Essential role of the carboxyl-terminus for proper rhodopsin trafficking and "enlightenment" to the pathway(s) causing retinal degeneration in a mouse model expressing a truncated rhodopsin mutant

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

ESSENTIAL ROLE OF THE CARBOXYL-TERMINUS FOR PROPER
RHODOPSIN TRAFFICKING AND “ENLIGHTENMENT” TO THE
PATHWAY(S) CAUSING RETINAL DEGENERATION IN A MOUSE MODEL
EXPRESSING A TRUNCATED RHODOPSIN MUTANT

by

Francis Avila Concepcion

____________________________________________________________________

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
(CELL AND NEUROBIOLOGY)

December 2007

Copyright 2007 Francis Avila Concepcion

Dedication

To especially my wife Jen, and my children David and Anabelle,

my parents Mario Sr. and Luisa,

my brothers and sister Marlon, Mario Jr., and Marilou,

my grandparents “Na-Nay and Ta-Tay”

and my Aunt Felicidad Gonzales (Aunt Fely) who made everything possible!

I couldn’t have done this without your support, patience, and especially your love!

ii
Acknowledgements

With much gratitude to my mentor, Dr. Jeannie Chen. I thank you for all the years

of tutorship, challenges, and support!

Without you, so many wonderful things in my life would not have happened:

My development as a scientist,

My excitement about vision research,

The friends I’ve made in southern California,

and my family, Jen, David, and Anabelle!

Thank you to my committee members:
Judy Garner, PhD
Robert Chow, MD, PhD
Curtis Okamoto, PhD
Your support and guidance have been invaluable! I am quite pleased with and
feel blessed for having a committee with a wide, yet insightful range of
perspectives within the scientific community. I also thank you for your
roles to my growth as a scientist!
Thank you Drs. Mendez and Roca (Ana & Angela) for all your help, advice, and
stimulating conversations in Dr. Chen’s laboratory since the beginning.

I also would like to acknowledge Dr. Cheryl Craft and Bruce Brown (who also have
been part of my USC life since the beginning).
iii
Table of Contents

Dedication ii
Acknowledgements iii

List of Figures vi
Abstract viii
Chapter 1 Overview of Photoreceptors, the Phototransduction 1
Cascade, and Rhodopsin Trafficking
1.1. Introduction. 1
1.2. Photoreceptor Outer Segments and Phototransduction. 5
1.3. Phototransduction Inactivation. 11
1.4. Rhodopsin Trafficking. 16
1.5. Drosophila Vision. 19
1.6. Thesis Outline. 25
Chapter 2 The Carboxyl-Terminal Domain is Essential for 28
Rhodopsin Transport in Rod Photoreceptors
2.Abstract 29
2.1. Introduction. 30
2.2. Materials and Methods. 33
2.3. Results. 38
2.4. Discussion. 50

Chapter 3 The Q344ter Rhodopsin Mutant (Q344ter) Causes 55
Mislocalization of Rhodopsin Molecules that are
Capable of Light-Activation
3.1. Introduction. 55
3.2. Materials and Methods. 62
3.3. Results. 71
3.4. Discussion. 95

Chapter 4 Photolyzed Rhodopsin/Arrestin Complex 104
Contributes to the Light-Accelerated Retinal
Degeneration in the Transgenic Q344ter Rhodopsin
Mutant Mouse Model
4.1. Introduction. 104
4.2. Materials and Methods. 107
4.3. Results. 110
4.4. Discussion. 123
iv

Chapter 5 Miscellaneous Personal Observations 126

5.1. Potential Correlation Between Penta-Phosphorylated 126
Q344ter Molecules and the Lowered Reproducibility
in Single Photon Responses in Rods Expressing this
Truncated Rhodopsin Mutant.
5.2. Unexpected Absence of [
35
S]GTP γS Loading in the 128
ROS of Nontransgenic
rho+/-
Retinal Sections after
5 min Light-Exposure.
5.3. The Presence of Mutant Rhodopsin Molecules in 132
Transgenic Q344ter
rho-/-
Retinas Lowers the Rate of
Degeneration Compared to Rho-/- Retinas.
5.4. Potential Contribution from β-Arrestin to the Light 134
Accelerated Damage in Transgenic Q344ter Retinas.
Bibliography 140

v
List of Figures
Figure 1.1. Schematic drawing of a magnified view of a cross- 3
sectional region of the retina.
Figure 1.2. Rod versus cone structures. 6
Figure 1.3. Phototransduction cascade in the rod cell. 8
Figure 1.4. Rhodopsin inactivation. 13
Figure 1.5. Rhodopsin transport within the polarized rod cell is 17
a multi-stage process.
Figure 1.6. Drosophila vision. 21
Figure 2.1. Comparison of transgene expression in S334ter and 39
S338A transgenic lines.
Figure 2.2. Expression of the control rhodopsin transgene 40
(S338A) restores rod outer segment formation
in rhodopsin -/- mice.
Figure 2.3. Retinal morphology of Ser334ter transgene-positive 42
versus negative control mice in the rhodopsin +/+, +/-
and -/- background.
Figure 2.4. In the absence of endogenous rhodopsin, Ser334ter 43
rhodopsin localizes to the outer nuclear layer and
inner segments of photoreceptors.
Figure 2.5. Western blot analysis of Ser334ter and S338A 45
rhodopsin in transgenic lines.
Figure 2.6. Endogenous rhodopsin mis-localizes to the inner 47
segments in the presence of Ser334ter rhodopsin.
Figure 2.7. Ser334ter rhodopsin expression does not lead to 49
cell death.

Figure 2.8. Schematic model showing Ser334ter and endogenous 53
rhodopsin being co-transported in post-Golgi vesicles.

vi
Figure 3.1. Generation of Q344ter transgenic mice. 72
Figure 3.2. Determination of Q344ter transgene expression 74
at the RNA and protein levels.
Figure 3.3. Q344ter rhodopsin mistrafficks and causes partial 80
mislocalization of endogenous rhodopsin molecules.
Figure 3.4. The QVAPA domain is essential for ROS formation. 83
Figure 3.5. Retinal degeneration caused by mislocalized 86
rhodopsin is accelerated by light.
Figure 3.6. Detection of light-dependent multi-phosphorylated 90
Q344ter species.
Figure 3.7. Detection of light-activation of mislocalized 92
rhodopsin.
Figure 3.8. Mislocalized rhodopsin molecules in the RIS 94
are capable of light-activation.
Figure 4.1. The R*/SAg complex contributes to retinal 113
degeneration in the transgenic Q344ter mouse model.
Figure 4.2. Light-dependent GTP γS loading (20 min exposure) 118
in transgenic Q344ter frozen retinal sections.
Figure 5.1. Light-dependent GTP γS loading (5 min exposure) 129
in transgenic Q344ter frozen retinal sections with
either the rho+/- or rho+/-, Tr α-/- genetic background.
Figure 5.2. Expression of truncated mutant rhodopsin decreases 133
progression of retinal degeneration in rho-/- mice.
Figure 5.3. Significant increase in β-arrestin translocation to the 136
membrane in light-exposed Q344ter
rho+/-, SAg-/-, Tr α-/-

retinas.

vii
viii
Abstract
Retinitis pigmentosa (RP) is a heterogeneous group of inherited eye diseases that is
typified by initial night blindness and loss of peripheral vision and eventually
culminates into total blindness. This retinal disorder afflicts around 1 in 4000 people
worldwide. Mutations in the carboxyl-terminus of rhodopsin have been linked to
autosomal dominant RP (ADRP). In this study, we have investigated two truncated
rhodopsin mutants, S334ter and Q344ter. Both mutants have demonstrated that the
presence of the QVAPA domain is necessary for proper rhodopsin localization and
ROS formation. However, the retention of the phosphorylation sites in the Q344ter
(and not in the S334ter) has allowed us to demonstrate the light-activation capability
of mislocalized rhodopsin. More pertinent to ADRP patients, the retention of these
phosphorylation sites in Q344ter also contributes to light-accelerated retinal
degeneration through a mechanism first described in Drosophila, the accumulation
of the rhodopsin/arrestin complexes in abnormal subcellular compartments.

Chapter 1
Overview of Photoreceptors, the Phototransduction Cascade,
and Rhodopsin Trafficking

1.1. Introduction.
Receiving and processing external stimuli are essential for not only mere
survival, but at least in human terms, enjoyment and appreciation of existence.
Whether it’s listening to a Beatles song on the radio, sharing a home-cooked meal
with the family, smelling your two-year old son’s freshly shampooed hair, holding
hands with your wife, or noticing her belly becoming larger while she’s pregnant
with a your first daughter, one relies on the five senses on nearly every aspect of life.
Of these senses sight is perhaps the most cherished. Thus, it is not surprising that
millions of dollars are spent each year to finance scientific research to preserve,
improve, and/or restore eyesight. On an individual level, this dissertation describes
years of personal scientific research under the guidance of Dr. Jeannie Chen and her
staff in an attempt to contribute but a small piece of the puzzle that is vision research.
Before I begin discussing the consequences of a rod-specific defect,
rhodopsin transport, this introductory chapter describes pertinent background
material on vision to reveal some of the intricacies involved that make sight possible.
Here, I will emphasize rod cell characteristics but include cone attributes when
applicable. The ultimate intent of this chapter is to convince the reader of the
1
importance of proper rhodopsin trafficking within the context of “our most cherished
sense”.
Human sight begins with specific light absorptions in the retina, which is the
laminar innermost tissue of the eye. The retina is multilayered and contains five
main types of neurons - photoreceptors, bipolar cells, horizontal cells, amacrine cells,
and ganglion cells (Fig. 1.1). The roles of this tissue are: to absorb photons from the
environment; to process this information; and to send these messages to the ganglion
cells whose axons extend to the lateral geniculate nucleus (LGN) of the brain.
Ultimately, the completed images are formed in the visual cortex.
On a cellular level, photon absorptions which initiate vision occur in two
types of photoreceptor cells - rods and cones, which are named after their respective
outer segment (OS) shape (Fig. 1.2). The other neuronal cell types are involved in
processing and propagating signals originating from the photoreceptors. Localized
next to the retinal pigment epithelium (RPE) layer, rods and cones together
determine our visual ranges. With respect to photon wavelengths, the visible
spectrum for humans spans the colors of the rainbow (400-750 nm). In terms of light
intensities, human vision operates under an immense range. The threshold of human
vision under dim light conditions transpires with the average of one photon
absorption every 85 minutes per rod cell; while the upper limits under bright light
conditions of human vision occur with as much as 1 x 10
6
photon absorptions every
second per cone cell (Rodieck, 1998). With this immense range in light intensity,
human vision is subdivided into three categories depending on the photoreceptors
2

Figure 1.1. Schematic drawing of a magnified view of a cross-sectional region of
the retina. (adapted from the following website:
www.stcsc.edu/anatomy/210/Chapter%2013%20part%201.ppt). The photoreceptor
cells are located in the outer layer of the retina, just below the retinal pigment
epithelium (RPE). The other neurons labeled above modulate and further process the
light-derived signals stemming from the photoreceptor layer. Ultimately, the
neuronal signals exit the retina via the axons of the ganglion cells (optic fiber layer).
Note the opposite directions of incoming light and outgoing visual messages.
(Insert) Schematic drawing displaying the overall structure of the eye. See text for a
more detailed explanation.
3
involved: scotopic vision occurs when only rod cells are stimulated; mesopic vision
occurs when both rods and cones are operating; while photopic vision occurs when
only cone cells are functioning due to the saturation of rods.
Along with providing color vision, cone-mediated responses also provide
visual acuity. Not surprisingly, we mainly use cones for daytime vision. However,
these cells only account for approximately 5% of the total photoreceptor population
(Curcio et al., 1990). Despite this scarce percentage, cone cells are able to mediate
our daytime vision is mediated by cone cells, in which the majority of these
photoreceptor cells densely populate the fovea. This region of the retina is located
directly opposite of the pupil and mainly conveys our “acute central vision”.
However, a smaller but significant proportion of cone cells are located in the retinal
periphery; these cones bestow us our peripheral sight during bright light conditions.
Composing the remaining 95% of total photoreceptor population, rods (as
many as 92 million per human retina; Curio et al., 1990) are responsible for dim light
vision and also sustain the existence of cones. This latter trait is supported by a
common observation in patients suffering from retinitis pigmentosa (RP; a detailed
explanation of this eye disease is described in the Introduction section of Chapter 3):
cone cell death occurring secondarily to rod cell apoptosis that is induced by rod-
specific mutations. It has been suggested that the secretion of an unidentified
paracrine agent(s) by rods is necessary for cone survival. This hypothesis is
supported by a simple experiment involving cultured retinal explants (Mohand-Said
et al., 1998). Cone cells within a rod-deprived retina (due to the rods expressing a
4
RP mutation) showed an increased survival in the presence of a second retina
without any genetic defects (wildtype; WT) as compared to the controls, such as the
absence of another retina or the presence of a second rod-deprived retina with the
same rod-specific RP mutation. The two retinas were physically separated but
shared the same medium. In summary, this latter role of rod cells exemplifies their
importance not only to dim light vision but also to the overall health of the retina.

1.2. Photoreceptor Outer Segments and Phototransduction.
The outer segments (OSs) of photoreceptors (Fig. 1.2) are the subcellular
structures specialized for the phototransduction cascade, the biochemical process by
which light signals are converted to neuronal signals. Each OS contains millions of
visual pigments, or photopigments, embedded within the membrane disks that are
stacked upon one another (approximately 1000 or so in rods). This arrangement of
the plethora of photopigments in disks of each OS increases the photoreceptor cell’s
ability to absorb the photons. (Disk formation will be addressed below in the
rhodopsin transport section.)
Photons are absorbed by the visual pigments. These proteins are prototypical
G protein-coupled receptors (GPCRs) with seven transmembrane α-helices.
Photopigments have two components: a chromophore called 11-cis-retinal that is
derived from vitamin A and used by all human photopigments; and an apoprotein
named opsin that is distinct to the type of photoreceptor cell. These two constituents
are linked together by a protonated Schiff base within the lipid bilayer section of the
photopigment. With rhodopsin, this linkage occurs at the 296 lysine residue (K296).
5

Figure 1.2. Rod versus cone structures. The shape of their outer segment (OS)
gives each type of photoreceptors its name, i.e. rods are rod-shaped, while cones are
cone-shaped. The metabolic and protein synthesis machineries in each cell type are
located in the inner segment (IS). Although not labeled, the collection of
mitochondria is termed the ellipsoid, while the intracellular region containing the
endoplasmic reticulum (ER) and Golgi network (GN) is called the myoid. The OS
and IS are joined together by the connecting cilium (CC). Below the IS region of
each photoreceptor cell is the nucleus followed by the synapse, where the cells make
contacts with other neuronal cell types. [The rod synapse is called a spherule, while
a cone synapse is known as a pedicle.] Note that the small dark lines within the
synapses represent synaptic ribbons. Rods are very sensitive and function in dim
light; while cone cells are less sensitive, function in bright light, and provide color
vision. Together, their functions largely are responsible for the ranges (in terms of
both wavelengths and light intensities) of vertebrate vision.
6
Additionally, this positively charged Schiff base linkage interacts with a negatively
charged counter ion, a glutamate at position 113 (E113), to form a salt bridge
(Sakmar et al., 1989). Because of these two factors (the intramolecular salt bridge
and the covalently attached 11-cis-retinal), the “dark state” rhodopsin is
exceptionally stable.
All rods express rhodopsin, which has its maximum spectral sensitivity at
498 nm. Because only one type of visual pigment is found in rods, scotopic vision
lacks color. Although each cone cell expresses only one type of photopigment, there
are three kinds of cone cells within the human retina depending on the cone opsin
expressed (with their maximum spectral sensitivity in parenthesis): blue opsin (420
nm), green opsin (535 nm), and red opsin (560 nm). Because of these three types of
cone cells, humans have trichromatic vision, and therefore, see in color. On the
other hand, mice exhibit dichromatic vision, since they only express two types of
cone opsin: short-wavelength opsin (S-opsin, 360 nm) and medium-wavelength
opsin (M-opsin, 510 nm).
When a photon is absorbed by a photopigment in such a manner to induce
conformational changes, the phototransduction cascade is initiated (Fig. 1.3). Much
of the details of this G-protein cascade have been revealed by experiments involving
rod cells, therefore elucidating rod phototransduction. Because cones have
analogous proteins to rods, i.e. cone opsin, cone transducin, cone phosphodiesterase,
etc., the phototransduction cascade at a general level operates by similar mechanisms
in both types of photoreceptors. More specifically, this photon absorption must
7
γ
 GDP
cGMP
5’ GMP
GTP
Ca
2+
K
+
Na
+
Ca
2+
Na
+
plasma
membrane
disc
membrane
cGMP
phospho-
diesterase
(PDE)
α
β
γ
CNG channel
(open)
+
β
γ
α
∗
Transducin
(Tr)
GTP
GDP
exchange
RGS9 G β5
β
γ
all trans-
retinal
R
R
∗
Rhodopsin
(R)
SAg
R
ATP
Pi
ADP RK
Pi
*
11 cis-
retinal
Light
Pi
Opsin α
*
G β5 RGS9
phtase
cGMP
cGMP
γ
β
Na
+
Ca
2+
X
R9AP
R9AP
γ
β
α*
GCs
GCAPs
CNG channel
(closed)
exchanger
RGS complex
(R9AP/RGS9/G β5)
GTP
GTP
γ
α
R
*

Figure 1.3. Phototransduction cascade in the rod cell. When the 11 cis-retinal
that is covalently linked to rhodopsin (R) absorbs a photon, it is converted into all
trans-retinal. This change ultimately activates R. This activated R molecule (R*)
stimulates the heterotrimeric G-protein, transducin (Tr), by serving as a guanine-
nucleotide exchange factor (GEF). Therefore, this interaction promotes the
exchange of GDP for GTP within the α-subunit of Tr (Tr α). The binding of GTP to
the Tr α subunit induces the release of Tr’s β γ complex. The freed GTP-bound Tr α
then activates cGMP phosphodiesterase (PDE, the effector molecule of this cascade)
by binding to one of PDE’s γ-subunit. This event, in turn, activates either PDE’s α-
or β- catalytic subunit (in the above case, α-subunit), which converts cGMP to GMP,
thereby lowering the intracellular [cGMP]. This condition causes cGMP molecules
to be released from cyclic nucleotide gated cation channels (CNGs), resulting in the
closure of these channels and thereby preventing further entry of Ca
2+
and Na
+
ions
at these sites. Because of continued activities of exchangers in the OS (pictured) and
Na
+
/K
+
pumps and K
+
pumps in the IS (not pictured), this situation creates a
hyperpolarization state across the entire plasma membrane of the cell.
Hyperpolarization ultimately signals the rod cell to decrease the rate of release of
glutamate, its neurotransmitter, to other neurons within the retina. See text for a
more detailed explanation on this cascade and for its inactivation steps involving
proteins not described here [i.e., RGS complex (R9Ap/RGS9/G β5) participating in
Tr inactivation; guanylyl cyclases (GCs) and guanylyl cyclase activating proteins
(GCAPs) participating in intracellular cGMP restoration levels]. This figure,
although modified, is courtesy of Dr. Mendez.
8
induce the chromophore of the photopigment, 11 cis-retinal, to isomerize to all
trans-retinal. This transformation of the chromophore creates a strain within the
protein component of the photopigment. This strain disrupts intramolecular
interactions (including the aforementioned salt bridge), which causes additional
conformational changes that lead to the active form of the visual pigment (in the case
of rhodopsin, metarhodopsin II). The activated receptor acts as a guanine-nucleotide
exchange factor (GEF) and during its lifetime stimulates hundreds of transducin (Tr)
molecules, a heterotrimeric G-protein, by catalyzing the substitution of Tr’s GDP for
GTP. This exchange subsequently releases the βγ complex from the α-subunit. Each
activated Tr α-subunit, in turn, interacts with only one γ-subunit of cGMP
phosphodiesterase (PDE), with PDE being the effector molecule of this cascade.
This Tr α-PDE γ interaction disrupts the inhibitory effect of the PDE γ, thereby
stimulating either the PDE α- or PDE β- subunit, depending on which PDE γ is
affected. In turn, this stimulation of the PDE’s catalytic subunit breaks down
cytoplasmic cGMP to GMP via hydrolysis, which lowers the concentration of the
intracellular “pool” of cGMP. This decrease in [cGMP] induces the release of
cGMP from specific cyclic nucleotide-gated ion (CNG) channels, causing them to
close. These closure events preclude the transport of Na
+
and Ca
2+
(and a lesser
degree with Mg
2+
) ions into the OS region of the photoreceptor cell. Concurrent
light-independent activities from Na
+
/Ca
2+
K
+
exchangers in the OS and Na
+
/K
+

pumps and K
+
pumps in the inner segment (IS) result in a net positive charge exiting
the photoreceptor cell compared to the cell’s “dark state”. This light-dependent
9
decrease in cations entering the cell is termed photocurrent, and this condition results
in the hyperpolarization of the entire cell.
At the synaptic terminus of the photoreceptor cell (termed a spherule in rods
and pedicle in cones), this hyperpolarization state triggers the closure of voltage-
gated Ca
2+
channels and subsequent lowering of the local Ca
2+
levels at this region.
Consequently, the release rate of glutamate-laden vesicles (which is Ca
2+
dependent)
is decreased, which lowers the concentration of glutamate molecules at the
photoreceptor synaptic terminal. It is this reduced availability of glutamates at this
microenvironment that serves as the chemical signal to the other neurons, i.e. bipolar
cells. This completes the transformation of light signals to chemical signals. Further
details concerning the phototransduction cascade can be found in Rodieck (1998)
and Yau (1994).
Despite the general similarities of the phototransduction process in both types
of photoreceptor cells, important peculiarities between rods and cones are observed
that contribute to the function of each type of photoreceptor. In terms of change in
photocurrent, the typical single photon response in rods is approximately 100X the
typical response in cones. Because rods are more sensitive, scotopic vision bestows
upon us images at much lower light levels than cones can provide. However, rod-
mediate responses also saturate at much lower light intensities than cone-mediated
responses. Moreover, cone responses recover dramatically faster, i.e. their response
recovery times are much shorter. These two cone attributes significantly account for
10
both the bright light intensity responses and the greater resolution in photopic vision,
thereby contributing to greater visual acuity (Rodieck, 1998).

1.3. Phototransduction Inactivation.
Of vital importance for vision is the proper termination of phototransduction
so that changes in our environment could be detected and that sharper images are
obtained, especially in the case of photopic vision. Again, the rod phototransduction
shutoff will serve as the general model with some details of cone phototransduction
shutoff provided for comparison purposes. Also, for simplicity, the following
descriptions for phototransduction shutoff assume that the initial light source that
activated the phototransduction cascade is turned off.
1.3.1. Transducin inactivation.
A key component of this process is the inactivation of Tr. The regulator of
G-protein signaling (RGS) complex - composed of RGS9-1, G β5L, and R9AP -
functions as the GTPase activating protein (GAP) for Tr by accelerating the inherent
GTPase activity of this G-protein (He et al., 1998; Hu and Wensel, 2002). Recently,
a experiment revealed that this activity of the RGS complex actually is the rate
limiting step in the single photon response in rod cells (Krispel et al., 2006). They
observed that overexpression of the RGS complex shortened the response recovery
time. It was also pointed out that the faster response recovery in mammalian cones
may be partly explained by their higher expression levels of the RGS complex
(Cowan et al., 1998; Zhang et al., 2003b; Krispel et al., 2006). Furthermore, several
studies show that the PDE γ-subunit bound to the activated Tr α-subunit actually
11
contributes to this GTPase accelerating effect by the RGS complex (Arshavsky and
Bownds, 1992; Angleson and Wensel, 1994; He et al., 1998; Tsang et al., 1998).

1.3.2. Rhodopsin inactivation.
The inactivation of the rhodopsin represents another factor to the overall
shutoff of phototransduction (Fig. 1.4). The first step of rhodopsin inactivation is the
phosphorylation of activated rhodopsin (R*) by rhodopsin kinase (Kuhn and Wilden,
1987; Chen et al., 1999a) at multiple serine/threonine residues located at the
receptors’ carboxyl-terminus (six in human and mouse retinas; seven in bovine
retinas). This number of potential phosphorylation sites appears to play an important
role in the single photon response. Doan et al. (2006) showed that rods expressing
wildtype (WT) rhodopsin molecules have greater reproducibility in the single photon
response than rods expressing mutant rhodopsin molecules with one or more
phosphorylation sites absent. However, it remains to be determined if the six
phosphorylation sites are actually all phosphorylated or if the simple availability of
all six sites is sufficient to account for the above results.
Despite phosphorylation, these rhodopsin molecules (R*-p) still have the
capacity for Tr activation albeit at a diminished rate (Chen et al., 1999b). Proper and
complete signaling termination of the R*-p molecule requires binding of visual
arrestin (SAg) (Kuhn et al., 1984; Xu et al., 1997). This protein belongs to the
arrestin family, in which members are involved with the inactivation, desensitization,
and/or turnover of various GPCR’s. Thus far, four members have been identified in
vertebrates: the aforementioned SAg, the cone-specific cone arrestin, and the
12
R
R
Pi
SAg
ATP
photon
ADP
11 cis-retinal
all trans-retinal
11 cis-retinal
all trans-retinal
ACTIVATION OF
TRANSDUCIN
Pi
RK
R
all trans-retinal
Pi
all trans-retinal
R
phtase
Opsin
SAg
*
*
*

toxic complex?
Figure 1.4. Rhodopsin inactivation. Rhodopsin (R) becomes activated when its
chromophore, 11 cis-retinal, absorbs a photon and is converted to all trans-retinal,
which induces a conformational change of the opsin component to the active form.
Activated rhodopsin (R*) then stimulates transducin molecules. In the first step of
inactivation rhodopsin kinase (RK) binds to and phosphorylates R*. (R* has a
diminished capacity to activate transducin molecules.) Arrestin (SAg) then binds to
R*, which precludes stimulation of additional transducin molecules. The all trans-
retinal is released from R*, which induces the dissociation of SAg from opsin. A yet
to be determined phosphatase then dephosphorylates opsin. The cycle is complete
when a newly regenerated 11 cis-retinal is linked to opsin, thereby reconstituting R.
Please note that the accumulation of the R*/SAg complexes has been shown to cause
retinal degeneration in both Drosophila and mouse models (Alloway et al., 2000;
Kiselev et al., 2000; Iakhine et al., 2004; Chen et al., 2006). One purpose of this
dissertation is to investigate the potential contribution of this complex in the Q344ter
mouse model (see Chapter 4). See text in this Introduction section for more
complete details on the rhodopsin inactivation process.
13
ubiquitously expressed β-arrestin1 and β-arrestin2. Interestingly, both β-arrestin
species are also expressed in rods, although their specific functions currently are
unknown (Nicolas-Leveque et al., 1999). Because of its key role in rhodopsin
inactivation, the abundant expression level of SAg in rods is essential to regulate the
potential activities from the enormous “pool” of rhodopsin molecules. Additionally,
several studies have demonstrated that the accumulation of the rhodopsin/SAg
complexes induce photoreceptor cell death (Alloway et al., 2000; Kiselev et al.,
2000; Chuang et al., 2004; Iakhine et al., 2004; Chen et al., 2006). Therefore, these
two factors (abundance and potential toxicity) delineate the importance of proper and
timely R*-p/SAg complex formation.
The current model for the R*-p/SAg complex formation involves a sequential
multisite binding between the two proteins (Gurevich and Benovic, 1993; Gurevich
et al., 1995; Vishnivetskiy et al., 1999; Gurevich and Gurevich, 2004; Vishnivetskiy
et al., 2004; Hanson and Gurevich, 2006). To increase the specificity of this
interaction, SAg has two “recognition” sites for the rhodopsin molecule: a
phosphorylation sensor that is located in the N-domain, and an activation sensor that
putatively is the interface of a specific interaction between the N-domain and C-
domain. Thus, when rhodopsin is both activated and phosphorylated, these two
sensor sites of SAg will recognize and bind to the R*-p molecule. This multibinding
event induces global conformational rearrangements within SAg by disrupting
intramolecular contacts (involving both electrostatic and hydrophobic interactions).
In this “active” conformational state, a polar core within SAg becomes exposed, and
14
SAg’s affinity for R*-p increases by a factor of 20. Furthermore, the binding of SAg
to R*-p completely terminates the signaling capacity of the R*-p molecule.
Referring back to rhodopsin inactivation, after the release of the all trans-
retinal, the cycle of rhodopsin renewal (Fig. 1.4) is completed upon the following
events: the dissociation of SAg; dephosphorylation by a presently unknown
phosphatase; and linkage to another 11 cis-retinal that was regenerated from a series
of RPE-mediated reactions known as the retinoid cycle.
Interestingly, perturbations to rhodopsin inactivation cause abnormally
prolonged responses. Deletion of either the RK enzyme or rhodopsin’s
phosphorylation sites not only lengthens the response time but also increases its
amplitude (Chen et al., 1995; Chen et al., 1999a). Because the time-to-peak
parameter of this response also is longer in the absence of rhodopsin
phosphorylation, this result shows that during the normal single photon response,
rhodopsin inactivation prevents the R* molecule from achieving maximum Tr
activation capacity. Furthermore, the absence of SAg molecules actually extends the
decay time of the response longer than that observed with the absence of rhodopsin
phosphorylation (Xu et al., 1997; Chen et al., 1999a).

