response amplitudes was fit as a sum of Gaussian
distributions whose areas were weighted by the Poisson probability of absorbing 0, 1, 2
… photons (Baylor et al., 1979; Field and Rieke, 2002; see Experimental Procedures).
Based on this fit we calculated the SNR in WT rods to be 1.69 + 0.15 (mean + SEM, n =
4; see Figure 2.6).
A superficial inspection of the dim flash histogram in GCAPs
-/-
rods (Figure 2.1D,
1.7 R* on average per flash) would suggest a higher SNR, as the peaks surrounding zero
current (i.e. noise), and ~ 3.5 pA (single photon response) are more separated than for
WT rods. Thus on individual trials, the response to a single absorbed photon could be
discriminated from dark noise more robustly, as was observed in previous studies that
allowed the determination of the discrete noise event rate (Burns et al., 2002). The
calculated SNR in GCAPs
-/-
rods was 2.51 + 0.08 (mean + SEM, n = 6, see Figure 2.6),
or ~ 1.5-fold higher than for WT rods. It should be noted that the higher SNR comes
with an increase in dark noise that is reflected in the wider noise distribution in the
GCAPs
-/-
rods (note difference in scale for Figure 2.1D, bottom). We estimated from
these recordings that the noise variance in the photocurrent of GCAPs
-/-
rods was > 5-fold
29
larger than WT rods, broadly consistent with the increase in noise seen in previous work
(Burns et al., 2002). The improvement of the SNR of GCAPs
-/-
rods compared to WT
would suggest that behavioral threshold would be lower, as the SNR of sensory receptors
has been considered to play a fundamental role in setting absolute visual threshold
(Barlow, 1956).
2.3.2 Elevated behavioral threshold in GCAPs
-/-
mice
To ascertain the influence of increased SNR of the rod photoresponse in GCAPs
-/-
rods on behavioral threshold, we tested the ability of WT and GCAPs
-/-
mice to escape
from a water maze. The water maze task has been previously used to characterize
changes in behavioral threshold for mice lacking the protein recoverin (Sampath et al.,
2005). Mice were tested in their ability to escape as the light level in the maze was
gradually reduced. In Figure 2.2A the time to escape from the maze was plotted as a
function of the background light intensity calculated in equivalent 501 nm photons (ϕ)
µm
-2
s
-1
(see Experimental Procedures). At bright light levels both WT and GCAPs
-/-
mice were able to identify the target and escaped quickly, but as light levels decreased the
time required for both mice to escape increased. At the lowest light levels both WT and
GCAPs
-/-
mice were not able to identify the target and thus searched randomly. Under
these circumstances their escape times matched those acquired in darkness (Figure 2.2A).
Light levels at the transition between where WT and GCAPs
-/-
mice can see the
target and where they cannot provide an estimate of the behavioral threshold for seeing.
Impairment in the ability of GCAPs
-/-
mice to find the maze target was apparent at an
intermediate light intensity (0.0073 ϕ µm
-2
s
-1
; Figure 2.2B), indicating that the
elimination of GCAPs increased behavioral threshold despite the higher SNR of the rod
30
1.0
0.8
0.6
0.4
0.2
Time/Time
dark
15
10
5
Escape Time (s)
0.001 0.01 0.1 1
Background Intensity (! µm
-2
s
-1
)
WT (n=8)
GCAPs
-/-
(n=8)
A. B.
**
**
! µm
-2
s
-1
0.00073 0.0073 1.45
0
Fig. 2.2. Behavioral estimate of absolute visual threshold in WT and GCAPs
-/-
mice.
(A) Escape time from the maze plotted versus light intensity in equivalent photons at the
peak absorption wavelength of mouse rhodopsin (501 nm), with total darkness plotted at
an intensity of 0 ϕ µm
-2
s
-1
. Mean + SEM at each intensity are plotted for WT mice (n =
8; black symbols) and GCAPs
-/-
mice (n = 8; grey symbols). The mean and the standard
error are collectively best fit with a Hill Equation where the parameters were allowed to
vary. In WT mice the Hill parameters were I
1/2
= 0.0085 ϕ µm
-2
s
-1
, and n = 2.2, and in
GCAPs
-/-
mice were I
1/2
= 0.013 ϕ µm
-2
s
-1
, and n = 2.9. (B) Differences in escape
behavior were more pronounced near behavioral threshold, which occurs in the transition
from higher light levels where mice can see the target and escape quickly, to lower light
levels where mice cannot see the target and search randomly. To facilitate the comparison
of escape times at these light levels the normalized time to escape on a mouse-by-mouse
basis is plotted at 0.00073 (low), 0.0073 (transition), and 1.45 (high) photons µm
-2
s
-1
. At
the transition background light level of 0.0073 photons µm
-2
s
-1
the difference in escape
times was statistically significant between WT and GCAPs
-/-
mice, p-value = 0.001 *
Student t-test p-value < 0.01.
photoresponse. We estimated behavioral threshold as the inflection point in the Hill
curve fit that characterizes the transition from higher light levels where the mice saw the
target and escaped quickly, and lower light levels where the mice could not see the target
and thus searched randomly. In WT mice the threshold background light level was ~
0.0085 ϕ µm
-2
s
-1
and in GCAPs
-/-
mice the threshold background light level was ~ 0.013
ϕ µm
-2
s
-1
. Given a 0.2 µm
2
collecting area per rod at the level of the mouse cornea
(Lyubarsky and Pugh, 1996) and the 0.2 s integration time of the rod photoresponse
31
(Walraven et al., 1990), behavioral threshold was ~ 1 R* per 2900 rods per rod
integration time for WT mice, and ~ 1 R* per 1900 rods per rod integration time for
GCAPs
-/-
mice. This measure of behavioral threshold thus reflects an ~ 1.5-fold elevation
in GCAPs
-/-
mice compared to WT mice. Thus, at these light levels behavior is
determined by the ability of the retinal cells to distinguish the single photon response in
few rods from the underlying dark noise of the majority.
2.3.3 Altered response properties of GCAPs
-/-
rod bipolar cells
The elevated behavioral threshold of GCAPs
-/-
mice, despite the higher SNR of
the rod photoresponse, indicates that SNR is degraded at some point downstream of the
rod photoresponse. Rod signals near absolute visual threshold are passed to rod bipolar
cells, the first stage in a specialized low-light level circuitry in the mammalian retina
called the rod bipolar pathway (Dacheux and Raviola, 1986; Smith et al., 1986). We
sought to understand how the larger but slower single photon response in GCAPs
-/-
rods,
as well as the increase in dark noise, impacted signaling in rod bipolar cells. Response
families from WT and GCAPs
-/-
rod bipolar cells are shown in Figure 2.3A. Despite the
large difference in the time-to-peaks of rod single photon responses, the time-to-peaks of
rod bipolar dim responses were similar between WT and GCAPs
-/-
rod bipolar cells
(Figure 2.3A, dashed). Rather, the slowing of GCAPs
-/-
rod responses appears to be
reflected in the slow recovery of the rod bipolar response.
In addition to the slowing of the rod bipolar response, the extent of nonlinearity in
signal transmission between rods and rod bipolar cells was reduced in GCAPs
-/-
rod
bipolar cells. The extent of nonlinearity has been estimated previously from the
relationship between the average response amplitude versus the flash strength, and is
32
directly related to the position on the nonlinear threshold between rods and rod bipolar
cells (Field and Rieke, 2002; Sampath and Rieke, 2004). In Figure 2.3B we plot the
normalized response amplitude versus the flash strength for WT (black) and GCAPs
-/-
(grey) rod bipolar cells. This relationship can be fit with a Hill equation, whose exponent
reveals greater linearity in signal transmission in GCAPs
-/-
rod bipolar (n ~ 1.25) cells
compared to WT (n ~ 1.55). For the linear summation of rod signals in rod bipolar cells
we predict a Hill exponent of 1, which will increase as the nonlinear threshold eliminates
more single photon responses (Field and Rieke, 2002). A reduction the Hill exponent
may be expected a priori by virtue of the greater SNR of the rod photoresponse in
GCAPs
-/-
mice. Thus the reduced nonlinearity may result from the more effective
separation of the signal and noise distributions that would lead to the loss of fewer single
photon responses (Field and Rieke, 2002; see Discussion).
2.3.4 Increased noise in GCAPs
-/-
rod bipolar cells
To determine the impact of higher cellular noise of GCAPs
-/-
rods on downstream
processing, we measured the dark noise in rod bipolar cells of WT and GCAPs
-/-
mice.
We determined the magnitude of the noise during whole-cell recordings (V
m
= -60 mV)
by dialyzing the cell with 25 µM GTP-γ-S, as shown in Figure 2.3C. As GTP-γ-S
dialyzes into the rod bipolar cell it will activate strongly the G
o
G-protein, resulting in the
closure of transduction channels (Sampath and Rieke, 2004). The magnitude of the dark
noise in the rod bipolar current can be estimated by comparing the noise variance
immediately after whole-cell break-in versus the noise when all of the transduction
channels are pinned closed with GTP-γ-S. Figure 2.3C demonstrates that in GCAPs
-/-
rod
bipolar cells considerably more noise was present upon whole-cell break-in than for WT
33
1.0 0.5 0.0
Time (s)
4
5
6
7
0.1
2
3
4
5
6
7
1
R/R
max
2 4 6 8
1
2 4 6 8
10
2
Flash Strength (Rh*/rod)
GCAPs
+/+
(n=22)
GCAPs
-/-
(n=10)
50 pA
50 pA
A. B.
GCAPs
+/+
GCAPs
-/-
0.001
0.01
0.1
1
Power (pA
2
/Hz)
0.1 1 10 100
Frequency (Hz)
GCAPs
+/+
(n=9)
GCAPs
-/-
(n=7)
-40
-20
0
-40
-20
0
1.0 0.8 0.6 0.4 0.2 0.0
Time (s)
C. D.
Holding Current (pA)
GCAPs
+/+
GCAPs
-/-
Fig. 2.3. Flash response families and dark noise in WT and GCAPs
-/-
rod bipolar
cells. (A) Light-evoked responses to 10 ms flashes (arrow) for voltage-clamped (V
m
= -
60 mV) rod bipolar cells. For the WT rod bipolar cell flash strengths were 0.084, 0.25,
0.59, 1.3, 2.6, 5.3, and 11 Rh*/rod, and for the GCAPs
-/-
rod bipolar cell flash strengths
were 0.084, 0.25, 0.59, 1.3, 2.6, 5.3 Rh*/rod. A vertical dashed line demonstrates the
initial time to peak of the rod bipolar responses are similar between WT (145 + 7.7 ms, n
= 8.2) and GCAPs
-/-
(169 + 19 ms, n = 5.8) rod bipolar cells (mean + SEM) despite the
slowed response of GCAPs
-/-
rods (see Figure 2.1). (B) Average response-intensity
relationship across all WT (n = 22) and GCAPs
-/-
(n = 10) rod bipolar cells. Data were fit
will a Hill equation with n = 1.55 and I
1/2
= 1.8 R*/rod for WT rod bipolar cells, and n =
1.25 and I
1/2
= 1.3 R*/rod for GCAPs
-/-
rod bipolar cells. (C) Noise variance was
measured in voltage-clamped rod bipolar cells immediately after whole-cell break in
(black) and after the dialysis of the cell with 25 µM GTP-γ-S (grey). The average change
in holding current following GTP-γ-S dialysis for WT mice was 10.2 + 0.7 pA (mean +
SEM; n = 10) and in GCAPs
-/-
mice was 15.2 + 2.2 pA (mean + SEM; n = 11), p-value =
0.045. (D) Power spectra of dark noise fluctuations for WT (filled symbols) and GCAPs
-/-
(open symbols) rod bipolar cells. Power spectra after the dialysis of GTP-γ-S (grey filled
and open symbols) was similar indicating that higher noise in GCAPs
-/-
rod bipolar cells
was not due to differences in the quality of the recordings.
34
cells, and following GTP-γ-S dialysis the level of instrumental noise remaining for each
cell type was nearly the same.
