University of Southern California Dissertations and Theses

Signaling cascades: a functional characterization of cone arrestin and a differential gene expression analysis of developing retinal ganglion cells

(USC Thesis Other)

Signaling cascades: a functional characterization of cone arrestin and a differential gene expression analysis of developing retinal ganglion cells

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

SIGNALING CASCADES: A FUNCTIONAL CHARACTERIZATION OF CONE
ARRESTIN AND A DIFFERENTIAL GENE EXPRESSION ANALYSIS OF
DEVELOPING RETINAL GANGLION CELLS

by

Sanny Kai-Wai Chan

____________________________________________________________________

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 Sanny Kai-Wai Chan

ii
Dedication

To my parents, Steven and Carol, who have sacrificed selflessly and
bestowed loving support; and to my sister, Candy, for always being there to help, in
person and with useful advice especially during my final year as a graduate student.
There are no words that can truly encompass my profound appreciation for
making this possible. Their initial and continued enthusiasm in the projects I pursue
from conception to completion is truly valued and so, I will simply say thank you.

iii
Acknowledgements
No one finishes a dissertation without the support and encouragement of
family, friends and faculty mentors. I would like to acknowledge the members of
my final defense committee. I would like to express my thanks for the promptness in
reviewing this work, their comments were seriously considered and suggestions
taken. As my mentor Dr. Chen gave me the wonderful opportunity to participate in
these exciting projects. Dr. Hinton has always been a strong advocate and supporter
during my training, and I want to thank him for his guidance and patience. Dr. Craft
kept challenging me about the implications of my results and her knowledge in this
field was especially valuable. Dr. Garner has lent precious guidance and
encouragement and whose comments and suggestions on my thesis are greatly
appreciated.
I would also like to thank friends, colleagues, and professors who have given
me so much during my training. Dr. Mendez's expertise and knowledge is greatly
appreciated. To Dr. Wu, who has been a tremendous help on this project, from
experimental design, result analysis to writing a paper. Special thanks to Dr. DF
Chen whose continued enthusiasm for this project has encouraged me and driven me
to clarify the written word. Finally, I would like to thank all the people from Drs.
Chen, Hinton, Craft, and Dr. Simon’s labs. I would not be able to complete my
thesis without their support and assistance.

iv
Table of Contents

Dedication
Acknowledgements
List of Tables
List of Figures
Abstract
Chapter I Introduction
Chapter I References
Chapter II Ectopic Cone Arrestin Rescues Photoreceptors from
Light Damage in a Dose Dependant Manner

Chapter II References
Chapter III Gene Expression Analysis During Perinatal Retinal
Ganglion Cell Development

Chapter III References
Chapter IV Conclusions and Implications
Chapter IV References
Bibliography
1
56
195
vi
x
iii
v
205
48
110
106
186
203
ii

v
List of Tables
Table 1.1 Mouse Rod and Cone Photoreceptors Express Many
Homologous Phototransduction Proteins.

Table 2.1 Rhodopsin Content is Similar in Arrestin Knockout
(Arr -/-) and Transgenic mCAR-high Mice.

Table 3.1 Comparison of the Effects of Alterations in the
Methods of Generating IVT Labeled Probes.

Table 3.2 Differentially Expressed Transcripts Grouped into
Functional Categories.

Table 3.3 Neurite Growth Related Genes Analyzed by
Microarray

141
150
135
21
94

vi
List of Figures
Figure 1.1 Micrograph of a Mature Mouse Retina Illustrates
the Well-Organized Laminar Nature of this Tissue.
Figure 1.2 Rod and Cone Photoreceptors Share Similar
Morphology.
Figure 1.3 The Main Components in the Activation and
Deactivation of the Phototransduction Cascade in Vertebrate
Rod Photoreceptor Cells.
Figure 1.4 A Representative Single Cell Recording of Rod
Photoreceptors in Response to a Single Photon.
Figure 1.5 Sequence Alignments of Homologous Salamander
Cone Arrestin (1SUJ), Mouse Cone Arrestin (mCAR), Bovine
Rod Arrestin, and Mouse Rod Arrestin.
Figure 2.1 Transgene Expression of mCAR in the Retina.
Figure 2.2 Dose Dependent Protection of Photoreceptors from
Light Damage.
Figure 2.3 Progressive Increase in Protection Against Light
Damage at Decreasing Light Intensity.
Figure 2.4 Comparison of mCAR and Arrestin Translocation
and Membrane Binding.
Figure 2.5 Average Single Photon Responses from Single Cell
Recording of Dim Light Response.

Figure 3.1 RGCs are Consistently Obtained at High Purity.
Figure 3.2 A Very Low Level of Auto-fluorescence in Isotype
Negative Control.
Figure 3.3 Co-localization of Cells Stained with
Neurofilament (an RGC Marker) and β-actin in Nearly All
Cells.
Figure 3.4 Developmental Stage Progression Marked by
Dramatic Decrease in RGC Neurite Growth.
5
70
127
128
74
79
87
92
130
6
10
28
15
127