1.3.3. Recovery to the dark-adapted state.
The restoration of the rod cell to the dark-adapted state involves additional
important steps. The lowered intracellular [Ca
2+
] (produced by the
phototransduction cascade) activates guanylyl cyclase activating proteins (GCAPs),
which in turn, stimulate guanylyl cyclases (GCs) to upregulate its cGMP synthesis.
15
Additionally, after Tr inactivation, the inhibitory interaction between PDE’s γ-
subunit and its catalytic subunit (either α or β) is re-established, and the low basal
activity of cGMP breakdown by PDE is restored. The combination of accelerated
GC activity and lowered PDE activity results in the overall rise in intracellular
[cGMP]. This condition allows more CNG channels to bind to cGMP molecules,
which causes more of these channels to reopen. The resulting influx of Na
+
and Ca
2+

(and Mg
2+
) decreases the rod cell’s transmembrane potential and eventually re-
establishes the depolarized “dark state”. In turn, this depolarization event causes the
aforementioned voltage-gated Ca
2+
channels at the rod cell’s synaptic terminus to
reopen. Consequently, this condition increases the release rate of glutamate-laden
vesicles at this cellular region and, ultimately, restores the rod cell to its “dark-
adapted state”.

1.4. Rhodopsin Trafficking.
Proper localization of the proteins involved in phototransduction is the result
of carefully regulated and directed trafficking events. Of particular interest to this
study is the proper trafficking of rhodopsin. It has been estimated that as much as 5
X 10
6
rhodopsin molecules are synthesized each day (Sung et al., 1994), which only
underscores the importance of regulated and efficient trafficking of this protein.
Rhodopsin transport is a multi-stage process (Fig. 1.5). It begins with the synthesis
and post-translational modifications, such as myristoylations and palmitoylations, of
rhodopsin in the rough endoplasmic reticulum and Golgi apparatus in the RIS (see
Fig. 1.2 for cell structure). Rhodopsin is released from the trans-Golgi network not
16

Figure 1.5. Rhodopsin transport within the polarized rod cell is a multi-stage
process. Rhodopsin synthesis transpires in the IS region, more specifically in the
myoid. Rhodopsin molecules bud off from the Golgi apparatus as large
pleiomorphic post-Golgi vesicles, or rhodopsin transport carriers (RTCs). These
rhodopsin bearing vesicles are transported to the basal body at the base of the
connecting cilium (CC). Whether they migrate to this site along microtubules or
microfilaments or both is a source of debate. After these RTCs fuse with the plasma
membrane, rhodopsin molecules move “up” the CC with concomitant evaginations
of the plasma membrane, which give rise to nascent rod disks. As the rhodopsin
molecules approach the distal end of the CC, these evaginations become larger. A
mature rod disk results from fusion between adjacent evaginations. Collectively,
these rod disks form the ROS structure. Displaced by newer disks, the older rod
disks travel towards the distal end of the ROS structure. There, mass phagocytosis
of the oldest rod disks by the retinal pigment epithelium (RPE) occurs. The QVAPA
domain within the C-terminus of rhodopsin serves as an essential trafficking signal
of this molecule. See text for further details. This figure was modified from Tai et
al. (1999). Cell 97 (7): p.877-87.
17
as single molecules but within the membranes of large pleiomorphic post-Golgi
vesicles (Deretic and Papermaster, 1991), with each of these vesicles containing as
much as 2000 rhodopsin molecules (Tam et al., 2000). These rhodopsin-containing
post-Golgi vesicles also are known as rhodopsin transport carriers (RTCs, termed by
Dr. Dusanka Deretic from U. of New Mexico School of Medicine). The formation
of RTCs involves several proteins that coordinate vesicle budding and fission. At
least three small GTPases have been shown to be involved with RTC budding: rab6,
rab11 (Deretic et al., 1996) and ARF-4 (Deretic et al., 2005). Of the three small
GTPases, thus far only ARF-4 has been demonstrated to have direct interaction with
rhodopsin, more specifically at the QVAPA domain of rhodopsin’s C-terminus
(Deretic et al., 2005).
After budding, the RTCs migrate toward the basal body region of the rod cell,
which is located at the base of the connecting cilium. The pathway by which RTCs
travel to the basal body is a source of controversy. One study has shown that
rhodopsin (again through its C-terminal domain) interacts with Tctex-1, a light chain
of the microtubule-dependent cytoplasmic motor protein, dynein (Tai et al., 1999).
However, the use of microtubule depolymerizing agents, such as nocodazole, did not
prevent rhodopsin trafficking to the ROS structure, while chemicals affecting F-
actin, such as cytochalasin D, disturbed ROS formation (Vaughan et al., 1989).
Once at the distal end of the RIS, several molecules and proteins [PIP
2
,
moesin, ezrin, actin, two other small GTPases (rac1 and rab8), etc.] regulate the
docking and fusion of the RTCs to the RIS plasma membrane near the base of the
18
connecting cilium (CC). The involvement of actin further substantiates RTC
transport along the microfilaments while in the RIS. At this stage, however, the
migration of rhodopsin molecules toward the distal end of the CC involves both
microtubules and microfilaments (Williams, 2002).
Concurrent to this mass migration, evaginations of the plasma membrane in
this cellular region induce the formation of nascent rod disks. These evaginations, in
turn, become larger as the rhodopsin molecules approach the distal end of the CC.
Ultimately, mature rod disks form as the result of fusion between adjacent
evaginations. Because no such membrane fusions occur between adjacent
evaginations in the cones, cone disks remain continuous with the plasma membrane
(see Fig. 1.2 for schematic drawing of OS structures). Collectively, these rod disks
shape the ROS structure. Older rod disks, in turn, are displaced towards the distal
end of the ROS structure by newer disks. When rod disks reach this distal end, they
are phagocytized by RPE cells. The life span of a rhodopsin molecule is about 10
days, and much energy by the rod cell is expended to the synthesis and degradation
of this molecule.

1.5. Drosophila Vision.
Even with the ancient divergence in the evolutionary scale between mammals
and invertebrates, discoveries within the Drosophila model significantly have
contributed to vision research. Illustratively, the original descriptions of a light-
dependent mechanism leading to retinal degeneration were first reported in
Drosophila (Alloway et al., 2000; Kiselev et al., 2000). As will be described in
19
Chapter 4, this mechanism was discovered to be conserved in mice. Therefore,
detailed descriptions of the Drosophila visual transduction mechanisms will assist
the understanding of this project.
1.5.1. Drosophila compound eye and phototransduction.
Unlike mammalian eyes, each Drosophila eye is actually a compound
structure composed of approximately 800 units, each termed an ommatidium (Fig.
1.6 A). These single eye units contain 20 cells, 8 of which are photoreceptor cells
(R1-R8). Of these photoreceptor cells, six (R1-R6) have rod-like properties and
remaining two (R7 and R8) have cone-like properties. These photoreceptor cells are
situated in each ommatidium with the two cone-like cells in the central region “on
top of one another” and with the six rod-like cells in the periphery. Therefore, most
cross sections of an ommatidium will reveal the R1-R6 cells and either a R7 or R8
cell (see cross section diagram in Fig. 1.6 A).
Each photoreceptor cell has two main regions: the cell body and the
rhabdomere, which is composed of approximately 30,000 units known as microvilli
(Fig. 1.6 B). The rhabdomere is the Drosophila equivalent of the ROS structure of
mammalian rod cells. However, instead of rod disks, the visual pigments and the
rest of the phototransduction machinery in Drosophila reside in each microvillus. In
the R1-R6 cells, the Drosophila phototransduction cascade (Fig. 1.6 C) is initiated
when the photopigment, also known as rhodopsin, absorbs a photon (most effective
wavelength at 480nm) and converts into the active form, metarhodopsin. This
excited molecule stimulates a G-protein (Gq) which then activates the effector
20
Figure 1.6. Drosophila vision. (A) Each compound Drosophila eye is composed of
approximately 800 simple eyes termed ommatidia. Each ommatidium contains 8
photoreceptor cells (R1-R8). R1-R6 cells are located in the periphery of each
ommatidium and have rod-like properties, while R7 and R8 cells are located “one on
top of the other” in the middle of the ommatidium and have cone-like properties.
(Adapted from the following website:
http://www.isa.org/InTechTemplate.cfm?Section=Article_Index1&template=/Conte
ntManagement/ContentDisplay.cfm&ContentID=53255.) (B) The Drosophila
photoreceptor cell has two main regions: the cell body and the rhabdomere, a
membranous stack of approximately 30,000 microvilli. The phototransduction
machineries are localized within each microvillus. The blue arrow denotes the
directions of both incoming light and nerve signals. (Adapted from the following
website: http://www.hhmi.swmed.edu/Labs/rr/individual/Ady/_private/intro.htm.)
(C) The phototransduction mechanism in the Drosophila system involves a typical
GPCR-mediated cascade in that an activated receptor (rhodopsin) stimulates a
heterotrimeric G-protein (Gq) that in turn leads to the upregulated activity of an
effector molecule (phospholipase C). The ultimate light response, however, is the
depolarization of the cell, as opposed to the hyperpolarization state in mammalian
photoreceptors. (D) Metarhodopsin in Drosophila can be converted to the inactive
form by the absorption of a photon of orange light (580 nm). The binding of a
Drosophila-specific arrestin (arrestin2) to metarhodopsin also prevents stimulation
of subsequent Gq molecules. This metarhodopsin/arrestin2 interaction is
21
independent of metarhodopsin phosphorylation, an event of unknown function.
Additionally, the endocytic activities of both arrestin2 and another Drosophila-
specific arrestin species (arrestin1) mediate rhodopsin turnover (Satoh and Ready,
2005; Orem et al., 2006). The mammalian counterparts in (C) and (D) are provided
for comparative purposes. [See text for more detailed explanations pertaining to all
subfigures.]
22

Figure 1.6. Drosophila vision.
23
molecule, phospholipase C (PLC). In turn, PLC hydrolyzes PIP
2
within the
microvillar membrane into IP
3
and diacylglycerol (DAG). Through a series of
events yet to be clearly defined, DAG (and perhaps IP
3
) causes both TRP and TRP-
like channels to open, which results in the subsequent influx of Ca
2+
. These events
cause the photoreceptor cell to become depolarized, as opposed to the
hyperpolarization outcome in mammalian photoreceptors.

1.5.2. Rhodopsin inactivation and turnover in Drosophila.
Another important feature within the Drosophila visual system is rhodopsin
inactivation (Fig. 1.6 D). Instead of the visual cycle that is necessary for the
regeneration of rhodopsin in the mammalian system, the Drosophila metarhodopsin
can be converted back to the inactive form by the absorption of a photon with a 580
nm wavelength (orange light). On the other hand (and similar to the mammalian
system), the binding of a Drosophila specific arrestin species, arrestin2 (Arr2), to
metarhodopsin also blocks stimulation of additional Gq molecules. Despite the
occurrence of metarhodopsin phosphorylation by a Drosophila-specific rhodopsin
kinase, this phosphorylation event is not required for the metarhodopsin/Arr2
interaction (Kiselev et al., 2000). Although the function of metarhodopsin
phosphorylation is presently unclear, persistent phosphorylation of metarhodopsin
induces retinal degeneration in flies (see Chapter 4, Kiselev et al., 2000). [Moreover,
the phosphorylation of Arr2 by a Ca
2+
-dependent kinase is necessary for the release
of Arr2 from metarhodopsin. Inhibition of the light-induced cell depolarization
event precludes this Arr2 phosphorylation, which in turn prevents the dissociation of
24
Arr2 from metarhodopsin. This condition also leads to retinal degeneration in
Drosophila (again see Chapter 4, Alloway et al., 2000; Orem and Dolph, 2002).]
Another Drosophila-specific visual trait concerns rhodopsin turnover (Fig.
1.6 D). Here, rhodopsin molecules are not phagocytized like those in the
mammalian system by RPE cells. Instead, these molecules are internalized into the
cell body, where they become degraded. Much of this endocytosis is mediated by
another Drosophila-specific arrestin species, arrestin1 (Satoh and Ready, 2005),
although another study showed that Arr2 also contributes (Orem et al., 2006). Thus,
the turnover of rhodopsin by these Drosophila-specific arrestin species is similar to
the turnover of other GPCRs that are mediated by the β-arrestins.

1.6. Thesis Outline.
Because of the complexity in (and importance to) rhodopsin transport, it is
not surprising that defects in the rod cell’s trafficking machinery of this protein lead
abnormal cellular structures and rod cell death. In particular, a cluster of genetic
mutations at rhodopsin’s C-terminus has been linked to a hereditary group of eye
diseases known as retinitis pigmentosa (RP). [Further explanation of RP will be
given in Chapter 3.] This dissertation investigates the consequences of the absence
of rhodopsin’s trafficking signal (the QVAPA domain), and the potential
mechanism(s) by which this genetic defect leads to rod cell death.
In Chapter 2, we used a previously generated transgenic mouse model that
expressed a truncated rhodopsin mutant missing the last 15 amino acids at its
carboxyl-end (S334ter). Although this mutation is not naturally occurring, the
25
transgenic S334ter mouse model enabled us to investigate the importance of the
QVAPA domain to proper rhodopsin trafficking, especially concerning the formation
of the ROS structure. In Chapter 3, we investigated the effects of a naturally
occurring rhodopsin mutant, Q344ter. We demonstrated that although retinal
degeneration occurs in the absence of light, light-exposure accelerates this
degeneration. Furthermore, to the best of my knowledge, we are the first to
demonstrate that mislocalized rhodopsin molecules are capable of light activation.
In Chapter 4, we show that the light-accelerated retinal degeneration involves
multiple pathways. In particular, we have provided further evidence that the R*-
p/SAg complex contributed to this light-dependent degeneration in the transgenic
Q344ter mouse model. Originally, this toxic effect of stabilized rhodopsin/SAg
complex to photoreceptor cells was described in the Drosophila model (Alloway et
al., 2000; Kiselev et al., 2000; Iakhine et al., 2004). Recent work with the transgenic
K296E rhodopsin mutant model in Dr. Jeannie Chen’s lab demonstrated that this
toxic complex is conserved in vertebrates (Chen et al., 2006). Together, all these
models reveal the pervasiveness of this mechanism to induce blindness.
In conclusion to this introductory chapter of my dissertation, “enlightenment”
came at three levels. One level was the lethal consequence of light-exposure to the
transgenic Q344ter mouse model. The second level was that the preservation of the
rod cells are important to the overall health of the retina, and therefore, to our sense
of sight, despite the narrow operating response range of these photoreceptor cells.
The third and last level is on a personal note in which I am quite thankful of my
26
learning experience and personal scientific “growth” as a scientist while being a
student in Dr. Jeannie Chen’s lab at The Keck School of Medicine of USC.
27
Chapter 2
The Carboxyl-Terminal Domain is Essential for Rhodopsin
Transport in Rod Photoreceptors
†

Francis Concepcion, Ana Mendez, Jeannie Chen*

The Mary D. Allen Laboratory for Vision Research, Beckman Macular Research
Center, Doheny Eye Institute & Departments of Ophthalmology and Cell and
Neurobiology, Keck School of Medicine of the University of Southern California, Los
Angeles, California 90089

Key Words: Rhodopsin Transport, Photoreceptors, Transgenic Mice, Retinal
Degeneration.

†
Permission was granted from Elsevier by Natalie David (Senior Rights Assistant) to
reprint the following article:
Vision Research (2002), Vol 42 (4). Francis Concepcion, Ana Mendez, Jeannie
Chen. "The Carboxyl-Terminal Domain is Essential for Rhodopsin Transport in Rod
Photoreceptors", pp. 417-426.
Modifications in this article were necessary for proper formatting to this dissertation.

*Correspondence to Jeannie Chen
Departments of Ophthalmology and Cell and Neurobiology
Keck School of Medicine of the University of Southern California
1333 San Pablo Street, Los Angeles, CA 90089-9112
Phone: (323) 442-6638
FAX: (323) 442-6655
E-mail: jeannie@hsc.usc.edu
28

Abstract

The role of the carboxyl-terminal domain in rhodopsin transport was investigated
using transgenic mice expressing a rhodopsin truncation mutant lacking the terminal
15 amino acids (S334ter). It was previously shown that S334ter translocates to the
outer segment in the presence of endogenous rhodopsin. We now show that in the
absence of endogenous rhodopsin S334ter mis-localizes to the plasma membrane and
fails to reconstitute outer segment structures. Surprisingly, this mis-localization does
not affect photoreceptor cell survival. These results provide further evidence on the
important role of the COOH-terminal domain in rhodopsin trafficking and
demonstrate an absolute requirement of this domain for correct vectorial transport of
rhodopsin.

Word count: 105.

29
2.1. Introduction.
The rod outer segment is an apical appendage of retinal rod photoreceptor
cells that is specialized for transducing the light signal. This specialization is
reflected by the extensive membranous disc stacks containing a near crystalline array
of the light-capturing molecule, rhodopsin. In each rod outer segment there are
approximately 5 x 10
7
molecules of rhodopsin embedded within the discs and the
plasma membrane. Rhodopsin comprises about 85% of the total protein in the rod
outer segment and is localized almost exclusively to this structure. Transducin,
phosphodiesterase and other proteins involved in phototransduction make up the
majority of other proteins found in the outer segment. How the proteins involved in
the signaling cascades are vectorially transported to this apical appendage of rod
cells is not well understood.
There are over a hundred naturally occurring mutations in the rhodopsin gene
that lead to autosomal dominant retinitis pigmentosa (Sohocki et al., 2001). Some of
these mutations are clustered around the highly conserved carboxyl-terminus of the
rhodopsin molecule, e.g., P347L, P347S, P347R, V345M, Q344ter and two
frameshift mutations fs341del (Horn et al., 1992). Expression of some of these
mutant proteins in COS-1 cells failed to reveal a biochemical defect (Sung et al.,
1991a; Kaushal and Khorana, 1994; Sung et al., 1994). However, the fact that these
mutations lead to retinal degeneration suggests that the carboxyl-terminal domain
participates in an important physiological process distinct from phototransduction.
30
A role of the carboxyl-terminal domain in rhodopsin transport was first
suggested by transgenic animal studies. Studies on transgenic mice expressing the
Q344ter (Sung et al., 1994), P347S (Li et al., 1996) and also transgenic rats
expressing S334ter (Green et al., 2000) in the presence of endogenous rhodopsin
showed extensive mis-localization of rhodopsin to the plasma membrane and inner
segments, where it is normally found in very low levels (Nir and Papermaster, 1983;
Hicks and Barnstable, 1987; Hicks et al., 1989). Analysis using specific antibodies
attributed this mis-localization exclusively to the mutant rhodopsins (Green et al.,
2000; Li et al., 1996; Sung et al., 1994). These observations suggest a defect in
transport or retention of these mutant rhodopsin molecules in the outer segment.
Nevertheless, in these studies the majority of the mutant rhodopsin molecules were
correctly localized to the outer segment along with the endogenous rhodopsin.
Therefore, it is uncertain whether the correct localization was due to co-transport of
mutant rhodopsin molecules with normal endogenous rhodopsin in post-Golgi
vesicles or whether there are other site(s) on the rhodopsin molecule distinct from the
carboxyl-terminus domain that also contain(s) signals for transport.
The in vivo role of the rhodopsin carboxyl-terminal domain has also been
investigated using transgenic Xenopus laevis. Tam et al. (2000) reported that the
terminal 44 amino acid cytoplasmic tail of rhodopsin redirects a GFP fusion protein
exclusively to the rod outer segment, but GFP fusion proteins bearing the naturally
occurring carboxyl-terminal mutations show GFP fluorescence at the outer segment,
plasma membrane and the inner segment. This experiment established that the
31
carboxyl-terminal domain contains a signal for transport to the outer segment, and
that it alone is sufficient to target a heterologous protein to the correct location.
The role of the carboxyl-terminus in rhodopsin transport is also supported by
recent in vitro evidence. Using a retinal cell-free system from frog retina, Deretic et
al. (1996) demonstrated that an antibody against the rhodopsin COOH-terminus
inhibits post-Golgi vesicle formation. A synthetic peptide corresponding to the
carboxyl-terminus of rhodopsin also inhibited rhodopsin trafficking in post-Golgi
vesicles, but peptides bearing the naturally occurring mutations, Q344ter, V345M,
and P347S, did not show this inhibition (Deretic et al., 1998). Interestingly, the
carboxyl-domain has been found to interact directly with Tctex-1, a dynein light
chain (Tai et al., 1999). This finding provides a mechanistic model by which
vesicles bearing rhodopsin molecules might be transported along the cytoskeleton
towards the apical surface of rod photoreceptor cells.
Together, the in vivo and in vitro evidence clearly implicate the carboxyl-
terminal domain in rhodopsin transport. However, since all of the in vivo
experiments were performed in the presence of normal endogenous rhodopsin, and
since all of the mutant rhodopsin molecules showed correct localization to the outer
segments, the question remains whether the rhodopsin carboxyl-terminus plays a
necessary and essential role. To address this question, we utilized a transgenic
mouse line that expressed a rhodopsin truncation mutant that lacks the carboxyl-
terminal 15 amino acid residues (S334ter). This transgene was introduced into the
rhodopsin +/+, +/- and -/- backgrounds to see whether rhodopsin transport can occur
32
in the absence of the carboxyl-terminal domain and in the absence of endogenous
rhodopsin.

2.2. Materials and Methods.
2.2.1. Generation of mouse lines.
All mice were treated in accordance with the ARVO Statement for the Use of
Animals in Ophthalmic and Vision Research. Transgenic mice expressing the
S334ter form of rhodopsin (Chen et al., 1995) or S338A rhodopsin (Mendez et al.,
2000a; Mendez et al., 2000b), a control line, were mated with wildtype or rhodopsin
knockout mice (rho-/-) (Lem et al., 1999) to generate mice that expressed the
transgene in the +/+, +/-, or -/- endogenous rhodopsin genetic background.
All mice were born and raised in constant darkness to avoid retinal
degeneration that might result from constitutive signaling caused by S334ter
rhodopsin (Chen et al., 1995).

2.2.2. Genotype analysis by PCR and Southern blotting.
Mice were genotyped by PCR amplification and by Southern blot analysis.
Genomic DNA was obtained from mouse-tail biopsy samples. The S334ter transgene
was detected with primers Rho2 (5’ TGGGAGATGACGACGCCTAA 3’) and Rho3
(5’ TGAGGGAGGGGTACAGATCC 3’), while S338A transgene was detected with
primers mRH7.2 (5’ GACGACGCCTCGGCCACCGTG 3’) and mRH5 (5’
GGAGCCTGCATGACCTCATCC 3’).
33
Transgenic mice bred to rhodopsin -/- mice were genotyped for the rod opsin
locus. PCR was performed to detect the presence of the rod opsin null allele by using
primers Rh1.1 (5’ GTGCCTGGAGTTGCGCTGTGGG 3’) and Neo3 (5’
CGGTGGATGTGGAATGTGTGCGAG 3’). To distinguish between hemizygous
(+/-) and homozygous (-/-) rhodopsin knockout mice, Southern blot analysis was
performed as described previously (Mendez et al., 2000b).

2.2.3. Determination of transgene expression level.
Transgene expression level was determined in each line by quantitating the
mutant-to-total transcript ratio in transgene-positive mice in the rhodopsin+/+
background at postnatal days 21-26 (P21-P26). Total RNA was isolated from
individual retinas using the Trizol Reagent (Invitrogen Corp., Carlsbad, CA), and
reverse transcription was performed to obtain single stranded cDNA. PCR
amplification of cDNA was performed with primers mRh4 (5’
GAGCTCTTCCATCTATAACCCGG 3’) and mRh6 (5’
GGCTGGAGCCACCTGGCTG 3’), which hybridize to identical sequences in the
endogenous and transgenic loci and, therefore, amplify both with equal efficiency.
Labeling of the PCR amplification product was carried out by addition of 10 µCi of
α-
32
P dCTP (3000 Ci/mmol) to the PCR reaction. PCR amplification products were
precipitated, washed, resuspended in dH
2
O, and divided into two equal aliquots. One
aliquot was digested with a restriction enzyme specific for the transgene (DdeI for
S334ter; EaeI for S338A), and the other one was mock digested. Both aliquots were
then loaded in a 12% Bis-Tris polyacrylamide gel (Invitrogen Corp., Carlsbad, CA).
34
The gel was dried, exposed and analyzed using the Storm 860 Phosphor Imager
software (Amersham Pharmacia Biotech Inc., Piscataway, NJ).

2.2.4. Western blot analysis.
Retinas were dissected out from dark-reared mice at postnatal days 21-26,
under infrared illumination. Individual retinas were homogenized in homogenization
buffer (80mM Tris-HCl, pH 8.0; 4mM MgCl
2
; protease inhibitor cocktail
(Boehringer Mannheim, Indianapolis, IN); and 0.5mM phenylmethylsulfonyl
fluoride) and incubated with DNase I (Boehringer Mannheim, Indianapolis, IN) for
30 min at room temperature. Laemmli buffer was added, and the indicated amounts
(in retina equivalents) were separated in a 12% Tris-glycine polyacrylamide gel
(Invitrogen Corp., Carlsbad, CA). Proteins were transferred onto a nitrocellulose
membrane and incubated with mAb R2-12N (a generous gift from Paul Hargrave,
University of Florida, Gainesville), and subsequently with goat anti-mouse IgG
conjugated to horseradish peroxidase (Invitrogen Corp., Carlsbad, CA).
Immunodetection was performed using the ECL system (Amersham Pharmacia
Biotech Inc., Piscataway, CA).

2.2.5.. Immunohistochemistry.
Before enucleation, the superior pole for each mouse eye was cauterized for
orientation. Eyecups from each transgenic line were prepared under infrared
illumination at two time points: P21-P26 and P45-P50. The eyecups were incubated
in fixative (4.0% paraformaldehyde, 0.5% glutaraldehyde in 0.1M cacodylate buffer,
35
pH 7.2) for 2 hours at room temperature, and subsequently embedded in 30% sucrose
in 0.1M cacodylate buffer, pH 7.2 for 14-18 hrs at 4
o
C. The eyecups were
hemisected, embedded in Tissue Tek
®
O.C.T. (Sakura Kinetek U.S.A. Inc.,
Torrance, CA), and quickly frozen using liquid N
2
. Using a Jung CM 3000 cryostat
machine (Leica Inc., Deerfield, IL), the tissues were sectioned at 10 µm thickness.
The sections were incubated for 1 hour in blocking solution (2.0% bovine serum
albumin (BSA), 0.3% Triton X-100, and 2% goat serum in phosphate buffer saline
(PBS)), and incubated with either 1:25 dilution of R2-12N or 1:200 dilution of 1D4
for 1 hr. Dilution buffer for primary antibody was 2.0% BSA, 2.0% goat serum in
PBS. Sections were washed for 3 x 5 min with plain dilution buffer and incubated
for 30-45 min with 1:100 dilution of FITC-conjugated rabbit anti-mouse IgG (Vector
Laboratories, Inc., Burlingame, CA). Sections were washed for 3 x 5 min with plain
dilution buffer and for 2 x 5 min with PBS. Sections were incubated for 5 min in
4.0% paraformaldehyde in PBS and washed for 2 x 5 min with PBS. A tiny drop of
Vectashield (Vector Laboratories Inc., Burlingame, CA) was placed on the sections,
which then were cover slipped. Prepared cryostat sections were viewed and analyzed
using a LSM 510 confocal microscope (Carl Zeiss Inc., Thornwood, NY).

2.2.6. Retinal morphometry.
Before enucleation, the superior pole was cauterized for orientation. Eyecups
from dark-reared mice were prepared under infrared light at two time points: P21-
P26 and P45-P50. The eyecups were fixed overnight at 4 °C in 1/2 Karnovsky buffer
(2.5% glutaraldehyde, 2.0% paraformaldehyde in 0.1M cacodylate buffer, pH 7.2)
36
during which the lenses were removed after the first 15 minutes. The fixed eyecups
were washed for 3 x 10 min with 0.1M cacodylate buffer, pH 7.2 and further fixed
with 1% osmium tetraoxide in 0.1M cacodylate buffer for 1.5-2.0 hrs at room
temperature. Eyecups were washed 2 x 10min with 0.1M cacodylate buffer, pH 7.2
and dehydrated in the following manner: 15 min each with 50%, 70%, 85%, and
95% ethanol; 3 x 10 minutes 100% ethanol; and 2 x 10 min 100% propylene oxide.
Infiltration with epon (28.7% w/w Epon 812, 14.4% w/w Epon 826, 4.8% w/w Epon
871, 46.4% w/w DDSA, 5.7% w/w NMA, and 3% BDMA) was performed as
follows: overnight incubation with 1:1 epon: propylene oxide; overnight incubation
with 2:1 epon: propylene oxide; and 6 hours to overnight incubation with 100%
epon. The eyecups were placed in molds filled with epon with the optic nerve in
back and the superior pole pointing at 12 o’clock. The samples were baked for 3-4
days at 55 °C.
Once hardened, the epon-embedded eyes were cut into 1 µm sections and
stained with Richardson stain (0.5% methylene blue, 0.5% Azine II, and 0.5% borax
in dH
2
O). Analysis by retinal morphometry was accomplished by counting the outer
nuclear layer thickness in four distinct regions of the retina using the SlideBook 3.0
software (Intelligent Imaging Innovations, Inc., Denver, CO). Comparisons were
made between mice expressing the S334ter rhodopsin and age-matched controls
expressing no transgene in an equivalent endogenous rhodopsin background.