In GCAPs
-/-
rod bipolar cells the greater noise is accompanied by an increase in
the holding current at -60 mV (see Figure 2.3C), indicating a larger number of
transduction channels are open in darkness. The mean change in holding current due to
channel closure by GTP-γ-S in darkness for WT rod bipolar cells was ~ 10 pA versus ~
15 pA in GCAPs
-/-
cells, indicating ~ 1.5-fold more transduction channels are open,
which in turn will report more cellular noise. Power spectra of the dark noise (Figure
2.3D) show that across all frequencies the noise variance of the rod bipolar current was
greater in GCAPs
-/-
cells compared with WT. The total increase in noise variance can be
estimated from the difference in the integral of the power spectrum, and reflects an ~ 4.5-
fold increase in rod bipolar cell noise in GCAPs
-/-
mice compared to WT. The noise in
GCAPs
-/-
rod bipolar cells had a significant high frequency component (i.e. > 10 Hz),
suggesting that it reflects noise in the synaptic glutamate concentration due to continuous
release of vesicles (Trifonov, 1968) or within the mGluR6 transduction cascade, and not
attributable to the slow rod photoresponse whose spectral density falls significantly at <
10 Hz (Sampath and Rieke, 2004).
2.3.5 Reduced SNR in GCAPs
-/-
rod bipolar cells
The greater dark noise in rod bipolar cells may influence the detection of single
photon responses, especially if the magnitude of the dark noise is of similar magnitude.
Previous studies have shown that a nonlinear threshold may improve the SNR of the
single photon response in rod bipolar cells over the linear summation of rod outputs near
absolute visual threshold (Berntson et al., 2004; Field and Rieke, 2002; van Rossum and
35
Smith, 1998). The improvement in the SNR can be attributed to the elimination of noise
caused by rod phototransduction and signal transfer to rod bipolar cells, and
mechanistically can be explained by postsynaptic saturation in darkness within the
mGluR6 cascade that controls the gating of the transduction channels in rod bipolar cells
(Sampath and Rieke, 2004).
Such improvement of the SNR in WT rod bipolar cells is shown in Figure 2.4,
and is consistent with previous studies (Field and Rieke, 2002). A dim flash that on
average activates ~ 0.064 R*/rod per trial yielded a series of responses where rod bipolar
cells produced a response with a reproducible amplitude, or no response at all (Figure
2.4A, top). This separation of signal and noise is evident in the amplitude histogram
(Figure 2.4B, top), which has been fit as the sum of two Gaussian distributions centered
at the peak corresponding to the noise and single photon response amplitudes,
respectively. The calculated SNR across 5 WT rod bipolar cells was ~ 2.91 + 0.11 (mean
+ SEM) reflecting an ~ 1.7-fold improvement from WT rods, compared to the ~ 4-fold
decrease in SNR predicted for the linear summation of 20 rods with the signal and noise
characteristics shown in Figure 2.1D (van Rossum and Smith, 1998; Field and Rieke,
2002).
Light-evoked responses measured from GCAPs
-/-
rod bipolar cells in response to
flashes that activate ~ 0.064 R*/rod per trial (Figure 2.4A, middle), on the other hand, are
more difficult to identify in individual trials. Thus the increase in cellular noise in
GCAPs
-/-
rod bipolar cells makes the detection of the single photon response more
problematic. This point is reflected in amplitude histograms from GCAPs
-/-
rod bipolar
36
15
10
5
0
3 2 1 0 -1
10
5
0
3 2 1 0 -1
15
10
5
0
3 2 1 0 -1
Normalized Amplitude
10 5 0
Time (s)
GCAPs
-/-
+ 0.2 µM APB
GCAPs
-/-
WT
A. B.
n = 70
n = 79
n = 74
10 pA
10 pA
10 pA
Fig. 2.4. Estimation of SNR in WT and GCAPs
-/-
rod bipolar cells.
(A) Perforated-patch voltage-clamp (V
m
= -60 mV) recordings from WT and GCAPs
-/-
rod
bipolar cells. Each epoch represents and individual trial where a 10 ms flash (down
triangle) was delivered at 0.2 s (WT and GCAPs
-/-
with 0.2 µM APB) or at 0.4 s (GCAPs
-/-
), and the grey record at the right is scaled to the average single photon response
amplitude for comparison. Flashes for the WT rod bipolar cell on average generated
0.064 R*/rod, for the GCAPs
-/-
rod bipolar cell on average generated 0.064 R*/rod, and
for the GCAPs
-/-
rod bipolar cell with 0.2 µM APB on average generated 0.033 R*/rod.
We assume the APB-dependent increase in synaptic saturation at the rod bipolar cell to be
predominantly a postsynaptic action, as the activation of presynaptic metabotropic
glutamate receptors at the rod spherule would reduce glutamate release and relieve
saturation, the opposite of the effect observed (see also Sampath and Rieke, 2004).
(B) SNRs were determined from the fits of histograms generated from the data in (A) (see
Materials and Methods). For ease of comparison amplitudes have been normalized to the
mean single photon response.
cells which show greater blurring, or overlap, between the peak corresponding to noise
and single photon response amplitudes (Figure 2.4B, middle). The greater overlap
between the peak corresponding to cellular noise and to single photon responses in
GCAPs
-/-
rod bipolar cells implies that responses falling in this range can’t be uniquely
identified as noise or as a single photon response, thereby impairing the ability to identify
37
signals near absolute threshold. The calculated SNR across 16 GCAPs
-/-
rod bipolar cells
was 2.49 + 0.06 (mean + SEM; Figure 2.6), showing no improvement compared to
GCAPs
-/-
rods, and an ~ 1.2-fold deterioration from WT rod bipolar cells. Thus despite
the increased SNR of the rod photoresponse in GCAPs
-/-
mice, the SNR of the
corresponding rod bipolar cells is degraded by virtue of the increase in noise.
2.3.6 Nonlinear threshold is poorly positioned in GCAPs
-/-
rod bipolar cells
The position of the nonlinear threshold in WT mice has been inferred to be close
to the crossing point between the distribution of dark noise and the single photon
response, and thus appears to be near optimally poised to separate the two (Field and
Rieke, 2002). As diagrammed in Figure 2.5A (left) the distributions of rod noise (i.e. the
Gaussian fit to the peak at 0 pA in Figure 2.1D) and signal (i.e. the Gaussian fit to the
single photon response peak in Figure 2.1D) are plotted on a log scale with the single
photon response peak scaled to reflect their distributions near absolute visual threshold
(see also Field et al., 2005). A dashed vertical line representing the nonlinear threshold
demarcates the parts of each distribution that are discarded (to the left) and those that are
retained (to the right), and is positioned close to the crossover of the signal and noise
Gaussian distributions (1.3 times the mean single photon response amplitude; Field and
Rieke, 2002). Simulated rod bipolar cell distributions assuming this threshold position
(Figure 2.5B; see legend for details) have signal and noise peaks that are more discrete
compared to those of rods (Figure 2.1D) and are similar to those measured for in rod
bipolar cells (Figure 2.4B). Thus such a threshold can account quantitatively for the
improved SNR of the single photon response in rod bipolar cells.
38
6 4 2 0 -2
Dark Noise
Light Response
Threshold (T)
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Probability
2 0
WT GCAPs
-/-
A.
2 1 0 2 1 0 2 1 0
1.0
0.5
0.0
Probability
2 1 0
T = 1.3 T = 1.3 T = 1.6 T = 2.7
Amplitude (pA)
Normalized Amplitude
B.
Fig. 2.5. Model for the nonlinear threshold at the rod-to-rod bipolar synapse.
(A) The probability density of dark noise and single photon (light) response amplitude in
WT and GCAPs
-/-
rods are plotted for conditions near absolute visual threshold (~ 0.0005
R*). Standard deviations for Gaussian distributions were averaged across all rod
photoreceptors and have been normalized to the mean single photon response amplitude
for WT (σ
D
/
!
A = 0.36, σ
A
/
!
A= 0.31), and GCAPs
-/-
rods (σ
D
/
!
A = 0.19, σ
A
/
!
A= 0.26)
which were multiplied by 3.4-fold difference in the mean single photon response
amplitude compared to WT rods (see Figure 2.1). Dashed vertical lines indicates the
position of the nonlinear threshold, plotted at 1.3 for WT (Field and Rieke, 2002), and at
1.3, 1.6, and 2.7 for GCAPs
-/-
rods. (B) Rod bipolar response distributions were simulated
assuming the convergence of 20 rods per rod bipolar cell and flash strengths of 0.1 R*/rod
and 0.04 R*/rod for WT and GCAPs
-/-
distributions, respectively. We assumed a sharp
threshold, where responses exceeding the threshold were transmitted and those below the
threshold were scaled by 0.15, to reproduce the observed variance in dark noise of the rod
bipolar cells in Figure 2.4. Simulated histograms resulting from 10,000 repeated flashes
were fit as the sum of Gaussian distributions centered at the absorption of 0, 1, 2 …
photons (see Experimental Procedures), and the response amplitude axis was normalized
to the mean single photon response amplitude.
As described previously the responses of GCAPs
-/-
rod bipolar cells differ in two
ways from WT cells. First, the relationship between average response amplitude and
flash strength is more linear in GCAPs
-/-
rod bipolar cells (Figure 2.3B). This suggests
that signal transfer from rods to rod bipolar cells eliminates fewer single photon
39
responses in GCAPs
-/-
rod bipolar cells. Second, GCAPs
-/-
rod bipolar cells exhibit
greater current noise in darkness (Figure 2.3D). A framework where the nonlinear
threshold is relatively stationary with respect to the rod bipolar cell current can explain
both these observations. If the position of the nonlinear threshold remains relatively
fixed (i.e. at 1.3), then as the distributions of signal and noise expand (such as for
GCAPs
-/-
rods) the nonlinear threshold no longer optimally separates signal from noise.
Figure 2.5A (right) demonstrates that a greater part of the signal distribution in GCAPs
-/-
rods would lie to the right of the nonlinear threshold, but a greater part of the noise
distribution would lie to the right as well. Under these conditions the simulated rod
bipolar histogram shows greater blurring of the rod single photon response and noise,
similar to measured histograms from GCAPs
-/-
rod bipolar cells (Figure 2.4B, middle).
These simulations indicate that the position of the nonlinear threshold is critical
for the separation of the rod single photon response and dark noise. In GCAPs
-/-
mice if a
nonlinear threshold is assumed to be at a slightly higher value (i.e. at 1.6; Figure 2.5B,
right), then a simulated rod bipolar histogram with greater separation between noise and
the single photon response is derived. This separation can be increased further if a
nonlinear threshold positioned near the crossover point between rod signal and noise in
GCAPs
-/-
mice is assumed (i.e. at 2.7; Figure 2.5B, right). The prediction from these
simulations is that even a subtle increase in the position of the nonlinear threshold with
respect to the rod bipolar current can markedly increase the SNR of the single photon
response in rod bipolar cells.
40
2.3.7 Increased postsynaptic saturation increases SNR in GCAPs
-/-
rod bipolar cells
To test the influence of increasing the relative position of the nonlinear threshold
on rod bipolar cell SNR in GCAPs
-/-
mice, we treated these cells with the Group III
metabotropic glutamate receptor agonist, APB (Slaughter and Miller, 1981). Treatment
with a low concentration (0.2 µM) of APB lightly activates the mGluR6 cascade resulting
in the closure of some transduction channels (Nawy and Jahr, 1991), and increasing
synaptic saturation. By pushing the transduction channels toward a condition reflecting
zero open probability, the effective position of the nonlinear threshold can be increased
(Sampath and Rieke, 2004).
Figure 2.4A (bottom) demonstrates the results for a recording from a rod bipolar
cell treated with APB. The activation of the mGluR6 cascade by APB reduces the width
of the noise distribution in the GCAPs
-/-
histogram (compare Figure 2.4B, middle vs.
bottom), thereby making the distribution of single photon responses and noise more
discrete. The SNR in the presence of APB increased to 2.94 + 0.17 (n=7; Figure 2.5), an
~ 1.2-fold improvement over GCAPs
-/-
rod bipolar cells in the absence of APB. Thus in
GCAPs
-/-
mice the rod bipolar cell is not effective at eliminating rod noise, and a subtle
increase in postsynaptic saturation (or the nonlinear threshold) can improve the SNR of
the single photon response. In WT mice postsynaptic saturation at the rod-to-rod bipolar
synapse (Sampath and Rieke, 2004) appears well tuned for eliminating synaptic noise and
separating it from the single photon response, and thus in lowering behavioral threshold.