vii
Figure 3.5 Scatter Plot of Gene Intensity and Correlation
Coefficients (R-value in Parenthesis in the Upper Left Corner
of Each Quadrant).
Figure 3.6 Bioanalyzer Analysis of the Integrity of Total RNA
Isolated from Purified RGCs Prior to Labeling.
Figure 3.7 Functional Categorization of Transcripts Defined as
Differentially Expressed (ANOVA Corrected p≤0.05 with at
Least 2^±1.5 Fold Difference).
Figure 3.8 Hierarchical Cluster Profile of Gene Transcripts
and Samples.
Figure 3.9 K-mean Cluster Analysis of Defined Differentially
Expressed Transcripts.
Figure 3.10 Expression of Transcripts Identified as
Differentially Expressed in RGC Were Confirmed.
Figure 3.11 Novel Transcripts Identified as Differentially
Expressed in RGC were Confirmed by Semi-quantitative Real-
time RT-PCR
Figure 3.12 Transcripts that Function in Growth, Signaling,
and Cell Cycle were Examined by Semi-quantitative RT-PCR
and ISH.
Figure 3.13 Transcripts of Structural and Synapse Associated
Genes were Examined by Semi-quantitative RT-PCR.
Figure 3.14 The Relative Expression Levels of Transcripts
Analyzed by Semi-quantitative RT-PCR were Compared to
Kruppel-like Factor 9 (the Lowest Expressed Mean Transcript)
at E16.
Figure 3.15 A Time Course In Situ Hybridization Study on
Purkinje Cell Protein 4 (Pcp4)
Figure 3.16 Low and High Power Views of ISH for
Transcription Factors Kruppel-like Factor 9 (kfl9) and Early
Growth Response-1 (egr1).
Figure 3.17 Low and High Power Views of ISH from P0
Retinas of Signaling and Cell Proliferation
133
139
154
156
146
147
159
137
152
158
162
163
164

viii
Figure 3.18 Low and High Power Views of ISH for Transcript
Involved in Adhesion and Synaptogenesis.
Figure 3.19 Differentially Expressed Transcripts from Cultured
RGCs at E16 and E18 were Examined by Semi-quantitative
PCR
Figure 3.20 Co-localization of Transfected Viable E16 RGCs
Expressing DsRed.

165
167
170

ix
Abstract

The mature retina is a specialized tissue critical for vision. Individual cell-
types that comprise the retina have specialized functional roles mediated by
differential gene expression. To identify the contribution of specific genes in
defining a cell-type, the function of the cone arrestin protein was analyzed in vivo in
photoreceptors and a differential transcript expression profile generated for perinatal
developing retinal ganglion cells.
Photoreceptors are responsible for converting photon inputs into neural
signals through a process called phototransduction. Rods and cones express unique
proteins that enable them to execute this function. The mechanism of rod
phototransduction has been well characterized, but little is known about cone
phototransduction. Since many homologous components of the rod
phototransduction cascade have been identified in cones, it has been suggested that
cone phototransduction is similar to rod phototransduction. We hypothesize that
CAR functions in cone photoreceptors similar to rod arrestin’s role in terminating
rod phototransduction.
Using a genetic approach, we expressed cone arrestin (CAR) in rod
photoreceptors without rod arrestin to directly compare the functional role of CAR
and rod arrestin in terminating the photoresponse. We provide light damage,
translocation and membrane association, and single cell electrophysiology evidence
that show CAR functions similarly to rod arrestin in the termination of the rod

x
photoresponse in rod photoreceptors in vivo although not as effectively as rod
arrestin. These data suggest that the role of CAR in cone phototransduction is akin
to rod arrestin’s role in quenching rod phototransduction.
In addition to analyzing the specific functional capabilities of CAR,
transcripts were screened from perinatal developing retinal ganglion cells (RGCs) for
differential expression in order to identify genes that may play a role in developing
functional RGCs. We hypothesize that there is uniform coordinated expression of
specific transcripts during perinatal RGC maturation that mediate neurite growth loss.
High-density microarray analysis of the differentially expressed transcripts from
CD90
+
RGCs was confirmed by semi-quantitative PCR and in situ hybridization for
temporal and spatial expression. Seventy-one transcripts out of ~26,000 genes were
identified as differentially expressed (ANOVA corrected p<0.05 with at least a
2^±1.5 fold difference) in RGCs from retinas of mice between embryonic day 16 and
postnatal day 5. These data help identify molecular factors that may play a
regulatory role in RGC physiology and development.

1
Chapter I: Introduction

Overview
Signaling cascades are a series of biochemical reactions that allow cells to
react and respond to the changing environment. Specific sets of genes encode for
proteins that define molecular pathways and determine a cell’s ability to respond to
external stimuli. Proper control of signaling pathways is essential for cell function
and survival.
The light activated phototransduction cascade, a process of converting photon
stimuli to electrical impulses, in rod photoreceptors has been well studied. Because
rods are much more abundant in vertebrates, significantly more is known about rod
than cone phototransduction. Our knowledge of cone phototransduction signal
termination is based on inferences drawn from studies of rod signal termination.
Rapid termination of rod signaling consists of rod arrestin (S-antigen, visual arrestin,
or arrestin 1) binding to the phosphorylated light activated G-protein-receptor
rhodopsin. Dysfunctional rod signal termination leads to rod photoreceptor cell
death (Chen J et al., 1999). Proper signal termination is critical for rod
photoreceptor survival.
Both rhodopsin and rod arrestin have cone photoreceptor homologous
counterparts (short- and medium-wavelength cone opsins and cone arrestin (CAR),
respectively). Important differences in physiological signal termination are already
known to exist and may be due to the unique properties of CAR. Recent in vitro