37
2.3. Results.
2.3.1. Transgene expression levels and rod outer segment formation.

Transgenic mice expressing S334ter (Fig. 2.1 A) were bred with wildtype and
rhodopsin knockout mice to obtain mice that expressed the S334ter transgene in the
rhodopsin +/+, +/- or -/- background (S334ter
rho+/+
, S334ter
rho+/-
, or S334ter
rho-/-
,
respectively). A transgenic line expressing full-length rhodopsin bearing an Ala
substitution, S338A, was chosen as a control based on its similar transgene mRNA
expression levels as the S334ter line (Fig. 2.1 B). Since constitutive signaling
through the phototransduction cascade has been implicated as a stimulus for
photoreceptor cell death (Chen et al., 1999a; Chen et al., 1999b) and to avoid this
condition caused by S334ter (Chen et al., 1995), all mice were raised in constant
darkness. Levels of transgene expression were determined by quantitating the
amount of transgene transcript in each line at postnatal day 21-26. Rod opsin mRNA
was amplified with primers that amplify both the endogenous and mutant rod opsin
genes and, subsequently, was digested with restriction enzymes specific for the
transgene (see Methods). Transgene transcript in S334ter and S338A mice was
determined to be ~10% of the total in the rhodopsin +/+ background.
Although expressed at only 10% level, the S338A transgene was able to fully
reconstitute the rod outer segments in the rhodopsin -/- background (Fig. 2.2,
compare B to A). Immunolocalization of rhodopsin using a monoclonal antibody
recognizing the rhodopsin amino terminus (R2-12N) showed correct rhodopsin
38

Figure 2.1. Comparison of transgene expression in S334ter and S338A
transgenic lines. A. The 11 kb Bam HI fragment of mouse rhodopsin gene. The
S334ter rhodopsin construct was generated by introducing a stop codon at position
334 (Chen et al., 1995). A transgenic line in which Ser 338 was substituted to Ala
was selected as the control line among a collection of rhodopsin mutants at the
phosphorylation sites (Mendez et al., 2000a) because it best matched S334ter level of
expression. B. Determination of the transgene-to-total rhodopsin transcript ratio, for
S334ter (left panel) and S338A rhodopsin (right panel). Total rhodopsin transcript
from transgenic retinas was amplified by RT-PCR and divided into two equal
fractions. One was loaded directly (left lane), the other was digested with a
restriction enzyme specific for the transgene and then loaded (right lane) into a 12%
polyacrylamide gel. Transgene level of expression was determined to be 10% in both
cases.
39

Figure 2.2. Expression of the control rhodopsin transgene (S338A) restores rod
outer segment formation in rhodopsin -/- mice. Shown are 1 µm-thick retinal
sections from rhodopsin -/- mice (A) or S338A
rho-/-
mice (B), at 45 postnatal days.
Immunolocalization of rhodopsin with mAb R2-12N shows localization of rhodopsin
to rod outer segments (C, dark field; D, bright field).
40
localization to the rod outer segments (Figs. 2.2, C and D), similar to the wildtype
control (Fig. 2.4 A). Therefore, S338A rhodopsin contains all the necessary
elements for correct rhodopsin transport and localization. This is consistent with our
previous finding that mutations at any of the Ser and Thr sites did not affect
rhodopsin transport (Mendez et al., 2000 a & b).

2.3.2. Transport of S334ter in the rhodopsin +/+ and the rhodopsin -/- genetic
background.
Retinal morphology of mice expressing S334ter in the rhodopsin +/+, +/- and
-/- background was examined to see whether S334ter was able to contribute to rod
outer segment formation. In contrast to S338A, rhodopsin-/- mice expressing
Ser334ter transcript to 10% of the normal rhodopsin mRNA levels failed to form rod
outer segments. As can be seen in Fig. 2.3, the overall retinal morphology was not
affected by the presence of S334ter transgene in any rhodopsin background. In the
rhodopsin +/+ or +/- background, no differences were observed at the outer nuclear
layer thickness or at the length of the rod outer segments between transgene-positive
or negative control mice. In the rhodopsin-/- background, no outer segment
formation was observed either in the absence or presence of the transgene. These
results indicate that the rhodopsin carboxyl-terminus is required for heterologous
rhodopsins to lead to rod outer segment formation.
Expression of the truncated rhodopsin in S334ter
rho-/-
mice was confirmed by
immunofluorescence staining using monoclonal antibody R2-12N (Fig. 2.4). In the
41

Figure 2.3. Retinal morphology of Ser334ter transgene-positive versus negative
control mice in the rhodopsin +/+, +/- and -/- background. Shown are 1 µm
retinal sections from rhodopsin +/+, +/- and -/- mice at 45-50 postnatal days (left
row), or corresponding littermates expressing the S334ter transgene (S334ter
rho+/+
,
S334ter
rho+/-
and S334ter
rho-/-
, right row). No differences in overall retinal
morphology were observed between transgene-positive mice and littermate controls.
42

Figure 2.4. In the absence of endogenous rhodopsin, Ser334ter rhodopsin
localizes to the outer nuclear layer and inner segments of photoreceptors.
Immunolocalization of rhodopsin with mAb R2-12N in frozen retinal sections from
21-26 day-old wildtype (A), Ser334ter
rho+/+
(B), and Ser334ter
rho-/-
(C). R2-12N mAb
recognizes residues 2-12 from the rhodopsin NH
2
-terminus, and therefore recognizes
both the endogenous and transgenic rhodopsins. In the rhodopsin -/- background,
S334ter fails to restore rod outer segments and is randomly distributed in the inner
segments (IS) and outer nuclear layer (ONL). Note also the increased signal at the
outer nuclear layer and inner segments in Ser334ter
rho+/+
sections.
43
rhodopsin-/- background where rod outer segments were absent, S334ter was mis-
localized to the outer nuclear layer and the inner segments (Fig. 2.4 C). Similar mis-
localization of S334ter in the rhodopsin +/+ background was also suggested by the
R2-12N staining pattern when compared to the wildtype control (Fig. 2.4, compare B
to A) with the signal in this case being additive from both S334ter and endogenous
rhodopsins. Taken together, these results show that in the absence of the carboxyl-
terminus and the absence of endogenous rhodopsin, S334ter rhodopsin molecules are
randomly distributed in the rod cell.
Expression of the truncated rhodopsin protein in transgenic retinas was also
confirmed by Western blot analysis (Fig. 2.5). Both the endogenous and the
truncated S334ter rhodopsin molecules were detected in the S334ter
rho+/+
sample,
whereas only the truncated rhodopsin was present in the S334ter
rho-/-
sample. As
expected, the control S338A rhodopsin was indistinguishable in size from the
wildtype rhodopsin. Despite their similar levels of transgene transcript, S334ter and
S338A transgenic lines showed a strikingly different amount of mutant protein. As
shown in Fig. 2.5, the amount of rhodopsin in S334ter
rho-/-
is dramatically lower than
the amount of rhodopsin present in the S338A
rho-/-
retina. This difference is likely
due to a lack of rod outer segment disks in the S334ter
rho-/-
retina, which precludes
the accommodation and accumulation of transported rhodopsin molecules. In the
absence of this structure, S334ter is likely targeted for degradation, resulting in
lowered protein concentration.
44

Figure 2.5. Western blot analysis of Ser334ter and S338A rhodopsin in transgenic
lines. The same amount of retinal homogenate (8 x 10
-5
retina equivalents) from
Ser334ter
rho+/+
and a wildtype littermate control were compared by 12% SDS-PAGE
after incubation with R2-12N mAb. Ser334ter rhodopsin is detected as a faster
migrating species, given its 15 amino acids difference in size. More sample (3.2 x
10
-4
retinal equivalents) was loaded from Ser334ter
rho-/-
and S338A
rho-/-
retinas in the
same gel to compare rhodopsin transgene expression. Although transgenic lines with
similar transcript expression levels were selected, there is a significant difference in
terms of protein expression. This difference can be explained by the presence of a
ROS layer in the S338A
rho-/-
retina, and its absence in the S334ter
rho-/-
retina.
45
Previous evidence indicate that some S334ter protein do traffic to the outer
segment when endogenous rhodopsin is present. This observation is based on
functional studies of S334ter in rod outer segments using single cell recordings and
immunoblot analysis of rhodopsin from isolated rod outer segment preparations from
transgenic mice expressing S334ter in the rhodopsin +/+ background (Chen et al.,
1995). The result that mutant rhodopsin lacking the correct trafficking signal is able
to transport to the rod outer segment in the presence of endogenous rhodopsin is
consistent with published studies from other laboratories (Green et al., 2000; Li et
al., 1996; Sung et al., 1994). Together, these results indicate that mutant rhodopsins
lacking the correct trafficking signal can be co-transported to the rod outer segment
along with normal rhodopsin. The mis-localization and the lack of outer segment
structures in the S334ter
rho-/-
retinas demonstrate that the last 15 amino acids of the
carboxyl-terminal domain contain a signal that is essential for rhodopsin transport.
2.3.3. Influence of S334ter on translocation of endogenous rhodopsin.
Because endogenous rhodopsin does affect the translocation of S334ter, we
explored the possibility that the reverse might also occur - that the presence of
S334ter causes improper localization of endogenous rhodopsin. Retinal sections
from S334ter
rho+/+
and age-matched non-transgenic control mice were stained with
the monoclonal antibody 1D4 that recognized the carboxyl-terminal eight amino
acids of rhodopsin (Hodges et al., 1988), and is therefore specific for the endogenous
rhodopsin (Fig. 2.6). In wildtype retina rhodopsin was detected predominantly in the
rod outer segment as expected. Light staining in the outer nuclear layer and a lack of
46

Figure 2.6. Endogenous rhodopsin mis-localizes to the inner segments in the
presence of Ser334ter rhodopsin. Retinal sections were incubated with mAb 1D4,
which recognizes rhodopsin last eight carboxyl-terminal amino acids. MAb 1D4
does not recognize the Ser334ter truncated rhodopsin, and is therefore specific for
the endogenous rhodopsin. A & C, S334ter
rho+/+
; B & D, wildtype. A & B, 40X
magnification; C & D, 126X magnification. At each magnification, the same
exposure time was used to take the pictures. Note the brighter signal in the outer
nuclear layer (ONL) and rod inner segment (RIS) of the S334ter
rho+/+
section (A)
than of the wildtype control (B). Arrows in C point to photoreceptor inner segments
that were immunoreactive to mAb 1D4. These results indicate that the presence of
Ser334ter causes the mis-localization and/or retention of the endogenous rhodopsin
in the ONL and IS to some extent.
47
staining in the inner segment indicate that rhodopsin is present in low amounts in the
outer nuclear layer and largely absent in the inner segment (Fig. 2.6 A). In
S334ter
rho+/+
retinas strong staining was also seen in the rod outer segment,
indicating that the majority of wildtype rhodopsin resides in this region. However,
increased signal was seen in the outer nuclear layer and the inner segment when
compared to the corresponding layers within the wildtype retina. This difference is
highlighted in Figs. 2.6, C & D, which show a higher magnification of the inner
segment and outer segment layers. Here, rhodopsin reactivity at the inner segment is
absent in the control retina (Fig. 2.6 D) but apparent in the S334ter
rho+/+
retina (Fig.
2.6 C, arrowheads). Together, these observations show that the presence of S334ter
does affect normal trafficking of endogenous rhodopsin albeit to a relatively small
degree.

2.3.4. S334ter mis-localization does not accelerate retinal degeneration.
Outer nuclear thickness was measured at four different retinal regions across
the span of the retina to compare the effect of S334ter mis-localization on
photoreceptor cell survival in rhodopsin +/+ and -/- backgrounds at two different
ages (Fig. 2.7). Measurements along the span of the retina were necessary because
the rhodopsin promoter tends to exhibit a gradient of expression, with the superior
pole of the retina showing a stronger expression pattern (Lem et al., 1991).
Consistent with previous reports, retinas from rhodopsin -/- mice showed a
progressive loss of photoreceptors as a function of age (compare the left panels with
the right panels) (Humphries et al., 1997; Lem et al., 1999). Measured outer nuclear
layer thickness from mice expressing S334ter (filled bars) were the same as the
48

Figure 2.7. Ser334ter rhodopsin expression does not lead to cell death. Retinal
outer nuclear layer (ONL) thickness was measured at 3.5 or 7 postnatal weeks in
Ser334ter-expressing mice (black bars) and negative littermate controls (white bars)
in the rhodopsin +/+ (left panels) or -/- (right panels) genetic background. ONL
thickness was measured at four different regions across a vertical section of the
retina at or near the optic nerve (A-D, retinal diagram; S = superior, I = inferior).
ONL thickness, expressed in µm, is the average of N = 4 to 6 determinations per
genotype per time point. Error bars represent the standard deviation.
49
corresponding rhodopsin +/+ and -/- controls (white bars) along the span of the retina
at both time points. These results demonstrate that rhodopsin mis-localization in
itself is not a strong stimulus for photoreceptor cell death.
2.4. Discussion.
Recent evidence from both in vitro and in vivo studies clearly demonstrates a
direct role of rhodopsin’s carboxyl-terminus in its vectorial transport. However, the
in vivo studies using transgenic animals do not address whether the carboxyl-
terminal domain is absolutely required for transport because the expressed mutant
rhodopsins do traffic to the outer segment in the presence of endogenous rhodopsin.
Is correct transport of mutant rhodopsin due to trafficking signals residing in other
domains distinct from the carboxyl-terminus? Or, are the mutant rhodopsin
molecules co-transported with the endogenous wildtype rhodopsin? We expressed a
carboxyl-terminal truncation mutant, S334ter, in the presence and absence of
endogenous rhodopsin to address these questions.
Normal retinal architecture can be obtained by expressing a normal human
rhodopsin transgene in the rhodopsin knockout mice (McNally et al., 1999). Our
result from S338A
rho-/-
also established that 10% expression level of a rhodopsin
phosphorylation site mutant transgene could restore normal retinal morphology in
rhodopsin knockout mice. In contrast, S334ter failed to restore an outer segment in
the rhodopsin -/- retina. Because rhodopsin also is required for outer segment
formation from a structural point of view, the lack of outer segments in S334ter
rho-/-
50
retinas might be due to its inability to maintain a stable disc structure. However, the
COOH-domain is cytoplasmic and highly dynamic (Langen et al., 1999) and
therefore is unlikely to contribute to structural stability. Rather, the lack of outer
segment formation is most likely due to transport defects of S334ter resulting from
the lack of the required trafficking signal. Thus, because rhodopsin-bearing vesicles
are not properly transported to the apical surface, the outer segment fails to form.
Recently, Tam et al. (2000) showed that the carboxyl-terminal 44 amino acid tail of
rhodopsin, when fused to GFP, targets this fusion protein exclusively to the rod outer
segment of transgenic frog retina. Our observation that S334ter rhodopsin fails to be
transported, together with the carboxyl-terminus alone being sufficient to direct
transport to the outer segment in the transgenic frog experiment, provides strong
evidence that the carboxyl-terminal domain is necessary and sufficient for rhodopsin
transport, and that there are no other trafficking signals residing in other domains
that might serve redundant functions.
Why did the previous transgenic animal studies show the mutant rhodopsins
to be localized to the outer segment? Our results suggest that the mutant rhodopsin
molecules are co-transported with the endogenous rhodopsin. The post-Golgi
vesicles containing rhodopsin are ~300 nm in diameter (Deretic and Papermaster,
1991), and it has been estimated that these vesicles could contain ~2,000 rhodopsin
molecules (Tam et al., 2000). When wildtype rhodopsin is present in the majority or
even at 50%, it can be expected that each vesicle will contain sufficient targeting
signals from the wildtype protein. By this mechanism, the mutant proteins follow
51
the bulk flow of normal rhodopsin to the outer segment (Fig. 2.8). The same
mechanism may also explain why S334ter affected localization of a small amount of
endogenous rhodopsin to the inner segment and outer nuclear layer. It is possible
that an over-threshold incorporation of the truncated rhodopsin in some vesicles
might have led to failed transport of both S334ter and endogenous rhodopsins to
these regions.
Failed rhodopsin transport may be a contributing factor to the pathogenesis of
rod photoreceptors in retinitis pigmentosa or macular degeneration (Sung and Tai,
2000). To see if this is the case, we performed careful morphometric measurements
on the outer nuclear layer thickness to see whether S334ter expression affected
photoreceptor cell survival in rhodopsin +/+ and -/- retinas. Surprisingly, no
difference in retinal thickness was observed between retinas expressing S334ter and
the corresponding control retinas at either an early time point (3 weeks), before
noticeable thinning of the outer nuclear layer was observed in the rhodopsin -/-
retina, or at a later time point (7 weeks), where the outer nuclear layer of the
rhodopsin -/- retina was halved in thickness. This is in contrast to the finding of
Green et al. (2000) who found that 10% expression level of the same murine S334ter
construct in transgenic rats caused retinal degeneration both in light-reared and dark-
reared animals. The discrepancy between the two different results may be due to a
difference in animal model used or difference in environmental factors such as
rearing conditions.
52

Figure 2.8. Schematic model showing Ser334ter and endogenous rhodopsin
being co-transported in post-Golgi vesicles. Rhodopsin is synthesized at the rough
endoplasmic reticulum in the proximal inner segment of rod photoreceptors. In its
journey to rod outer segments, where phototransduction takes place, it is first
transported in post-Golgi membrane vesicles to the base of the connecting cilium.
These 300 nm diameter-vesicles contain up to 2000 rhodopsin molecules. Ser334ter
and endogenous rhodopsin could therefore coexist in these rhodopsin-laden vesicles.
Given that the rhodopsin carboxyl terminal domain has been involved in the sorting
of these vesicles through the inner segment, co-localization of both rhodopsins
within the same vesicles would explain both the presence of S334ter at rod outer
segments in the rhodopsin +/+ background, and the mis-localization of endogenous
rhodopsin at the inner segments observed to some degree in the presence of the
truncated rhodopsin.
53
Rhodopsin mutations at the carboxyl-terminus cause a dominantly inherited
form of retinal degeneration in humans. Therefore, only one affected allele is
sufficient to cause the disease phenotype, so it can be expected that in human
disease, mutant rhodopsins will be expressed at 50% of the total rhodopsin
population. The lack of effect by S334ter in our studies may be due to the relatively
low expression level of this transgene. Doubling the gene dosage by breeding the
S334ter transgene to homozygosity should bring the expression level to 20%, which
is more in line with the expression levels in human disease. Future experiments will
be conducted to see whether 20% expression of S334ter will have an effect on
photoreceptor cell survival.

Acknowledgements
This work was supported by NIH grant EY12155 and the Beckman Macular
Research Center (JC). We thank Dr. Janis Lem for providing the rhodopsin
knockout mice; Dr. Paul Hargrave for providing the rhodopsin monoclonal antibody
R2-12N; and Dr. Robert Molday for providing the rhodopsin monoclonal antibody
1D4. We also thank the Specialized Imaging Core of the Doheny Eye Institute (NEI
grant Ey03040) for their technical support and guidance.
54
Chapter 3
The Q344ter Rhodopsin Mutant (Q344ter) Causes Mislocalization of
Rhodopsin Molecules that are Capable of Light-Activation

3.1. Introduction.
Retinitis pigmentosa (RP) comprises a group of inherited retinal disorders
typified by initial night blindness and a progressive loss of peripheral vision which
eventually compromises visual acuity and culminates into total blindness. This
hereditary disease is diagnosed by a reduction in rod sensitivity, as shown by
electroretinogram (ERG) responses, and its hallmark, the presence of pigmented
deposits in the retina. In most cases, RP is initiated by the death of rod
photoreceptors, but its progression eventually affects cones, leading to total vision
loss. Moreover, thus far, all observed photoreceptor cell death occurs through
apoptosis, or programmed cell death (Portera-Cailliau et al., 1994; Wenzel et al.,
2005). Epidemiological studies have revealed that RP is heterogeneous both
genetically and clinically, and afflicts around 1 in every 3500 to 5000 persons
worldwide (Rivolta et al., 2002; Hamel, 2006). The prevalence of RP can be
attributed to the multitude of mutations in numerous genes that have been linked to
this disease (RetNet, http://www.sph.uth.tmc.edu/RetNet/, University of Texas-
Houston Health Science Center). Accordingly, RP has been categorized by the
associated mode of inheritance: autosomal dominant (ADRP), 15 – 20%; autosomal
55
recessive (ARRP), 20 – 25; X-linked (XLRP), 10 -15%; unclassified, 40 – 55%
(including mitochondrial DNA and rare digenic mutations) (Wang et al., 2005).
Although some RP-related mutations occur in the retinal pigment epithelium
(RPE) cells, the majority of genetic defects causing RP are rod photoreceptor-
specific, affecting proteins involved in the rod phototransduction cascade, rod outer
segment (ROS) formation and structure, vectorial intracellular trafficking, etc. Over
100 different mutations in the rhodopsin (or rod opsin) gene alone have been linked
to RP. Moreover, nearly all RP-related rhodopsin mutations are autosomal dominant
and collectively have accounted for approximately 30% of all ADRP cases (Gregory-
Evans and Bhattacharya, 1998; Sohocki et al., 2001; Wang et al., 2005; Hamel,
2006). Based on observations when expressed in cultured mammalian cells (293S
cells - Sung et al., 1991b; COS-1 cells - Kaushal and Khorana, 1994), ADRP-related
rhodopsin mutations were classified into two main types: Class I (15%) and Class II
(85%). Interestingly, Class I mutants have no obvious defective traits, for they
closely resembled wild-type (WT) rhodopsin in terms of expression levels,
regeneration by binding to 11-cis retinal, and localization to the plasma membrane.
On the other hand, Class II mutants have characteristics distinct from WT rhodopsin:
their expression levels were markedly lowered; they failed to or poorly regenerated
with 11-cis retinal; and in varying degrees they were retained in the endoplasmic
reticulum (ER). These empirical properties were attributed to protein misfolding
(Sung et al., 1991b; Sung et al., 1993; Kaushal and Khorana, 1994; Mendes et al.,
2005). Additionally, due to these defect variations of Class II mutants, Sung et al.
56
(1991) further subdivided this class into IIa and IIb, while Kaushal and Khorana
(1994) actually separated these mutants into Class II and Class III categories. [More
recently, Mendes et al. (2005) updated the classification of disease-causing
rhodopsin mutations. Class I mutant criteria generally remained intact, while Class
II mutants were further subdivided on the basis of their biochemical and cellular
defects observed in cultured cells, mouse models, and clinical patients.]
The lack of significant biochemical abnormalities in Class I mutants when
expressed in cultured mammalian cells indicates that these rhodopsin mutants are
properly folded and capable of forming a light-absorbing pigment (Sung et al.,
1991b; Sung et al., 1993; Kaushal and Khorana, 1994; Mendes et al., 2005). Thus,
in this type of mutant, another aspect involving rhodopsin must lead to rod
photoreceptor death. Further investigations have revealed that the majority of Class I
mutants are clustered at the rhodopsin carboxyl-terminus (C-terminus), which is
essential not only for rhodopsin’s proper trafficking to the ROS structure (Sung et
al., 1994; Deretic et al., 1996; Li et al., 1996; Deretic et al., 1998; Tai et al., 1999;
Sung and Tai, 2000; Tam et al., 2000; Deretic et al., 2005) but also for ROS
formation itself (Concepcion et al., 2002; Shi et al., 2004). Therefore, photoreceptor
death by Class I rhodopsin mutants is the consequence of defective rhodopsin
polarized trafficking in rod cells. [As for the lack of differences between Class I
mutants and WT rhodopsin when ectopically expressed in cultured mammalian cells,
two speculative reasons are: (i) the proper folding of these mutants, which preserves
their inherent affinity for the plasma membrane; and (ii) the absence of any
57
specialized compartment for these rhodopsin molecules in the cultured cells, i.e. the
ROS structure. Thus, any mutations in the C-terminal targeting signal would not
perturb the “default” plasma membrane localization of the rhodopsin species in these
cultured cells.]
The importance of preserving rhodopsin’s trafficking signal is illustrated by
the Q344ter rhodopsin mutation. This naturally occurring genetic defect is a Class I
member that causes a severe form of ADRP. In the Q344ter rhodopsin mutant
(Q344ter), codon 344, which normally encodes for glutamine, is converted into an
early stop codon, thereby resulting in the absence of the QVAPA domain. These
five amino acids have been shown to be the minimal sorting signal for the proper
budding and trafficking of rhodopsin-bearing transport carriers (RTCs) in a retinal
cell-free assay (Deretic et al., 1996; Deretic et al., 1998). Thus, any disruption to this
sorting signal, including its deletion, leads to rhodopsin mistrafficking (and even
inhibition of ROS formation in the complete absence of this trafficking motif). [It
must also be noted that additional amino acids to rhodopsin’s C-terminus
(Ter349Glu) has been linked to a severe form of retinal degeneration (Bessant et al.,
1999). Although the QVAPA domain is still present in this mutant, the putative
extra 51 amino acids probably disrupts the proper vectorial trafficking of rhodopsin
by obstructing its accessibility to other proteins involved in normal rhodopsin
transport. This add-on rhodopsin mutant, therefore, only emphasizes the importance
of preserving the overall integrity of its C-terminal domain.]
58
Previously, Sung et al. (1994) eloquently revealed several intrinsic properties
of the Q344ter rhodopsin mutant (Q344ter), which were quite comparable to WT
rhodopsin properties. Biochemically, no significant differences were observed when
comparing Q344ter and WT rhodopsin in their interactions with transducin (Tr; the
G-protein in the rod phototransduction cascade) and rhodopsin kinase [RK; the
kinase which phosphorylates activated rhodopsin (Kuhn and Wilden, 1987; Chen et
al., 1999a)]. WT rhodopsin-like sub-cellular localization to the plasma membrane
was confirmed when Q344ter was transfected in 293S cells. Additionally, when
transgenic mice expressing Q344ter in rod cells were analyzed, single rod cell
recordings of the membrane current at the ROS revealed that the electrophysiological
properties of Q344ter transgenic rods generally were similar to WT rods. However,
some minor discrepancies were observed. Compared to WT rods, the responses to
dim light flashes of Q344ter expressing rods: (a) were slightly more sensitive; (b)
had a slightly longer time to amplitude peak; (c) were slightly longer in duration; and
(d) showed a greater variation in exponential decay constant. Nevertheless, these
slight perturbations in the electrophysiological properties of Q344ter transgenic rods
were unlikely to account for the fast-progressing retinal degeneration observed in
ADRP patients expressing this truncated rhodopsin mutant.
The most distinct feature in the retinas of Q344ter transgenic mice was
rhodopsin mistrafficking (Sung et al., 1994). Here, rhodopsin molecules were not
only observed in the ROS but also mislocalized in the rod inner segment (RIS) and
outer nuclear layer (ONL). As expected, no such rhodopsin mislocalization was
59
observed in control retinas from nontransgenic mice. Because each rod
photoreceptor normally synthesizes approximately 5 x 10
6
rhodopsin molecules per
day, the trafficking of rhodopsin-laden vesicles to the ROS must be efficient, and any
disturbance in the vectorial transport of this protein, such as the absence of the
trafficking signal QVAPA, will result in the undesired accumulation of rhodopsin in
“un-natural” compartments of the rod cell. Therefore, it is most likely rhodopsin
mistrafficking (and subsequent mislocalization) that induces retinal degeneration in
patients inheriting the Q344ter rhodopsin mutation.
We previously elaborated on the importance of the QVAPA domain for
proper rhodopsin trafficking. When expressing another truncated rhodopsin mutant
(S334ter, which is missing the last 15 amino acids at the C-terminus) in transgenic
mice in the absence of endogenous rhodopsin (S334ter
rho-/-
), S334ter was observed
only in the RIS and ONL compartments, and the conspicuous absence of ROS
formation was noted. Additionally, we also showed that S334ter caused the
mislocalization of WT rhodopsin molecules when the two rhodopsin species were
co-expressed (S334ter
rho+/+
) (Concepcion et al., 2002). We further characterized the
rhodopsin sorting motif in a recent study, in which short-wave cone opsin (S-opsin)
was ectopically expressed in mouse rod photoreceptors in the absence of endogenous
rhodopsin (S-opsin
rho-/-
) (Shi et al., 2004). Strikingly, in these mouse retinas, S-opsin
was able to reconstitute ROS-like structures. Because the last five amino acids at S-
opsin’s C-terminus are KVGPH, we were able to narrow down rhodopsin’s
trafficking motif to VXPX. This particular motif together with the membrane
60
association comprises the essential requirements for directing proper rhodopsin
vectorial transport (Tam et al., 2000; Shi et al., 2004).
Interestingly, the aforementioned transgenic S334ter mouse model did not
show any significant retinal degeneration, which could be explained by the low
transgene expression level (10% of total rhodopsin content, Concepcion et al., 2002).
Since our transgenic Q344ter mouse line has a higher transgene expression level
compared to the transgenic S334ter mice (24% versus 10%), we expected to observe
retinal degeneration, which would confirm our transgenic Q344ter mice serving as
an ADRP mouse model that expresses a naturally occurring rhodopsin mutation
(unlike S334ter). Furthermore, as mentioned previously, RP is heterogeneous not
only genetically but also phenotypically. Therefore, patients with the same RP-
related mutation may show variations in disease progression at comparable ages
(Rivolta et al., 2002). This discrepancy indicates that environmental conditions also
contribute to the severity of ADRP (but was not addressed in the previous study
involving transgenic Q344ter mice). Light-exposure is one such environmental
candidate that we investigated for potential exacerbation of ongoing retinal
degeneration in our transgenic Q344ter mouse line. Additionally, because Q344ter
retains its six potential phosphorylation sites, unlike S334ter, we have provided
evidence for the light-activation of mislocalized rhodopsin through experiments
involving light-dependent rhodopsin phosphorylation. By revealing this capability
of mislocalized rhodopsin molecules, we may have provided a key step towards
discovering a potential cell signaling cascade that triggers rod cell death. Ultimately,
61
we hope that the results from this study “shed light” onto the development of
therapies to alleviate, or even cure, this blinding disease in patients expressing this
rhodopsin mutant.