2.4 Discussion
The study of absolute visual threshold in humans has a rich history that has
highlighted the ability of our sensory systems to detect stimuli that approach the physical
41
limit (reviewed by Field et al., 2005). Early psychophysical work has shown that the
visual system is capable of detecting few photon absorptions among hundreds or
thousands of rods (Hecht et al., 1942; van der Velden, 1946), but the unambiguous
identification of biophysical mechanisms that control our perception remain elusive. The
general strategy in these experiments was to use targeted genetics to alter the properties
of the phototransduction cascade in mammalian rods and study the full range of how
retinal signal processing impacts behavioral threshold. The use of knockout mice
provides additional leverage in understanding how absolute visual threshold is set in
humans, as the cells and circuits involved are largely conserved (Bloomfield and
Dacheux, 2001; Dacheux and Raviola, 1986; Smith et al., 1986; Tsukamoto et al., 2001).
In particular GCAPs
-/-
mice allow us to consider the impact of dark noise (Baylor et al.,
1979; Chichilnisky and Rieke, 2005; Rieke and Baylor, 1996), and thus the signal-to-
noise ratio (SNR), in the processing of the single photon response in rods and rod bipolar
cells, and how it impacts behavioral threshold.
2.4.1 Influence of rod and rod bipolar SNR on behavior
Figure 2.6 shows how the SNR of the single photon response evolves as it passes
from rods to rod bipolar cells. In WT mice we observe an increase in the SNR of the
single photon response of ~ 1.5-fold as it passes from the rod outer segment to the rod
bipolar cell. This increase is attributable to the nonlinear threshold at the rod-to-rod
bipolar synapse, which is designed to eliminate continuous noise from the array of rod
photoreceptors under conditions where rods are rarely absorbing a photon (Baylor et al.,
1984; Field and Rieke, 2002; van Rossum and Smith, 1998). In GCAPs
-/-
rods the
elimination of Ca
2+
feedback on cGMP synthesis increases significantly the
42
discriminability of the single photon response. One may expect that the greater
separation of the single photon response from the underlying noise should decrease
detection threshold in these animals, but we found to the contrary that behavioral
threshold was elevated by ~ 1.5-fold. This increase in behavioral threshold instead
appears more qualitatively consistent with the reduction of the SNR of the single photon
2.5
2.0
1.5
3.5
3.0
2.5
2.0
Signal-to-Noise Ratio (SNR)
GCAPs
-/-
with APB
Rod Photoreceptors Rod Bipolar Cells
GCAPs
-/-
GCAPs
+/+
*
**
**
Fig. 2.6. Comparison of SNRs of rods and rod bipolar cells.
(A) SNRs as determined for rods and rod bipolar cells (see Experimental Procedures).
Individual cells are plotted as open symbols, with the mean and SEM included as filled
symbols. As shown in Figure 2.1 the SNR for WT rods was 1.69 + 0.15 (mean + SEM;
n=4) compared to GCAPs
-/-
rods where the SNR was 2.51 + 0.08 (mean + SEM; n = 6), p-
value = 0.0054. As shown in Figure 2.4 the SNR for WT rod bipolar cells was 2.91 +
0.11 (mean + SEM; n = 5) compared to GCAPs
-/-
rod bipolar cells where the SNR fell to
2.49 + 0.06 (mean + SEM; n = 16), p-value = 0.015. In GCAPs
-/-
rod bipolar cells treated
with 0.2 µM APB the SNR increased to 2.94 + 0.17 (mean + SEM; n = 7), p-value =
0.041 compared to GCAPs
-/-
. Variability in the SNR may result from variability in
individual cells, or from retinal slice to retinal slice. For GCAPs
-/-
rod bipolar recordings
with and without APB that were done in the same slice, we compared the SNR by plotting
a line between the mean values for both conditions. When more than one cell from each
condition were derived from the same retinal slice, the line extended to the average value
from those cells. Note the upward trend of every line to indicate an improvement in rod
bipolar cell SNR in the presence of APB. * indicates Student’s t-test p-value < 0.01, and
** indicates Student’ t-test p-value < 0.05.
43
response in the rod bipolar cells of ~ 1.2-fold, rather than the improvement of ~ 1.5-fold
seen for rods. This observation goes against the hypothesis that the SNR of sensory
receptors solely can explain the limits of behavior.
The role of GCAPs in the control of light-driven signals in the retina has largely
been relegated to phototransduction, where it speeds the temporal properties of the light
response and extends the operating range of the rod by preventing response saturation
(Mendez et al., 2001). The fast and powerful Ca
2+
feedback onto cGMP synthesis also
minimizes fluctuations in cGMP that create the ‘continuous’ noise (Rieke and Baylor,
1996; Burns et al., 2002). The magnitude of continuous noise appears well matched to
the position of the nonlinear threshold in the rod bipolar cell dendrites (Figure 2.5A;
Field and Rieke, 2002; van Rossum and Smith, 1998). Thus GCAPs places the
continuous noise into a range where it can be eliminated effectively. This elimination of
noise at the level of rod bipolar cells appears well tuned to the rod SNR (c.f. Figure 2.4A,
top), and thus appears to be critical in setting behavioral sensitivity.
2.4.2 Can the nonlinear threshold adapt to accommodate changing SNR?
A common feature of sensory systems is their ability to alter their properties
based on the stimulus history. Such adaptation is critical for maintaining sensitivity over
a wide range of stimulus strengths, thus maximizing the information captured. Near
visual threshold, given the sparse density of photons in the array of rod photoreceptors,
the retina must be maximally sensitive to absorbed photons (reviewed by Field et al.,
2005). A consequence of this arrangement is that the system is also maximally sensitive
to the intrinsic noise. The nonlinear threshold imposed by the mGluR6 signaling cascade
in the rod bipolar cell dendrites is critical for improving the SNR of the rod
44
photoresponse near visual threshold, but it must balance the elimination of noise in
phototransduction and signal transmission with the loss of some single photon responses
(Berntson et al., 2004; Field and Rieke, 2002; van Rossum and Smith, 1998). Thus near
visual threshold the precise positioning of the nonlinear threshold with respect to the rod
signal and noise is critical. This raises the question; near visual threshold can the position
of the nonlinear threshold adjust dynamically to the distributions of rod signal and noise?
Evidence exists that the extent of nonlinearity in rod-to-rod bipolar cell signal
transfer can be modulated. Application of APB, LY341495, and background light
change the extent of postsynaptic saturation and thus nonlinearity measured in flash
families, and in the presence of LY341495 or background light the system can be
linearized (Sampath and Rieke, 2004). However, these experiments don’t necessarily
address the adaptive properties of the nonlinear threshold, but may instead reflect the
movement of the mGluR6 transduction current with respect to a fixed threshold. For
weak background light previous work has suggested the optimal positioning of the
nonlinear threshold could provide some improvement of rod SNR (Field and Rieke,
2002). But, near visual threshold how far can the position of the nonlinear threshold
itself adapt?
The present work suggests the lack of an adaptive mechanism that would optimize
the position of the nonlinear threshold between rods and rod bipolar cells, since the
position of the nonlinear threshold appears relatively stationary between WT and GCAPs
-
/-
rod bipolar cells (Figure 2.5B). Experiments with a low concentration of APB show
that even the subtle activation of the mGluR6 cascade can improve the SNR of the single
photon response (Figure 2.4B, bottom). This would suggest that the ability to manipulate
45
the position of the nonlinear threshold might be valuable in optimizing the output of the
system. A requirement for such a mechanism would be the ability to sense the
distributions of rod signal and noise near visual threshold, and provide feedback to the
mGluR6 signaling cascade. For instance, the degree of saturation within the signaling
cascade may be manipulated postsynaptically by altering the stochiometry of the limiting
mGluR6 signaling component. If such a mechanism were positioned optimally in
GCAPs
-/-
mice the prediction would be that they would have a higher sensitivity for rod
vision, as the nonlinear threshold would be acting on rods that already have a higher SNR
compared to WT. Thus it appears the extent of saturation within the rod bipolar signaling
cascade isn’t controlled dynamically, but instead may be evolutionarily ‘hardwired’.
46
Chapter 3
Cooperative control of sensitivity by two splice
variants of Gα
o
in retinal On-bipolar cells
3.1 Introduction
G-proteins are essential signaling molecules that connect hundreds of G-protein-
coupled receptors with a relatively limited number of downstream effectors (reviewed by
Wettschureck and Offermanns, 2005). They form heterotrimers consisting of α, β and γ
subunits, and GDP to GTP exchange on the α subunit leads to the dissociation of the α
subunit from the βγ complex, both of which then can activate specific downstream
effectors. Each subunit has several isoforms originating from different genes. In general,
different isoforms of α subunits engage in different signaling cascades due to the
coupling specificity with receptors or effectors. In contrast, β and γ subunits are relatively
promiscuous, and the function of one isoform can often be substituted by other isoforms.
In addition to the segregation of different signaling cascades by various isoforms
of α subunits, several α subunits are known to have multiple splice variants, which could
serve as another mechanism for differentiating signaling cascades. In fact, it has been
shown that two major types of splice variants for the G
o
α subunit, Gα
o1
and Gα
o2
, couple
with different receptors in rat pituitary GH3 cells, suggesting that they play an important
role in the regulation of hormone secretion in response to different upstream signals
(Kleuss et al., 1991; Chen and Clarke, 1996; Degtiar et al., 1997). However, there is also
evidence for redundancy in the cascade selection by multiple splice variants (Mattera et
47
al., 1989; Schmidt et al., 1991). One possible role of these redundant splice variants may
be to adjust cooperatively the stoichiometry of signaling cascade to optimize the shared
function.
Retinal On-bipolar cells receive glutamatergic inputs from photoreceptors, and the
transduction of this signal is generated through the mGluR6 receptor and the G-protein
G
o
to close nonselective cation channels (Masu et al., 1995; Dhingra et al., 2000). The
remaining components of the signaling cascade including the transduction channel
remain largely elusive. The expression of G
o
in the retina is restricted to On-bipolar cells
(Vardi 1993, Vardi 1998), and two splice variants of the G
o
α subunit (Gα
o1
and Gα
o2
) are
found in mouse On-bipolar cells (Dhingra et al, 2002). However, the expression of Gα
o1
dominates over that of Gα
o2
, and electroretinography (ERG) from knockout mice for each
splice variant has demonstrated that On-bipolar responses appear to have require Gα
o1
but
not Gα
o2
(Dhingra et al., 2002). To investigate the role of Gα
o1
and Gα
o2
, as well as other
putative G-proteins, in the mGluR6 signaling cascade we chose to study a subtype of On-
bipolar cells, rod bipolar cells. Rod bipolar cells receive input exclusively from rod
photoreceptors, which are capable of informing the absorption of single photons (Field
and Rieke, 2002). These cells are pooled together in a specialized circuitry know as the
Rod Bipolar Pathway (reviewed by Field et al., 2005), thought to underlie our most
sensitive vision (Dacheaux and Raviola, 1986; Smith et al., 1986). The optimal detection
of single photons requires rod bipolar cells to process rod inputs with a thresholding
nonlinearity (Field and Rieke, 2002), which is achieved by postsynaptic saturation
somewhere downstream of mGluR6 (Sampath and Rieke, 2004). Thus, a delicate
balancing of the stoichiometry of the mGluR6 - G
o
signaling cascade may be crucial to
48
set the level of saturation at the optimal point (reviewed by Field et al., 2005; Okawa and
Sampath, 2007).
Here, contrary to the previous study by Dhingra et. al. (2002), we found that the
On-response in rod bipolar cells is partially mediated by Gα
o2
, Gα
o2
,
is also controlled by
mGluR6, and thus Gα
o1
and Gα
o2
play a redundant role. Although Gα
o2
-mediated
responses were much smaller than normal rod bipolar responses, Gα
o2
-/-
rod bipolar cells
exhibited reduced light sensitivity. This was not attributable to the reduction in the total
level of Gα
o
protein since a 50% reduction in the total amount of Gα
o
using heterozygous
mice for Gα
o
did not alter the light sensitivity. We propose that two splice variants of G
o
α act in a cooperative manner to fine-tune the stoichiometry of the signaling cascade, and
improve the light sensitivity of rod bipolar cells. This cooperative action by multiple
splice variants in the same signaling cascade may be a generalized mechanism by which
cells adjust the stoichiometry of G-protein signaling cascades to optimize their function.