2
evidence suggests that binding stability of CAR with its natural cone opsin target is
less stable than rod arrestin with rhodopsin (Sutton et al., 2005).
To characterize the unique biochemical and functional properties of CAR,
ectopic expression of CAR in the well defined rod phototransduction cascade lacking
native rod arrestin was used to compare CAR with rod arrestin. These studies are in
vivo functional characterizations of CAR. We hypothesize that CAR is capable of
terminating a photoresponse in vivo and that the CAR-rhodopsin complex is less
stable than the physiological rod arrestin-rhodopsin complex.
Remarkable accomplishments have occurred during the last 40 years in
characterizing the rod phototransduction cascade from the initial stimuli (a photon),
activation of the receptor (rhodopsin), identifying and characterizing the components
in the integration and amplification step (G-proteins), understanding the response
(electrical impulses), and ultimately characterizing the termination of signaling
(receptor phosphorylation and arrestin binding). It is thought that homologous
counterparts identified in cone photoreceptors function in a similar manner.
The advances in understanding molecular functions within photoreceptor
have not been mirrored in other retinal neurons. Many physiological processes are
mediated by unknown biochemical pathways. For vision, retinal ganglion cells
(RGCs) must carry the electrical signals from photoreceptors to the brain. During
perinatal development, these neurons like all central nervous system (CNS) neurons
rapidly switch to being unable to grow long axons. Unlike in the peripheral nervous
system that can regenerate axons when lesioned, any injury to the CNS such as

3
trauma or stroke leads to irreversible loss of function.
The entire signaling cascade responsible for irreversible restriction of axon
growth from the initiating stimuli to the molecular effectors remains unclear.
Although recent evidence suggests that amacrine cells in the developing retina act as
the signal responsible for loss of axonal growth ability (Goldberg et al., 2002), there
is still debate of whether it is an internal process or externally signaled.
Physiological axonal growth loss is transcription dependent (Smith and Skene, 1997;
Cai et al., 2002). Additionally, collapse of the growth cone and the inability of
neurites to cross the glial scar occur in lesioned adult CNS neurons. Axonal growth
loss is likely due to a contribution from internal cell signals as well the hostile
external milieu. This pathway is ill defined and the components of the signaling
pathway remain obscure. To investigate the internal molecular events that mediate
the loss of axonal growth ability, differential gene expression in developing perinatal
retinal ganglion cells were profiled. We hypothesis that transcriptional factors
control key differentially expressed molecular factors that mediate the physiological
loss of axon growth in RGCs.

Retinal Morphology
Vision is an integral part in perceiving and responding effectively to changes
in the environment. The retina is a specialized tissue critical for vision. This laminar
sheet of neural tissue that lines the back of the eye is responsible for photoreception,
the process of converting light stimuli into nerve action potentials. The invariant

4
division of cellular layers and cell types vertically separates the retina into the
classically defined layers as shown by a stained transverse section of a mature adult
mouse retina (Figure 1.1). The retinal pigment epithelial (RPE) layer is the
outermost cellular layer and contains cells responsible for the phagocytosis of
photoreceptor outer disks membranes. Photons of light are absorbed by rod and cone
photoreceptors in the distal compartment called the outer segment (OS). The
photoreceptor inner segment (IS) contains the metabolic molecular machinery while
the outer nuclear layer (ONL) contains the photoreceptor cell bodies. Photoreceptors
synapse to two interneuron types, bipolar and horizontal cells. These synaptic
connections are made in the outer plexiform layer (OPL) with the cell bodies of
interneurons residing in the inner nuclear layer (INL). Bipolar and amacrine cells
synapse in the inner plexiform layer (IPL) with each other and with ganglion cell
neurons in the ganglion cell (GC) layer. These ganglion cells send all of the
processed information through afferent fibers that exit the retina at the optic disc to
the brain via the optic nerve.
The sub-cellular compartments of photoreceptors are stratified into distinct
regions that serve specialized functions. The morphology of the rod and cone
photoreceptor is similar and consists of outer segment (OS), connecting cilium, inner
segment (IS), nucleus, axon, and synaptic terminal (Figure 1.2). In rods, the OS is a
set of membranous discs derived from the plasma membrane, but is no longer
connected with one another or the surface. In cones, the OS is about 2/3 the length
of rods, composed of discs that remain attached to the plasma membrane, and tapers

5
Figure 1.1

Figure 1.1. Micrograph of a mature mouse retina illustrates the well-organized laminar nature of this
tissue. The outermost layer at the bottom of the figure consists of the retinal pigment epithelial (RPE)
layer that forms a cell layer that rests atop Bruch membrane. The outer segment (OS) consists of rod
and cone processes where phototransduction takes place. The inner segment (IS) contains the
molecular machinery for proper protein synthesis and mitochondria which supply energy for the rest
of the cell. The cell bodies of rod and cone photoreceptors reside in the outer nuclear layer (ONL).
The short axons of photoreceptors synapse to integrating neurons in the outer plexiform layer (OPL)
whose cell bodies lie in the inner nuclear layer (INL). Integrating neurons make synaptic connections
with dendrites of ganglion cell neurons in the inner plexiform layer (IPL). The cell bodies of ganglion
cell neurons whose axons make up the afferent fibers of the optic nerve tract compose the ganglion
cell (GC) layer.
RPE
OS
ONL
OPL
INL
IPL
GC
IS
RPE
OS
ONL
OPL
INL
IPL
GC
IS

6

Figure 1.2

Figure 1.2. Rod and cone photoreceptors share similar morphology. The outer segment is where
phototransduction occurs. The connecting cilium is composed of microtubules with surrounding
cytoplasm that bridges the outer segment with the inner segment, in cones it is often very small or
absent. The inner segment contains the cells metabolic machinery including mitochondria,
endoplasmic reticulum, Golgi apparatus, and ribosomes. The nucleus contains the genetic
information of the cell. Specific gene expression defines the unique signaling cascade of a cell and
how it responds to stimuli. The axon connects the cell to the synaptic terminal, in rods it is termed the
spherule, and in cones, the pedicle.
Outer
segment
Connecting
cilium
Inner
segment
Nucleus
Synaptic
Terminal
Axon
Outer
segment
Connecting
cilium
Inner
segment
Nucleus
Synaptic
Terminal
Axon