3.2. Materials and Methods.
3.2.1. Generation of transgenic Q344ter mice and genotype analysis by PCR and
Southern blotting.
The Q344ter rod opsin mutation (along with two silent mutations which
generated an AvrII restriction site designed for genotyping purposes) was introduced
into an 11 kb BamHI-flanked genomic clone of the murine opsin gene (Zack et al.,
1991) that was inserted into the pBluescript KS II vector (Stratagene, La Jolla, CA;
Fig. 3.1). These mutations were introduced by PCR-mediated mutagenesis in a
three-step process. (Please note the following concerning the involved primers:
underlined nucleotides signify mutation sites, and the numbers 5940 and 6974 refer
to the nucleotide numbering according to PubMed accession # M55171.) First, a 184
bp PCR product was generated with the primers Rho5940
(5’GTGAGGGGACATGCTGGAGGTGAGGC3’) and Q344ter(-)
(5’TGGAGCCACCTAGGAGGTCTCCGTCTTGG3’) using the 11 kb BamHI-
flanked genomic murine opsin gene clone as the DNA template. Next, a 874 bp PCR
product was generated with the primer pair Q344ter(+)
(5’CGGAGACCTCCTAGGTGGCTCCAGCCT3’) and Rho6974
(5’GCAAAAGCCTACTAAGGCTGAGG3’) again using the aforementioned opsin
gene clone as the DNA template. These two PCR products were combined to create
62
the last PCR product (1.034 kb) using the primers Rho5940 and Rho6974. This final
PCR product was digested by BspEI and PacI restriction enzymes, and via
subcloning, replaced its corresponding region in the aforementioned genomic clone
of the murine opsin gene. The construct was purified by the CsCl
2
gradient method,
and the mutated rod opsin gene was released from its vector by BamHI digestion.
The digested DNA fragments were separated electrophorectically in a 0.8% agarose
gel, and the BamHI-flanked Q344ter gene fragment was gel-extracted by using the
QIAEXII kit (Qiagen, Valencia, CA). After further purification with an Elutip-D
column (Whatman Schleicher & Schuell, Sanford, ME), this digested Q344ter gene
product was microinjected into fertilized eggs of donor B6D2F1 females to generate
transgenic Q344ter mice (Norris Transgenic Core facility, Keck School of Medicine
of USC, Los Angeles, CA). All mice were treated in accordance with the ARVO
Statement for the Use of Animals in Ophthalmic and Vision Research. All
transgenic Q344ter mice and their negative littermate controls were dark-reared
(except when noted) to prevent potential undesired light-dependent retinal
degeneration (Noell et al., 1966; Noell, 1980; Organisciak et al., 2003; recent review
Wenzel et al., 2005).
For genotyping, mouse-tail biopsy samples were used to extract genomic
DNA, from which a 376 bp PCR product was generated by using the primer pair
FACmRho6020 (5’TCCGGAACTGTATGCTCACCAC3’) and Rho3
(5’TGAGGGAGGGGTACAGATCC3’). This amplified product then was digested
63
by AvrII. If the mouse possessed the Q344ter transgene, two fragments (92 bp and
284 bp) would result.
Q344ter transgenic mice were bred to rho -/- mice (Lem et al., 1999) to
generate transgenic mice with the endogenous rhodopsin +/- and -/- genetic
background (rho+/- and rho-/-, respectively). PCR was performed to detect the
presence of the rod opsin null allele by using primers Rh1.1 (5’
GTGCCTGGAGTTGCGCTGTGGG 3’) and Neo3 (5’
CGGTGGATGTGGAATGTGTGCGAG 3’). To distinguish between hemizygous
(+/-) and homozygous (-/-) rhodopsin knockout mice, we performed Southern blot
analysis based on a previous protocol (Mendez et al., 2000b). Briefly, genomic
DNA was digested by NcoI, run in a 0.8% agarose gel, and transferred to Zetaprobe
blotting membrane (Bio-Rad Laboratories, Hercules, CA) via capillary action with
0.4 N NaOH. The membrane was cross-linked in a Stratalinker UV 2400
(Stratagene), and probed with a P
32
- labeled 1.7 kb fragment of the rod opsin gene
(flanked by SphI and EcoRI) in Church buffer (7% SDS, 1 mM EDTA, 500 mM
phosphate buffer pH 6.8). In mice with the rho+/- background two bands of 7.0 kb
and 7.5 kb would be present, while in mice with the rho-/- only the 7.5 kb band
would be present. (Please note that mice with the polymorphic “extra” NcoI site in
WT rod opsin gene were not included in these studies, because these mice would not
produce the desired rod opsin WT 7.0 kb fragment DNA.)

64
3.2.2. Determination of transgene expression level by RT-PCR.
Q344ter transgene expression level was determined by quantitating the
mutant-to-total transcript ratio in transgenic Q344ter
rho+/-
mice at postnatal days 28-
31 (P28-31). As controls, this assay included mice with the following genetic
backgrounds: rho+/-; Q344ter
rho-/-
; and rho-/-. From the various dark-reared mice,
total RNA was isolated from individual retinas by using the Trizol Reagent
(Invitrogen Corp., Carlsbad, CA), and reverse transcription with random primers was
performed to obtain single stranded cDNA molecules. These cDNA products served
as templates, in which a 250 bp subregion common to both WT and transgenic rod
opsin transcript species - beginning at the 3’ of exon 4 and ending within exon 5
beyond the sites of mutagenesis – was amplified by PCR with the primers
FACRhoEx4A (5’ GGTCATCTACATCATGTTGAACAAGC 3’) and mRh5 (5’
TGAGGGAGCCTGCATGACCTCATCC 3’). To label the amplification product,
10 µCi of α-
32
P dCTP (3000 Ci/mmol (GE Healthcare, Piscataway, NJ)) were added
to the PCR buffer. These amplified products were precipitated, washed, resuspended
in dH
2
O, and divided into two equal aliquots. One aliquot of 7.0 µl was digested with
AvrII for Q344ter, and the other aliquot of equivalent volume was mock digested.
(AvrII digestion of transgene transcripts would result in two fragments – 122 bp and
128 bp.) Both aliquots were then loaded in a 3% 3:1 Nusieve agarose gel (ISC
BioExpress, Kaysville, UT). The PCR fragments were transferred to Zetaprobe
blotting membrane through capillary action with 0.4 N NaOH, and the intensities of
the 250 bp radioactive bands were analyzed using the Storm 860 Phosphor Imager
65
software (GE Healthcare). Mathematically, the Q344ter transgene expression level
was determined as follows: 1 – (intensity of “leftover” intact radioactive PCR
product in the digested aliquot divided by the intensity of the “total” intact
radioactive PCR product in the undigested aliquot).

3.2.3. Western blot analysis.
Retinas were dissected out from dark-reared mice at P28-31 under infrared
illumination. Individual retinas were homogenized in homogenization buffer [80
mM Tris-HCl pH 8.0; 4 mM MgCl
2
; protease inhibitor cocktail (Roche Diagnostics,
Indianapolis, IN); and 0.5 mM phenylmethylsulfonyl fluoride (PMSF)], and
subsequently incubated with DNase I (Roche Diagnostics) for 30 - 45 min at room
temperature. 2X protein sample loading buffer (100 mM Tris, pH 6.8, 0.2 M
dithiothreitol (DTT), 8% SDS, 20% glycerol, dash of bromophenol blue) was added,
and the equivalent amounts of retina per sample (1/800) were loaded and run in a
12% Tris-glycine polyacrylamide gel (Invitrogen Corp.). The protein samples then
were transferred onto nitrocellulose membrane (Whatman Schleicher & Schuell) and
were incubated with either the anti-N-terminal rhodopsin monoclonal antibody
(mAb), R2-12N (1:10,000) that recognizes residues 2 – 12 (Adamus et al., 1991); or
the anti-C-terminal rhodopsin mAb, 1D4 (1:20,000) that recognizes residues 340 –
348 (Hodges et al., 1988). Goat anti-mouse IgG conjugated horseradish peroxidase
(HRP; 1:10,000; Vector Laboratories, Burlingame, CA) was used as the secondary
antibody (Ab). [All Abs were diluted with TBS-T (20 mM Tris, 1.37 M NaCl, 0.1%
66
Tween-20, pH 7.6) and contained 1% bovine serum albumin (BSA).]
Immunodetection was performed by using the ECL system (GE Healthcare).

3.2.4. Rhodopsin trafficking and light-dependent phosphorylation by
immunohistochemistry (IHC).
For IHC experiments involving rhodopsin mislocalization, all mice were
dark-reared and sacrificed at P28-31. For IHC experiments involving
phosphorylated rhodopsin, mice were either dark-reared or light-exposed (3000 lux
for 0.5 hr with dilated pupils) and sacrificed also at P28-31. Before enucleation, the
superior pole for each mouse eye was cauterized for orientation. All eyecups from
dark-reared mice were prepared under infrared light until removal of the cornea.
Eyecup preparation proceeded as follows: after cauterization, the mouse eye was
fixed in fixative solution (4.0% paraformaldehyde, 0.5% glutaraldehyde in 0.1M
cacodylate buffer pH 7.2) for 2 hours at room temperature (RT), during which the
cornea was removed after 5 min, and the lens was removed 10 min later. The
eyecups were infiltrated with 30% sucrose in 0.1M cacodylate buffer pH 7.2 (14-18
hrs at 4
o
C), hemisected, embedded in Tissue Tek
®
O.C.T. (Sakura Kinetek U.S.A.
Inc., Torrance, CA), and quickly frozen by using liquid N
2
. The frozen eyecups were
sectioned (10 µm) with a Jung CM 3000 cryostat machine (Leica Inc., Deerfield, IL).
The retinal sections were incubated for 1 hour in blocking solution [2.0% BSA, 0.3%
Triton X-100, and 2% goat serum in phosphate buffer saline (PBS)]. This blocking
solution also served as the dilution solution for all involved antibodies. These
sections were incubated with one of the following mAbs: R2-12N (1:100), 1D4
67
(1:100), or A11-82P (1:50), which recognizes multi-phosphorylated rhodopsin
species (Adamus et al., 1988; Adamus et al., 1991). After washing with blocking
solution, the sections were incubated with a 1:100 dilution of FITC-conjugated rabbit
anti-mouse IgG (Vector Laboratories, Inc., Burlingame, CA). After a series of
washing and a short fix (5 min in 4.0% paraformaldehyde in PBS), the sections were
mounted with Vectashield (Vector Laboratories, Inc.), cover-slipped, and analyzed
with an AxioPlan 2 imaging microscope (Carl Zeiss, Inc., Goettingen, Germany).

3.2.5. Retinal morphometry.
At P28-P31, we sacrificed the Q344ter
rho+/-
mice and their negative control
littermates. These mice were either dark-reared only or dark-reared and exposed to
five days of continuous light (3000 lux with undilated pupils) preceding their
sacrifice. [Only darkly pigmented mice (black or agouti) were light-exposed.] After
excising the mouse eyes, they were fixed and embedded as eyecups in an epoxy resin
(epon) as previously described (Concepcion et al., 2002). These eyecups were
sectioned at or near the vertical meridian as determined by the optic nerve. We
measured retinal degeneration by retinal morphometry based on a previously
described method (Danciger et al., 2000).
Briefly, retinal section was viewed by a microscope (40X objective) attached
with a camera lucida; and measurements were taken with the aid of a graphics tablet
(WACOM, Vancouver, WA) and the Axiovision LE Rel. 4.1. imaging software (Carl
Zeiss Inc., Germany). Calibration of measurements was accomplished by using a
stage micrometer (Klarmann Rulings, Litchfield, NH). Each hemisphere -
68
determined by the optic nerve - was divided into ten equal segments from the optic
nerve to either the superior or inferior tip. (Due to the thinness of the ONL at the
optic nerve location, determination of the ten equal segments for each hemisphere
excluded the first 100 µm from the optic nerve site.) Within a particular segment the
average of three recorded measurements of the ONL thickness was inversely related
to the degree of retinal degeneration in that segment. Furthermore, because a
subregion within the superior half and near the optic nerve was revealed to be the
most sensitive to light damage in previous studies (Danciger et al., 2000; Vaughan et
al., 2003), ONL thickness measurements obtained from this particular subregion
represented the statistically significant extent of retinal degeneration occurring
within a particular retina. Statistically, when comparing sample populations to
determine significant differences, t-tests were used with α = 0.05.

3.2.6. Light-dependent phosphorylation as shown by isoelectric focusing (IEF).
To observe light-dependent phosphorylation of rhodopsin, we processed P28-
31 mice with the following genetic backgrounds: Q344ter
rho+/-
; Q344ter
rho-/-
; and their
negative littermate controls. For mice that were subjected to light-exposure, their
pupils were dilated and exposed to 3000 lux for 0.5 hr. Retinas (one or two per
sample) were collected and quickly frozen. (Afterwards, all proceeding steps were
performed under infrared illumination until the end of “focusing” run.) The retinal
samples were homogenized with a PT 1200 C polytron (Kinematica, Switzerland) in
400 µl homogenization buffer [25 mM Hepes pH 7.5, 100 mM EDTA, 50 mM NaF,
5 mM adenosine, 1 mM PMSF, and protease inhibitors (Roche Diagnostics)] and
69
centrifuged at 19,000 x g (4°C, 17 min). After washing with 10 mM Hepes pH 7.5,
the pellet was resuspended in 1 ml regeneration buffer (10 mM Hepes pH 7.5, 0.1
mM EDTA, 50 mM NaF, 5 mM adenosine, 1 mM PMSF, 1 mM MgCl
2
, 2% BSA,
protease inhibitors, and approx. 1000 pmol 11-cis retinal) and incubated overnight
(O/N) at 4°C. The samples were spun down at 19,000 x g and washed with 10 mM
Hepes pH 7.5. The pellets were incubated in 50 µl or 100 µl solubilization buffer [20
mM Hepes pH 7.5, 0.1 mM EDTA, 50 mM NaF, 5 mM adenosine, 1 mM PMSF, 1
mM MgCl
2
, 10 mM NaCl, 1% dodecyl-maltoside (DM), 1 mM dithiothreitol (DTT),
protease inhibitors] for 3 – 24 hrs (4°C). Glycerol was added to the solubilized pellet
samples, which were loaded onto an acrylamide gel [5% acrylamide, 0.01% DM,
13.33% glycerol, 3.8% Pharmalyte pH 2.5 – 5 (GE Healthcare), 2.53% Pharmalyte
pH 5 -8 (GE Healthcare), catalyzed by ammonium persulfate and TEMED]. The
sample amounts (fraction of a retina) are as follows: rho+/- (1/20); Q344ter
rho+/-

(1/10); rho-/- and Q344ter
rho-/-
(1/5). The samples were run at a constant 23 W (with
0.04 M glutamic acid as the anode solution and 1.0 M NaOH as the cathode solution)
on a Pharmacia Flat Bed Apparatus FBE300 (GE Healthcare) at 10°C for 2hrs.
Afterwards, the “focused” proteins were transferred onto nitrocellulose membrane by
capillary action with PBS. The membranes were subjected to immunoblotting
analysis (as previously described) with mAbs 1D4 and 4D2 (another anti-N-terminal
rhodopsin mAb (1:10,000 dilution)).

70
3.3. Results.
3.3.1. Generation of transgenic Q344ter rhodopsin mutant (Q344ter) mice.
To investigate the pathway(s) leading to the development of ADRP in
patients with the Q344ter rod opsin (Q344ter) mutation, we generated a line of
transgenic mice expressing Q344ter. For this transgenic expression, a mutation at
codon 344 was introduced in an 11 kb genomic fragment comprising the whole
coding sequence of murine rod opsin and its upstream regulatory regions (Zack et al.,
1991). This rod opsin mutation substituted the glutamine at codon 344 to an early
stop signal (Fig. 3.1, Methods and Materials). Therefore, the last five amino acids,
QVAPA, are absent from this rhodopsin mutant, while the six putative light-
dependent phosphorylation sites are retained.
Q344ter transgenic mice were generated and mated with endogenous
rhodopsin homozygous knockout (rho-/-) mice (Lem et al., 1999) to produce Q344ter
transgenic mice with either rho+/- or rho-/- genetic background (Q344ter
rho+/-
and
Q344ter
rho-/-
, respectively). At least one copy of the endogenous rod opsin gene was
deleted to minimize retinal degeneration that may be induced by rhodopsin
overexpression (Olsson et al., 1992; Tan et al., 2001). Additionally, because the
Q344ter-mediated mechanism(s) leading to retinal degeneration is largely unknown,
Q344ter transgenic mice were raised in the dark (except when noted) to avoid any
uncontrolled light-dependent retinal degeneration.

71

Figure 3.1. Generation of Q344ter transgenic mice. The glutamine at codon 344
was mutated to an early stop signal in an 11-kb BamHI-flanked genomic clone
containing the murine rod opsin gene. Thus, the QVAPA domain is absent in this
truncated form of rhodopsin. All six potential phosphorylation sites of rod opsin’s
carboxyl terminus (underlined) are retained in this transgenic construct. An AvrII
site located near the Q344ter point mutation also was generated by two silent
mutations for genotyping purposes. The three point mutations are denoted by
capitalized nucleotides.
72
3.3.2. Q344ter transgene expression level by RT-PCR and confirmation of
successful Q344ter transgene expression by Western blot analysis.
The level of Q344ter transgene expression was quantified by RT- PCR. Total
RNA was isolated from the retinas of dark-reared transgenic Q344ter mice and their
negative littermate controls in rho+/- and rho-/- backgrounds. By RT-PCR in the
presence of
32
P-labeled dCTP, a pair of primers (see Materials & Methods) mapping
at exon 4 and 5 (and flanking intron 4) was used to amplify a 250 bp RNA region
common to both WT and Q344ter rod opsin transcripts. This primer pair amplifies
with equal efficiency the same 250 bp subregion in both the endogenous and
transgenic rod opsin gene products. Therefore, the fraction of RT-PCR product
derived from the transgenic transcript subpopulation reflected the relative amount of
transgenic transcripts from the total transcript population. In other words, the
difference in degree of amplification between the two transcripts depends only in the
relative starting amounts of each species. Radioactively labeled RT-PCR products
from total RNA from each retina was digested to completion with AvrII, which cuts
specifically the amplified Q344ter transgene product into 122 bp and 128 bp
fragments. An equal amount of radioactively labeled RT-PCR product was mock-
digested under identical conditions. To determine transgene expression levels, we
compared intensities between the “leftover” RT-PCR product band in the AvrII-
digested fraction versus the same-sized band in the mock-digested fraction. Almost
the entire amplified transcript product from the Q344ter
rho-/-
retinas was digested by
AvrII (Fig. 3.2 A, leftmost set), indicating that the vast majority of rod opsin
73
Figure 3.2. Determination of Q344ter transgene expression at the RNA and
protein levels. (A) Determination of the percentage of transgene transcript from the
total transcript population. Total RNA was isolated from retinas of dark-reared
Q344ter
rho+/-
and Q344ter
rho-/-
mice and negative littermate controls. RT-PCR was
used to amplify a common region to both endogenous and transgenic rod opsin
transcripts in the presence of
32
P-labeled dCTP (insert). The restriction enzyme,
AvrII, was used to specifically digest the amplified transgenic pool, and we
compared digested versus undigested samples from each retina to determine
transgene expression levels. From the Q344ter
rho-/-
retinal samples, nearly the entire
RT-PCR product was digested, showing that the majority of rod opsin transcripts
originated from the Q344ter transgene (leftmost set). From the nontransgenic rho+/-
and rho-/- retinal samples, we detected no AvrII digestion of the RT-PCR products,
as expected (three right sets). From the two Q344ter
rho+/-
retinal samples, we
observed partial digestion of the RT-PCR samples (second and third sets from left).
The average reduction of the original 250 bp RT-PCR products in these two retinal
samples was ~24%, which corresponded to the level of transgene expression in
Q344ter
rho+/-
retinas. (B) Detection of Q344ter rhodopsin species in transgene
positive mice. Q344ter lacks the QVAPA domain and, therefore, is recognized by
R2-12N mAb (anti-N-terminal) but not recognized by 1D4 mAb (anti-C-terminal).
[Please note that in addition to the rhodopsin monomer (38 kD), the rhodopsin dimer
(66kD) was present and most likely resulted from the preparation of these retinal
homogenate samples.] In Q344ter
rho-/-
homogenates (*), we observed rhodopsin
74
bands with R2-12N that was absent with 1D4, which confirmed the presence of
Q344ter species. As expected, no rhodopsin was detected in rho-/- retina, whereas
rhodopsin was detected by both mAbs (R2-12N and 1D4) in both Q344ter
rho+/-
and
rho+/- retinas. Q344ter and endogenous rhodopsin molecules in the Q344ter
rho+/-

retina could not be discriminated by size with mAb R2-12N. Note that the presence
of Q344ter appeared to decrease the total amount of rhodopsin despite equivalent
loading of retina for each sample (1/800) - with the least amount of rhodopsin in
Q344ter
rho-/-
retinas. This observation is indicative of the ongoing retinal
degeneration caused by this truncated rhodopsin mutant (see Fig. 3.5 and related
text). All mice were sacrificed between P28-31.
75

Figure 3.2. Determination of Q344ter transgene expression at the RNA and
protein levels.
76
transcripts originated from the Q344ter transgene. The amplified transcripts from
nontransgenic rho+/- and rho-/- retinas were not recognized by AvrII (Fig. 3.2 A,
three right sets), and no differences were detected in digested versus mock-digested
samples, as expected. As for the two Q344ter
rho+/-
retinal samples, we observed an
average of ~24% partial digestion of the RT-PCR samples (Fig. 3.2 A, second and
third sets).Thus, the level of transgenic gene expression in Q344ter
rho+/-
retinas was
determined to be ~24% of total rod opsin transcripts.
At the protein level, it is difficult to distinguish between WT rhodopsin and
Q344ter by Western blot analysis when using only a single anti-rhodopsin
monoclonal antibody (mAb), because Q344ter is missing only five amino acids.
This scenario differs from mice expressing another truncated rhodopsin S334ter,
which is missing the last 15 amino acids and can be easily differentiated from WT
rhodopsin (Concepcion et al., 2002). Adding difficulty towards identifying Q344ter
molecules in these transgenic mice, we did not possess an antibody which
specifically recognizes this rhodopsin mutant.
To confirm the expression of Q344ter in transgene positive mice, Western
blot analysis was performed with retinal homogenates from dark-reared Q344ter
rho+/-
and Q344ter
rho-/-
mice (and their respective negative littermate controls). Two anti-
rhodopsin mAbs were used: R2-12N and 1D4. R2-12N recognizes rhodopsin’s N-
terminal region [residues 2 -12] (Adamus et al., 1991), and therefore binds to both
WT rhodopsin and Q344ter; 1D4 recognizes rhodopsin’s C-terminal region
[residues 340 -348] (Hodges et al., 1988), and therefore binds to WT rhodopsin but
77
not to Q344ter. (Due to the preparation of these retinal samples, both the monomeric
(38kD) and dimeric (66kD) forms of rhodopsin were present in homogenates
whenever rhodopsin molecules were expressed.) In retinal homogenates from mice
with the rho+/- background (Q344ter
rho+/-
and nontransgenic
rho+/-
retinas), rhodopsin
was detected by both R2-12N and 1D4 mAbs (Fig. 3.2 B). As previously mentioned,
the two rhodopsin species in the Q344ter
rho+/-
retinas could not be distinguished by
size in this assay. Worth noting, however, is the rhodopsin content appeared to be
lower in Q344ter
rho+/-
retinas than rho+/- retinas with both R2-12N and 1D4
immunostainings, despite loading equal amounts of retinal homogenates (1/800 of a
retina per sample). This observation suggests the occurrence of degeneration in
these transgenic Q344ter retinas (see Fig. 3.5).
Confirmation of Q344ter expression was achieved by using retinal
homogenates from mice with the rho-/- background (Q344ter
rho-/-
and
nontransgenic
rho-/-
). In both types of mice, endogenous rhodopsin is absent.
Rhodopsin was detected with R2-12N mAb in the transgenic Q344ter
rho-/-
retinas but
not in the negative littermate control retinas (Fig. 3.2 B). This outcome showed that
a rhodopsin species was present in the Q344ter
rho-/-
retinas, whereas rhodopsin was
completely absent in rho-/- retinas (Lem et al., 1999). When we incubated the same
retinal homogenate samples with the anti-C-terminal rhodopsin mAb 1D4 (Fig. 3.2
B), rhodopsin staining was not observed in either retinal types (Q344ter
rho-/-
and rho-
/- retinas). This result indicated that the rhodopsin species present in the Q344ter
rho-/-

retinas was missing a portion of its C-terminus (i.e. the QVAPA domain).
78
Altogether, these observations confirmed successful protein expression of Q344ter in
these transgenic mice.