3.2 Materials & Methods
3.1.1 Mice and Preparation
Gα
o
-/-
, Gα
o1
-/-
and Gα
o1
-/-
Cx36
-/-
mice rarely survived more than 4 weeks. Since
their retinas mature by 3 weeks based on retinal morphology and ERG (Dhingra et al.,
2000), we used these mice at the age of 3 – 4 weeks. The WT, Cx36
-/-
, Gα
o
+/-
and Gα
o2
-/-
mice were used at the age between 6 weeks to 3 months. The preparation of retinal slices
was performed under infrared illumination as described previously (Sampath et al.,
2005). Briefly, mice were dark-adapted overnight and sacrificed according to protocols
approved by the Institutional Animal Care and Use Committee of the University of
Southern California. Eyes were enucleated, and the corneas and the lenses were removed.
49
Then, the retinas were isolated and kept in Ames’ buffer equilibrated with 5% CO
2
/95%
O
2
at 32 °C. A piece of retina was cut and embedded in agar. Retinal slices were cut with
a vibratome, transfered into a recording chamber inside a dark Faraday cage, and
superfused with Ames’ buffer at 35-37 °C.
3.2.2 Electrophysiology and Light Stimulation
Light-evoked currents in rod bipolar cells and AIIACs were recorded by whole-
cell patch-clamp techniques, holding the voltage at -60 mV. Intracellular solution for
bipolar cells consisted of (in mM): 125 K-Aspartate, 10 KCl, 10 HEPES, 5 NMG-
HEDTA, 0.5 CaCl
2
, 1 ATP-Mg, 0.2 GTP-Mg; pH was adjusted to 7.2 with NMG-OH.
ATP-Mg and GTP-Mg were excluded for the experiments in Figure 3.1. Intracellular
solution for AIIACs consisted of (in mM): 110 Cs-Methansulfate, 20 TEA-Cl, 10
HEPES, 10 EGTA, 2 QX-314, 1 ATP-Mg, 0.2 GTP-Mg; pH was adjusted to 7.2 with Cs-
OH. Cell types were generally identified by the location of cell somas within the inner
plexiform layer and their response properties. However, when these criteria were hard to
apply such as cells in Gα
o
-/-
and Gα
o1
-/-
mice, the cell types were confirmed by the
morphology, visualized with 100-200 µM Alexa 750 (Invitrogen) added in the internal
solution. Full-field 10 ms flashes were delivered from blue LED (λ
max
~ 470 nm, FWHM
~ 30 nm) and focused onto the retinal slice with 20X 0.75NA objective (Nikon), Light-
evoked currents were low-pass filtered at 300 Hz with an 8-pole Bessel filter and
digitized at 1 kHz. Light intensity was calibrated as an effective photon flux at the peak
of spectral sensitivity for mouse rhodopsin (λ
max
~ 501 nm) by convolving the power-
scaled LED spectrum with the normalized spectral sensitivity curve for mouse rhodopsin.
The number of rhodopsin photoisomerization per rod for the flash was calculated by
50
multiplying the effective photon flux with the collecting area of mouse rod for our setup,
which was estimated as 0.18 µm
2
(Cao et al., 2008, Okawa et al., 2008).
3.2.3 Western blotting
Retinas from WT and Gα
o
+/-
mice where homogenized in lysis buffer containing
Protease inhibitor (Roche), 50 mM Tris-HCl (pH 8), 150 mM NaCl, 1 % NP-40, 0.5 %
sodium deoxycholate and 0.1 % SDS. The homogenate was treated with 100 units/ml
DNase for 30 min at room temperature. The protein concentration was checked using a
BCA Protein Quantification Assay (Thermo Fisher Scientific). The extracted protein was
run on a 10% NuPAGE™ Bis-Tris gel (Invitrogen) and transferred to a nitrocellulose
membrane using a Transphor Electrophoresis Unit (Hoefer). The membrane was blocked
in 10% milk in TBST for 1 h at room temperature, and incubated in Gα
o
rabbit
polyclonal antibody (Santa Cruz Biotechnology, Inc.) in TBST (1:200) at 4 °C overnight.
The membrane was washed with TBST and incubated with IRDye 800 CW anti-rabbit
antibody (LI-COR) in TBST (1:20,000) for 1h at room temperature, then washed with
TBST. The positive bands were detected with an Odyssey® Infrared Image System (LI-
COR).
3.3 Results
3.3.1 The default state of rod bipolar transduction channels is closed
To confirm that the transduction cascade in On-bipolar cells acts through G
o
to
close channels, we dialyzed rod bipolar cells with 30 µM GTP-γ-S, a nonhydrolyzable
GTP analogue, to constitutively activate G-proteins. As previously found by Sampath and
Rieke (2004), this eliminated the saturating responses within 40 seconds on average
without changing the holding current (Figure 3.1A and 3.1B), much faster than the
51
typical response washout seen with control patch-pipette solution, which took about 3
minutes on average (Figure 3.1E). This is in a good agreement with the current view of
the On-bipolar cell transduction cascade, in which mGluR6 activation leads to the closure
of transduction channels via the activation of G
o
(Nawy, 1999; Dhingra et al., 2000).
Next, to further confirm the action of G
o
, we dialyzed rod bipolar cells with 500
µM GDP-β-S, a nonhydrolyzable GDP analogue, to constitutively inactivate G-proteins.
According to the current scheme of On-bipolar transduction cascade, this would open the
transduction channels. However, GDP-β-S did not open the channels as shown in the
quick decay of saturating response amplitude from 40 sec to 60 sec without a change in
the holding current (Figure 3.1C and 3.1D). Therefore, the default state of transduction
channels without G-protein activity seems to be closed. The longer effect of GDP-β-S
compared to GTP-γ-S is probably because all the GTP bound G
o
α subunits must be
turned over to bind GDP-β-S to be inactivated. Since Gα
o
acts toward closing
transduction channels, another pathway must be required for opening channels. This
other pathway is also likely to be GDP-β-S-sensitive, possibly a second G-protein,
otherwise the GDP-β-S treatment would have opened transduction channels if the
channel-opening pathway remained intact. If this is true, the result of GTP-γ-S
experiment can be explained if the closing action by constitutively active G
o
cascade won
over the opening action by constitutively active second G-protein cascade. The
alternative explanation is that the βγ complex released from the α subunit acts toward
opening transduction channels. Then, this action was inhibited by the inactive α subunits
bound to GDP-β-S, and consequently to the βγ complex. Similarly to the former
52
explanation, the result of GTP-γ-S experiment would be reasonable if constitutively
active Gα
o
cascade won over the released βγ complex.
Normalized Amplitude
sec
Control
GTP-!-S
GDP-"-S
pA
sec
pre
GDP-"-S
post
GDP-"-S
pA
sec
I
dark
I
min
pA
sec
I
dark
I
min
pA
sec
pre
GTP-!-S
post
GTP-!-S
0 0
0 0
-200 -200
-400 -400
A. B.
C. D.
0
1
E.
0 100
100 0
-0.2
-0.2 0.6
0.6
0 100 200
Fig. 3.1. The default state of transduction channels in rod bipolar cells is closed
(A and C) Following the establishment of whole-cell recordings with patch pipettes
containing 30 µM GTP-γ-S (A) or 500 µM GDP-β-S (C), the amplitude of the saturating
flash response (open circle) and the dark current (filled circle) were monitored. Saturating
flash strengths producing 15 Rh*/rod in (A) and 19 Rh*/rod in (C) were delivered every
1.2 second. (B and D) Saturating flash responses before and after the decay are shown.
Response times are indicated by arrows in (A) and (C). (E) Each recording was
normalized to the amplitude of the first saturating response and averaged across cells for
GTP-γ-S (n=5), GDP-β-S (n=6) and control (n=5). Error bars show mean + SEM.
53
On-bipolar transduction channels were previously thought to be gated by cGMP
in a similar manner as cGMP-gated channels in the photoreceptor cells (Nawy and Jahr,
1990; Nawy and Jahr, 1991), but later studies refuted this interpretation (Nawy, 1999).
Presently the role of cGMP in On-bipolar transduction cascade is known to increase the
gain of On-response rather than mediating it (Snellman and Nawy, 2004; Shiells and
Falk, 2002). We tested the influence of 1mM cGMP for rod bipolar response while
eliminating all the G-protein activity with 500 µM GDP-β-S, and the results were not
significantly different from the dialysis of GDP-β-S alone (data not shown). Thus cGMP
appears not to mediate light responses through the On-bipolar transduction cascade.
3.3.2 Gα
o1
-/-
rod bipolar cells have weak On-response mediated by Gα
o2
The lack of increase in holding current (i.e. run open channels) in the absence of
G-protein activity following the dialysis of GDP-β-S suggests G-protein activity is
required also to open channels. To test if a second G-protein or the βγ complex is
opening transduction channels, we used knockout mice for G
o
. Since it was reported that
Gα
o1
but not Gα
o2
, mediates the On-bipolar response (Dhingra et al., 2002), we first
studied Gα
o1
-/-
mice. If the channel-opening pathway is constitutively active regardless of
mGluR6 activity, a large initial holding current is expected unless the cell is heavily
adapted. If mGluR6 is bistable, and the channel-opening pathway becomes active only
when mGluR6 is unbound to glutamate, the light response is expected to remain.
However, this possibility is unlikely because ERG from Gα
o1
-/-
mice lacked the b-wave
caused by On-bipolar cells (Dhingra et al., 2002).
54
Figure 3.2B shows a representative recording from a Gα
o1
-/-
rod bipolar cell
visualized in Figure 3.2A. Surprisingly, the On-response persisted in the absence of Gα
o1
.
The remaining On-response decayed quickly after establishing whole-cell configuration
Fig. 3.2. Rod bipolar response is partially mediated by Gα
o2
(A and B) A representative Gα
o1
-/-
rod bipolar cell visualized with Alexa 750 in (A)
showed a small flash response immediately after initiating the recoding (0 sec), which
decayed in 15 sec. Flash strength was 15 Rh*/rod, a strength that saturates WT rod bipolar
cells. (C and D) A representative Gα
o
-/-
rod bipolar cell visualized with Alexa 750 in (C)
did not have any response to flashes producing 28 Rh*/rod. (E) A Gα
o
-/-
Off-bipolar cell
located near the rod bipolar cell in the same slice shown in (C) was visualized with Alexa
750 in (E), and showed normal response family. Flash strengths were doubled between
0.44-28 Rh*/rod.
55
(Figure 3.2B) and was much smaller than usual WT rod bipolar responses, which
routinely exceed several hundred pA. The largest response among all the recorded Gα
o1
-/-
rod bipolar cells was ~ 10 pA. Thus, the ERG seems to have failed to detect this small
remaining On-response (see Dhingra et al., 2002). The holding current was not
significantly different from WT rod bipolar cells, and did not change as the response
washed out (data not shown). Thus, the assumed channel-opening pathway could become
active when mGluR6 was unbound to glutamate, but the response appears too small to
support this interpretation. Alternatively, Gα
o2
, the other splice variant of G
o
α subunit,
might mediate this response.
To test this possibility, we recorded rod bipolar responses from knockout mice for
both Gα
o1
and Gα
o2
(Gα
o
-/-
). As shown in a representative rod bipolar cell in Figure 3.2C
and 3.2D, the On-response was completely lost from all the Gα
o
-/-
On-bipolar cells tested,
while neighboring Off-bipolar cells demonstrated a response in the same retinal slices
(Figure 3.2E and 3.2F). Thus, the remaining On-response in Gα
o1
-/-
rod bipolar cells is
most likely mediated by Gα
o2
. Similarly to Gα
o1
-/-
rod bipolar cells, the holding currents
of Gα
o
-/-
rod bipolar cells were normal without any significant change during recordings
(data not shown). Although the G
o
pathway that closes channels was completely
eliminated by the knockout of both Gα
o1
and Gα
o2
, the mechanism assumed to open
channels remains unclear.
3.3.3 The sensitivity of Gα
o2
-mediated On-response
The Gα
o2
-mediated On-response in Gα
o1
-/-
rod bipolar cells was too small and
decayed too quickly to be characterized. We circumvented this problem by recording the
readout of an array of rod bipolar inputs from the postsynaptic AII amacrine cells
56
(AIIAC). AIIACs are expected to have larger responses than rod bipolar cells due to the
convergent inputs from several rod bipolar cells (Tsukamoto et al., 2000). Furthermore,
the AIIAC response does not decay because it is mediated by ionotropic glutamate
receptors. To isolate only feed-forward rod bipolar inputs, we eliminated feedback inputs
through electrical synapses between AIIACs, and also between AIIACs and cone On-
bipolar cells by crossing Gα
o1
lines with Cx36
-/-
mice (Deans et al., 2002).
pA
sec
before
APB
wash
pA
sec
Normalized amplitude
Rh*/rod
pA
sec
pA
sec
0
-400
-0.2 0.6
0
-150
-0.2 0.6
1
0.1
0.1 10 1
0
-150
-0.2 0.6
A. B.
D. E.
C.