7
lightly at the apex. The rod outer segment disc membrane contains a visual pigment
that uses the chromophore 11-cis retinal and is termed rhodopsin. The disc
membrane incorporates the seven-transmembrane domain of rhodopsin and is the
site of phototransduction. In murine cone photoreceptors, the disc membranes
contain visual pigments homologous to rhodopsin (medium- and short-wavelength
opsin) with a phototransduction mechanism hypothesized to be the same. These
membranous discs are replaced every two weeks (Rodieck, 1998). In rods, the
connecting cilium is a thin cytoplasmic bridge composed of nine-microtubule
doublets similar to motile cilium but without the inner pair of microtubules that
connects the OS and the IS. In cones, the body is directly attached to the inner
segment and lies immediately deep to the OS. The inner segment is where proteins
are synthesized and where mitochondria supply energy for the rest of the cell.
Deep to the IS lies the remaining photoreceptor structures. The nucleus is in
the middle of the cell body and contains the genetic code that determines the proteins
and enzymes that are synthesized. The axon connects the cell body to the synaptic
terminal. The synaptic terminal of rods, termed the spherule, contain an invagination
of the cell membrane where the rod cell synapses with horizontal and rod bipolar
cells. The large synaptic terminals of cone, termed pedicle, contain multiple
invaginations which synapse to multiple sets of horizontal and bipolar neurons.
These specialized structures are all a result of proper gene expression during
development and help the photoreceptors execute their responsibility in responding
to light stimuli.

8

Cellular Response
How a cell responds to external stimuli depends upon the genes expressed.
Neural ectoderm cells (retinal progenitors) originally lining the neural tube form the
retinal ventricular zone early in retinal development. Proper gene expression is
critical for the generation of a mature functional retina from a single pool of cells
during the complex process of correct cell division, migration to the proper layer,
differentiation, synapse formation, and cellular maturation. Cell specific gene
expression confers the specialized response of each cell to external stimuli. The
ability of photoreceptors to convert light signals to neural signals reflects their
specific expression of components involved in responding to light stimuli; the
phototransduction cascade. Photoreceptors are specialized to respond to light inputs
by giving rise to a photoresponse.
The phototransduction cascade in the rod photoreceptor represents a classical
G-protein-coupled receptor (GPCR) cascade. This superfamily of receptors is
composed of over 1,000 different receptors characterized by a signature seven-
transmembrane configuration coupled to signal transduction via a GTP binding
protein (G-protein). These receptors mediate signaling for a multitude of hormones,
neurotransmitters, chemokines, and even calcium ions. GPCRs are involved in
responding to many stimuli from light (through rhodopsin), adrenaline,
gastrointestinal peptides (secretin, glucagons, vasoactive intestinal peptide (VIP)),
growth hormone releasing hormone, corticotrophin releasing hormone, calcitonin,

9
parathyroid hormone, and GABA
B
(Pierce et al., 2002). They play major
physiological roles in development (through the Wnt receptor, Frizzle, and the Sonic
Hedgehog receptor, Smoothened), sensory recognition, and metabolic maintenance.
GPCRs signal through similar pathways. GPCR activation allows the 2
nd
and
3
rd
intracellular loop of the activated receptor to bind to a heterotrimeric G-protein
consisting of three subunits: G
α
, G
β
, and G
γ
(Hamm and Gilchrist, 1996). The α
subunit contains a GTP/GDP binding site, which in its inactive state is occupied by
GDP. Activation of the G-protein by a GPCR catalyzes the exchange of GDP for
GTP resulting in disassociation of α from the βγ subunits and subsequent activation
of downstream effectors (Gilman, 1987). A single activated GPCR can activate
hundreds of G-proteins, and these in turn, activate many effector molecules, thereby
amplifying the original signal.

Phototransduction
In rod photoreceptors, the GPCR cascade responsible for the initial steps in
converting light stimuli to neural signals is well characterized. Rod cells mediate
dark and light perception in dim light environments. Its cascade is illustrated
(Figure 1.3). A photon of light is absorbed by the chromophore 11-cis-retinal
embedded within the opsin protein which undergoes photoisomerization to all-trans-
retinal (Wald, 1968). This induces a conformational change yielding light activated
rhodopsin (R*). R* like other GPCR binds to and catalyzes GTP-GDP exchange in

10

Figure 1.3. The main components in the activation and deactivation of the phototransduction cascade
in vertebrate rod photoreceptor cells. The chromophore, 11-cis-retinal is embedded within the opsin
protein and together compose rhodopsin (R). The 11-cis-retinal absorbs a photon and is converted
into all-trans retinal whereby it induces a conformational change resulting in light activated rhodopsin
(R*). R* binds and catalyzes the exchange of GDP for GTP on the α-subunit of the G-protein-
coupled receptor protein, transducin (G
t
α). R* activates multiple transducin molecules and thereby
amplifies the light signal. G
t
α is released from the rest of the transducin βγ complex. GTP-G
t
α then
binds the γ subunits of multiple cGMP phosphodiesterase (PDE), which in turn exposes the catalytic
sites on PDE αβ subunits that converts cGMP to GMP. The signal is further amplified when PDE
converts multiple cGMP to GMP, thereby lowering the intracellular [cGMP]. A drop in [cGMP]
causes cGMP-gated (CNGs) cation channels to close and prevents the entry of Ca
2+
and Na
+
. The
Ca
2+
/Na
+
/K
+
cation exchanger remain active, hyperpolarizing the entire rod cell. Voltage-gated ion
channels near the synaptic terminal close and intracellular Ca
2+
levels drop. Decrease Ca
2+
levels
result in decreasing glutamate release to other neurons within the retina.
The deactivation of the photoresponse requires the shut-off of R* and the return of [cGMP]
levels. The termination of R* occurs with the binding of rhodopsin kinase (RK) that utilizes ATP to
phosphorylate R*. Rod arrestin (Arr or Arr1) binds R*-P with high affinity and terminates R*
signaling. The all-trans retinal is released and converted by retinal dehydrogenase to all-trans-retinol.
Retinyl-ester isomerase converts the all-trans-retinol to 11-cis-retinol, which is converted by 11-cis-
retinol dehydrogenase to 11-cis-retinal through the retinoid cycle in the RPE (not shown). The opsin
is dephosphorylated by the pp2A phosphatase and is reconstituted with 11-cis-retinal to form
regenerated rhodopsin. Figure adapted from a figure by Dr. Ana Mendez with modifications.