3.3.3. Observations of defective rhodopsin trafficking in the Q344ter transgenic
mice.
Previously, we showed the effects of rhodopsin mistrafficking in our S334ter
mouse model (Concepcion et al., 2002). However, in addition to the absence of the
VXPX domain, S334ter also lacks the six murine phosphorylation sites, which may
have contributed to the defective rhodopsin trafficking and/or inhibition of ROS
formation. Although proper rhodopsin localization has been demonstrated in a
transgenic mouse line expressing a rhodopsin species in which all six
phosphorylation sites were mutated to alanine residues in the absence of endogenous
rhodopsin molecules (Mendez et al., 2000a), it has not been shown that the absence
of the VXPX domain alone is sufficient to induce rhodopsin mistrafficking. To
isolate the importance of the VXPX domain (and exclude contributions from the
region containing these phosphorylation sites) towards proper rhodopsin trafficking,
we performed immunohistochemistry (IHC) on our transgenic Q344ter mouse line
with the rho+/- (Fig. 3.3) and rho-/- (Fig. 3.4) genetic backgrounds involving the
same two aforementioned mAbs, R2-12N and 1D4. When immunostaining frozen
retinal sections from Q344ter
rho+/-
mice with the anti-N-terminal rhodopsin mAb R2-
12N, rhodopsin was detected not only in the rod outer segment (ROS) but also in the
rod inner segment (RIS) and outer nuclear layer (ONL). This pattern is indicative of
rhodopsin mislocalization (Fig. 3.3 A) and consistent with previous studies
79

Figure 3.3. Q344ter rhodopsin mistrafficks and causes partial mislocalization
of endogenous rhodopsin molecules. Q344ter
rho+/-
and rho+/- littermate control
mice were dark-reared and sacrificed at P28-P31. Frozen retinal sections were
immunostained with the anti-N-terminal mAb R2-12N (A and B), which recognizes
both endogenous and Q344ter rhodopsin species. R2-12N immunostaining was
restricted to the rod outer segment (ROS) in rho+/- sections (B), but extended to the
rod inner segment (RIS) and outer nuclear layer (ONL) in Q344ter
rho+/-
sections (A),
indicating rhodopsin mistrafficking to proximal compartments in transgene positive
retinas. Frozen sections from the same mice were immunostained with the anti-C-
terminal mAb 1D4 (C and D), which recognizes only endogenous rhodopsin
molecules. In the Q344ter
rho+/-
retina (C), rhodopsin staining is primarily located in
the ROS. Therefore, the mislocalized rhodopsin molecules observed in the
Q344ter
rho+/-
retinal section (A) are mainly Q344ter. However, we observed a small
degree of rhodopsin staining in the RIS and ONL as well, which indicated that
Q344ter molecules also caused mistrafficking of some endogenous WT rhodopsin
molecules. In the rho+/- retina (D), rhodopsin staining only occurred in the ROS, as
expected. Additionally, the “outlining” of the cell bodies with R2-12N
immunostaining in (A) and 1D4 immunostaining in (C) indicates plasma membrane-
association of the mislocalized rhodopsin molecules. (Scale bar = 20 µm)
80
(Sung et al., 1994; Shi et al., 2004). Furthermore, because the R2-12N
immunostaining appeared to “outline” the cell bodies, mislocalized rhodopsin
molecules most likely were associated with plasma membrane, which can be
explained by the “default” localization to the plasma membrane of properly folded
rhodopsin molecules with defective trafficking signals (see Introduction). Moreover,
this plasma membrane-association by these mislocalized rhodopsin molecules was
reported previously (Sung et al., 1994; Sung and Chuang, 2004). As a control, in
retinal sections from negative littermate (nontransgenic
rho+/-
) mice, R2-12N
immunostaining showed rhodopsin present only in the ROS (Fig. 3.3 B), which
illustrated proper rhodopsin localization.
When retinal sections from both Q344ter
rho+/-
transgenic mice and their
negative littermate controls were immunostained with the anti-C-terminal rhodopsin
mAb 1D4 (Figs. 3.3, C and D respectively), rhodopsin was observed only in the ROS
in rho+/- retinas and predominantly in the ROS in Q344ter
rho+/-
retinas. This
immunostaining pattern of WT rhodopsin in Q344ter
rho+/-
transgenic retinas revealed
that the mislocalized rhodopsin molecules in these retinas are primarily Q344ter.
Unlike Sung et al. (1994), we also noticed a detectable level of WT rhodopsin
immunostaining in the ONL and RIS in frozen sections from Q344ter
rho+/-
retinas but
not in rho+/- retinas (Figs. 3.3, C and D respectively). Therefore, truncated
rhodopsin molecules lacking the QVAPA domain also caused a slight degree of
mislocalization of WT rhodopsin molecules, which had been reported earlier
81
(Concepcion et al., 2002; Shi et al., 2004). [Moreover, the mislocalized WT
rhodopsin molecules again appear to be plasma membrane bound (Fig. 3.3 C).]
Thus far, we have provided evidence that the VXPX domain by itself is the
essential signaling motif for proper rhodopsin trafficking. In addition to directing
proper vectorial rhodopsin transport, we previously showed that this domain also is a
crucial element for ROS morphogenesis (Concepcion et al., 2002). When S334ter
molecules were expressed in the absence of endogenous rhodopsin (S334ter
rho-/-

retinas), no ROS formation was observed. In order to confirm that the VXPX
domain is the essential region in rhodopsin’s C-terminus for ROS formation, we
examined again by IHC the morphology of retinas from both Q344ter transgenic
mice and their negative littermate controls in the absence of endogenous rhodopsin
molecules (Q344ter
rho-/-
and rho-/- retinas, respectively; Fig. 3.4). When frozen
sections from rho-/- retinas were immunostained with the anti-N-terminal rhodopsin
mAb R2-12N, neither ROS formation nor the presence of rhodopsin itself was
detected (Fig. 3.4 B), which was consistent with a previous study regarding rho-/-
mice (Lem et al., 1999). When Q344ter
rho-/-
retinal sections were immunostained
with R2-12N mAb, rhodopsin was detected in the RIS and ONL, but the ROS
structure was noticeably absent (Fig. 3.4 A), which was consistent with another
earlier study (Shi et al., 2004). (As before, mislocalized Q344ter in the Q344ter
rho-/-

retinas appeared to be associated with the plasma membrane (Fig. 3.4 A). Therefore,
these results collectively show that despite the presence of a rhodopsin species, the
82

Figure 3.4. The QVAPA domain is essential for ROS formation. With the rho-/-
genetic background, Q344ter transgenic (A) and nontransgenic mice (B) were dark-
reared and sacrificed at P28-P31. Consequently, neither of the two retinas expressed
endogenous WT rhodopsin. Frozen retinal sections were immunostained with the
anti-N-terminal mAb R2-12N. (A) ROS formation was not detected in the
Q344ter
rho-/-
frozen retinal section, despite the presence of a rhodopsin species.
Because Q344ter molecules lack the QVAPA domain, together with the absence of
endogenous rhodopsin, proper polarized trafficking of Q344ter molecules was
precluded, resulting in inhibition of ROS morphogenesis in the Q344ter
rho-/-
retinas.
This result is consistent with earlier studies of the VXPX domain (Concepcion et al.,
2002; Shi et al., 2004). Additionally, the mislocalized Q344ter molecules in the
ONL appeared to be plasma membrane bound. (B) In the rho-/- frozen retinal
section, neither rhodopsin staining nor ROS formation was detected, as previously
observed (Lem et al., 1999). Although not shown, no 1D4 immunostaining was
detected with either retina, as expected. (Scale bar = 20 µm)
83
presence of the VXPX domain is the essential signaling motif for ROS
morphogenesis.

3.3.4. Q344ter initiates retinal degeneration that is accelerated by light-exposure.
Because the previously studied dark-reared S334ter mouse model did not
experience any significant retinal degeneration (Concepcion et al., 2002), we
attributed this observation to a low transgene expression level. However, when
comparing transgene expression levels between Q344ter and S334ter mice, the
Q344ter transgene expression level was more than double the S334ter transgene
expression level (24% versus 10%, respectively). Therefore, we expect to observe
retinal degeneration in our dark-reared transgenic Q344ter mouse line, which would
be consistent with previous animal models (Green et al., 2000; Vaughan et al., 2003;
Tam et al., 2006). Additionally, as mentioned earlier, the heterogenous disease
progressions among ADRP patients inheriting the same rhodopsin mutation indicate
that environmental conditions contributes to the severity of this disease. One
potential environmental candidate is light-exposure. By rearing our transgenic
Q344ter mice and their negative littermate controls under two conditions (constant
darkness versus 3000 lux constant light-exposure for 5 days), we could investigate
the degree of retinal degeneration caused by rhodopsin mistrafficking alone and
together with light-exposure.
We measured the degree of degeneration by retinal morphometry based on a
previous method (Danciger et al., 2000). Measuring the thinning of ONL is a quick
and suitable protocol to determine the extent of retinal degeneration (Michon et al.,
84
1991). As before, only darkly pigmented mice (black or agouti) were light-exposed
to minimize potential degeneration caused by high light intensity (Noell et al., 1966;
Noell, 1980; Hao et al., 2002; recent review Wenzel et al., 2005). Additionally, for
statistical analysis between sample populations, ONL thickness measurements were
recorded within a subregion of the superior half and near the optic nerve of each
retina (Fig. 3.5, green bar with asterisk), which previously was revealed to be the
most sensitive region to light damage (Danciger et al., 2000; Vaughan et al., 2003).
This narrowing of the retinal area also would reduce the inherent variability of ONL
thicknesses observed within a whole retina (see Material & Methods).
In retinas from Q344ter transgenic and nontransgenic control mice with the
rho+/-genetic background, we observed three instances of retinal degeneration (Fig.
3.5). In the first case of degeneration, a slight but statistically significant thinning of
the ONL occurred in light-exposed negative littermate control retinas when
compared to their dark-reared counterparts. This nominal degeneration was
somewhat expected, for although excessive light can cause light damage in WT mice
(Noell et al., 1966; Noell, 1980; Hao et al., 2002; recent review Wenzel et al., 2005) ,
ocular pigmentation significantly reduces the light intensity by approximately two
orders of magnitude (Rapp and Williams, 1980; LaVail and Gorrin, 1987). By light-
exposing only heavily pigmented mice (black or agouti), we could minimize this
potential light effect in the control littermates and, therefore, isolate the potential role
of Q344ter towards retinal degeneration in light-exposed transgenic mice.
85

Figure 3.5. Retinal degeneration caused by mislocalized rhodopsin is
accelerated by light. At P28-31, we sacrificed Q344ter
rho+/-
and nontransgenic
control mice that were either dark-reared only or exposed to continuous light (3000
lux with undilated pupils) during their last five days. One copy of endogenous rod
opsin gene was deleted to minimize retinal degeneration that can be caused by
rhodopsin overexpression (Olsson et al., 1992; Tan et al., 2001). Eyecups were
embedded in epon and sectioned near the optic nerve. The diagram displays the
ONL thickness along the entire span of the retina, and the legend denotes the
genotype and rearing conditions of the involved mice. When measuring
degeneration by retinal morphometry, we focused on a light-sensitive region in the
superior half near the optic nerve, as denoted by a green asterisk with bars (Danciger
et al., 2000; Vaughan et al., 2003). From each of the four groups represented in the
“spidergraph”, a collection of light microscopy pictures within this light-sensitive
region is displayed in the bottom half (letters in pictures correspond to letters in
legend; scale bar = 20 µm). A slight but statistically significant ONL thinning
occurred in light-exposed nontransgenic mice when compared to their dark-reared
counterparts. In the presence of Q344ter, two cases of retinal degeneration emerged:
(i) rhodopsin mistrafficking caused a moderate level of degeneration in the absence
of light - compare dark-reared Q344ter transgenic retinas with dark-reared
nontransgenic retinas; and (ii) light exposure induced a severe form of degeneration
in Q344ter transgenic retinas when compared to both dark-reared Q344ter transgenic
retinas and light-exposed nontransgenic retinas. (For all statistically significant
differences, p ≤ 0.05.)

86
The other two cases of retinal degeneration observed in these mice involved
the Q344ter molecule (Fig. 3.5). When comparing dark-reared Q344ter transgenic
retinas to dark-reared nontransgenic retinas, the ONLs of Q344ter transgenic retinas
were moderately thinner; this provides an explanation for the decreased rhodopsin
content in retinal homogenates from Q344ter transgenic mice (see Fig. 3.2 B). This
thinning of dark-reared Q344ter transgenic retinas was statistically more significant
than the retinal thinning found in the first case above. Moreover, when comparing
light-exposed Q344ter transgenic retinas to both light-exposed nontransgenic retinas
and dark-reared Q344ter, we observed severe ONL thinning of the light-exposed
Q344ter transgenic retinas in the highly sensitive region within the superior half.
These observations indicate that although rhodopsin mistrafficking causes retinal
degeneration in the absence of light, Q344ter dependent retinal degeneration was
accelerated by light-exposure.

3.3.5. Mislocalized rhodopsin is capable of light-activation.
Because light-exposure exacerbated the severity of retinal degeneration in the
Q344ter transgenic mice and that activated mislocalized rhodopsin potentially could
interact with “alternative” proteins, it is conceivable that these properly folded
mislocalized rhodopsin molecules may trigger rod cell death by initiating a light-
dependent cell signaling cascade distinct from phototransduction. Numerous
investigations have reported that such novel signaling cascades can occur (although
not always involving mislocalized rhodopsin). These studies used Drosophila
models (Alloway et al., 2000; Kiselev et al., 2000; Iakhine et al., 2004); cultured
87
mammalian cells (Alfinito and Townes-Anderson, 2002; Chuang et al., 2004); and
constitutively active rhodopsin mutants (Chuang et al., 2004; Chen et al., 2006).
Yet, none of these studies addressed whether mislocalized rhodopsin molecules are
capable of light-activation in intact mammalian retinas. Substantiating this property
may prove to be a crucial first step towards discovering the mechanism(s) that leads
to the observed light-accelerated retinal degeneration in our transgenic Q344ter
mouse model.
Because we could not test directly for light-dependent activation of
mislocalized rhodopsin, we examined this possibility by exploiting the light-
dependent rhodopsin phosphorylation by rhodopsin kinase (RK) (Kuhn and Wilden,
1987; recent review Metaye et al., 2005). Various well-established assays are
available to investigate this important step in rhodopsin inactivation:
32
P-labeled in
situ

assay, mass-spectrometry, IHC involving the A11-82P mAb (see below), or
isoelectric focusing (IEF). By using IEF electrophoresis, in which native proteins
are separated by their isoelectric point in a pH-gradient gel, it is possible to detect
different phosphorylated rhodopsin species in retinas (McDowell et al., 2000). In
this study, samples from both dark reared and light-exposed (3000 lux with dilated
pupils) retinas from both Q344ter
rho+/-
and rho+/- mice were loaded into a gel with a
pH gradient [3 – 8]. The various nonphosphorylated and phosphorylated rhodopsin
species were separated according to their isoelectric properties and transferred onto a
nitrocellulose membrane.
88
Afterwards, we performed two successive Western blots: first with 4D2 (Fig.
3.6 A), another anti-N-terminal rhodopsin mAb (Laird and Molday, 1988) that
recognizes both WT rhodopsin and Q344ter; and second with the anti-C-terminal
rhodopsin mAb 1D4 (Fig. 3.6 B). As shown in Figs. 3.6 A & B, only
nonphosphorylated rhodopsin and opsin molecules were detected in retinas from
dark-reared mice irrespective of transgene inheritance. These nonphosphorylated
species also existed in light-exposed retinas, which indicate that not all rhodopsin
molecules in the retinas became phosphorylated under the described light conditions.
In Fig. 3.6 A, 4D2 immunostaining revealed six light-activated phosphorylated
rhodopsin (R*-p) species that were consistently generated in both types of light-
exposed retinas (numbered 1-6). Furthermore, four additional R*-p species were
detected in the light-exposed Q344ter
rho+/-
retinas (Fig. 3.6 A, red arrowheads) that
were not detected in light-exposed rho+/-retinas. These results provide evidence for
the capacity of light-dependent phosphorylation of Q344ter molecules.
To confirm that these four additional R*-p species originate from the Q344ter
population, we subsequently incubated the same above membrane with the mAb 1D4
(Fig. 3.6 B). We reasoned that the four additional R*-p species are phosphorylated
Q344ter molecules and, therefore, would not be recognized by 1D4. Again, R*-p
species common to both types of light-exposed retinas were observed by 1D4
immunostaining. Interestingly, 1D4 did not recognize the hexa-phosphorylated
rhodopsin species in both retinal types. This result indicated the presence of multiple
phosphorylated endogenous rhodopsin species in both light-exposed retinal types.
89

Figure 3.6. Detection of light-dependent multi-phosphorylated Q344ter species.
Rho+/- mice expressing Q344ter and negative littermate controls were sacrificed at
P28-P31. Mice were dark-reared or exposed to 3000 lux with dilated pupils for 30
minutes immediately before sacrifice (P28-31). With the isoelectric focusing (IEF)
assay, retinal samples were loaded as fractions of retinas (rho+/- mice: 1/20;
Q344ter
rho+/-
mice: 1/10) in a pH-gradient gel (pH range 3 – 8), and proteins were
separated according to their isoelectric point. Hence, rhodopsin species with
increasing phosphorylated residues migrate towards more acidic regions
(corresponding to bottom section of the membrane). The numbers to the right of
each membrane image corresponds to the number of phosphates bound to the
rhodopsin (or opsin) species. (A) Western blot analysis involving the anti-N-
terminal rhodopsin mAb 4D2 (Laird and Molday, 1988) of phosphorylated rhodopsin
species separated by IEF. Irregardless of genetic background, only
nonphosphorylated rhodopsin (and opsin) species were detected in retinas from dark-
reared mice. Light-exposure, however, produced multiple light-dependent
phosphorylated rhodopsin species in both types of retinas. Moreover, four “extra”
bands (red arrowheads) were detected in light-exposed Q344ter
rho+/-
retinas which
were noticeably absent in light-exposed rho+/- mouse retinas. (B) Western blot
analysis involving the anti-C-terminal rhodopsin mAb 1D4 (Hodges et al., 1988) of
phosphorylated rhodopsin species separated by IEF. When the same membrane was
subsequently incubated with 1D4 mAb, the most striking result was the absence of
detection of the four “extra” bands in the light-exposed Q344ter
rho+/-
retinas. Thus,
these “extra” bands are multi-phosphorylated Q344ter molecules that initially were
light-excited. Note that the hexa-phosphorylated rhodopsin species was not
recognized by the 1D4 mAb in either retinal type. [White arrows demarcate bands
resulting from improper shifting/”lifting” of the nitrocellulose membrane during
transfer step.]
90
Furthermore, the four R*-p species unique to the light exposed Q344ter
rho+/-
retinas
were notably absent. This lack of detection of the four R*-p species by 1D4 verified
light-dependent Q344ter phosphorylation, which indirectly confirms Q344ter’s
capability for light-activation.
As previously alluded to, truncated rhodopsin mutants are present in the
ROS when expressed with endogenous rhodopsin (Sung et al., 1994; Chen et al.,
1995). Thus, the previous phosphorylation assay only provides evidence for light-
dependent excitation of Q344ter but does not distinguish the location of these
activated mutants. Thus, the original question of whether mislocalized Q344ter
molecules in the RIS and ONL are capable of light-activation was not addressed.
To investigate this hypothesis more directly, we took advantage of the
absence of ROS in the Q344ter
rho-/-
retinas. (Therefore, all Q344ter molecules would
be mislocalized in the RIS and ONL.) As shown in Fig. 3.7, we performed IEF and
Western blot analysis on retinas from Q344ter
rho-/-
and the necessary control mice
under dark-reared and light-exposure (3000 lux with dilated pupils) conditions. No
bands were observed in rho-/- mice regardless of rearing conditions, as expected.
Immunostaining with the anti-N-terminal mAb, R2-12N, (Fig. 3.7 A) again revealed
only nonphosphorylated rhodopsin in dark-reared retinas and multiple R*-p species
in light-exposed retinas. In the Q344ter
rho-/-
retina, however, only five R*-p species
were identified. (Perhaps the S343 residue of the Q344ter molecule cannot be
phosphorylated by rhodopsin kinase. Since S343 became the last amino acid in the
Q344ter molecule, it was this residue that experienced the most drastic change in
91

Figure 3.7. Detection of light-activation of mislocalized rhodopsin. In Q344ter
rho-
/-
retinas, essentially all Q344ter molecules are mislocalized due to the absence of the
rod outer segment. These Q344ter
rho-/-
retinas were included in the isoelectric
focusing assay (IEF; see Fig. 3.5) to analyze light-dependent phosphorylation of
mistrafficked rhodopsin. Mice were dark-reared or exposed to 3000 lux with dilated
pupils for 30 minutes immediately before sacrifice (P28-31). The numbers to the
right of each membrane image reflects the number of phosphates bound to the
rhodopsin (or opsin) species. Retinal homogenates from rho+/- and rho-/- mice
served as controls (A and B). (A) Western blot analysis with the anti-N-terminal
rhodopsin mAb, R2-12N, (Adamus et al., 1991) of phosphorylated rhodopsin species
separated by IEF. Multiple light-dependent phosphorylated rhodopsin (R*-p)
species were detected in light-exposed Q344ter
rho-/-
and rho+/- retinas. When
comparing the bi- through penta-R*-p species (marked with *) between the two
light-exposed retinas, slight differences in isoelectric points were observed, further
indicating that the “extra” bands in Q344ter
rho+/-
retinas were phosphorylated
Q344ter molecules (Fig. 3.6). (B) Western blot analysis involving the anti-C-
terminal rhodopsin mAb 1D4 (Hodges et al., 1988) of phosphorylated rhodopsin
species separated by IEF. No R*-p species from the Q344ter
rho-/-
retina were
detected, as expected. These results indicated that mislocalized rhodopsin molecules
become phosphorylated only after light-exposure, and therefore, are capable of light
excitation. Again, 1D4 mAb did not recognize the hexa-phosphorylated R* species
in the nontransgenic
rho+/-
retina (see Fig. 3.6 B). [Fractions of retina loaded per
sample are as follows: rho+/- mice (1/20); rho-/- and Q344ter
rho-/-
mice (1/5).]
92
surroundings when compared to the other potential phosphorylation sites.) Serving
as another control, the light-exposed rho+/- retina produced the previously described
six R*-p species. When comparing the R*-p species from light-exposed retinas in
Q344ter
rho-/-
and rho+/- mice, the phosphorylated Q344ter species appeared to have
slightly different isoelectric points than their phosphorylated endogenous rhodopsin
counterparts. This comparison further validated the detection of light-dependent
phosphorylation of Q344ter molecules in light-exposed Q344ter
rho+/-
retinas (Fig. 3.6
A).
To corroborate that the five R*-p species identified in the light-exposed
Q344ter
rho-/-
retina were phosphorylated truncated rhodopsin molecules, we
subsequently incubated the same membrane with the anti-C-terminal rhodopsin mAb
1D4 (Fig. 3.7 B). As expected, no R*-p species were identified in the light-exposed
Q344ter
rho-/-
retina, since 1D4 does not bind to Q344ter molecules. By
demonstrating in the Q344ter
rho-/-
retina that mislocalized rhodopsin undergoes light-
dependent phosphorylation, we have provided evidence for the capacity of
mislocalized rhodopsin to be light-activated.
Thus far, it has been shown that mislocalized rhodopsin in the absence of
ROS is capable of light-excitation. To investigate whether mislocalized rhodopsin
molecules in the presence of ROS are capable of light-activation, we performed IHC
on frozen retinal sections from mice with the rho+/- genetic background using the
mAb, A11-82P (Fig. 3.8). This particular mAb recognizes multi-phosphorylated
rhodopsin molecules (Adamus et al., 1988; Adamus et al., 1991). Although not
93

Figure 3.8. Mislocalized rhodopsin molecules in the RIS are capable of light-
activation. (A) Q344ter
rho+/-
and (B) rho+/- mice were exposed to light (3000 lux)
for 30 minutes with dilated pupils and immediately sacrificed (P28-31). Frozen
retinal sections from these mice were incubated with A11-82P, a mAb which binds
to multi-phosphorylated rhodopsin species (Adamus et al., 1988; Adamus et al.,
1991). (A) In the Q344ter
rho+/-
frozen retinal section, light-dependent multi-
phosphorylated rhodopsin (R*-p) molecules were observed not only in the rod outer
segment (ROS) but also in the rod inner segment (RIS, green arrow) and only
slightly in the distal region of the ONL (white arrowheads). (B) When Q344ter was
not expressed (rho+/- retinal section), the R*-p molecules were observed only in the
ROS. As expected, R*-p molecules were not detected in the RIS (white arrow) nor
ONL. Although not shown, A11-82P immunostaining of R*-p molecules was never
observed from dark-reared only mice regardless of genotype. These results indicate
that mislocalized rhodopsin in the RIS (and to a lesser degree in the ONL) were
phosphorylated upon light-exposure, and therefore, are capable of light-activation.
(Scale bar = 20 µm)
94
shown, no A11-82P immunostaining was observed in retinal sections from dark-
reared mice regardless of transgenic inheritance. In frozen retinal sections from
light-exposed (3000 lux) rho+/- mice, only the ROS was significantly stained,
indicating that most, if not all, of the light-dependent R*-p species reside in the ROS
(Fig. 3.8 B). However, when Q344ter
rho+/-
retinas were light-exposed (Fig. 3.8 A),
A11-82P immunostaining revealed the presence of R*-p species in the ROS as well
as the RIS and only faintly in the ONL (white arrowheads in Fig. 3.8 A).
Presumably, this slight immunostaining of R*-p species in the ONL compartment
could be explained by: the lower availability of 11-cis retinal, the lowered amount of
RK (Zhao et al., 1998), or both. Nevertheless, these results substantiated the light-
excitation capability of mislocalized rhodopsin molecules even in the presence of
ROS.

3.4. Discussion.
Polarized rhodopsin transport is a multistage process. After synthesis and
post-translational modifications in the rough endoplasmic reticulum and the Golgi
network, rhodopsin molecules are transported in the membranes of post-Golgi
vesicles (also termed rhodopsin-bearing transport carriers, RTCs) towards the distal
end of the inner segment (Deretic and Papermaster, 1991). After docking and fusing
to the plasma membrane at base of the connecting cilium, rhodopsin molecules are
transported towards the distal end of this subcellular structure. At this site, nascent
ROS disks emerge, rapidly mature, and collectively, form the ROS structure.

95
Because of the complexity in RTC transport, it is not surprising that at
various stages of this process rhodopsin directly interacts with number of different
proteins, including ARF-4, a small GTPase (Deretic, 2004) and the controversial
Tctex-1, a light-chain subunit of dynein (Tai et al., 1999). Its ability to interact with
multiple proteins could be explained by rhodopsin’s freely accessible C-terminal
domain, as demonstrated by its “highly disordered and dynamic” structure (Langen
et al., 1999). Because of this attribute, it is at the level of the amino acid sequence of
this C-terminal region, however, which dictates rhodopsin’s interactions with the
other proteins during its vectorial trafficking. Thus, mutations in this region of
rhodopsin result in its aberrant localization - most likely in the first stage of vectorial
transport, in which RTCs are trafficked to the distal end of the inner segment where
the base of the connecting cilium is located.
In this investigation we have successfully generated a line of transgenic
Q344ter mice to serve as a mouse model for ADRP patients inheriting this rod opsin
mutation. Our line of transgenic Q344ter mice showed rhodopsin mistrafficking
even in the presence of endogenous rhodopsin and is consistent with a the previous
study (Sung et al., 1994). Immunostaining with R2-12N mAb revealed that
rhodopsin localized not only in the ROS but also abnormally accumulated in the RIS
and ONL compartments (Fig. 3.3 A). This observed defect in vectorial rhodopsin
transport were caused by the rod opsin mutation which ablated the sorting motif,
QVAPA.