0
-200
0 300
APB
Cx36-/-
G!
"1
-/-
Cx36-/-
Cx36-/-
G!
"1
-/-
Cx36-/-
Fig. 3.3. The property of the Gα
o2
-mediated response measured in Gα
o1
-/-
Cx36
-/-
AIIACs
(A and B) Flash response families were recorded in a Cx36
-/-
(Gα
o1
+/+
Cx36
-/-
littermate)
AIIAC (A) and a Gα
o1
-/-
Cx36
-/-
AIIAC (B). Flash strengths were roughly doubled between
0.04-3.8 Rh*/rod in (A) and between 0.21-26.5 Rh*/rod in (B). Arrows indicate the lack of
initial spikes in saturating flash responses in the Gα
o1
-/-
Cx36
-/-
AIIAC. (C) Normalized
response amplitudes from individual families were averaged across cells for Cx36
-/-
AIIACs
(n=10) and Gα
o1
-/-
Cx36
-/-
AIIACs (n=9) and plotted as a function of flash intensities. Half
saturating flash intensities estimated from the Hill curve fits were 0.17+0.01 vs. 2.56+0.13
Rh*/rod (mean+SEM) for Cx36
-/-
vs. Gα
o1
-/-
Cx36
-/-
AIIACs. (D and E) The amplitude of the
saturating flash response (26.5 Rh*/rod) in a Gα
o1
-/-
Cx36
-/-
AIIAC was monitored while 8µM
APB was applied for the duration indicated by the bar in (E). Responses measured before,
during and after the washout of APB are shown in (D). Response times are indicated by
arrows in (E).
57
Figure 3.3A and 3.3B shows response families from Gα
o1
+/+
Cx36
-/-
and Gα
o1
-/-
Cx36
-/-
AIIACs. The maximum response amplitude among all the Gα
o1
-/-
Cx36
-/-
AIIACs
tested was 200 pA, much larger than Gα
o1
-/-
rod bipolar cells, suggesting that the Gα
o2
-
mediated rod bipolar response may not be too small for the downstream processing. In
Figure 3.3C, we plot the normalized response amplitude versus the flash strength. Gα
o1
-/-
Cx36
-/-
AIIACs were almost 10 times less sensitive than Gα
o1
+/+
Cx36
-/-
AIIACs. This
implicates that the rod bipolar response mediated by Gα
o2
is 10 times less sensitive than
that mediated by Gα
o1
, assuming that sensitivity transforms from rod bipolar cells to
AIIACs in the same ratio for two genotypes. However, this assumption may be wrong if
the saturating response amplitudes of rod bipolar cells are different between the two
genotypes. In dark-adapted WT mice, AIIACs are almost 10 times more sensitive than
rod bipolar cells (Dunn et al., 2006). Therefore, the rod bipolar – AIIAC synapse
saturates at light levels much lower than where the rod bipolar response saturates, thus
coding only the most sensitive part of the dynamic range of the rod bipolar response.
Since the Gα
o2
-mediated rod bipolar response was much smaller than the WT rod bipolar
response, it is probable that Gα
o2
-mediated rod bipolar responses did not saturate rod
bipolar – AIIAC synapses, resulting in the seemingly low AIIAC sensitivity. Indeed,
Gα
o1
-/-
Cx36
-/-
AIIACs did not show the initial spikes in saturating flash responses, which
is characteristic for WT AIIAC, but the initial spike was observed in Gα
o1
+/+
Cx36
-/-
AIIACs (Figure 3.3A). In addition, the sensitivity of Gα
o1
-/-
Cx36
-/-
AIIACs was
relatively close to that of Cx36
-/-
rod bipolar cells (half-saturating flash intensity 2.56 +
0.13 Rh*/rod for Gα
o1
-/-
Cx36
-/-
AIIACs vs. 3.43 + 0.24 Rh*/rod for Cx36
-/-
rod bipolar
cells, data not shown). Thus, it is conceivable that Gα
o2
-mediated rod bipolar responses
58
had similar sensitivity, but less efficiency, compared to Gα
o1
-mediated rod bipolar
response, and that rod bipolar cells and AIIACs in Gα
o1
-/-
Cx36
-/-
mice had similar
sensitivity simply because Gα
o2
-mediated rod bipolar responses did not saturate rod
bipolar – AIIAC synapses.
3.3.4 mGluR6 controls both Gα
o1
and Gα
o2
To determine if Gα
o2
-mediated rod bipolar responses are controlled by mGluR6
activity, we studied the light response of Gα
o1
-/-
Cx36
-/-
AIIACs during the application of
the mGluR6 agonist APB (8 µM). APB completely suppressed the response in Gα
o1
-/-
Cx36
-/-
AIIACs, and the response recovered after its washout (Figure 3.3D and 3.3E).
Since the other types of mGluRs expressed by rods and rod bipolar cells are not sensitive
to this concentration of APB, and AIIAC response is mediated by ionotropic glutamate
receptors, the effect of APB on the AIIAC response was specific to mGluR6 in rod
bipolar cells. Thus, both the Gα
o1
and the Gα
o2
pathway are controlled by mGluR6 on the
rod bipolar dendrites.
3.3.5 Reduced gain and sensitivity in Gα
o2
-/-
rod bipolar cells
Gα
o1
and Gα
o2
thus appear to work in a redundant manner, and the Gα
o2
-mediated
response is much smaller than the Gα
o1
-mediated response, which casts doubt on its
functional role. To investigate the role of Gα
o2
, we recorded flash response families from
Gα
o2
-/-
rod bipolar cells (Figure 3.4A). The overall response kinetics and response
amplitude looked similar to rod bipolar cells recorded from WT littermates (Figure 3.4A).
Dim response kinetics was also similar to WT rod bipolar cells (Figure 3.4B). However,
the loss of Gα
o2
caused a reduction in the gain of the Gα
o2
-/-
dim flash responses (Figure
59
pA pA
R/R
max
Rh*/rod
WT
G
o
2! -/-
sec
WT
G!
"2
-/-
A.
B.
C. 0
-150
0
-200
10 %
-0.2 0.6
1
0.1
1 10
WT
G!
"2
-/-
Fig. 3.4. Gα
o2
-/-
rod bipolar cells exhibited reduced gain and light sensitivity
(A) Responses to a family of flashes producing 0.29-18.8 Rh*/rod were recorded in a WT
(Gα
o2
+/+ littermate) rod bipolar cell and a Gα
o2
-/-
rod bipolar cell. (B) Normalized rod
bipolar response to flashes producing 1 Rh*/rod was estimated by averaging normalized
responses to dim flashes causing 5-25 % of saturating responses and dividing it by the
average dim flash strength. WT response is the average of 332 dim flash responses across
15 cells from 8 mice and Gα
o2
-/-
response is the average of 321 dim flash responses across
16 cells from 6 mice. (C) Normalized response amplitudes from individual families were
averaged across cells for WT rod bipolar cells (n=15) and Gα
o2
-/-
rod bipolar cells (n=16)
and plotted as a function of flash intensities. Half saturating flash intensities estimated
from the Hill curve fits were 2.18 + 0.16 vs. 2.63 + 0.20 Rh*/rod and the Hill exponents
were 1.51 + 0.04 vs. 1.55 + 0.06 (mean + SEM, WT vs. Gα
o2
-/- rod bipolar cells).
3.4B), which led to an overall reduction of light sensitivity of rod bipolar cells as seen in
the plot of normalized response amplitude versus flash intensity (Figure 3.4C).
3.3.6 Reducing the total amount of Gα
o
by half does not alter the property of rod
bipolar response
Reduced sensitivity in Gα
o2
-/-
rod bipolar cells may be simply due to the decrease
in the total amount of Gα
o
protein rather than the specific role played by Gα
o2
. To test this
60
possibility, we recorded rod bipolar response from heterozygous mice for Gα
o
. As shown
in Figure 3.5C, Gα
o
+/-
retinae had reduced Gα
o
expression by ~50% compared to WT.
This reduction will be larger than the loss of Gα
o2
in Gα
o2
-/-
rod bipolar cells since Gα
o1
expression dominates over Gα
o2
expression in WT On-bipolar cells, including rod bipolar
Fig. 3.5. Reduced Gα
o
expression in Gα
o
+/-
retina does not alter rod bipolar response
(A) Responses to a family of flashes producing 0.59-18.8 Rh*/rod were recorded in a WT
(Gα
o
+/+
littermate) rod bipolar cell and a Gα
o
+/-
rod bipolar cell. (B) Normalized rod
bipolar response to flashes producing 1 Rh*/rod was estimated by averaging normalized
responses to dim flashes causing 5-25 % of saturating responses and dividing it by the
average dim flash strength. WT response is the average of 437 dim flash responses across
14 cells from 3 mice and Gα
o
+/-
response is the average of 271 dim flash responses across
15 cells from 3 mice. (C) The total amount of Gα
o
proteins in WT and Gα
o
+/-
retinas were
compared by Western blot analysis. The protein level of Gα
o
+/-
retina was normalized to
that of WT retina for every pair of WT and Gα
o
+/-
retinas used in one experiment, and the
result of three experiments is shown in the bar graph. The error bar shows SEM. (D)
Normalized response amplitudes from individual families were averaged across cells for
WT rod bipolar cells (n=14) and Gα
o
+/- rod bipolar cells (n=15) and plotted as a function
of flash intensities. Half saturating flash intensities estimated from the Hill curve fits were
2.53+0.13 vs. 2.52+0.17 Rh*/rod and the Hill exponents were 1.54+0.03 vs. 1.62+0.06
(mean+SEM, WT vs. Gα
o
+/- rod bipolar cells).
61
cells (Dhingra et al., 2002). However, the overall response kinetics and the sensitivity of
Gα
o
+/-
rod bipolar cells remained similar to WT (Figure 3.5A and 3.5D). Dim response
kinetics was also similar to WT rod bipolar cells (Figure 3.5B). Thus, reduced sensitivity
in Gα
o2
-/-
rod bipolar cells must be due to a specific role of Gα
o2
, and not to a reduction in
the overall protein level. This result demonstrates that the two splice variants of Gα
o
work
in a redundant, but cooperative, manner to improve the light sensitivity of rod bipolar
cells.
Rod bipolar cells process rod inputs with a thresholding nonlinearity to optimize
photon detection near absolute visual threshold (Field and Rieke, 2002), which is caused
by the saturation of the transduction cascade somewhere downstream of mGluR6
(Sampath and Rieke, 2004). The extent of saturation is reflected in a nonlinear
relationship between flash strength and response amplitude, which is represented by the
slope of Hill curve fit to the plot of this relationship (Sampath and Rieke, 2004). If the
nonlinearity is caused by the saturation of Gα
o
, we would expect a change in the Hill
slope in Gα
o
+/-
rod bipolar cells. However, the nonlinearity observed in these cells was
similar to WT as seen in the similar slopes of the two Hill curves in Figure 3.5D. Thus,
the nonlinear threshold is likely to occur somewhere downstream of Gα
o
activation.
3.4 Discussion
We have investigated the role of G-proteins in the transduction cascade
connecting the mGluR6 receptor to the nonselective cation channels in retinal On-bipolar
cells. The main conclusions are i) the default state of the nonselective cation channel in
the G-protein transduction cascade of rod bipolar cells is closed, ii) the coordinated action
of two splice variants of Gα
o
(Gα
o1
and Gα
o2
) are needed for control of the sensitivity in
62
rod bipolar cells, and iii) the nonlinear threshold due to saturation of the transduction
cascade in rod bipolar cells resides somewhere downstream of Gα
o
activation.