11

Figure 1.3. Continued

all-trans
retinal
plasma
membrane
disc
membrane
 GDP
cGMP
5’ GMP
GTP
Ca
2+
 K
+
Exch
Na
+
Ca
2+
Na
+
cGMP
PDE
α
β
γ
CNG channel
(open)
CNG channel
(closed)
+
β
γ
GTP
α
∗
GTP
Transducin
β∗
α∗
γ
γ
GDP
GTP
exchange
β
α
β
γ
GC
GC
GCAP
GCAP
R
R*
Arr
R*
ATP
Pi
ADP RK
R*
Pi
11-cis-
retinal
Light
Pi
R
* α
*
phtase
cGMP
cGMP
11-cis-retinal
all-trans retinal
all-trans
retinal
plasma
membrane
disc
membrane
 GDP
cGMP
5’ GMP
GTP
Ca
2+
 K
+
Exch
Na
+
Ca
2+
Na
+
cGMP
PDE
α
β
γ
CNG channel
(open)
CNG channel
(closed)
+
β
γ
GTP
α
∗
GTP
Transducin
β∗
α∗
γ
γ
GDP
GTP
exchange
β
α
β
γ
GC
GC
GC
GC
GCAP
GCAP
GCAP
GCAP
R
R*
Arr
R*
ATP
Pi
ADP RK
R*
Pi
11-cis-
retinal
Light
Pi
R
* α
*
phtase
cGMP
cGMP
11-cis-retinal
all-trans retinal

12
the α-subunit of the G-protein transducin (G
t
α). Release of GTP-loaded G
t
α from
the transducin βγ dimer allows it to bind to the γ subunits cGMP phosphodiesterase
(PDE). Release of the inhibitory γ subunits of PDE exposes the catalytic sites on
PDE αβ subunits that convert cGMP to GMP. The intracellular [cGMP] decreases
causing cGMP molecules to be released from cGMP-gated (CNG) cation channels.
High levels of cGMP are required to keep the CNG cation channels open. A drop in
[cGMP] causes these channels to close and prevents the entry of Ca
2+
and Na
+
.
In the dark, when rod photoreceptors are in a state of depolarization without
R*, there is a dark current of cation flowing into the OS. This influx of Ca
2+
and
Na
+
cations into the OS is balanced by the efflux of K
+
cations leaving via K
+

selective channels in the IS, thus maintaining a stable membrane potential. Light
stimulation leads to CNG cation channels closure and a drop in Ca
2+
and Na
+
influx
into the OS. The reduced influx of Ca
2+
and Na
+
in the OS remains coupled to the
continual K
+
cation efflux in the IS, as a result, the transmembrane potential across
the entire cell hyperpolarizes. Voltage-gated ion channels located near the
presynaptic junctions close and Ca
2+
levels in the synaptic terminal falls. Ca
2+

mediates the fusion of synaptic vesicles to the plasma membrane and the concurrent
release of the vesicle-sequestered glutamate neurotransmitter into the synaptic cleft.
Neurotransmitter release by photoreceptors is reduced in response to light.

Phototransduction Signal Termination
Termination of the photoresponse requires the inactivation of R* and

13
transducin, as well as the return of cGMP levels. The level of Ca
2+
also acts as a
second messenger and needs to return to baseline levels. The termination of the
catalytic ability of R* is initiated by the binding of rhodopsin kinase (GRK1) that
phosphorylates R* at multiple sites at the C-terminal end followed by rod arrestin
binding (Wilden and Kühn, 1982; Wilden et al., 1986; Chen et al., 1995; Xu et al.,
1997; Chen et al., 1999). In the dark at high levels of Ca
2+
, Ca
2+
-bound-recoverin
binds to and inhibits RK phosphorylation of rhodopsin (Chen et al., 1995). In the
light, [Ca
2+
] is low and thus, RK is able to function. Transducin signaling is
quenched by enhanced intrinsic GTPase activity through binding of GTPase
accelerator protein RGS9-1 and γ subunit of PDE (Arshavsky and Bownds, 1992;
Chen et al., 2000).
The decreased cation influx during phototransduction activates Ca
2+
sensitive
proteins including recoverin, guanylyl cyclase activating proteins (GCAP1 and 2),
and calmodulin. With the drop in Ca
2+
levels in the OS during phototransduction,
the Ca
2+
sensitive guanylyl cyclase activating proteins (GCAPs) bind and increase
the activity of guanylyl cyclases (GCs), which in turn synthesizes cGMP faster
(Gorczyca et al., 1994; Gorczyca et al., 1995, Palczewski et al., 1994). GCAPs
which are situated on the disc membrane have three binding sites for Ca
2+
. For
activation of GCs, GCAPs must bind. In the dark, GCAPs binding are prevented
when two Ca
2+
ions are bound thereby blocking GCs association. The resulting drop
in Ca
2+
levels in the light removes bound Ca
2+
from GCAPs, free GCAPs then bind
to and activate GCs. The low [Ca
2+
] in the light also mediates calmodulin binding to