96
Although we did not directly identify Q344ter molecules in the ROS, Sung et
al. (1994) provided convincing evidence for this occurrence through single cell
recordings. Q344ter’s presence in the ROS supports our hypothesis for the co-
transport of both WT and Q344ter molecules in the same post-Golgi vesicles, or
RTCs (Concepcion et al., 2002; Shi et al., 2004). Thus, a threshold for the number
or concentration of VXPX domains in these vesicles likely needs to be overcome to
properly direct vectorial rhodopsin trafficking. Moreover, the converse of our co-
transport hypothesis would predict the mislocalization of WT rhodopsin molecules.
Thus, the detectable presence of endogenous WT rhodopsin molecules in the RIS
and ONL compartments (Fig. 3.3 C) of Q344ter
rho+/-
retinas may have resulted from
WT rhodopsin molecules present in RTCs that did not contain sufficient amounts of
VXPX domains. (Although not measured, the degree of WT rhodopsin
mislocalization appeared to be more extensive in the Q344ter mouse retinas than that
observed in the previously studied S334ter mouse retinas (Concepcion et al., 2002)
and could be a consequence of the greater transgene expression level in Q344ter
mice versus the S334ter mice.)
The significance of the QVAPA domain (or more specifically VXPX) was
further demonstrated in transgenic Q344ter
rho-/-
retinas. Because endogenous
rhodopsin was not expressed, no VXPX domains were present to properly direct
trafficking of the RTCs. As a result, the ROS compartment was conspicuously
absent. As with the previous studies (Concepcion et al., 2002; Shi et al., 2004), this
observation indicated that this sorting motif is the crucial element not only for proper
97
rhodopsin trafficking but also for the morphogenesis of ROS, which in turn, is
essential for efficient and sensitive phototransduction cascade activity. Moreover,
this inability to form ROS structures in the transgenic Q344ter
rho-/-
retinas suggests
that the mislocalization of rhodopsin mainly was due to rhodopsin mistrafficking and
not the impairment in retention of these proteins in the ROS, which has been
previously postulated (Sung et al., 1994). If the latter was true, we would expect to
observe some basal forms of the ROS structures. Perhaps, examination of younger
transgenic Q344ter mice for ROS formation is warranted.
Also, we currently are expressing WT and Q344ter rhodopsin species fused
with a timer protein (Terskikh et al., 2000) in mouse retinas in an attempt to
distinguish between these two possible causes for rhodopsin mislocalization. We
believe that if the nascent Q344ter-timer fusion proteins (as would be revealed by
their green fluorescence) are mislocalized, then rhodopsin mislocalization chiefly
would be caused by rhodopsin mistrafficking; if only “older” Q344ter-timer fusion
proteins (as would be revealed by their red fluorescence) are mislocalized, then
rhodopsin mislocalization chiefly would be caused by lack of retention of rhodopsin
molecules in the ROS structure. [Promising to this experiment, a previous study
fusing the atrial natriuretic factor protein with this fluorescent timer successfully was
able to identify an age-dependency of this protein release through exocytotic
activities in bovine adrenal chromaffin cells (Duncan et al., 2003).] However,
mounting evidence (such as works done by Dr. Deretic, see Introduction) generally
supports rhodopsin mistrafficking as the main cause for rhodopsin mislocalization.
98
Interestingly, another rhodopsin mutant, P23H opsin (P23H), also caused
rhodopsin mislocalization in ADRP animal models (Olsson et al., 1992; Roof et al.,
1994; Tam and Moritz, 2006). One important distinction, however, exists between
the two types of rhodopsin mislocalization: mislocalized P23H molecules were
mainly within the somata, while Q344ter-induced mislocalized rhodopsin molecules
generally were restricted to the plasma membrane (Figs. 3.3, A and C; Fig. 3.4 A).
This difference is attributed to the folding-deficiencies of Class II mutants (such as
P23H) compared to the normal folding properties of Class I mutants (such as
Q344ter). Therefore, despite the convergence to apoptosis associated with both
rhodopsin mutants, each mutant probably initiates retinal degeneration by divergent
pathways.
We have also provided evidence for Q344ter-induced retinal degeneration
which occurred in the absence of light but was accelerated by light-exposure (Fig.
3.5). [This light-independent degeneration was not caused by rhodopsin
overexpression. Since Q344ter composed of 24% of the total rhodopsin population
but that only one copy of the endogenous rhodopsin was intact in these transgenic
Q344ter retinas, the overall rhodopsin expression level of these transgenic mice was
well below levels known to induce retinal degeneration (500%, Olsson et al., 1992;
123%, Tan et al., 2001).] One explanation for the light-independent retinal
degeneration (resulting from rhodopsin mistrafficking alone) is the possible
disruption to the rod cells’ overall “physiological balance” by mislocalized
rhodopsin. Because of the high expression of rhodopsin, these cells may be deficient
99
in the total removal of these aberrantly trafficked proteins, which results in the
observed plasma membrane-bound rhodopsin molecules in the RIS and ONL. The
mere presence of rhodopsin molecules in these undesired compartments may cause
perturbations to normal activities and processes occurring at the plasma membrane of
these subcellular regions and lead to ill-fated consequences. Future studies are
needed to identify the specific mechanism(s) by which rhodopsin mistrafficking
alone (and therefore independent of light-exposure) leads to retinal dystrophy in this
ADRP mouse model. Additionally, experiments involving the aforementioned
Q344ter-timer fusion protein may provide insights into the efficiency of
removal/turnover durations of mislocalized rhodopsin molecules at the plasma
membrane.
Interestingly, the significant light-accelerated retinal degeneration was not
reported in either Q350ter transgenic Xenopus laevis (Q344ter equivalent in frogs,
Tam et al., 2006) or S334ter transgenic mice (Green et al., 2000). These different
observations most likely were attributed to different light-exposure conditions
(Wenzel et al., 2005), different animal model (in the case of the transgenic frogs),
and the absence of phosphorylation sites (in the case of S334ter mutation).
Concerning the light treatment of the transgenic Q350 frogs and S334ter mice, these
animal models were reared in 12 hr dark/moderately intense light cycles, while we
light-exposed our mice to 5 days of continuous white fluorescent light (3000 lux
intensity). When only the light-exposed Q344ter transgenic retinas experienced
severe degeneration (and not the light-exposed nontransgenic controls), we believed
100
that our light-exposure conditions were acceptable for our investigation. We
reasoned that this light-exposure treatment was within the limits to accelerate
Q344ter-induced retinal degeneration in a controlled manner. Furthermore, because
it is quite possible that different intensities and/or durations of light-exposure may
induce differing degrees of retinal degeneration, these variations in light-exposure
effects may actually reflect the observed heterogeneity of disease progression in
ADRP patients inheriting this truncated rhodopsin mutant.
Additionally, our current study reveals that mislocalized rhodopsin is capable
of light-activation. This finding allows us to formulate the hypothesis that
mislocalized rhodopsin may initiate directly a light-dependent response(s) that
lead(s) to accelerated degeneration. Such examples could be the activation of
transducin or an unknown G-protein in the RIS (Alfinito and Townes-Anderson,
2002) and/or the accumulation of the rhodopsin/arrestin complexes (Alloway et al.,
2000; Kiselev et al., 2000; Chuang et al., 2004; Iakhine et al., 2004; Chen et al.,
2006). These responses are possible, given that mislocalized Q344ter molecules in
the RIS and ONL would become accessible (directly or indirectly) to proteins not
normally encountered by rhodopsin. Also, these “abnormal” compartments may be
more conducive to the internalization of rhodopsin-associated complexes (as
opposed to the crowded ROS compartment), which was necessary for the initiation
of rod cell death (Alloway et al., 2000; Kiselev et al., 2000). Further work is
required to explore the potential contributions of these pathways (see Chapter 4).
101
Light-exposure also has been reported to accelerate retinal degeneration in
animal models expressing P23H (Naash et al., 1996; Organisciak et al., 2003;
Vaughan et al., 2003; Jozwick et al., 2006). In these situations, P23H most likely
perturbs the normal light-induced processes in rod cells without itself becoming
photolyzed, since P23H is misfolded (and therefore unable to be light-activated).
However, studies have shown that patients inheriting Class I rhodopsin mutants, i.e.
Q344ter, tend to have more severe cases of ADRP than patients inheriting Class II
rhodopsin mutants, i.e. P23H (Berson et al., 2002; Oh et al., 2003). Because patients
are never restricted to dark-only conditions, the above observation may be attributed
to the activation of an unknown yet deleterious light-dependent cell signaling
pathway(s) specific to the Class I rhodopsin mutants. Again, this hypothesis is
possible with our observation that Q344ter-dependent mistrafficked rhodopsin
molecules are capable of light-excitation.
Lastly, this effect of rhodopsin mistrafficking causing retinal degeneration
also has been reported in mice deficient in expression of tubby (tub) or tubby-like
protein 1 (tulp1) or both. [Tub-/- mice were studied by Kong et al. (2006); while
(tulp1-/-) and (tub-/-, tulp1-/-) mice were examined by Hagstrom et al.(2001).] The
functions of these two proteins (which are both localized in the RIS layer) remain
unclear. Although rhodopsin was present in the ROS, the absence of each protein
induced both: (i) mislocalization of WT rhodopsin molecules to the RIS and ONL
(and OPL); and (ii) retinal degeneration. Also, light-exposure of tub-/- mice
increased the rate of ongoing retinal degeneration. When both proteins were ablated
102
in the mice
tub-/-, tulp1-/-
, their retinas showed: (i) lack of ROS formation and (ii) a rapid
form of retinal degeneration that was more severe than the degenerations transpiring
when only one of the proteins was deleted. These similar characteristics between
transgenic Q344ter mice and the tub-/- and/or tulp1-/- mice support shared
mechanisms that lead to apoptotic photoreceptor cell death. Thus, results from
future investigations involving these mouse models not only should be compared but
also combined to gain insights into the pathways by which rhodopsin mistrafficking
induces retinal degeneration.
103
Chapter 4
Photolyzed Rhodopsin/Arrestin Complex Contributes to the
Light-Accelerated Retinal Degeneration in the Transgenic
Q344ter Rhodopsin Mutant Mouse Model

4.1. Introduction.
With the absence of the last five amino acids (QVAPA) at the carboxyl-
terminal (C-terminal) domain, the Q344ter rhodopsin mutant (Q344ter) causes
plasma membrane associated rhodopsin mistrafficking and induces retinitis
pigmentosa in humans (Chapter 3, Portera-Cailliau et al., 1994; Sung et al., 1994).
Previous studies analyzing the consequences from truncated rhodopsin mutations
reported that light had no effect on the rate of retinal degeneration in two animal
models [transgenic S334ter rats and transgenic Q350ter frogs] (Green et al., 2000;
Tam et al., 2006). These studies implied that rhodopsin mistrafficking alone
accounted for the observed retinal degeneration. However, in Chapter 3, we have
provided evidence not only for light-accelerated retinal degeneration but also for
light-activation of mislocalized rhodopsin molecules.
The next phase of this project was to investigate a potential causal
relationship between the activation of mislocalized rhodopsin molecules and the
light-accelerated photoreceptor cell death. One potential mechanism is the
accumulation of rhodopsin/arrestin complexes in abnormal cellular compartments.
Three Drosophila models showed that genetic mutations which created conditions
104
conducive for persistent and stable binding between light-activated rhodopsin (R*)
and a Drosophila-specific arrestin species, arrestin2 (Arr2), induced retinal
degeneration (Alloway et al., 2000; Kiselev et al., 2000; Iakhine et al., 2004). The
first Drosophila model expressed a loss-of-function mutation in the norpA gene
(Alloway et al., 2000). This gene encodes for phospholipase C (PLC), the effector
molecule of the Drosophila phototransduction cascade. The absence of PLC activity
precludes the light-dependent photoreceptor cell depolarization, which is a requisite
event for Arr2 phosphorylation. Since Arr2 phosphorylation is necessary for Arr2
release from R*, this norpA mutation ultimately results in formation of stable
rhodopsin/arrestin complexes. In the second Drosophila model, a loss-of-function
mutation in the rdgC gene, which encodes the phosphatase that targets
phosphorylated rhodopsin, results in the persistent phosphorylated state of the
rhodopsin molecules. This condition also generates a situation for stable
rhodopsin/arrestin complexes (Kiselev et al., 2000). The third Drosophila model
creates a similar condition for the stability of these complexes by expressing a
constitutively active form of rhodopsin, NinaE
pp100
(Iakhine et al., 2004). In all three
fly models, the accumulation and subsequent internalization of these R*/arrestin2
complexes led to apoptosis within the photoreceptor cells.
Interestingly, a recent study by our lab (Chen et al., 2006) revealed that a
similar mechanism also occurred in transgenic mice expressing a constitutively
active opsin mutant (K296E). Therefore, this mouse model is analogous to the
Drosophila model expressing NinaE
pp100
. Here, we observed mistrafficking of this
105
opsin mutant in the rod inner segment (RIS) and outer nuclear layer (ONL) regions
in addition to its correct localization to the rod outer segment (ROS). This opsin
mislocalization pattern is similar to that of Q344ter-induced rhodopsin
mislocalizations. Moreover, the K296E molecules were shown to: (A) interact with
visual arrestin (SAg, the vertebrate counterpart to arrestin2) independent of light due
to K296E’s constitutively hyperphosphorylated state); and (B) induce a more severe
form of retinal degeneration in the presence of SAg. All these observations together
support a toxic effect from the K296E/SAg complexes, since these interactions not
only would be stable (and therefore allow accumulation of these complexes) but also
that these complexes are formed in the RIS and ONL regions (and therefore may
become accessible to other proteins, particularly those that initiate apoptosis).
With respect to Q344ter, its induction of rhodopsin mislocalization creates a
similar condition to the transgenic K296E mouse model with one important
difference: in the K296E mouse retinas, the formation of the rhodopsin/SAg
complexes is independent of light; while in the Q344ter mouse retinas, the formation
of these complexes is light-dependent. Additionally, because Q344ter retains the
phosphorylation sites (unlike the S334ter mutant), the accumulation of the
rhodopsin/arrestin complexes in abnormal cellular compartments is a viable
hypothesis to account for the light-enhanced retinal damage observed in the
transgenic Q344ter mouse model (Fig. 3.5).
Another potential light-dependent mechanism of photoreceptor cell death
investigated in this study is the activation of transducin and/or an unknown G-
106
protein(s) in the rod inner segment (RIS) layer as proposed by Alfinito and Townes-
Anderson (2002). By using tissue-cultured salamander rod cells, they demonstrated
that mislocalized R* molecules initiated an adenylate cyclase-dependent signaling
cascade leading to rod cell apoptosis. Furthermore, the partial rescue in the presence
of pertussis toxin [which inhibits G
i
, G
o
, and transducin (Tr)] implicated the
participation of transducin and another unknown G-protein(s) in the activation of the
above signaling cascade. By employing a [
35
S]GTP γS assay on various frozen
retinal sections from transgenic Q344ter mice and the necessary controls, we
investigated whether Q344ter molecules stimulates G-proteins in the RIS and/or
ONL regions of the retina, which would support another mode of light-accelerated
retinal degeneration in this mouse model (Fig. 3.5).

4.2. Materials and Methods.
4.2.1. Generation of mouse lines.
All mice were treated in accordance with the ARVO Statement for the Use of
Animals in Ophthalmic and Vision Research. Transgenic Q344ter
rho+/-
mice (Chapter
3) further were crossbred into the following genetic backgrounds: [1] rho+/-,
transducin α-subunit (Tr α)-/-; and [2] rho+/-, rhodopsin kinase (RK) -/-, Tr α -/-.

4.2.2. Retinal morphometry.
Retinal degeneration was measured by retinal morphometry as previously
described in Chapter 3 (Materials and Methods). Transgenic mice and their negative
littermate controls with the rho+/-, RK -/-, Tr α -/- genetic background were either
107
dark-reared only or dark-reared and exposed to five days of continuous light (3000
lux, undilated pupils) preceding their sacrifice (P28-31). As before, only darkly
pigmented mice (black or agouti) were light-exposed to minimize potential
degeneration caused by high light intensity in mice lacking the protective effect of
pigmented irises (Noell et al., 1966; Noell, 1980; Hao et al., 2002; recent review
Wenzel et al., 2005). Additionally, as mentioned in Chapter 3, “because a subregion
within the superior half near the optic nerve was revealed to be the most sensitive to
light damage in previous studies (Danciger et al., 2000; Vaughan et al., 2003), ONL
thickness measurements obtained from this particular subregion represented the
statistically significant extent of retinal degeneration occurring within a particular
retina.” T-tests of sample populations were used with α = 0.05.

4.2.3. [
35
S]GTP γS (Guanosine 5’-O-( γ-thio) triphosphate) in situ assay.
To obtain dark-adapted frozen retinal slices of Q344ter mice with the two
following genetic backgrounds: [A] rho+/-; and [B] rho+/-, Tr α-/-, mice were dark-
reared and sacrificed at P28-31. Unless otherwise stated, all work (including
cryosectioning) was performed under infrared light with night vision goggles (model
# E8600; Excalibur Electro Optics Inc., Fogelsville, PA). After cauterizing the
superior end of each eye for orientation and removing the cornea and lens, the
eyecup was embedded in 3% low-melting agarose (Sigma-Aldrich, St. Louis, MO)
dissolved in Ames’s like solution [10 mM HEPES pH 7.4, 2 mM NaHCO
3
, 110 mM
NaCl, 2.5 mM KCl, 1.0 ml CaCl
2
, 1.6 mM MgCl
2
, 10 mM glucose). The agarose-
embedded eyecups were frozen in Tissue Tek
®
O.C.T. compound (Sakura Kinetek
108
U.S.A. Inc., Torrance, CA) by using liquid N
2
. The frozen eyecups were sectioned
(10 µm) with a Jung CM 3000 cryostat machine (Leica Inc., Deerfield, IL).
This [
35
S]GTP γS assay was based on previously described protocols with
modifications (Taylor et al., 2000; Sim-Selley and Childers, 2002). After allowing
frozen sections to reach room temperature (RT), the eyecup sections were incubated
(10 min) in rod outer segment (ROS) buffer [20 mM HEPES pH 7.4, 120 mM KCl, 5
mM MgCl
2
, 1 mM dithiothreitol (DTT), 100 µm phenylmethylsulfonyl fluoride
(PMSF)] to remove the surrounding mounting medium. To equilibrate the sections
to assay conditions, the tissue samples were incubated for 1 hour in preincubation
buffer [ROS buffer, 100 µm guanosine – 5’-O- diphosphate (GDP, disodium salt
form, MP Biomedicals, Irvine, CA), 2 mM β-nicotinamide adenine dinucleotide
phosphate (NADPH, reduced tetra(cyclohexylammonium) salt form, Sigma-
Aldrich)]. The tissue sections were then incubated with either the ‘hot” reaction
buffer [preincubation buffer, 100 nM GTP γS (Roche Diagnostics, Indianapolis, IN),
20 nM [
35
S]GTP γS (1000 Ci/mmol; GE Healthcare, Piscataway, NJ)] or “hot/cold”
reaction buffer (preincubation buffer, 20 µm GTP γS, 20 nM [
35
S]GTP γS) for 20 min
with the dark sections remaining under infrared illumination and “lighted” sections
exposed to bright white light illumination (> 3000 lux). After incubation with either
of the [
35
S]GTP γS reaction buffers, all tissue samples were transferred back to the
dark condition. The sections were washed 4 X 5 min with ROS buffer and 1 X 30
sec with H
2
O, air dried (20 min), submerged in autoradiography emulsion NTB
109
(Eastman Kodak Co., Rochester, NY), and allowed to dry (30 min). Afterwards, the
sample slides were stored in a light-tight container at -80˚C for 3 days.
To develop the film, the sample slides were submerged (3.5 min) in
Developer-19 solution (Eastman Kodak Co.), rinsed with H
2
O, and submerged (5
min) in Kodak fixer solution. After drying the slides for 20 minutes, the samples
were stained with 0.4% Toluidine Blue O solution (Sigma-Aldrich) and washed with
phosphate buffer saline (PBS). This staining of nuclei provided general orientation
of the retinal cell layers. Dehydration was accomplished by the following steps: 1 X
10 min with 50%, 70%, 90% ETOH, 2 X 10 min with 100% ETOH, 2 X 10 min
xylene. The sections were viewed and photographed with an AxioPlan 2 imaging
system (Carl Zeiss, Inc., Goettingen, Germany).

4.3. Results.
4.3.1. The accumulation of R*/SAg complexes contributes to retinal degeneration in
transgenic Q344ter mice.
Since we have shown that mislocalized rhodopsin molecules are capable of
light-activation (Chapter 3), these mislocalized R* may be internalized by
endocytosis through interactions with SAg molecules, which are also present in the
RIS and ONL (Whelan and McGinnis, 1988; Mendez et al., 2003; Zhang et al.,
2003a; Elias et al., 2004). To determine if the R*/SAg complexes were involved, we
further crossbred the Q344ter
rho+/-
mice and their negative littermate controls into the
rho+/-, rhodopsin kinase (RK)-/-, Tr α-/- genetic background. Removing Tr α was
necessary when deleting RK, since the absence of RK leads to prolonged
110
phototransduction in the presence of dim light. This constitutive activity of the
phototransduction cascade in itself causes retinal degeneration (Chen et al., 1999a;
Chen et al., 1999b; Hao et al., 2002). With this genetic background, we would
expect to significantly reduce R*/SAg complex formation in the light-exposed
transgenic Q344ter retinas by inhibiting RK expression, since rhodopsin
phosphorylation is catalyzed by RK (Chen et al., 1999a) and is a general requisite for
SAg binding (Kuhn et al., 1984; Kuhn and Wilden, 1987; recent review Arshavsky,
2002). [The genetic background (rho+/-, SAg-/-, Tr α-/-) would provide a more
direct test for potential toxicity from mislocalized R*/Sag complexes, since this
background would inhibit completely this complex formation. A previous study,
however, has reported evidence for an unknown genetic factor that appears to make a
subpopulation of retinas with the SAg-/-, Tr-/- background susceptible to severe
degeneration even at low light intensities and independent of any transgene
expression (Hao et al., 2002). Therefore, any rescue or severe retinal degenerations
observed in mouse retinas with this particular genetic background would not be able
to discriminate effects originating from the transgenic Q344ter expression or from
this unknown genetic factor.]
If the R*/SAg complex contributed to the severe retinal degeneration in the
light-exposed transgenic Q344ter
rho+/-
background (Fig. 3.5), we would expect rescue
(as shown by greater ONL thickness) in the light-exposed transgenic retinas with the
rho+/-, RK-/-, Tr α-/- genetic background compared with the light-exposed transgenic
Q344ter
rho+/-
retinas. At P28-P31, we sacrificed the crossbred mice that were either
111
dark-reared only or dark-reared and exposed to five days of continuous light (3000
lux, undilated pupils). We measured the degree of degeneration by retinal
morphometry based on a previous method (Danciger et al., 2000). Measuring the
thinning of ONL is a quick and suitable protocol to determine the extent of retinal
degeneration (Michon et al., 1991). As before, only darkly pigmented mice (black or
agouti) were light-exposed to minimize potential degeneration caused by high light
intensity (Noell et al., 1966; Noell, 1980; Hao et al., 2002; recent review Wenzel et
al., 2005). Additionally, for statistical analysis between sample populations, ONL
thickness measurements were recorded within a subregion of the superior half and
near the optic nerve of each retina (Fig. 4.1, green bar with asterisk) which was
revealed to be the most sensitive to light damage in previous studies (Danciger et al.,
2000; Vaughan et al., 2003). This narrowing of the retinal area also would reduce
the inherent variability of ONL thicknesses observed within a whole retina (see
Materials and Methods).
As shown in Fig. 4.1 A, multiple retinal degenerations were observed with
rho+/-, RK-/-, Tr α-/- genetic background: (i) Light-exposure of nontransgenic retinas
caused a slight but statistically significant ONL thinning when compared with dark-
reared nontransgenic retinas; (ii) dark-reared Q344ter transgenic retinas showed
moderate ONL thinning when compared with dark-reared nontransgenic retinas
(again demonstrating that Q344ter-induced rhodopsin mistrafficking by itself caused
retinal degeneration); and (iii) light-exposed transgenic Q344ter retinas revealed the
greatest degree of ONL thinning in this genetic background. However, in this latter
112
Figure 4.1. The R*/SAg complex contributes to retinal degeneration in the
transgenic Q344ter mouse model. (A) To investigate the possible role of R*/SAg
complexes to light-accelerated degeneration in transgenic retinas, we disrupted this
complex formation by preventing R* phosphorylation with the RK-/- background
(Kuhn et al., 1984; Kuhn and Wilden, 1987; Chen et al., 1999a; Arshavsky, 2002).
Transgenic Q344ter mice and their negative littermate controls were either dark-
reared only or dark-reared and exposed to 3000 lux continuous light for 5 days. All
mice were sacrificed between P28-31. Additionally, only darkly pigmented mice
(black or agouti) were light-exposed. The extent of retinal degeneration was
determined by measuring the ONL thicknesses within a region in the superior half
near the optic nerve site (demarcated by green bar with asterisk). This particular
region is known to be susceptible to light damage (see text and Materials and
Methods section). The four pictures to the right are representatives of the ONL
within this light sensitive region (scale bar = 20 µm). [For all statistically significant
differences, p ≤ 0.05.] (B) Comparison of retinal degenerations of transgenic
Q344ter mice between rho+/- and rho+/-, RK-/-, Tr α-/- backgrounds. The colored
bars indicate reductions in ONL thickness when comparing two sample populations.
Because the preclusion of rhodopsin phosphorylation reduced the severity of retinal
degeneration (compare yellow and aqua green bars from the two genetic
backgrounds), this observation provides evidence that the R*/SAg complex is toxic
in the transgenic Q344ter retinas. Also note that in both genetic backgrounds, (i)
light-exposure does induce a slight albeit statistically significant degeneration in
113
negative littermate controls (blue bars); and (ii) rhodopsin mistrafficking alone (and
therefore independent of light) causes ONL thinning.
114

Figure 4.1. The R*/SAg complex contributes to retinal degeneration in the
transgenic Q344ter mouse model.
115
case, retinal degeneration was not as severe as the corresponding retinal degeneration
in the light-exposed transgenic Q344ter
rho+/-
mice (compare yellow and aqua green
bars in Fig. 4.1 B). This intergroup comparison of transgenic Q344ter retinas with
different genetic backgrounds (rho+/- versus rho+/-, RK-/-, Tr α-/-) revealed that the
reduction in mislocalized R*/SAg complex formation corresponded to less ONL
thinning. Therefore, this diminished severity in the light-dependent retinal
degeneration of Q344ter
rho+/-,RK-/-, Tr α-/-
retinas (compared to light-exposed
Q344ter
rho+/-
retinas) shows that the R*/SAg complex formation has a toxic effect in
the transgenic Q344ter mouse model.

4.3.2. Light-dependent GTP loading in transgenic Q344ter retinas and their
nontransgenic counterparts.
Because diminished formation of mislocalized R*/SAg complexes did not
produce a full rescue (in terms of ONL thickness), other light-dependent mechanisms
appear to be contributing to the light-accelerated retinal degeneration in this
truncated rhodopsin mouse model. Another potential mechanism is the activation of
an unknown G-protein and/or transducin in the inner segment by photolyzed,
mislocalized rhodopsin molecules as shown in cultured salamander rod cells
(Alfinito and Townes-Anderson, 2002). Previous in vitro studies also have
demonstrated rhodopsin’s capacity to interact with other G-proteins (especially those
in the Gi/o family) or to activate pertussis toxin-sensitive, G-protein-mediated
cascades in systems where Tr is absent (Kanaho et al., 1984; Tsai et al., 1987; Weiss
et al., 1990). Because other G-proteins most likely are present in rod cells and of this
116
potential “cross-talk”, mislocalized R* molecules could interact with various G-
proteins (including Tr) in the RIS and ONL regions. In turn, this situation may
activate a yet unknown signaling cascade that leads to photoreceptor cell death.
By using a GTP loading assay involving a radioactive and nonhydrolyzable
GTP analog, [
35
S]GTP γS (Fig. 4.2), we attempted to detect R*-catalyzed GTP
loading (presumably of G-proteins) in unfixed frozen retinal sections from transgenic
Q344ter mice. This assay included the following genetic backgrounds: (i) rho+/-;
and (ii) rho+/-, Tr α-/-. With the rho+/- background (Figs. 4.2, A-F), we expected to
observe both a light-dependent [
35
S]GTP γS loading of transducin in the ROS and a
difference in light-dependent [
35
S]GTP γS loading patterns between retinas from
transgenic Q344ter mice and nontransgenic controls. More specifically, this
difference would be an increased light-dependent [
35
S]GTP γS loading in the RIS
region of transgenic retinas due to mislocalized rhodopsin molecules capable of
light-activation compared to their negative littermate controls. With the rho+/-, Tr α-
/- background (Figs. 4.2, G-L), we hoped to attribute any differences in light-
dependent [
35
S]GTP γS loading patterns to G-proteins other than transducin.
To observe basal activity and serve as a control, a set of all involved retinal
sections were treated in the “dark-only” condition (Fig. 4.2, left column), which
showed minimal [
35
S]GTP γS loading. This loading can be perceived as the black
granules resulting from radioactivity within the [
35
S]GTP γS molecules that was
exposed an autoradiography emulsion film. Furthermore, Toluidine Blue O staining
was necessary for general orientation of the retinal cell layers. The upper purple-
117
Figure 4.2. Light-dependent GTP γS loading (20 min exposure) in transgenic
Q344ter frozen retinal sections. A [
35
S]GTP γS loading assay was performed on
unfixed frozen retinal sections from mice with the genetic backgrounds: rho+/- (A-
F); and rho+/-, Tr α-/- (G-L). For orientation in each retinal section, the upper
purple-bluish staining represents the ONL, while the lower purple-bluish staining
reveals the INL. Because we currently could not clearly distinguish the border
between the ROS and RIS, we assigned the region between the RPE and ONL as the
ROS/RIS bilayer with the upper half being the ROS and the lower half being the
RIS. Within each background, nontransgenic/control retinas and transgenic Q344ter
retinas are displayed as the top and bottom rows, respectively. The three
experimental conditions were: dark only (left column); 20 min bright-light exposure
(middle column); and 20 min bright-light exposure in the presence of “cold” GTP γS
(right column). As expected, a baseline [
35
S]GTP γS loading (as shown as black
granules) in the dark-only condition occurred in all retinal sections regardless of
genotype (left column). Upon light-exposure, both nontransgenic (B) and transgenic
Q344ter (E) retinas with the rho+/- genetic background displayed [
35
S]GTP γS
loading throughout the ROS/RIS region, which is located between the retinal
pigment epithelial (RPE) layer and the ONL (upper purplish layer). Concerning the
(rho+/-, Tr α-/-) genetic backgrounds, the light-exposed nontransgenic retinal sections
(H) consistently displayed greater [
35
S]GTP γS staining in the RIS region, while
light-exposed transgenic Q344ter retinal sections (K) consistently showed
[
35
S]GTP γS loading throughout the ROS/RIS bilayer. As a control indicating
118
specificity, the vast majority of [
35
S]GTP γS loading was competitively inhibited by
the addition of nonradioactive GTP γS (right column). Taking into account the
overall limitations of this assay (such as the inability to demarcate distinctly the
border between the ROS and RIS), this difference in [
35
S]GTP γS loading patterns
can be explained by the often thin ROS layer observed in retinas expressing Q344ter
(see text). Also note the following: the red arrows display an artifactual tear in the
Q344ter
rho+/-
retinal sections; and the light-dependent increased [
35
S]GTP γS loading
in the plexiform layers (border between the ONL and INL and the lower border of
the INL) were noted but not addressed in this study. (All mice were sacrificed
between P28-31; scale bars = 20 µm)
119

Figure 4.2. Light-dependent GTP γS loading (20 min exposure) in transgenic
Q344ter frozen retinal sections.
120
bluish staining demarcates the ONL, while the lower purple-bluish staining reveals
the inner nuclear layer (INL). (In some sections, we also observed a lower third
stained layer, which was the ganglion cell nuclear layer.) Concerning the ROS and
RIS, these two layers are located between the RPE and ONL regions (see Fig. 4.2
labeling). However, at the present time we could not distinctly delineate the fine
border between the ROS and RIS. Thus, albeit crude, within the ROS/RIS bilayer
we designated the upper half as the ROS and the lower half as the RIS.
With the 20 min bright light-exposure (Fig. 4.2, middle column), this time
period would be more than sufficient for the light-dependent [
35
S]GTP γS loading of
all available protein candidates to compensate for potential lowered efficiency in
molecular activity that may be compromised during preparation of the frozen retinal
samples. Moreover, this duration of light-exposure was necessary, since a 5 min
light-exposure failed to reveal any significant [
35
S]GTP γS loading in the ROS (see
Fig. 5.1). With the rho+/- background (Figs. 4.2, A-F), we observed light-dependent
increased [
35
S]GTP γS loading (as shown by black granules) in the ROS and RIS in
both nontransgenic (Fig. 4.2, compare B to A) and transgenic Q344ter (Fig. 4.2,
compare E to D) retinal sections. This increased [
35
S]GTP γS loading in the ROS
was attributed to activation of rod transducin molecules by light-excited rhodopsin
molecules. As for the increased [
35
S]GTP γS loading in the RIS, this observation
probably was the result of cone transducin activation, since cone outer segments
have been observed within the RIS layer (Zhu et al., 2002; Le et al., 2004). Also,
because of light-dependent translocation of Tr α (Mendez et al., 2003; Zhang et al.,
121
2003a; Elias et al., 2004), the uptake of [
35
S]GTP γS by Tr α and subsequent light-
dependent translocation also may have contributed to this radioactive labeling within
the RIS. Interestingly, we also observed greater [
35
S]GTP γS loading in the outer
plexiform layer (between the ONL and INL) and in the inner plexiform layer (bottom
of INL) that was independent of transgene expression. The cause and significance of
these increased light-dependent [
35
S]GTP γS loading presently are unknown.
With the rho+/-, Tr α-/- background (Figs. 4.2, G-L), the nontransgenic
control retinal sections showed only the increased light-dependent [
35
S]GTP γS
loading in the RIS (again most likely due to cone transducin activation; Fig. 4.2,
compare H to G); while the transgenic Q344ter retinal section produced an increased
light-dependent [
35
S]GTP γS loading throughout the ROS/RIS bilayer (Fig. 4.2,
compare K to J). The absence of [
35
S]GTP γS loading in the ROS of nontransgenic
light-exposed retinas with the (rho+/-, Tr α-/-) genetic background (Fig. 4.2 H) was
due to the absence of Trα molecules. Concerning the transgenic Q344ter retinal
sections (which consistently showed light-dependent increased [
35
S]GTP γS loading
throughout the ROS/RIS bilayer in both genetic backgrounds [Figs. 4.2, E & K]), a
simple explanation for the radioactively labeled GTP γS loading patterns in the
transgenic retinas involved their often thin ROS layer (see Fig. 3.3). We suspect that
the [
35
S]GTP γS loading in the transgenic retinal sections largely was occurring in the
RIS region similar to that of nontransgenic retinal sections. Hence, this
interpretation would signify no real difference in the [
35
S]GTP γS loading patterns
between the transgenic and nontransgenic control retinal sections. Improvements in
122
this particular assay (especially refinements in delineating the various retinal
subcellular layers) are warranted to better assess the potential contribution of the
activation of Tr or an unknown G-protein(s) in the RIS towards the light-accelerated
degeneration in these transgenic mice.