3.4.1 The default state of the transduction channel in the G-protein cascade is closed
Glutamate released by photoreceptors onto retinal On-bipolar cells are sensed by
the metabotropic mGluR6 receptor near the tip of the dendrites (Nawy and Jahr, 1990;
Shiells and Falk, 1990; Vardi et al., 2000). This G-protein-coupled receptor mediates its
signal via poorly understood G-protein cascade where mGluR6 activates a G-protein
(G
o
), which by unknown mechanisms closes nonselective cation channels that are also
unknown. Gα
o
is the most abundant form of G-protein in the brain (Dolphin, 1998), but
surprisingly little is known about signaling downstream of Gα
o.
In On-bipolar cells the
light activated decrease of glutamate from photoreceptors decreases the amount of active
Gα
o
,
which opens nonselective cation channels and depolarizes the cell (Nawy, 1999;
Dhingra et al., 2000). Much to our surprise the abolishment of Gα
o
activity by GDP-β-S
in the rod bipolar cells did not render the transduction channels open, but they remained
closed, as seen by the lack of change in the holding current (see Figure 3.1C and 3.1D).
Thus, the default state of the nonselective cation channels is closed, and a different GTP
dependent mechanism is critical for channel opening. This implies that another G-protein,
other than G
o
could mediate the opening of the transduction cascade in rod bipolar cells.
Moreover, the effect of the G-proteins is not only mediated through the α subunit, but
Gβγ also regulates effectors such as muscarinic K
+
channels (GIRKs) and voltage
dependent Ca2
+
channels (Wettschureck and Offermanns, 2005). If the default state of
the channel is closed and the effect is GTP mediated, then one possible candidate could
even be Gβγ. The closed default state of the transduction channels can be critical for the
63
need to minimize noise and a competing mechanism to open these channels allows more
control for signal transduction.
3.4.2 Gα
o1
and Gα
o2
work in cooperation to set the sensitivity in rod-bipolar cells
The α subunit of G
o
is expressed as two splice variants (Gα
o1
and Gα
o2
) that
differ by 26 amino acids in the GTPase domain near the C terminal end (Hsu et al., 1990;
Strathmann et al., 1990; Tsukamoto et al., 1991; Horn and Latchman, 1993), which links
the α
o
subunit to its effector (for review see Clapham and Neer, 1997). Gα
o
mediates
signals from the receptors to effectors and ion channels, and the different splice variants
have usually been assigned with different or redundant functions. For instance, in the rat
pituitary GH
3
cells Gα
o1
and Gα
o2
mediate Ca
2+
channel inhibition through muscarinic
and somastotatin receptors, respectively (Kleuss et al., 1991; Kleuss et al., 1993). Both
subunits are expressed in On-bipolar cells, and it was previously shown that the specific
splice variant Gα
o1
, but not Gα
o2
was needed for the light response (Dhingra et al., 2002).
Our results, however, shows that the rod bipolar cells in Gα
o1
-/-
mice have a small
residual light response, but this response is gone in total Gα
o
-/-
mice. Moreover, the gain
in mice lacking Gα
o2
is smaller than that in WT (see Figure 3.4B). Thus, Gα
o2
must
mediate the small residual response seen in rod bipolar cells of Gα
o1
-/-
mice, and although
Gα
o2
by itself is 10-fold less sensitive than WT at the level of AII amacrine cells Gα
o2
generates significant downstream activity (see Figure 3.3B). The function of Gα
o2
in the
rod bipolar cells could thus be to fine-tune the sensitivity for dim light responses. Both
splice variants are needed for setting the sensitivity, but what is the relative contribution
of Gα
o1
and Gα
o2
to the sensitivity of rod bipolar cells? If Gα
o1
and Gα
o2
cooperatively
adjust the stoichiometry of signaling cascade, or if the efficiency of the Gα
o2
pathway is
64
less than that for Gα
o1
, this could relieve the channels from saturation by a small fraction
and thus render the cells more sensitive to small decreases in glutamate release from the
photoreceptors. The relative contribution of Gα
o1
and Gα
o2
could then
determine the
optimal trade-off between noise and sensitivity. To our knowledge this is the first time it
has been shown that two splice variants of the α subunit of a G-protein work in
cooperation acting on the same function.
3.4.3 Nonlinear thresholding of rod signals is mediated downstream of Gα
o
activation
The responses from rod photoreceptors, capable of informing the absorption of
single photons, are pooled in a specialized circuitry known as the Rod Bipolar Pathway,
thought to underlie our most sensitive vision (Dacheaux and Raviola, 1986; Smith et al.,
1986). The optimization of signal transfer in this pathway requires a thresholding
nonlinearity in rod bipolar cells that separates the single photon response from noise
(Field and Rieke, 2002). This nonlinearity is generated by a saturation of the post-
synaptic transduction cascade in the rod bipolar cell dendrites and not by mGluR6
receptor saturation (Sampath and Rieke, 2004). The molecular mechanisms that underlie
this are not understood, partly because the components of the transduction cascade
remain unknown. Here we show that the nonlinearity resides somewhere downstream of
Gα
o
activation. Gα
o2
improves sensitivity by increasing the gain, as seen in Gα
o2
vs. WT
mice in Figure 3.4. This is not simply an effect of different protein levels, since the
sensitivity of Gα
o
+/-
was identical to WT, although the protein level in the former was
reduced to ~50%. If the nonlinearity is a consequence of saturation of Gα
o
, then reducing
the protein level by half in the Gα
o
+/-
would be expected to shift the slope of the Hill
65
curve. Since the Hill coefficient of Gα
o
+/-
was almost identical to WT (Figure 3.5), this
suggests that the nonlinearity must reside somewhere downstream of Gα
o
activation. This
could be caused by saturation of another component of the transduction cascade or by
saturation of the transduction channels themselves (cf. Sampath and Rieke, 2004). For the
threshold to be optimally positioned, it must be high enough to eliminate most of the
continuous noise produced by spontaneous PDE activation, but low enough to allow the
single photon response to pass. Thus a delicate trade-off between noise and sensitivity
must exist, since too high a proportion of Gα
o2
will decrease the nonlinearity to a point
where the increase in noise rises to an unacceptable level.
3.4.4 A model for the mGluR6-signaling pathway
The mGluR6 receptor of rod bipolar cells receives glutamatergic input from rod
photoreceptors. The activation of the receptor triggers the change of GDP for GTP on the
α subunit of the G-protein G
o
. The action of the activated subunits Gα
o1
and Gα
o2
closes
non-specific cation channel by unknown effector mechanisms. The coordinated work of
the two splice variants Gα
o1
and Gα
o2
fine-tunes the sensitivity of the rod bipolar cells
and the relative contribution of Gα
o1
and Gα
o2
could determine the optimal trade-off
between noise and sensitivity. The mechanism by which the two splice variants engage
their effect could be mediated through different downstream signaling pathways or by the
same effector and transduction channel. However, it is clear that the α subunits of G
o
in
rod bipolar cells work in cooperation acting on the same function. The nonlinear
threshold in rod bipolar cells is generated somewhere downstream of Gα
o
activation by
saturation of another component of the transduction cascade or by saturation of the
transduction channels. The default state of the transduction channel mediating the light
66
Fig. 3.6. A revised model of the signal transduction in rod bipolar cells
GDP-bound inactive Gα
o1
and Gα
o2
compete for active mGluR6 receptors. The activation
of both splice variants leads to the closure of nonselective cation channels through
unknown downstream cascade. The efficiency of Gα
o2
pathway is lower than that of Gα
o1
pathway as represented by the thin arrow. This cooperative action by Gα
o1
and Gα
o2
fine-
tunes the stoichiometry of the signaling cascade. A saturation that underlies the
thresholding nonlinearity resides somewhere downstream of Gα
o
. The default state of
channels is closed, and another G-protein or βγ subunit may be required to open the
channels.
response through Gα
o
in rod bipolar cells is closed, and a GTP dependent mechanism
other than that of Gα
o1
and Gα
o2
is needed for channel opening. This could be
accomplished through a different G-protein cascade or through the competing action of
the activated Gβγ subunit, which dissociates from the GTP bound Gα. The closed default
state of the transduction channels can be critical for the need to minimize noise for
67
optimal signal processing and a competing mechanism to open these channels can allow
more control of the transduction cascade (Figure 3.6).
68
Chapter 4
Conclusions
We can navigate ourselves on a moonless night when the number of available
photons is limited. This exquisite sensitivity may not be well appreciated in our daily
lives, but it could make difference in survival in the wild. The importance of this high
sensitivity is reflected in the great amount of energy consumed by rod photoreceptors in
darkness, which must pump out cations flowing into the outer segment through open
channels. However, as light hyperpolarizes rods, their energy consumption is estimated to
fall to less than 25% of that in darkness (Okawa et al., 2008). Thus, rods spend an
enormous energetic cost to remain depolarized at low light levels. Why would rod
phototransduction be set in this manner? Part of the reason is that this arrangement will
allow single photon responses to be faithfully transmitted to downstream cells. For
example, calcium channels at the rod terminals are more sensitive to the small changes in
the membrane voltage at the rod resting potential compared to hyperpolarized potentials
(Figure 1.2). Through this costly effort, rods are perhaps optimized to inform the arrival
of photons to the downstream retinal circuit. Then, our ability to see in dim light depends
on how the downstream circuit interprets the signals presented by the array of rods.
The optimal downstream processor would (1) know the amplitude distribution and
the frequency characteristics of both noise and single photon signals in rods, (2) know the
probability of photon absorptions by individual rods at a given light level and (3) apply
an appropriate filter to maximize the probability of extracting single photon signals over
69
noise. This optimal processing can be easily performed by a computer, but biological
systems must achieve this task using available sets of molecules under a variety of
constraints. It is also possible that the biological systems may not be necessarily
processing signals in the way we think is optimal, which suggests we may not always
understand how a system is optimized. A previous study suggested that the rod-to-rod
bipolar transmission in the WT mouse retina is in fact explained largely by the optimal
processing of rod inputs near absolute visual threshold (Field and Rieke, 2002). This
optimal processing is biologically realized by saturation of the signaling cascade in rod
bipolar cell dendrites (Sampath and Rieke, 2004), and the level of the saturation must be
set precisely to separate ideally the single photon response from the underlying noise.
This requires delicate control in the stoichiometry of the rod bipolar signaling cascade.
Such control may be genetically encoded, so that the mice happen to have the optimal
thresholding of rod inputs as they grow, or may be adjusted by experience during
development. My use of GCAPs
-/-
mice provided an excellent model to test this idea.
Rod photoresponses in GCAPs
-/-
mice showed an increased amplitude and the
slowed time course for both single photon responses and noise due to the lack of Ca
2+
feedback to the synthesis of cGMP (Mendes et al., 2001; Burns et al., 2002). If we set a
filter to the average time course of the single photon response in GCAPs
-/-
rods and match
the filter to individual trials, we would observe a higher SNR for the single photon
response compared to WT rods. If the GCAPs
-/-
retina uses this filter matching the
response time course, and further processes the rod inputs optimally in the subsequent
circuit, they would have improved sensitivity near absolute threshold since their rod SNR
is higher than WT to begin with. However, a behavioral assay demonstrated that the
70
GCAPs
-/-
mice have a reduced sensitivity near absolute threshold. This behavioral
deterioration was explained by the rod-to-rod bipolar transmission being poorly-tuned to
the altered rod SNR in GCAPs
-/-
mice. More specifically, the postsynaptic saturation was
not adjusted to the right position for GCAPs
-/-
rod inputs, which led to a large amount of
noise obscuring the detection of single photon responses in the rod bipolar cells.
These results indicate that the level of saturation in the rod bipolar signaling
cascade is not adjustable or limited to a certain extent and cannot adjust to eliminate
noise for the large change in GCAPs
-/-
rod SNR. Instead, my modeling work suggests that
the level of saturation is likely hard-wired. The optimal level of saturation observed in
WT rod bipolar cells may have been attained through evolution. The inflexibility or the
limited capacity of flexibility in the postsynaptic saturation may stem from biophysical
limitations in the signaling cascade, or from the lack of feedback that reports the rod
single photon response and noise to rod bipolar cells. In either case, there are biological
constraints that hinder the visual system from adjusting to and optimally processing the
altered rod inputs. If the downstream circuit fails to process the rod inputs optimally, the
visual system loses the benefit of the improved rod SNR, which even leads to the
deteriorated behavioral performance.