14
the CNG cation channels. Calmodulin has four binding sites for Ca
2+
and becomes
active when three or four are occupied. The active Ca
2+
-calmodulin complex has a
positively charged α helix structure that interacts with other proteins. The β subunit
of the CNG cation channels contains the site for calmodulin interaction. Binding of
Ca
2+
-calmodulin lowers the binding affinity of CNG channels for cGMP. Low Ca
2+

levels in the light cause the disassociation of Ca
2+
-calmodulin from the CNG cation
channels therefore higher affinity binding of cGMP can occur at lower
concentrations. The channels open, returning Ca
2+
and Na
+
influx into the OS (Hsu
and Molday, 1993).
Biochemical analysis defining interacting components in phototransduction
combined with electrophysiological studies has lead to a greater understanding of rod
phototransduction. Electrophysiological single cell recordings of individually
isolated cells allow biochemical reactions to be followed in vivo under real-time
kinetics. This is especially useful in conjunction with gene ablation (knockout) and
over-expression (transgene) studies targeting specific components in the biochemical
and functional characterization of individual components in the phototransduction
cascade. Suction electrodes measure the current change in the OS in response to
light stimuli. Current consists of a net movement of positive charges and is usually
measured from the interior of the cell; by convention it is negative if it passes into
the cell. The electrophysiological waveform depicting the current in rod
photoreceptors during a response to a photon consists of an activation phase until it

15

Figure 1.4

Figure 1.4. A representative single cell recording of rod photoreceptors in response to a single
photon. It contains an activation phase that reaches a peak and a recovery phase. The peak, time to
peak (tp), and time to recovery (tr) are illustrated. The single photon response is commonly
characterized by the amplitude of the peak, the time to peak (tp), and the integration time calculated
by (tr)/amplitude. The time to recovery is from the initial light flash at “0” until the photocurrent
returns to original dark levels.
wild-type
Peak
Time (s)
1.0
0.5
0.0
pA
1
0
tr
tp
wild-type wild-type
Peak
Time (s)
1.0
0.5
0.0
pA
1
0
tr
tp

16
reaches a peak, and a recovery phase from which ion movements return to normal
(Figure 1.4). Important for quantification and comparison: the peak amplitude, time
to peak (tp), and the integration time calculated by tr/amplitude (time for recovery
(tr)) are often used to describe a single photon response. They represent the
integrated molecular events of amplification and deactivation in the
phototransduction cascade. Phototransduction of a single R* leads to a decrease in
the current due to CNG channel closures as indicated by a recording peak. Proper
termination occurs when the current returns to baseline conditions.

Dysfunctional Signal Termination Leads to Photoreceptor Death
Termination of the phototransduction cascade is critical for cell survival.
Understanding the molecular mechanism of photoreceptor cell death may help to
identify similar mechanisms in retinitis pigmentosa (Lisman and Fain, 1995), age-
related macular degeneration, and other retinal dystrophies. Two pathways exist for
R* mediated phototransduction signaling to terminate. The first, a slow recovery,
involves the gradual thermal decay of R*. The second involves a 2 step mechanism
where R* is phosphorylated by a receptor kinase resulting in R*-P. This is followed
by rod arrestin binding and blockade of the domains essential for signal propagation.
Essentially, rod arrestin sterically hinders the ability of G-proteins to interact with
the rhodopsin receptor and thus quickly terminates signaling.
Defects in rod phototransduction can lead to continuous signaling equivalent
to that produced by light. Constitutive activation of phototransduction is thought to

17
ultimately lead to photoreceptor cell death (Fain and Lisman, 1993). Photoreceptor
degeneration by this proposed mechanism has been studied with both genetic
mutants and prolonged light exposure. Continual activation of transducin by opsin
unbound to chromophore in vitamin A deprived mice lacking the 11-cis-retinal
chromophore and mutants in rhodopsin interfere with arrestin binding and leads to
photoreceptor cell death (Fain and Lisman, 1993). Single cell recordings from null
mutants for proteins involved in terminating R* activity like arrestin (Xu et al.,
1997) and rhodopsin kinase (RK) (Chen CK et al., 1999) do not exhibit a proper
return to baseline kinetics. The inability to phosphorylated R* in RK knockout mice
allow R* to continually initiate signaling. Without arrestin binding, R*-P continues
to initiate inappropriately prolonged signaling.
Photoreceptors that cannot terminate rhodopsin initiated signaling are at
increased susceptibility to cell damage and ultimately cell death. The degree of cell
death can be quantified through morphologic measurements. Morphological
assessments in arrestin1 null retinas after light exposure effectively demonstrate the
loss of rod photoreceptor cells (Chen J et al., 1999). Importantly, dark reared null
arrestin1 and RK mutants do not exhibit increased photoreceptor cell loss when
compared to controls. Yet when light exposed, these null mutants do lose
photoreceptors (Xu et al., 1997 and Chen J et al., 1999). In null mutants of
transducin, the G-protein necessary to propagate the rhodopsin receptor signal,
photoreceptor cell death was not observed (Calvert et al., 2000). This mutant
effectively uncouples signal transduction and is protective from inappropriate

18
phototransduction signaling. Together, these studies provide evidence that
constitutive light activation by abnormally prolonged transducin mediated
phototransduction signaling leads to cell death.
Indeed, the theory of constitutive light activation is not the only pathway for
light induced cell death. In an effort to identify molecular factors prior to transducin
in the cascade that may initiate apoptosis, a study using a genetic approach with null
mutants of transducin and null mutants of c-Fos (a component of the AP-1
transcription complex), identified two apoptotic pathways of light induced
photoreceptor cell death. Induction of the AP-1 transcription factor has been
demonstrated to precede bright light induced cell death. The authors concluded that
high intensity bright light triggered apoptosis requires rhodopsin activation and the
transcriptional factor AP-1, but not transducin. Dim light triggered photoreceptor
apoptosis requires transducin, but not AP-1 (Hao et al., 2002). They suggest that
divergent pathways exist which lead to light induced cell death depending on the
type of light exposure; either high intensity for short duration or low intensity for
longer periods. The signaling cascade linking constitutive light activation and
photoreceptor apoptosis remains unknown. It is clear that dysfunctional signal
termination leads to photoreceptor cell death. Restoration of proper
phototransduction termination will lead to recovery of photoreceptors.