4.4. Discussion.
Our current study substantiated that light-exposure accelerates Q344ter-
dependent retinal degeneration. Furthermore, we have demonstrated that the toxicity
caused by the R*/SAg complexes observed in both Drosophila and in the K296E
mouse model also occurs in the transgenic Q344ter mouse model (Alloway et al.,
2000; Kiselev et al., 2000; Iakhine et al., 2004; Chen et al., 2006). An important
distinction between the transgenic K296E and Q344ter mouse models, however, is
that the R*/SAg toxic effect is light-dependent in the transgenic Q344ter mice (due
to the light-activation capability of mislocalized rhodopsin) and light-independent in
the transgenic K296E mice (due to the constitutive activity of this mutant).
Regardless of the light-dependency, the transgenic Q344ter mouse model shows
“enlightenment” on the pervasiveness of the potential toxic effect from the R*/SAg
complex formation.
Furthermore, another study showed that ectopic co-expression of another
constitutively active opsin mutant (R135L) together with SAg in mammalian
cultured cells (i.e., HEK cells) disrupted the endocytic pathways, which was believed
to eventually cause apoptosis (Chuang et al., 2004). This study also ectopically
expressed R135L in mouse retinas via in vivo electroporation. Here, SAg
123
localizations were consistent with formation of R*/SAg complexes within the RIS
and ONL, since Chuang et al. observed localizations of endogenous SAg molecules
in the RIS and ONL of R135-transfected cells that also expressed GFP. No such
SAg staining patterns, however, were observed in the retinas of mice ectopically
expressing Q344ter. Again, the levels of light intensities may not have been
sufficient to produce a phenotype in the light-exposed retinas transfected with
Q344ter mutant gene. Nevertheless, Chuang et al. (2004) potentially may have
identified downstream events in the pathway by which the accumulation of R*/SAg
complexes in the wrong cellular compartments leads to apoptosis in photoreceptor
cells. [Consequently, current work in Dr. Jeannie Chen’s laboratory is investigating
the endocytic capability of SAg molecules, since the internalization of the
R*/arrestin complexes was necessary to induce cell death in the Drosophila
photoreceptors (Alloway et al., 2000; Kiselev et al., 2000; Orem and Dolph, 2002).]
Because inhibiting formation of Q344ter/SAg complexes did not fully rescue
retinal degeneration (as shown in the rho+/-, RK-/-, Tr α-/- genetic background),
additional investigations are necessary to elucidate these additional mechanisms.
Although results from the [
35
S]GTP γS loading assay were inconclusive, refinements
to this procedure are essential to determine (or discount) the contribution of the
mechanism involving G-protein activation in the RIS within the transgenic Q344ter
mouse model. The additional deletion of the cone Tr α gene also may improve this
assay by removing any involvement of the cone Tr α protein to the observed RIS
[
35
S]GTP γS loading patterns. Perhaps, this scenario would further isolate the light-
124
dependent activities within the rod cells, so that differences between transgenic
Q344ter and nontransgenic retinas would be detected.
Besides the light-accelerated pathways, still another aspect of this ADRP
mouse model is elucidating the photoreceptor cell death caused by rhodopsin
mislocalization alone (and therefore independent of light-exposure). This
dissertation only observed but did not investigate the mechanism(s) that underlie this
mode of degeneration. However, a hypothesis was proposed within the Discussion
section of Chapter 3 involving the disruption by mislocalized rhodopsin molecules of
normal cellular activities within the RIS region, where much of the metabolic
machinery exists in rods.
Also mentioned previously in Chapter 3, two other mouse models are known
to cause rhodopsin mislocalization and subsequent retinal degeneration, tubby-/-
and/or tubby-like protein1 (tulp1)-/- (Hagstrom et al., 2001; Kong et al., 2006).
Interestingly, increased activities of caspase-3, -9, and cytosolic cytochrome c were
observed in light-exposed tubby-/- retinas experiencing light-accelerated retinal
degeneration. Since tubby-/- and transgenic Q344ter mice share many common
features regarding retinal degeneration (Chapter 3), these downstream events also
may be applicable to the transgenic Q344ter mouse model. Future studies involving
tubby/tulp1 and Q344ter will provide a broader and clearer “picture” to the
mechanisms by which rhodopsin mistrafficking induces rod cell death.

125
Chapter 5
Miscellaneous Personal Observations

Throughout this rhodopsin transport project, I have made certain observations
that at this stage only should be viewed as tangents to the main storyline of this
project. In this somewhat informal supplemental section of my dissertation, I wish to
present these observations, whose interpretations are either highly speculative or part
of preliminary experiments. Since the study of rhodopsin mistrafficking is far from
over, further developments in this field (or perhaps protein trafficking in general)
someday may incorporate these presently tangential data.

5.1. Potential Correlation Between Penta-Phosphorylated Q344ter Molecules
and the Lowered Reproducibility in Single Photon Responses in Rods
Expressing this Truncated Rhodopsin Mutant.
The observation that the penta-phosphorylated Q344ter molecule was the
maximum phosphorylated rhodopsin species in the Q344ter transgenic retinas caught
my attention. This result may explain a previous study concerning Q344ter
properties (Sung et al., 1994). Single-cell recordings showed that Q344ter-induced
single photon responses were quite similar to WT rhodopsin-induced responses.
Moreover, closer analysis revealed that the parameter of “exponential decay constant
for the recovery of dim flash response” was not statistically significant between the
two types of rod species. The main reason was the greater variability in the average
126
responses in Q344ter transgenic rods as shown by the higher standard deviation. In
other words, the reproducibility of the dim light flashes was lower from transgenic-
expressing rods than from WT rods.
This lowered reproducibility taken together with the maximum penta-
phosphorylated species (instead of the hexa-phosphorylated species) could be
elucidated by recent findings from Doan et al. (2006). They showed that the
reproducibility of dim flash responses in WT murine rods could be explained by the
presence of six “capable” phosphorylated sites in rhodopsin. When one of these sites
was mutated in a manner that prevented phosphorylation, i.e. either S343A or S338A
rhodopsin mutant, the reproducibility was lowered. Additionally, mutating two or
more phosphorylation sites further decreased this reproducibility value.
Our IEF assays indicated that the Q344ter mutation somehow rendered one of
the phosphorylation sites incapable of being phosphorylated, which would cause a
lowered reproducibility in (or more varied) dim flash responses. Unfortunately, my
observation cannot be tested in vivo, since the studies by Doan et al. (2006) requires
a “pure” population of rhodopsin species in detecting the subtle differences among
various parameters. [In contrast, the mutant rod cells studied by Sung et al. (1994)
expressed both WT and Q344ter molecules.] As previously described, a pure
population of Q344ter molecules precludes obvious ROS formation (Fig. 3.4), which
renders the single-cell recording of these mutant rods improbable. Additionally, the
greater variability observed in Sung et al.’s study (1994) also may be attributed to
ongoing retinal degeneration and not to the absence of a phosphorylation site in the
127
remaining C-terminus of Q344ter molecules. Hopefully, future innovative
techniques can address further the extent of phosphorylation of this truncated
rhodopsin mutant and its possible influence on reproducibility at the level of the
single photon response.

5.2. Unexpected Absence of [
35
S]GTP γS Loading in the ROS of
Nontransgenic
rho+/-
Retinal Sections after 5 min Light-Exposure.
As mentioned in Chapter 4, a 20 min light-exposure in the [
35
S]GTP γS
loading assay was necessary to produce significant radioactive labeling in the rod
outer segment (ROS) of nontransgenic
rho+/-
frozen retinal sections. When the light-
exposure was only 5 min, the levels of [
35
S]GTP γS loading in the ROS of the
nontransgenic
rho+/-
frozen retinal sections conspicuously was lower than expected
(Fig. 5.1, compare B to A; green arrowhead in B). This 5 min light-exposure is more
than ample time to activate transducin by GTP/GDP exchange, given that in vivo
experiments clearly have demonstrated that the activation of transducin molecules by
light-excited rhodopsin (R*) is in the milliseconds time frame (Rodieck, 1998).
Also, the argument that this 5 min light-incubation could be sufficient time for the
subsequent inactivation of transducin and/or other G-proteins by their inherent
GTPase activity (and consequently result in the absence of radioactive labeling) can
be countered by two reasons: (i) ROS radiolabeling actually required a longer (and
not shorter) incubation period (20 min); and (ii) [
35
S]GTP γS is a nonhydrolyzable
analog of GTP.
128
Figure 5.1. Light-dependent GTP γS loading (5 min exposure) in transgenic
Q344ter frozen retinal sections with either the rho+/- or rho+/-, Tr α-/- genetic
background. The [
35
S]GTP γS loading assay was repeated with the same conditions
as Fig. 4.2 with one exception – the light-exposure was only 5 min instead of 20 min.
With the previous 20 min light-exposure, both ROS and RIS regions showed
increased light-dependent [
35
S]GTP γS loading in the control nontransgenic
rho+/-

retinal sections. (See Chapter 4 for Materials & Methods and for cell layer
orientation.) This radioactive labeling pattern could be attributed to the light-
dependent translocation of [
35
S]GTP γS-tagged Tr α molecules from the ROS to the
RIS (Mendez et al., 2003; Zhang et al., 2003a; Elias et al., 2004) or the activation of
both rod and cone Tr α species. By reducing the light-exposure time, we predicted
that the first explanation, if occurring, would be minimized. [As before, the left
column (A, D, G, J) depicts basal radioactive labeling in dark sections; the middle
column (B, E, H, K) shows light-dependent increases in [
35
S]GTP γS loading; and
the right column (C, F, I, L) displays the control condition in which radioactive
labelng was decreased competitively by the presence of “cold” GTP γS.]
Surprisingly, 5 min light-exposure of control rho+/- retinal sections resulted in
greater radioactive labeling in the RIS than in the expected ROS (green arrowhead;
compare B to A). This observation indicated that upon light-exposure molecules in
the RIS (most likely cone Tr α) were binding to the freely available [
35
S]GTP γS
before the transducin molecules in the ROS. Similar to Fig. 4.2, this 5 min light-
exposure displayed increased [
35
S]GTP γS loading in the region between the RPE
129
layer and ONL in transgenic Q344ter
rho+/-
retinal sections (compare E to D). Results
from retinal sections with the rho+/-, Tr α-/- genetic background (G-L) were
consistent with those from the rho+/- background of this figure and from the 20 min
light exposure with the same rho+/-, Tr α-/- genetic background (Fig. 4.2): light-
exposed nontransgenic retinal sections produced greater [
35
S]GTP γS in the RIS than
in the ROS (G and H), while light-exposed Q344ter retinal sections showed similar
increased radioactive labeling between the RPE and ONL layers (J and K). As
before, one explanation for the light-increased radioactive loading patterns in the
transgenic Q344ter retinal sections is the consistent narrowing of ROS region (see
Chapter 4 - [
35
S]GTP γS loading), which implies that the radioactive labeling patterns
in transgenic retinal sections would be similar to those in the nontransgenic retinal
sections.
130

Figure 5.1. Light-dependent GTP γS loading (5 min exposure) in transgenic
Q344ter frozen retinal sections with either the rho+/- or rho+/-, Tr α-/- genetic
background.

131
As mentioned previously, the sample preparation for this GTP γS loading
assay most likely accounted for the slower kinetics occurring within these retinal
slices. Unfortunately, no explanation can be provided to account for the greatly
exaggerated light-exposure time (20 min) required for the observed ROS [
35
S]
radiolabeling. However, because RIS [
35
S] radiolabeling was achieved with the 5
min light-exposure, this observation suggests [
35
S]GTP γS loading of cone transducin
molecules, which is consistent with the much faster cone cell-mediated light
responses (Rodieck, 1998).
As for the detection of G-protein activation in the RIS by mislocalized R*
molecules with this 5 min light-exposure, interpretations of the radioactive labeling
patterns were consistent with those from the previously described 20 min light-
exposure. Namely, no real differences in [
35
S]GTP γS loading patterns were detected
between transgenic Q344ter and nontransgenic retinas when taking into account the
shortened ROS observed in the transgenic retinas.

5.3. The Presence of Mutant Rhodopsin Molecules in Transgenic Q344ter
rho-/-
Retinas Lowers the Rate of Degeneration Compared to Rho-/- Retinas.
We performed a simpler variation of retinal morphometry on retinas from
dark-reared aged P57-59 Q344ter
rho-/-
mice and their aged matched negative
littermate controls (Fig. 5.2). Here, only four regions of the retina were measured
(Concepcion et al., 2002). Interestingly, the two regions adjacent to the optic nerve
showed a statistically significant greater ONL thickness in the transgenic Q344
retinas. This result showed that the mere presence of a rhodopsin species (despite its
132

Figure 5.2. Expression of truncated mutant rhodopsin decreases progression of
retinal degeneration in rho-/- mice. Retinal morphometry was performed on
retinal sections from P57-59 dark-reared Q344ter
rho-/-
and their age-matched negative
littermate controls. (For each group, n = 3.) Specifically, we measured the outer
nuclear layer (ONL) thickness at four distinct regions of the retina (see figure insert,
Concepcion et al., 2002). Although no statistically significant differences in ONL
thickness measurements were observed between the two groups of mice in retinal
regions A & D (p > 0.05), the ONL thickness measurements were statistically
significantly greater in Q344ter
rho-/-
retinas than in rho-/- retinas in regions B (p =
0.0088) & C (p = 0.0014). Therefore, the presence of Q344ter appeared to delay the
rate of degeneration occurring in rho-/- mice (Lem et al., 1999).
133
mutation) somehow prolonged photoreceptor viability when compared to retinas
completely devoid of any rhodopsin species (rho-/- retinas). Additional
investigations are essential to elucidate the mechanisms accounting for this
observation. At this time, no explanation can be provided for this phenomenon.

5.4. Potential Contribution from β-Arrestin to the Light Accelerated Damage in
Transgenic Q344ter Retinas.
Because the R*/SAg complex could not completely account for the light-
accelerated ONL thinning in transgenic Q344ter mice and that the GTP γS loading
assay essentially was inconclusive, one intriguing possibility is the participation of β-
arrestin molecules to the light-dependent retinal degeneration through a variation of
the toxic R*/arrestin complex accumulation – the R*/ β-arrestin complex. Nicolas-
Leveque et al. (1999) had shown that β-arrestin molecules are predominantly
localized in the inner segments of rod cells, although presently, the specific
function(s) of β-arrestin in rod cells is unknown. Interestingly, in vitro experiments
by Gurevich et al. (1995) had demonstrated significant affinity for R* molecules by
the β-arrestins. Moreover, since SAg not only has greater affinity than the
β-arrestins for R* molecules but also is much more abundant compared to the
β-arrestins in rod cells, we could isolate potential β-arrestin interaction with
mislocalized R* within transgenic Q344ter retinas by removing SAg molecules with
the rho+/-, SAg-/-, Tr α-/- background.
To indirectly detect complex formation between the β-arrestins and R*
molecules in transgenic Q344ter retinas, we performed a supernatant versus pellet
134
assay observing the translocation of β-arrestins in light-exposed retinas from
transgenic Q344ter
rho+/-, SAg-/-, Tr α-/-
mice and their negative littermate controls (Fig.
5.3). Because inactive arrestins are cytoplasmic, we would expect a light-dependent
translocation of β-arrestins to the cell membrane fraction if they interact with
mislocalized R* molecules. Thus, these mice were either dark-reared only or light-
exposed to 3000 lux for 0.5 hr with dilated pupils. Ultracentrifugation separated the
cytosolic (supernatant) fraction from the cell membrane (pellet) fraction. Detection
of β-arrestins in these samples was achieved by subsequent Western blotting with a
rabbit polyclonal anti- β-arrestin mAb, (A1CT, Kohout et al., 2001) on a
nitrocellulose membrane containing both retinal homogenates sample sets. (See
below for details of this assay.)
Within the supernatant fraction set, similar amounts of β-arrestins were
detected in the cytoplasm regardless of transgene presence and light conditions.
Within the pellet fraction set, however, varying levels of β-arrestins were observed.
Interestingly, the greatest amounts of β-arrestins within this fraction set were
discerned from light-exposed transgenic Q344ter retinal homogenates (marked by
red asterisks in Fig. 5.3). Thus, the most extensive translocation to the cellular
membrane fraction by β-arrestins was dependent on both light-exposure and the
presence of Q344ter. Although these results do not provide direct evidence for
mislocalized R*/ β-arrestin complex formation in Q344ter
rho+/-, SAg-/-, Tr α-/-
retinas, one
interpretation is that upon light-exposure of retinas expressing Q344ter, β-arrestin
molecules translocate to the plasma membrane within the RIS region and bind to
135

Figure 5.3. Significant increase in β-arrestin translocation to the membrane in
light-exposed Q344ter
rho+/-, SAg-/-, Tr α-/-
retinas. Q344ter transgenic mice and their
negative littermate controls with rho+/-, SAg-/-, Tr α-/- genetic background were
dark-reared only (D) or exposed to 0.5 hr of 3000 lux with dilated pupils (L). (All
mice were sacrificed between P29-P31.) Retinal homogenates from these mice were
centrifuged to separate the cytoplasmic (supernatant) and cell membrane (pellet)
fractions. After separation and extensive washing of pellet fraction, both fraction
sets were loaded in a 12% Bis-Tris polyacrylamide gel (1/16 of a retina for each
sample), transferred to a nitrocellulose membrane, and incubated with A1CT, a
rabbit polyclonal antibody recognizing β-arrestin molecules (Kohout et al., 2001). In
the supernatant fraction set, comparable amounts of β-arrestins were present in all
the samples. (The slightly decreased amount observed from the dark-reared
nontransgenic retina most likely resulted from loading error.) In the pellet fraction
set, differences in amount of β-arrestins were detected. Although a slight increase in
the β-arrestins was observed in the light-exposed nontransgenic retinal homogenates
(compared to their dark-reared counterpart), a more significant increase in these
proteins was observed in the light-exposed transgenic Q344ter retinas (denoted by
the red asterisks). Thus, the greatest translocation of β-arrestins to the cell
membrane fraction (and perhaps to R* molecules) was dependent on both light-
exposure and the presence of Q344ter. References to sizes of proteins are provided
by the numbers on the sides of the picture. The two bands between 45 and 66 kD
represent β-arretin1 (upper band) and β-arrestin 2 (lower band). Because the lower
band displayed greater light- and Q344ter-dependent translocation, β-arrestin 2 is the
more likely participant in the proposed toxic R*/ β-arrestin complex formation.
136
light-excited mislocalized rhodopsin molecules. Since both β-arrestin 1 and 2
species were detected in these retinal homogenates (the two bands in the pellet set),
the increased staining of the lower band in the light-exposed transgenic Q344ter
pellet samples points to β-arrestin 2 (the smaller of the two β-arrestins) to be
involved in this translocation. Clearly, future experiments, such as co-
immunoprecipitation, are needed to further investigate the role of β-arrestin
molecules in the light-accelerated retinal degeneration in these transgenic Q344ter
retinas.

Protocol for β-arrestin localization (cytoplasmic versus cell membrane association):
Determining the localizations of β-arrestins (cytoplasm versus cell membrane) was
achieved by an assay based on a previously described method (Li et al., 1995).
Separation of cytoplasmic and cell membrane fractions. Transgenic
Q344ter
rho+/-, SAg-/-, Tr α-/-
mice and their negative littermate controls (P29-31) were
either dark-reared only or dark-reared and exposed to 30 min bright light (3000 lux
with dilated pupils). After sacrificing, each retina was frozen by using liquid N
2
, but
only one retina per mouse was sufficient for this assay. [All subsequent preparation
of retinal samples was accomplished in the dark under infrared illumination with the
aid of night vision goggles (model # E8600) until loading in the polyacrylamide gel.]
The retinal samples were thawed and homogenized in 100 µl of homogenization
buffer A [80 mM Tris-HCl pH 8.0, 4 mM MgCl
2
, protease inhibitor cocktail (Roche
Diagnostics), 0.5 mM PMSF, 10 µm calpain II inhibitor (Roche Diagnostics), 50
mM NaF, 5 mM adenosine (Sigma-Aldrich)]. Each homogenate was transferred to
137
an ultracentrifuge tube (Beckman Coulter, Inc., Fullerton, CA) and was spun-down
at 163,640 x g (4˚C, 33 min). 80 µl of the supernatant fraction was transferred to a
microcentrifuge tube, and protein sample loading buffer was added (supernatant set).
The pellet fractions were washed 3X by the following steps: 1 ml of hypotonic buffer
(5.0 mM Tris pH 7.5, 0.5 mM MgCl
2
, 0.5 mM DTT, 0.5 mM PMSF, protease
inhibitor cocktail, 10 µm calpain II inhibitor, 50 mM NaF, 5 mM adenosine) was
added; the pellet was resuspended by using the PT 1200 C polytron (Kinematica,
Switzerland); the retinal samples were centrifuged at 129,900 x g (4 ˚C, 17 min); and
the supernatant was discarded. After the washings, we added 100 µl of solubilization
buffer [10 mM HEPES pH 7.5, 2 mM MgCl
2
, 2 mM CaCl
2
, 150 mM NaCl, 3% n-
dodecyl- β-D-maltoside (DM, EMD Biosciences, San Diego, CA), 0.5 mM PMSF,
protease inhibitor cocktail, 10 µm calpain II inhibitor, 50 mM NaF, 5 mM adenosine)
to the pellet fractions, and samples were solubilized O/N at 4 ˚C. Afterwards, we
added protein sample loading buffer (membrane set).
Western blots analyses. All prepared retinal homogenate samples were
boiled for 5 min and loaded (1/16 retina equivalents) in 12% Bis-Tris
polyacrylamide gels (Invitrogen Corp., Carlsbad, CA). The samples were run in 1X
MOPS buffer (Invitrogen Corp.), and transferred to a nitrocellulose membrane
(Whatman Schleicher & Schuell, Sanford, ME). The nitrocellulose membrane was
blocked with 10% milk; incubated with a 1:5000 dilution of the rabbit polyclonal
anti- β-arrestin antibody, A1CT (Kohout et al., 2001); incubated with a 1:5000
dilution of peroxidase labeled anti-rabbit IgG antibody (Vector Laboratories, Inc.,
138
Burlingame, CA); and washed. [All antibodies were diluted with TBS-T (20 mM
Tris, 1.37 M NaCl, 0.1% Tween-20, pH 7.6) and contained 10% milk.]
Immunodetection was performed by using the ECL system (GE Healthcare,
Piscataway, NJ).
139
Bibliography