What molecular events underlie this postsynaptic saturation? Unfortunately, little
is known about molecular identities in the rod bipolar signaling cascade. We know that
mGluR6 is the sensor of glutamate released from rod terminals, and that the mGluR6
relays the signal to nonselective cation channels through G-protein, G
o
. Recent progress
suggests that the nonselective cation channel is likely TRPM1 (Bellone et al., 2008), and
others are beginning to identify the other components in the signaling cascade. Since one
71
of the most important functions of the rod bipolar signaling cascade is to set the right
level of saturation for vision near absolute threshold, which will require a delicate control
of signaling molecules, it is critical to study the cascade not only by identifying unknown
signaling molecules but also by assessing the role of each molecule quantitatively. My
study adds a few important contributions in this regard. First, two splice variants of G
o
alpha subunits control the signaling cascade in a cooperative manner. The G
o2
splice
variant, which failed to cause a large response by itself, was found to increase the gain of
dim flash responses in WT rod bipolar cells. I raised a possibility that rod bipolar cells
may use two splice variants with different signaling efficiency to adjust the stoichiometry
of the cascade and set saturation at the optimal level. Second, the saturation in the
signaling cascade occurs somewhere downstream of G
o
. If the saturation of G
o
is
thresholding rod inputs, the reduction of G
o
by half in G
o
+/-
rod bipolar cells would
change the extent of nonlinear relationship between the flash intensity and the response
amplitude compared to WT rod bipolar cells. However, I found no difference in the
sensitivity and the nonlinearity between WT and G
o
+/-
rod bipolar cells. Thus, the
saturation must reside downstream of G
o
. Further investigations have to await the
identification of signaling molecules downstream of G
o
.
Collectively, my studies shed light on the importance of the optimal processing of
photoreceptor signals by the retinal circuit. The significance of these studies is not limited
to the visual neuroscience. Similar receptor-circuit matching will be crucial for the other
sensory systems to detect environmental cues effectively. A number of previous studies
suggested the importance of such optimal processing by the post-receptor circuits. My
studies occupy a unique position in that it directly demonstrates that the non-optimal
72
post-receptor processing can deteriorate the behavioral performance even when the
receptor performance is improved. To gain the benefit from the improved receptor
performance, the animals have to modify the circuit to be tuned to the improved receptor
performance.
73
References
Armstrong-Gold CE and Rieke F. Bandpass filtering at the rod to second-order cell
synapse in salamander (Ambystoma tigrinum) retina. J Neurosci 23, 3796-3806,
2003.
Ashmore JF and Falk G. Responses of rod bipolar cells in the dark-adapted retina of the
dogfish, Scyliorhinus canicula. J Physiol 300: 115-150, 1980.
Attwell D, Borges S, Wu SM, and Wilson M. Signal clipping by the rod output synapse.
Nature 328: 522-524, 1987.
Bader CR, Bertrand D, and Schwartz EA. Voltage-activated and calcium-activated
currents studied in solitary rod inner segments from the salamander retina. J
Physiol 331: 253-284, 1982.
Barlow HB. Retinal noise and absolute threshold. J Opt Soc Am 46: 634-639, 1956.
Barlow HB, Levick WR, and Yoon M. Responses to single quanta of light in retinal
ganglion cells of the cat. Vision Res Suppl 3: 87-101, 1971.
Baylor DA, Lamb TD, and Yau KW. Responses of retinal rods to single photons. J
Physiol 288: 613-634, 1979.
Baylor DA, Matthews G, and Yau KW. Two components of electrical dark noise in toad
retinal rod outer segments. J Physiol 309: 591-621, 1980.
Baylor DA and Nunn BJ. Electrical properties of the light-sensitive conductance of rods
of the salamander Ambystoma tigrinum. J Physiol 371: 115-145, 1986.
Baylor DA, Nunn BJ, and Schnapf JL. The photocurrent, noise and spectral sensitivity of
rods of the monkey Macaca fascicularis. J Physiol 357: 575-607, 1984.
Belgum JH and Copenhagen DR. Synaptic transfer of rod signals to horizontal and
bipolar cells in the retina of the toad (Bufo marinus). J Physiol 396: 225-245,
1988.
Bellone RR, Brooks SA, Sandmeyer L, Murphy BA, Forsyth G, Archer S, Bailey E, and
Grahn B. Differential gene expression of TRPM1, the potential cause of
congenital stationary night blindness and coat spotting patterns (LP) in the
Appaloosa horse (Equus caballus). Genetics 179, 1861-70, 2008.
74
Berntson A, Smith RG, and Taylor WR. Transmission of single photon signals through a
binary synapse in the mammalian retina. Vis Neurosci 21, 693-702, 2004
Bhandawat V, Reisert J, and Yau KW. Elementary response of olfactory receptor neurons
to odorants. Science 308, 1931-1934, 2005.
Bialek W and Owen WG. Temporal filtering in retinal bipolar cells. Elements of an
optimal computation? Biophys J 58: 1227-1233, 1990.
Bloomfield SA and Dacheux RF. Rod vision: pathways and processing in the mammalian
retina. Prog Retin Eye Res 20: 351-384, 2001.
Boycott KM, Maybaum TA, Naylor MJ, Weleber RG, Robitaille J, Miyake Y, Bergen
AA, Pierpont ME, Pearce WG, and Bech-Hansen NT. A summary of 20
CACNA1F mutations identified in 36 families with incomplete X-linked
congenital stationary night blindness, and characterization of splice variants. Hum
Genet 108: 91-97, 2001.
Burns ME and Baylor DA. Activation, deactivation, and adaptation in vertebrate
photoreceptor cells. Annu Rev Neurosci 24: 779-805, 2001.
Cao Y, Song H, Okawa H, Sampath AP, Sokolov M, and Martemyanov KA. Targeting of
RGS7/Gbeta5 to the dendritic tips of ON-bipolar cells is independent of its
association with membrane anchor R7BP. J Neurosci 28, 10443-9, 2008.
Capovilla M, Hare WA, and Owen WG. Voltage gain of signal transfer from retinal rods
to bipolar cells in the tiger salamander. J Physiol 391: 125-140, 1987.
Chen C, Clarke IJ. G
o
-2 protein mediates the reduction in Ca
2+
currents by somatostatin
in cultured ovine somatotrophs. J Physiol 491, 21-9, 1996
Chichilnisky EJ and Rieke F. Detection sensitivity and temporal resolution of visual
signals near absolute threshold in the salamander retina. J Neurosci 25, 318-330,
2005
Clapham DE and Neer EJ. G protein beta gamma subunits. Annu Rev Pharmacol Toxicol
37, 167-203, 1997.
Clark PT and van Rossum MC. The optimal synapse for sparse, binary signals in the rod
pathway. Neural Comput 18: 26-44, 2006.
Dacheux RF and Raviola E. The rod pathway in the rabbit retina: a depolarizing bipolar
and amacrine cell. J Neurosci 6: 331-345, 1986.
75
Deans MR, Volgyi B, Goodenough DA, Bloomfield SA, and Paul DL. Connexin36 is
essential for transmission of rod-mediated visual signals in the mammalian retina.
Neuron 36, 703-12, 2002.
Degtiar VE, Harhammer R, and Nürnberg B. Receptors couple to L-type calcium
channels via distinct G
o
proteins in rat neuroendocrine cell lines. J Physiol 502,
321-33, 1997.
Dhingra A, Lyubarsky A, Jiang M, Pugh EN Jr, Birnbaumer L, Sterling P, and Vardi N.
The light response of ON bipolar neurons requires G[alpha]o. J Neurosci. 2000
Dec 15;20(24):9053-8.
Dhingra A, Jiang M, Wang TL, Lyubarsky A, Savchenko A, Bar-Yehuda T, Sterling P,
Birnbaumer L, and Vardi N. Light response of retinal ON bipolar cells requires a
specific splice variant of Gα
o
. J Neurosci 22, 4878-84, 2002.
Doering CJ, Hamid J, Simms B, McRory JE, and Zamponi GW. Cav1.4 encodes a
calcium channel with low open probability and unitary conductance. Biophys J
89: 3042-3048, 2005.
Dolphin AC. Mechanisms of modulation of voltage-dependent calcium channels by G
proteins. J Physiol 506, 3-11, 1998.
Dunn FA, Doan T, Sampath AP, and Rieke F. Controlling the gain of rod-mediated
signals in the Mammalian retina. J Neurosci 26, 3959-3970, 2006.
Fain GL, Matthews HR, Cornwall MC, and Koutalos Y. Adaptation in vertebrate
photoreceptors. Physiol Rev 81: 117-151, 2001.
Fain GL, Quandt FN, Bastian BL, and Gerschenfeld HM. Contribution of a caesium-
sensitive conductance increase to the rod photoresponse. Nature 272: 466-469,
1978.
Field GD and Rieke F. Mechanisms regulating variability of the single photon responses
of mammalian rod photoreceptors. Neuron 35: 733-747, 2002.
Field GD and Rieke F. Nonlinear signal transfer from mouse rods to bipolar cells and
implications for visual sensitivity. Neuron 34: 773-785, 2002.
Field GD, Sampath AP, and Rieke F. Retinal processing near absolute threshold: from
behavior to mechanism. Annu Rev Physiol 67: 491-514, 2005.
Haeseleer F, Imanishi Y, Maeda T, Possin DE, Maeda A, Lee A, Rieke F, and Palczewski
K. Essential role of Ca2+-binding protein 4, a Cav1.4 channel regulator, in
photoreceptor synaptic function. Nat Neurosci 7: 1079-1087, 2004.
76
Hasegawa J, Obara T, Tanaka K, and Tachibana M. High-density presynaptic
transporters are required for glutamate removal from the first visual synapse.
Neuron 50: 63-74, 2006.
Hecht S, Shlaer S, and Pirenne MH. Energy, quanta, and vision. J Gen Physiol 25: 819-
840, 1942.
Hemara-Wahanui A, Berjukow S, Hope CI, Dearden PK, Wu SB, Wilson-Wheeler J,
Sharp DM, Lundon-Treweek P, Clover GM, Hoda JC, Striessnig J, Marksteiner
R, Hering S, and Maw MA. A CACNA1F mutation identified in an X-linked
retinal disorder shifts the voltage dependence of Cav1.4 channel activation. Proc
Natl Acad Sci U S A 102: 7553-7558, 2005.
Hille B. Ion channels of excitable membranes. Sunderland, MA: Sinauer Associates,
2001.
Horn DA and Latchman DS, Alternative splicing of the mRNA encoding the alpha
subunits of the G
o
GTP-binding protein during brain development and in neuronal
cell lines. Neurosci Lett 155, 57-60, 1993.
Hornstein EP, Verweij J, Li PH, and Schnapf JL. Gap-junctional coupling and absolute
sensitivity of photoreceptors in macaque retina. J Neurosci 25: 11201-11209,
2005.
Hsu SF, Augustine GJ, and Jackson MB. Adaptation of Ca(2+)-triggered exocytosis in
presynaptic terminals. Neuron 17: 501-512, 1996.
Hudspeth AJ and Corey DP. Sensitivity, polarity, and conductance change in the
response of vertebrate hair cells to controlled mechanical stimuli. Proc Natl Acad
Sci U S A 74, 2407-2411, 1977.
Kang H and Schuman EM. A requirement for local protein synthesis in neurotrophin-
induced hippocampal synaptic plasticity. Science 273: 1402-1406, 1996.
Kleuss C, Hescheler J, Ewel C, Rosenthal W, Schultz G, and Wittig B. Assignment of G-
protein subtypes to specific receptors inducing inhibition of calcium currents.
Nature 353, 43-8, 1991.
Kleuss C, Scherübl H, Hescheler J, Schultz G, and Wittig B. Selectivity in signal
transduction determined by gamma subunits of heterotrimeric G proteins. Science
259, 832-4, 1993.
77
Koschak A, Reimer D, Walter D, Hoda JC, Heinzle T, Grabner M, and Striessnig J.
Cav1.4alpha1 subunits can form slowly inactivating dihydropyridine-sensitive L-
type Ca2+ channels lacking Ca2+-dependent inactivation. J Neurosci 23: 6041-
6049, 2003.
Lamb TD and Pugh EN, Jr. Phototransduction, dark adaptation, and rhodopsin
regeneration the proctor lecture. Invest Ophthalmol Vis Sci 47: 5137-5152, 2006.
Lyubarsky AL and Pugh EN Jr. Recovery phase of the murine rod photoresponse
reconstructed from electroretinographic recordings. J Neurosci 16, 563-571, 1996.