Rod and Cone Photoreceptors
The abundance of rod photoreceptors allows in depth biochemical and

19
electrophysiological analysis of specific gene contributions to cell physiology, yet
much less is known about the genes that determine cone photoreceptor function. In
the adult mouse retina, photoreceptors account for ~70% of all retinal cells (Young,
1985; Drager and Olsen, 1980) and only about 3% of these are cones. This
represents on average about 6.4 million rods and about 180,000 cones per retina
(Carter-Dawson and LaVail, 1979; Jeon et al., 1998).
The importance of cone cell mediated high acuity color vision required for
most daily activities cannot be trivialized. The scarcity of these cells has made them
difficult to study despite their importance. Since the density of cones in the mouse
retina is on the same order of magnitude as cone density in monkey retinas 3-4 mm
from the fovea (Wikler and Rakic, 1990), the mouse retina is also a valuable tool in
studying macular degenerative related disease that affect cone-mediated vision.

Photoresponse in Rods and Cones
Though both rods and cones are responsible for vision, they posses unique
characteristics. Rods are functionally limited in their dynamic range because high
light intensity can saturate their response. Rods can respond to a single photon but
saturate at above 400 photoisomerizations of the chromophore per second while
cones do not (Rodieck et al., 1998). Rods therefore function predominately in dim
light while cones mediate sight during the day. Cones can limit the number of
receptors photoisomerizing by limiting the rate of chromophore regeneration. The
chromophore is usually sequestered in the retinal pigment epithelial layer away from

20
the photoreceptor until needed to form functional receptors. It normally takes about
2 minutes to regenerate a chromophore on average in rods. In cones, a second
alternative pathway independent of the RPE layer speeds chromophore regeneration.
The single cell recordings of the photoresponse in rod and cone photoreceptor
cells are distinctive. As mentioned, in the dark there is a resting influx of cation
flowing into both the rod and cone photoreceptor outer segment creating what is
termed a dark current. This recorded photoresponse is quantitative in measuring the
response to a photon of light. The rod photoresponse is clear and represents nearly
2% of the total dark current. On the other hand, the cone photoresponse to a single
photon is only 5% in amplitude of a typical rod response. This small response in
cones increases the noise to signal ratio contributing to the difficulty in studying the
cone phototransduction cascade. Essentially, the cone response may be buried in
noise. Not only is the relative rarity of cone cells a barrier to characterizing the cone
phototransduction cascade, the small cone photoresponse is an additional challenge.
Rods and cones respond to photons of lights differently. The cone
photoresponse is faster. Normal cones take about 50 ms to peak while in rods, it
takes 125 ms. Unique to cone photoreceptors is a biphasic waveform. Cones
overshoot their baseline dark current levels during the recovery phase. Falling
cGMP levels cause guanylate cyclase activity to increase cGMP levels, resulting in
the opening of some of the cation channels thereby counteracting a fraction of the
closing CNG channels. Though this occurs physiologically in both rods and cones,
in cones it is sufficiently stronger and faster to create a biphasic waveform. These

21
Table 1.1
Proteins
Rod
Transduction
Proteins
(Amino
Acids)
Cone
Transduction
Proteins
(Amino Acids)
Homology
(%)
Visual Pigment 348 359 (M-opsin) 42.3
346 (S-opsin) 44

Transducin α1 350 354 79.9

PDEγ 87 83 94.2

cGMP-gated Channel α 684 631 71.5

Arrestin 403 381 70

Rhodopsin Kinase 564 same

RGS29/Gβ5 675/395 same

GCAP1 202 same
GCAP2 201 same

Recoverin 202 same

Table 1.1. Mouse rod and cone photoreceptors express many homologous phototransduction proteins.
Homologous phototransduction cascade proteins suggest a highly comparable functional cascade in
both rod and cone photoreceptors. Figure adapted from a figure by Dr. Guang Shi with modifications.

22
physiological differences are most likely mediated by the unique biochemical
characteristics of interacting cone phototransduction factors.
Electrophysiological single cell recordings represent the concerted
interactions of all the molecules in the phototransduction cascade. As opposed to
rods, the cone phototransduction cascade is not well characterized. Homologous
components to those in the rod phototransduction cascade exist (Table 1.1). It has
been suggested that these homologous cone proteins function as their rod
counterparts do in rod phototransduction. Most of our conclusions about the cone
phototransduction cascade are based on experiments with rod signaling factors. Cell
specific molecules and their unique interactions contribute to specialized cone
functions. Of particular interest to us is cone arrestin (CAR) which is hypothesized
to function in the termination of the cone photoresponse akin to rod arrestin’s role.
Its gene expression pattern has been well characterized: it is developmentally
regulated and is expressed in the cone photoreceptors as well as pinealocytes
(Murakami et al. 1993, Craft et al. 1994, Zhu et al., 2002). Structurally, CAR is
70% homologous to rod arrestin.