Adamus, G., Z. S. Zam, A. Arendt, K. Palczewski, J. H. McDowell and P. A.
Hargrave (1991). "Anti-rhodopsin monoclonal antibodies of defined
specificity: characterization and application." Vision Res 31(1): 17-31.
Adamus, G., Z. S. Zam, J. H. McDowell, G. P. Shaw and P. A. Hargrave (1988). "A
monoclonal antibody specific for the phosphorylated epitope of rhodopsin:
comparison with other anti-phosphoprotein antibodies." Hybridoma 7(3):
237-47.
Alfinito, P. D. and E. Townes-Anderson (2002). "Activation of mislocalized opsin
kills rod cells: a novel mechanism for rod cell death in retinal disease." Proc
Natl Acad Sci U S A 99(8): 5655-60.
Alloway, P. G., L. Howard and P. J. Dolph (2000). "The formation of stable
rhodopsin-arrestin complexes induces apoptosis and photoreceptor cell
degeneration." Neuron 28(1): 129-38.
Angleson, J. K. and T. G. Wensel (1994). "Enhancement of rod outer segment
GTPase accelerating protein activity by the inhibitory subunit of cGMP
phosphodiesterase." J Biol Chem 269(23): 16290-6.
Arshavsky, V. and M. D. Bownds (1992). "Regulation of deactivation of
photoreceptor G protein by its target enzyme and cGMP." Nature 357(6377):
416-7.
Arshavsky, V. Y. (2002). "Rhodopsin phosphorylation: from terminating single
photon responses to photoreceptor dark adaptation." Trends Neurosci 25(3):
124-6.
Berson, E. L., B. Rosner, C. Weigel-DiFranco, T. P. Dryja and M. A. Sandberg
(2002). "Disease progression in patients with dominant retinitis pigmentosa
and rhodopsin mutations." Invest Ophthalmol Vis Sci 43(9): 3027-36.
Bessant, D. A., S. Khaliq, A. Hameed, K. Anwar, A. M. Payne, S. Q. Mehdi and S.
S. Bhattacharya (1999). "Severe autosomal dominant retinitis pigmentosa
caused by a novel rhodopsin mutation (Ter349Glu). Mutations in brief no.
208. Online." Hum Mutat 13(1): 83.
Chen, C. K., M. E. Burns, M. Spencer, G. A. Niemi, J. Chen, J. B. Hurley, D. A.
Baylor and M. I. Simon (1999a). "Abnormal photoresponses and light-
induced apoptosis in rods lacking rhodopsin kinase." Proc Natl Acad Sci U S
A 96(7): 3718-22.
140
Chen, J., C. L. Makino, N. S. Peachey, D. A. Baylor and M. I. Simon (1995).
"Mechanisms of rhodopsin inactivation in vivo as revealed by a COOH-
terminal truncation mutant." Science 267(5196): 374-7.
Chen, J., G. Shi, F. A. Concepcion, G. Xie, D. Oprian and J. Chen (2006). "Stable
Rhodopsin/Arrestin Complex Leads to Retinal Degeneration in a Transgenic
Mouse Model of Autosomal Dominant Retinitis Pigmentosa." J Neurosci
26(46): 11929-11937.
Chen, J., M. I. Simon, M. T. Matthes, D. Yasumura and M. M. LaVail (1999b).
"Increased susceptibility to light damage in an arrestin knockout mouse
model of Oguchi disease (stationary night blindness)." Invest Ophthalmol Vis
Sci 40(12): 2978-82.
Chuang, J. Z., C. Vega, W. Jun and C. H. Sung (2004). "Structural and functional
impairment of endocytic pathways by retinitis pigmentosa mutant rhodopsin-
arrestin complexes." J Clin Invest 114(1): 131-40.
Concepcion, F., A. Mendez and J. Chen (2002). "The carboxyl-terminal domain is
essential for rhodopsin transport in rod photoreceptors." Vision Res 42(4):
417-26.
Cowan, C. W., R. N. Fariss, I. Sokal, K. Palczewski and T. G. Wensel (1998). "High
expression levels in cones of RGS9, the predominant GTPase accelerating
protein of rods." Proc Natl Acad Sci U S A 95(9): 5351-6.
Curcio, C. A., K. R. Sloan, R. E. Kalina and A. E. Hendrickson (1990). "Human
photoreceptor topography." J Comp Neurol 292(4): 497-523.
Danciger, M., M. T. Matthes, D. Yasamura, N. B. Akhmedov, T. Rickabaugh, S.
Gentleman, T. M. Redmond, M. M. La Vail and D. B. Farber (2000). "A
QTL on distal chromosome 3 that influences the severity of light-induced
damage to mouse photoreceptors." Mamm Genome 11(6): 422-7.
Deretic, D. (2004). From the golgi to the rod outer segment: formation, movement,
docking and fusion of rhodopsin transport carriers. Photoreceptor Cell
Biology and Inherited Retinal Degenerations. D. S. Williams. Singapore,
World Scientific. 10: 29-64.
Deretic, D. and D. S. Papermaster (1991). "Polarized sorting of rhodopsin on post-
Golgi membranes in frog retinal photoreceptor cells." J Cell Biol 113(6):
1281-93.
Deretic, D., B. Puleo-Scheppke and C. Trippe (1996). "Cytoplasmic domain of
rhodopsin is essential for post-Golgi vesicle formation in a retinal cell-free
system." J Biol Chem 271(4): 2279-86.
141
Deretic, D., S. Schmerl, P. A. Hargrave, A. Arendt and J. H. McDowell (1998).
"Regulation of sorting and post-Golgi trafficking of rhodopsin by its C-
terminal sequence QVS(A)PA." Proc Natl Acad Sci U S A 95(18): 10620-5.
Deretic, D., A. H. Williams, N. Ransom, V. Morel, P. A. Hargrave and A. Arendt
(2005). "Rhodopsin C terminus, the site of mutations causing retinal disease,
regulates trafficking by binding to ADP-ribosylation factor 4 (ARF4)." Proc
Natl Acad Sci U S A 102(9): 3301-6.
Doan, T., A. Mendez, P. B. Detwiler, J. Chen and F. Rieke (2006). "Multiple
phosphorylation sites confer reproducibility of the rod's single-photon
responses." Science 313(5786): 530-3.
Duncan, R. R., J. Greaves, U. K. Wiegand, I. Matskevich, G. Bodammer, D. K.
Apps, M. J. Shipston and R. H. Chow (2003). "Functional and spatial
segregation of secretory vesicle pools according to vesicle age." Nature
422(6928): 176-80.
Elias, R. V., S. S. Sezate, W. Cao and J. F. McGinnis (2004). "Temporal kinetics of
the light/dark translocation and compartmentation of arrestin and alpha-
transducin in mouse photoreceptor cells." Mol Vis 10: 672-81.
Green, E. S., M. D. Menz, M. M. LaVail and J. G. Flannery (2000).
"Characterization of rhodopsin mis-sorting and constitutive activation in a
transgenic rat model of retinitis pigmentosa." Invest Ophthalmol Vis Sci
41(6): 1546-53.
Gregory-Evans, K. and S. S. Bhattacharya (1998). "Genetic blindness: current
concepts in the pathogenesis of human outer retinal dystrophies." Trends
Genet 14(3): 103-8.
Gurevich, V. V. and J. L. Benovic (1993). "Visual arrestin interaction with
rhodopsin. Sequential multisite binding ensures strict selectivity toward light-
activated phosphorylated rhodopsin." J Biol Chem 268(16): 11628-38.
Gurevich, V. V., S. B. Dion, J. J. Onorato, J. Ptasienski, C. M. Kim, R. Sterne-Marr,
M. M. Hosey and J. L. Benovic (1995). "Arrestin interactions with G protein-
coupled receptors. Direct binding studies of wild type and mutant arrestins
with rhodopsin, beta 2-adrenergic, and m2 muscarinic cholinergic receptors."
J Biol Chem 270(2): 720-31.
Gurevich, V. V. and E. V. Gurevich (2004). "The molecular acrobatics of arrestin
activation." Trends Pharmacol Sci 25(2): 105-11.
142
Hagstrom, S. A., M. Adamian, M. Scimeca, B. S. Pawlyk, G. Yue and T. Li (2001).
"A role for the Tubby-like protein 1 in rhodopsin transport." Invest
Ophthalmol Vis Sci 42(9): 1955-62.
Hamel, C. (2006). "Retinitis pigmentosa." Orphanet J Rare Dis 1: 40.
Hanson, S. M. and V. V. Gurevich (2006). "The differential engagement of arrestin
surface charges by the various functional forms of the receptor." J Biol Chem
281(6): 3458-62.
Hao, W., A. Wenzel, M. S. Obin, C. K. Chen, E. Brill, N. V. Krasnoperova, P.
Eversole-Cire, Y. Kleyner, A. Taylor, M. I. Simon, C. Grimm, C. E. Reme
and J. Lem (2002). "Evidence for two apoptotic pathways in light-induced
retinal degeneration." Nat Genet 32(2): 254-60.
He, W., C. W. Cowan and T. G. Wensel (1998). "RGS9, a GTPase accelerator for
phototransduction." Neuron 20(1): 95-102.
Hicks, D. and C. J. Barnstable (1987). "Different rhodopsin monoclonal antibodies
reveal different binding patterns on developing and adult rat retina." J
Histochem Cytochem 35(11): 1317-28.
Hicks, D., J. Sparrow and C. J. Barnstable (1989). "Immunoelectron microscopical
examination of the surface distribution of opsin in rat rod photoreceptor
cells." Exp Eye Res 49(1): 13-29.
Hodges, R. S., R. J. Heaton, J. M. Parker, L. Molday and R. S. Molday (1988).
"Antigen-antibody interaction. Synthetic peptides define linear antigenic
determinants recognized by monoclonal antibodies directed to the
cytoplasmic carboxyl terminus of rhodopsin." J Biol Chem 263(24): 11768-
75.
Horn, M., P. Humphries, M. Kunisch, C. Marchese, E. Apfelstedt-Sylla, L. Fugi, E.
Zrenner, P. Kenna, A. Gal and J. Farrar (1992). "Deletions in exon 5 of the
human rhodopsin gene causing a shift in the reading frame and autosomal
dominant retinitis pigmentosa." Hum Genet 90(3): 255-7.
Hu, G. and T. G. Wensel (2002). "R9AP, a membrane anchor for the photoreceptor
GTPase accelerating protein, RGS9-1." Proc Natl Acad Sci U S A 99(15):
9755-60.
Humphries, M. M., D. Rancourt, G. J. Farrar, P. Kenna, M. Hazel, R. A. Bush, P. A.
Sieving, D. M. Sheils, N. McNally, P. Creighton, A. Erven, A. Boros, K.
Gulya, M. R. Capecchi and P. Humphries (1997). "Retinopathy induced in
mice by targeted disruption of the rhodopsin gene." Nat Genet 15(2): 216-9.
143
Iakhine, R., I. Chorna-Ornan, T. Zars, N. Elia, Y. Cheng, Z. Selinger, B. Minke and
D. R. Hyde (2004). "Novel dominant rhodopsin mutation triggers two
mechanisms of retinal degeneration and photoreceptor desensitization." J
Neurosci 24(10): 2516-26.
Jozwick, C., K. Valter and J. Stone (2006). "Reversal of functional loss in the P23H-
3 rat retina by management of ambient light." Exp Eye Res 83(5): 1074-80.
Kanaho, Y., S. C. Tsai, R. Adamik, E. L. Hewlett, J. Moss and M. Vaughan (1984).
"Rhodopsin-enhanced GTPase activity of the inhibitory GTP-binding protein
of adenylate cyclase." J Biol Chem 259(12): 7378-81.
Kaushal, S. and H. G. Khorana (1994). "Structure and function in rhodopsin. 7. Point
mutations associated with autosomal dominant retinitis pigmentosa."
Biochemistry 33(20): 6121-8.
Kiselev, A., M. Socolich, J. Vinos, R. W. Hardy, C. S. Zuker and R. Ranganathan
(2000). "A molecular pathway for light-dependent photoreceptor apoptosis in
Drosophila." Neuron 28(1): 139-52.
Kohout, T. A., F. S. Lin, S. J. Perry, D. A. Conner and R. J. Lefkowitz (2001). "beta-
Arrestin 1 and 2 differentially regulate heptahelical receptor signaling and
trafficking." Proc Natl Acad Sci U S A 98(4): 1601-6.
Kong, L., F. Li, C. E. Soleman, S. Li, R. V. Elias, X. Zhou, D. A. Lewis, J. F.
McGinnis and W. Cao (2006). "Bright cyclic light accelerates photoreceptor
cell degeneration in tubby mice." Neurobiol Dis 21(3): 468-77.
Krispel, C. M., D. Chen, N. Melling, Y. J. Chen, K. A. Martemyanov, N. Quillinan,
V. Y. Arshavsky, T. G. Wensel, C. K. Chen and M. E. Burns (2006). "RGS
expression rate-limits recovery of rod photoresponses." Neuron 51(4): 409-
16.
Kuhn, H., S. W. Hall and U. Wilden (1984). "Light-induced binding of 48-kDa
protein to photoreceptor membranes is highly enhanced by phosphorylation
of rhodopsin." FEBS Lett 176(2): 473-8.
Kuhn, H. and U. Wilden (1987). "Deactivation of photoactivated rhodopsin by
rhodopsin-kinase and arrestin." J Recept Res 7(1-4): 283-98.
Laird, D. W. and R. S. Molday (1988). "Evidence against the role of rhodopsin in
rod outer segment binding to RPE cells." Invest Ophthalmol Vis Sci 29(3):
419-28.
144
Langen, R., K. Cai, C. Altenbach, H. G. Khorana and W. L. Hubbell (1999).
"Structural features of the C-terminal domain of bovine rhodopsin: a site-
directed spin-labeling study." Biochemistry 38(25): 7918-24.
LaVail, M. M. and G. M. Gorrin (1987). "Protection from light damage by ocular
pigmentation: analysis using experimental chimeras and translocation mice."
Exp Eye Res 44(6): 877-89.
Le, Y. Z., J. D. Ash, M. R. Al-Ubaidi, Y. Chen, J. X. Ma and R. E. Anderson (2004).
"Targeted expression of Cre recombinase to cone photoreceptors in
transgenic mice." Mol Vis 10: 1011-8.
Lem, J., M. L. Applebury, J. D. Falk, J. G. Flannery and M. I. Simon (1991).
"Tissue-specific and developmental regulation of rod opsin chimeric genes in
transgenic mice." Neuron 6(2): 201-10.
Lem, J., N. V. Krasnoperova, P. D. Calvert, B. Kosaras, D. A. Cameron, M. Nicolo,
C. L. Makino and R. L. Sidman (1999). "Morphological, physiological, and
biochemical changes in rhodopsin knockout mice." Proc Natl Acad Sci U S A
96(2): 736-41.
Li, T., W. K. Franson, J. W. Gordon, E. L. Berson and T. P. Dryja (1995).
"Constitutive activation of phototransduction by K296E opsin is not a cause
of photoreceptor degeneration." Proc Natl Acad Sci U S A 92(8): 3551-5.
Li, T., W. K. Snyder, J. E. Olsson and T. P. Dryja (1996). "Transgenic mice carrying
the dominant rhodopsin mutation P347S: evidence for defective vectorial
transport of rhodopsin to the outer segments." Proc Natl Acad Sci U S A
93(24): 14176-81.
McDowell, J. H., J. P. Nawrocki and P. A. Hargrave (2000). "Isolation of isoelectric
species of phosphorylated rhodopsin." Methods Enzymol 315: 70-6.
McNally, N., P. Kenna, M. M. Humphries, A. H. Hobson, N. W. Khan, R. A. Bush,
P. A. Sieving, P. Humphries and G. J. Farrar (1999). "Structural and
functional rescue of murine rod photoreceptors by human rhodopsin
transgene." Hum Mol Genet 8(7): 1309-12.
Mendes, H. F., J. van der Spuy, J. P. Chapple and M. E. Cheetham (2005).
"Mechanisms of cell death in rhodopsin retinitis pigmentosa: implications for
therapy." Trends Mol Med 11(4): 177-85.
Mendez, A., M. E. Burns, A. Roca, J. Lem, L. W. Wu, M. I. Simon, D. A. Baylor
and J. Chen (2000a). "Rapid and reproducible deactivation of rhodopsin
requires multiple phosphorylation sites." Neuron 28(1): 153-64.
145
Mendez, A., N. V. Krasnoperova, J. Lem and J. Chen (2000b). "Functional study of
rhodopsin phosphorylation in vivo." Methods Enzymol 316: 167-85.
Mendez, A., J. Lem, M. Simon and J. Chen (2003). "Light-dependent translocation
of arrestin in the absence of rhodopsin phosphorylation and transducin
signaling." J Neurosci 23(8): 3124-9.
Metaye, T., H. Gibelin, R. Perdrisot and J. L. Kraimps (2005). "Pathophysiological
roles of G-protein-coupled receptor kinases." Cell Signal 17(8): 917-28.
Michon, J. J., Z. L. Li, N. Shioura, R. J. Anderson and M. O. Tso (1991). "A
comparative study of methods of photoreceptor morphometry." Invest
Ophthalmol Vis Sci 32(2): 280-4.
Mohand-Said, S., A. Deudon-Combe, D. Hicks, M. Simonutti, V. Forster, A. C.
Fintz, T. Leveillard, H. Dreyfus and J. A. Sahel (1998). "Normal retina
releases a diffusible factor stimulating cone survival in the retinal
degeneration mouse." Proc Natl Acad Sci U S A 95(14): 8357-62.
Naash, M. L., N. S. Peachey, Z. Y. Li, C. C. Gryczan, Y. Goto, J. Blanks, A. H.
Milam and H. Ripps (1996). "Light-induced acceleration of photoreceptor
degeneration in transgenic mice expressing mutant rhodopsin." Invest
Ophthalmol Vis Sci 37(5): 775-82.
Nicolas-Leveque, C., I. Ghedira, J. P. Faure and M. Mirshahi (1999). "Beta-arrestin-
related proteins in ocular tissues." Invest Ophthalmol Vis Sci 40(8): 1812-8.
Nir, I. and D. S. Papermaster (1983). "Differential distribution of opsin in the plasma
membrane of frog photoreceptors: an immunocytochemical study." Invest
Ophthalmol Vis Sci 24(7): 868-78.
Noell, W. K. (1980). "Possible mechanisms of photoreceptor damage by light in
mammalian eyes." Vision Res 20(12): 1163-71.
Noell, W. K., V. S. Walker, B. S. Kang and S. Berman (1966). "Retinal damage by
light in rats." Invest Ophthalmol 5(5): 450-73.
Oh, K. T., R. Longmuir, D. M. Oh, E. M. Stone, K. Kopp, J. Brown, G. A. Fishman,
P. Sonkin, K. M. Gehrs and R. G. Weleber (2003). "Comparison of the
clinical expression of retinitis pigmentosa associated with rhodopsin
mutations at codon 347 and codon 23." Am J Ophthalmol 136(2): 306-13.
Olsson, J. E., J. W. Gordon, B. S. Pawlyk, D. Roof, A. Hayes, R. S. Molday, S.
Mukai, G. S. Cowley, E. L. Berson and T. P. Dryja (1992). "Transgenic mice
with a rhodopsin mutation (Pro23His): a mouse model of autosomal
dominant retinitis pigmentosa." Neuron 9(5): 815-30.
146
Orem, N. R. and P. J. Dolph (2002). "Loss of the phospholipase C gene product
induces massive endocytosis of rhodopsin and arrestin in Drosophila
photoreceptors." Vision Res 42(4): 497-505.
Orem, N. R., L. Xia and P. J. Dolph (2006). "An essential role for endocytosis of
rhodopsin through interaction of visual arrestin with the AP-2 adaptor." J Cell
Sci 119(Pt 15): 3141-8.
Organisciak, D. T., R. M. Darrow, L. Barsalou, R. K. Kutty and B. Wiggert (2003).
"Susceptibility to retinal light damage in transgenic rats with rhodopsin
mutations." Invest Ophthalmol Vis Sci 44(2): 486-92.
Portera-Cailliau, C., C. H. Sung, J. Nathans and R. Adler (1994). "Apoptotic
photoreceptor cell death in mouse models of retinitis pigmentosa." Proc Natl
Acad Sci U S A 91(3): 974-8.
Rapp, L. M. and T. P. Williams (1980). "The role of ocular pigmentation in
protecting against retinal light damage." Vision Res 20(12): 1127-31.
Rivolta, C., D. Sharon, M. M. DeAngelis and T. P. Dryja (2002). "Retinitis
pigmentosa and allied diseases: numerous diseases, genes, and inheritance
patterns." Hum Mol Genet 11(10): 1219-27.
Rodieck, R. W. (1998). The First Steps in Seeing. Sunderland, MA, Sinauer
Associates, Inc.
Roof, D. J., M. Adamian and A. Hayes (1994). "Rhodopsin accumulation at
abnormal sites in retinas of mice with a human P23H rhodopsin transgene."
Invest Ophthalmol Vis Sci 35(12): 4049-62.
Sakmar, T. P., R. R. Franke and H. G. Khorana (1989). "Glutamic acid-113 serves as
the retinylidene Schiff base counterion in bovine rhodopsin." Proc Natl Acad
Sci U S A 86(21): 8309-13.
Satoh, A. K. and D. F. Ready (2005). "Arrestin1 mediates light-dependent rhodopsin
endocytosis and cell survival." Curr Biol 15(19): 1722-33.
Shi, G., F. A. Concepcion and J. Chen (2004). Targeting of visual pigments to rod
outer segment in rhodopsin knockout mice. Photoreceptor Cell Biology and
Inherited Retinal Degenerations. D. S. Williams. Singapore, World Scientific.
10: 93-107.
Sim-Selley, L. J. and S. R. Childers (2002). "Neuroanatomical localization of
receptor-activated G proteins in brain." Methods Enzymol 344: 42-58.
147
Sohocki, M. M., S. P. Daiger, S. J. Bowne, J. A. Rodriquez, H. Northrup, J. R.
Heckenlively, D. G. Birch, H. Mintz-Hittner, R. S. Ruiz, R. A. Lewis, D. A.
Saperstein and L. S. Sullivan (2001). "Prevalence of mutations causing
retinitis pigmentosa and other inherited retinopathies." Hum Mutat 17(1): 42-
51.
Sung, C. H. and J. Z. Chuang (2004). Insights on retinal degeneration: molecular
basis of polarized rhodopsin transport. Photoreceptor Cell Biology and
Inherited Retinal Degenerations. D. S. Williams. Singapore, World Scientific.
10: 65-92.
Sung, C. H., C. M. Davenport, J. C. Hennessey, I. H. Maumenee, S. G. Jacobson, J.
R. Heckenlively, R. Nowakowski, G. Fishman, P. Gouras and J. Nathans
(1991a). "Rhodopsin mutations in autosomal dominant retinitis pigmentosa."
Proc Natl Acad Sci U S A 88(15): 6481-5.
Sung, C. H., C. M. Davenport and J. Nathans (1993). "Rhodopsin mutations
responsible for autosomal dominant retinitis pigmentosa. Clustering of
functional classes along the polypeptide chain." J Biol Chem 268(35): 26645-
9.
Sung, C. H., C. Makino, D. Baylor and J. Nathans (1994). "A rhodopsin gene
mutation responsible for autosomal dominant retinitis pigmentosa results in a
protein that is defective in localization to the photoreceptor outer segment." J
Neurosci 14(10): 5818-33.
Sung, C. H., B. G. Schneider, N. Agarwal, D. S. Papermaster and J. Nathans
(1991b). "Functional heterogeneity of mutant rhodopsins responsible for
autosomal dominant retinitis pigmentosa." Proc Natl Acad Sci U S A 88(19):
8840-4.
Sung, C. H. and A. W. Tai (2000). "Rhodopsin trafficking and its role in retinal
dystrophies." Int Rev Cytol 195: 215-67.
Tai, A. W., J. Z. Chuang, C. Bode, U. Wolfrum and C. H. Sung (1999). "Rhodopsin's
carboxy-terminal cytoplasmic tail acts as a membrane receptor for
cytoplasmic dynein by binding to the dynein light chain Tctex-1." Cell 97(7):
877-87.
Tam, B. M. and O. L. Moritz (2006). "Characterization of rhodopsin P23H-induced
retinal degeneration in a Xenopus laevis model of retinitis pigmentosa."
Invest Ophthalmol Vis Sci 47(8): 3234-41.
148
Tam, B. M., O. L. Moritz, L. B. Hurd and D. S. Papermaster (2000). "Identification
of an outer segment targeting signal in the COOH terminus of rhodopsin
using transgenic Xenopus laevis." J Cell Biol 151(7): 1369-80.
Tam, B. M., G. Xie, D. D. Oprian and O. L. Moritz (2006). "Mislocalized rhodopsin
does not require activation to cause retinal degeneration and neurite
outgrowth in Xenopus laevis." J Neurosci 26(1): 203-9.
Tan, E., Q. Wang, A. B. Quiambao, X. Xu, N. M. Qtaishat, N. S. Peachey, J. Lem, S.
J. Fliesler, D. R. Pepperberg, M. I. Naash and M. R. Al-Ubaidi (2001). "The
relationship between opsin overexpression and photoreceptor degeneration."
Invest Ophthalmol Vis Sci 42(3): 589-600.
Taylor, M. R., H. A. Van Epps, M. J. Kennedy, J. C. Saari, J. B. Hurley and S. E.
Brockerhoff (2000). "Biochemical analysis of phototransduction and visual
cycle in zebrafish larvae." Methods Enzymol 316: 536-57.
Terskikh, A., A. Fradkov, G. Ermakova, A. Zaraisky, P. Tan, A. V. Kajava, X. Zhao,
S. Lukyanov, M. Matz, S. Kim, I. Weissman and P. Siebert (2000).
""Fluorescent timer": protein that changes color with time." Science
290(5496): 1585-8.
Tsai, S. C., R. Adamik, Y. Kanaho, J. L. Halpern and J. Moss (1987).
"Immunological and biochemical differentiation of guanyl nucleotide binding
proteins: interaction of Go alpha with rhodopsin, anti-Go alpha polyclonal
antibodies, and a monoclonal antibody against transducin alpha subunit and
Gi alpha." Biochemistry 26(15): 4728-33.
Tsang, S. H., M. E. Burns, P. D. Calvert, P. Gouras, D. A. Baylor, S. P. Goff and V.
Y. Arshavsky (1998). "Role for the target enzyme in deactivation of
photoreceptor G protein in vivo." Science 282(5386): 117-21.
Vaughan, D. K., S. F. Coulibaly, R. M. Darrow and D. T. Organisciak (2003). "A
morphometric study of light-induced damage in transgenic rat models of
retinitis pigmentosa." Invest Ophthalmol Vis Sci 44(2): 848-55.
Vaughan, D. K., S. K. Fisher, S. A. Bernstein, I. L. Hale, K. A. Linberg and B.
Matsumoto (1989). "Evidence that microtubules do not mediate opsin vesicle
transport in photoreceptors." J Cell Biol 109(6 Pt 1): 3053-62.
Vishnivetskiy, S. A., M. M. Hosey, J. L. Benovic and V. V. Gurevich (2004).
"Mapping the arrestin-receptor interface. Structural elements responsible for
receptor specificity of arrestin proteins." J Biol Chem 279(2): 1262-8.
Vishnivetskiy, S. A., C. L. Paz, C. Schubert, J. A. Hirsch, P. B. Sigler and V. V.
Gurevich (1999). "How does arrestin respond to the phosphorylated state of
rhodopsin?" J Biol Chem 274(17): 11451-4.
149
150
Wang, D. Y., W. M. Chan, P. O. Tam, L. Baum, D. S. Lam, K. K. Chong, B. J. Fan
and C. P. Pang (2005). "Gene mutations in retinitis pigmentosa and their
clinical implications." Clin Chim Acta 351(1-2): 5-16.
Weiss, E. R., R. A. Heller-Harrison, E. Diez, M. Crasnier, C. C. Malbon and G. L.
Johnson (1990). "Rhodopsin expressed in Chinese hamster ovary cells
regulates adenylyl cyclase activity." J Mol Endocrinol 4(1): 71-9.
Wenzel, A., C. Grimm, M. Samardzija and C. E. Reme (2005). "Molecular
mechanisms of light-induced photoreceptor apoptosis and neuroprotection for
retinal degeneration." Prog Retin Eye Res 24(2): 275-306.
Whelan, J. P. and J. F. McGinnis (1988). "Light-dependent subcellular movement of
photoreceptor proteins." J Neurosci Res 20(2): 263-70.
Williams, D. S. (2002). "Transport to the photoreceptor outer segment by myosin
VIIa and kinesin II." Vision Res 42(4): 455-62.
Xu, J., R. L. Dodd, C. L. Makino, M. I. Simon, D. A. Baylor and J. Chen (1997).
"Prolonged photoresponses in transgenic mouse rods lacking arrestin."
Nature 389(6650): 505-9.
Yau, K. W. (1994). "Phototransduction mechanism in retinal rods and cones. The
Friedenwald Lecture." Invest Ophthalmol Vis Sci 35(1): 9-32.
Zack, D. J., J. Bennett, Y. Wang, C. Davenport, B. Klaunberg, J. Gearhart and J.
Nathans (1991). "Unusual topography of bovine rhodopsin promoter-lacZ
fusion gene expression in transgenic mouse retinas." Neuron 6(2): 187-99.
Zhang, H., W. Huang, X. Zhu, C. M. Craft, W. Baehr and C. K. Chen (2003a).
"Light-dependent redistribution of visual arrestins and transducin subunits in
mice with defective phototransduction." Mol Vis 9: 231-7.
Zhang, X., T. G. Wensel and T. W. Kraft (2003b). "GTPase regulators and
photoresponses in cones of the eastern chipmunk." J Neurosci 23(4): 1287-
97.
Zhao, X., J. Huang, S. C. Khani and K. Palczewski (1998). "Molecular forms of
human rhodopsin kinase (GRK1)." J Biol Chem 273(9): 5124-31.
Zhu, X., A. Li, B. Brown, E. R. Weiss, S. Osawa and C. M. Craft (2002). "Mouse
cone arrestin expression pattern: light induced translocation in cone
photoreceptors." Mol Vis 8: 462-71.

Abstract (if available)

Abstract Retinitis pigmentosa (RP) is a heterogeneous group of inherited eye diseases that is typified by initial night blindness and loss of peripheral vision and eventually culminates into total blindness. This retinal disorder afflicts around 1 in 4000 people worldwide. Mutations in the carboxyl-terminus of rhodopsin have been linked to autosomal dominant RP (ADRP). In this study, we have investigated two truncated rhodopsin mutants, S334ter and Q344ter. Both mutants have demonstrated that the presence of the QVAPA domain is necessary for proper rhodopsin localization and ROS formation. However, the retention of the phosphorylation sites in the Q344ter (and not in the S334ter) has allowed us to demonstrate the light-activation capability of mislocalized rhodopsin. More pertinent to ADRP patients, the retention of these phosphorylation sites in Q344ter also contributes to light-accelerated retinal degeneration through a mechanism first described in Drosophila, the accumulation of the rhodopsin/arrestin complexes in abnormal subcellular compartments.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

New understanding of rhodopsin in retinal degeneration and high gain phosphorylation

PDF

Exploring alternative roles of visual arrestin 1 in photoreceptor synaptic regulation and deciphering the molecular pathway of retinal degeneration using mouse knockout technology

Asset Metadata

Creator Concepcion, Francis Avila (author)

Core Title Essential role of the carboxyl-terminus for proper rhodopsin trafficking and "enlightenment" to the pathway(s) causing retinal degeneration in a mouse model expressing a truncated rhodopsin mutant

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Cell and Neurobiology

Publication Date 09/05/2007

Defense Date 08/13/2007

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

Tag ADRP,OAI-PMH Harvest,Q344ter,Retinal degeneration,rhodopsin transport,rod photoreceptor,S334ter

Language English

Advisor Chen, Jeannie (committee chair), Chow, Robert HP. (committee member), Garner, Judy A. (committee member), Okamoto, Curtis Toshio (committee member)

Creator Email faconcepcion@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m806

Unique identifier UC1119917

Identifier etd-Concepcion-20070905 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-548940 (legacy record id),usctheses-m806 (legacy record id)

Legacy Identifier etd-Concepcion-20070905.pdf

Dmrecord 548940

Document Type Dissertation

Rights Concepcion, Francis Avila

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