Mansergh F, Orton NC, Vessey JP, Lalonde MR, Stell WK, Tremblay F, Barnes S,
Rancourt DE, and Bech-Hansen NT. Mutation of the calcium channel gene
Cacna1f disrupts calcium signaling, synaptic transmission and cellular
organization in mouse retina. Hum Mol Genet 14: 3035-3046, 2005.
Masu M, Iwakabe H, Tagawa Y, Miyoshi T, Yamashita M, Fukuda Y, Sasaki H, Hiroi K,
Nakamura Y, Shigemoto R, Takada M, Nakamura K, Nakao K, Katsuki M, and
Nakanishi S. Specific deficit of the ON response in visual transmission by
targeted disruption of the mGluR6 gene. Cell 80, 757-65, 1995.
Mattera R, Graziano MP, Yatani A, Zhou Z, Graf R, Codina J, Birnbaumer L, Gilman
AG, and Brown AM. Splice variants of the alpha subunit of the G protein G
s
activate both adenylyl cyclase and calcium channels. Science 243, 804-7, 1989.
Mendez A, Burns ME, Sokal I, Dizhoor AM, Baehr W, Palczewski K, Baylor DA, and
Chen J. Role of guanylate cyclase-activating proteins (GCAPs) in setting the flash
sensitivity of rod photoreceptors. Proc Natl Acad Sci U S A 98, 9948-9953, 2001.
Morgans CW. Localization of the alpha(1F) calcium channel subunit in the rat retina.
Invest Ophthalmol Vis Sci 42: 2414-2418, 2001.
Nawy S and Jahr CE. Suppression by glutamate of cGMP-activated conductance in
retinal bipolar cells. Nature 346, 269-71, 1990.
Nawy S and Jahr CE. cGMP-gated conductance in retinal bipolar cells is suppressed by
the photoreceptor transmitter. Neuron 7, 677-683, 1991.
Nawy S. The metabotropic receptor mGluR6 may signal through G
o
, but not
phosphodiesterase, in retinal bipolar cells. J Neurosci 19, 2938-44, 1999.
O'Connor E, Allen LE, Bradshaw K, Boylan J, Moore AT, and Trump D. Congenital
stationary night blindness associated with mutations in GRM6 encoding
glutamate receptor MGluR6. Br J Ophthalmol 90: 653-654, 2006.
78
Okawa H and Sampath AP. Optimization of single-photon response transmission at the
rod-to-rod bipolar synapse. Physiology (Bethesda) 22: 279-86, 2007.
Okawa H, Sampath AP, Laughlin SB, and Fain GL. ATP consumption by mammalian
rod photoreceptors in darkness and in light. Curr Biol 18, 1917-1921, 2008.
Rao VR, Cohen GB, and Oprian DD. Rhodopsin mutation G90D and a molecular
mechanism for congenital night blindness. Nature 367: 639-642, 1994.
Rao-Mirotznik R, Buchsbaum G, and Sterling P. Transmitter concentration at a three-
dimensional synapse. J Neurophysiol 80: 3163-3172, 1998.
Rao-Mirotznik R, Harkins AB, Buchsbaum G, and Sterling P. Mammalian rod terminal:
architecture of a binary synapse. Neuron 14: 561-569, 1995.
Rieke F and Baylor DA. Molecular origin of continuous dark noise in rod photoreceptors.
Biophys J 71: 2553-2572, 1996.
Rieke F and Baylor DA. Single-photon detection by rod cells of the retina. Rev Mod
Phys 70, 1998.
Rieke F and Schwartz EA. Asynchronous transmitter release: control of exocytosis and
endocytosis at the salamander rod synapse. J Physiol 493 ( Pt 1): 1-8, 1996.
Sampath AP and Rieke F. Selective transmission of single photon responses by saturation
at the rod-to-rod bipolar synapse. Neuron 41: 431-443, 2004.
Sampath AP, Strissel KJ, Elias R, Arshavsky VY, McGinnis JF, Chen J, Kawamura S,
Rieke F, and Hurley JB. Recoverin improves rod-mediated vision by enhancing
signal transmission in the mouse retina. Neuron 46: 413-420, 2005.
Schein S and Ahmad KM. A clockwork hypothesis: synaptic release by rod
photoreceptors must be regular. Biophys J 89: 3931-3949, 2005.
Schmidt A, Hescheler J, Offermanns S, Spicher K, Hinsch KD, Klinz FJ, Codina J,
Birnbaumer L, Gausepohl H, Frank R, Schultz G, and Rosenthal W. Involvement
of pertussis toxin-sensitive G-proteins in the hormonal inhibition of
dihydropyridine-sensitive Ca
2+
currents in an insulin-secreting cell line
(RINm5F). J Biol Chem 266, 18025-33, 1991.
Schneeweis DM and Schnapf JL. Photovoltage of rods and cones in the macaque retina.
Science 268: 1053-1056, 1995.
Shiells RA and Falk G. Glutamate receptors of rod bipolar cells are linked to a cyclic
GMP cascade via a G-protein. Proc Biol Sci 242, 91-4, 1990.
79
Shiells RA and Falk G. Responses of rod bipolar cells isolated from dogfish retinal slices
to concentration-jumps of glutamate. Vis Neurosci 11: 1175-1183, 1994.
Shiells RA and Falk G. Potentiation of 'on' bipolar cell flash responses by dim
background light and cGMP in dogfish retinal slices. J Physiol 542, 211-20, 2002.
Sjostrand FS. The ultrastructure of the innersegments of the retinal rods of the guinea pig
eye as revealed by electron microscopy. J Cell Physiol 42: 45-70, 1953.
Smith RG, Freed MA, and Sterling P. Microcircuitry of the dark-adapted cat retina:
functional architecture of the rod-cone network. J Neurosci 6: 3505-3517, 1986.
Snellman J and Nawy S. cGMP-dependent kinase regulates response sensitivity of the
mouse on bipolar cell. J Neurosci 24, 6621-8, 2004.
Strathmann M, Wilkie TM, and Simon MI. Alternative splicing produces transcripts
encoding two forms of the alpha subunit of GTP-binding protein G
o
. Proc Natl
Acad Sci U S A 87, 6477-81, 1990.
Taylor WR and Smith RG. Transmission of scotopic signals from the rod to rod-bipolar
cell in the mammalian retina. Vision Res 44: 3269-3276, 2004.
Teich MC, Prucnal PR, Vannucci G, Breton ME, and McGill WJ. Multiplication noise in
the human visual system at threshold: 1. Quantum fluctuations and minimum
detectable energy. J Opt Soc Am 72, 419-431, 1982.
Thoreson WB, Rabl K, Townes-Anderson E, and Heidelberger R. A highly Ca2+-
sensitive pool of vesicles contributes to linearity at the rod photoreceptor ribbon
synapse. Neuron 42: 595-605, 2004.
Trifonov IU. [Study of synaptic transmission between photoreceptor and horizontal cell
by electric stimulations of the retina]. Biofizika 13: 809-817, 1968.
Tsukamoto T, Toyama R, Itoh H, Kozasa T, Matsuoka M, and Kaziro Y. Structure of the
human gene and two rat cDNAs encoding the alpha chain of GTP-binding
regulatory protein Go: two different mRNAs are generated by alternative splicing.
Proc Natl Acad Sci U S A 88, 2974-8, 1991.
Tsukamoto Y, Morigiwa K, Ueda M, and Sterling P. Microcircuits for night vision in
mouse retina. J Neurosci 21, 8616-8623, 2001.
van der Velden HA. The number of quanta necessary for the perception of light in the
human eye. Opthalmologica 111: 321-331, 1946.
van Rossum MC and Smith RG. Noise removal at the rod synapse of mammalian retina.
Vis Neurosci 15: 809-821, 1998.
80
Vardi N, Duvoisin R, Wu G, and Sterling P. Localization of mGluR6 to dendrites of ON
bipolar cells in primate retina. J Comp Neurol 423: 402-412, 2000.
Wahl-Schott C, Baumann L, Cuny H, Eckert C, Griessmeier K, and Biel M. Switching
off calcium-dependent inactivation in L-type calcium channels by an
autoinhibitory domain. Proc Natl Acad Sci U S A 103: 15657-15662, 2006.
Walraven J, Enroth-Cugell C, Hood DC, Dia ML, and Schnapf JL. The control of visual
sensitivity. San Diego: Academic, 1990.
Wettschureck N and Offermanns S. Mammalian G proteins and their cell type specific
functions. Physiol Rev 85, 1159-204, 2005.
Yau KW, Matthews G, and Baylor DA. Thermal activation of the visual transduction
mechanism in retinal rods. Nature 279: 806-807, 1979.
Yee R and Liebman PA. Light-activated phosphodiesterase of the rod outer segment.
Kinetics and parameters of activation and deactivation. J Biol Chem 253: 8902-
8909, 1978.
Zeitz C, Kloeckener-Gruissem B, Forster U, Kohl S, Magyar I, Wissinger B, Matyas G,
Borruat FX, Schorderet DF, Zrenner E, Munier FL, and Berger W. Mutations in
CABP4, the gene encoding the Ca2+-binding protein 4, cause autosomal
recessive night blindness. Am J Hum Genet 79: 657-667, 2006.
Zenisek D, Steyer JA, and Almers W. Transport, capture and exocytosis of single
synaptic vesicles at active zones. Nature 406: 849-854, 2000.
Abstract (if available)
Abstract
When few photons are available, such as on a moonless night, our vision is limited by the probability of photon absorption in rod photoreceptors, and the reliable transmission of light-evoked signals through the retina. Dark-adapted rod photoreceptors are capable of reliably signaling single photon absorptions. However, these single photon responses are challenged by neural noise arising from rod photoreceptors themselves, and the subsequent retinal circuitry. Rod bipolar cells, the second-order neurons specialized for low light level vision, pool 20 - 100 rod inputs. This convergence imposes on rod bipolar cells the difficult task of identifying sparse single photon responses among the majority that generate noise. It has been shown that saturation within the postsynaptic signaling cascade in the rod bipolar cells can effectively eliminate signals from rods whose amplitudes are statistically dominated by noise near absolute visual threshold. My first study emphasizes the delicate match between the signal-to-noise ratio of the rod photoresponse and the properties of the signaling cascade in the postsynaptic rod bipolar cells. In particular, I provide evidence that the improved signal-to-noise ratio in rod photoreceptors does not guarantee improved behavioral performance if the rod bipolar signaling cascade is not well-tuned to the rod signal-to-noise ratio. In my second study, I show that two splice variants of the heterotrimeric G-protein alpha subunit, Goα, are involved in the signaling cascade of On-bipolar cells. In rod (On) bipolar cells these subunits may together subserve the fine-tuning of the stoichiometry in the signaling cascade to improve the detection of light near absolute threshold. Collectively, my work demonstrates how the signaling cascade in the rod bipolar cells can be fine-tuned to the rod inputs to improve vision near absolute threshold, and how such adjustments are limited by biological constraints.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Sensitivity and dynamic range of rod pathways in the mammalian retina
PDF
Synaptic mechanisms for basic auditory processing
PDF
Visual arrestin interactions with clathrin adaptor AP-2 regulate photoreceptor survival
PDF
An investigation of dark adaptation: the role of metabolism and alternative rod pathways in shaping visual sensitivity following bright light
PDF
Visual information processing by the feedforward and feedback inputs to the lateral geniculate nucleus
PDF
Contextual modulation of sensory processing via the pulvinar nucleus
PDF
Functional models of fMRI BOLD signal in the visual cortex
Asset Metadata
Creator
Okawa, Haruhisa
(author)
Core Title
Signal processing by the mammalian retina near absolute visual threshold
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Neuroscience
Publication Date
07/08/2010
Defense Date
05/22/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
absolute visual threshold,OAI-PMH Harvest,Physiology,reitna
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Sampath, Alapakkam P. (
committee chair
), Chen, Jeannie (
committee member
), Hinton, David R. (
committee member
), Hirsch, Judith (
committee member
), Zhang, Li (
committee member
)
Creator Email
harusam@hotmail.com,hokawa@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2334
Unique identifier
UC155142
Identifier
etd-Okawa-2989 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-405497 (legacy record id),usctheses-m2334 (legacy record id)
Legacy Identifier
etd-Okawa-2989.pdf
Dmrecord
405497
Document Type
Dissertation
Rights
Okawa, Haruhisa
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
absolute visual threshold
reitna