Arrestins
To date, four mammalian arrestins have been identified (Craft and Whitmore,
1995). Structurally, they are highly homologous and exhibit 44-84% identity.
Functionally, the members of this family of proteins are thought to act in a similar
manner. The classical mechanism of arrestin function in receptor desensitization

23
includes arrestin binding to their respective phosphorylated signal activated receptor
targets. This binding blocks receptor interactions with the heterotrimeric G protein
and thus arrests signal transduction.
The arrestin superfamily is composed of four homologous members. Rod
arrestin, localized in photoreceptors, is the second most abundant protein consisting
of ~3mM in the retina (Shinohara et al. 1987) while rhodopsin is the most abundant
at ~5mM. Rod arrestin is involved in quenching phototransduction. The non-visual
arrestins, β-arrestin (arrestin-2) and β-arrestin-2 (arrestin-3), are ubiquitously
expressed and regulate hundreds of different GPCRs. An additional segment within
the C terminal binding domain in these non-visual arrestins mediates clathrin
dependent internalization of many GPCR and also serves as a scaffold to commit
GPCR to alternative signaling pathways including MAP kinases. Arrestin-2 is
involved in desensitization of the β
2
-adrenoceptor (β
2
AR) while arrestin-3 interacts
with β
2
AR and odorant receptors. The fourth arrestin, cone arrestin (CAR) has been
proposed to terminate cone opsin initiated phototransduction in cone
phototransduction. Both visual arrestin and cone arrestin lack the C-terminal clathrin
binding site found in β-arrestin that is necessary for internalization of a receptor-
arrestin complex.

Rod Arrestin
Rod arrestin is a critical part of the phototransduction cascade. It is the
protein that rapidly terminates signaling when the external light stimulus is no longer

24
present. A large amount of protein termed S-antigen protein was identified in the
bovine retina (Wacker et al., 1977). This 48-kDa protein, later found to be rod
arrestin, remains soluble in rod photoreceptors in the dark but associates with the
disk membranes in the light (Kühn et al., 1984). The authors show it binds to R*-P
and, for the first time, that excessive transducin can displace arrestin from the
phosphorylated rod outer segment membrane. This study is instrumental in
demonstrating the sequential termination of phototransduction. It also suggests the
mechanism whereby arrestin binding sterically hinders the propagation of the G-
protein coupled receptor signaling cascade in the retina. Subsequent studies
quantitatively characterized the competitive nature of arrestin and transducin for R*
and R*-P (Krupnick et al., 1997). There is a limited effect of phosphorylation to
transducin binding and signal propagation. Using a competitive binding assay, they
conclude that the major role of rhodopsin phosphorylation is to promote high
specificity and high affinity arrestin binding. Single cell recordings from arrestin
knockout mice demonstrated the requirement of arrestin for rapid termination of the
photoresponse (Xu et al., 1997).
The interaction between R*-P and rod arrestin has been characterized. To
identify the functional domains of rod arrestin, cell free in vitro assays of rod arrestin
along with two different truncated arrestin mutants were bound to R*-P (Gurevich
VV and Benovic, 1992). They demonstrate that full length rod arrestin has a 10-12
fold greater affinity for R*-P than either R* or R alone. Most importantly, they
delineate in broad terms the functional domains of rod arrestin: the C-terminal

25
domain is important for the specificity of arrestin to recognize R*-Pi. Further studies
using thirty-three different truncated and deleted mutations of bovine arrestin from
69-391 of the 404 amino acid long protein in similar assays identified multiple
functional domains (Gurevich and Benovic, 1993).

Rod Arrestin’s Structure
The resolution of the crystal structure of bovine rod arrestin to 2.8 Å (Hirsch
et al., 1999) placed into perspective the functional domains and the significance of
specific residues. Bovine rod arrestin is separated into 4 separate regions: N domain
(residues 8-180), C domain (residue 188-362), which is joined by a Linker region
(residues 362-371), and a C tail (residue 372-404). A polar core is embedded
between the N and C domain in the fulcrum of the rod arrestin molecule. It
comprises charged residues from the very N terminus (Asp-30), the body of the N
domain (Arg-175 and Lys-176), the interfacial loop of the C domain (Asp-296 and
Asp-303), and the C tail (Arg-382) (Figure 1.5). Amino acid mutations to these
residues alters the ability to recognize and bind R*-P (Gurevich and Benovic, 1997).
Mutational analysis of Arg-175 (R175E) abolishes arrestin’s ability to distinguish
between R*-P and R* implicating Arg 175 as a critical phosphate sensitive trigger.
Another mutant D296R binds R* at levels similar to R175E. When the mutations at
each residue are combined, R175E/D296R restores wild-type activity (Vishnivetskiy
et al., 1999). These residues interact cooperatively to give rod arrestin its particular
binding properties.

26
The C tail also plays a crucial role in discriminating against R* and R*-P. A
splice variant where the last 35 amino acids are replaced with an alanine residue
resulted in a 44 kDa protein (p44). Compared to the more abundant 48 kDa bovine
rod arrestin, p44 binds to R* as well as R*-P. The C tail interacts with Arg-382, Asp
30 and Asp 303 in the polar core (Hirsch et al., 1999). These residues are highly
conserved in all arrestins and are crucial for proper protein function. Together they
exhibit an elaborate set of hydrophobic and hydrogen bonds that regulate the affinity
of rod arrestin to its phosphorylated receptor. Intrusion of the phosphate moiety on
the C-terminal end of the rhodopsin receptor into arrestin’s polar core region disrupts
the electrostatic interactions and leads to a rearrangement in the structure of arrestin
into an active state enabling receptor binding. To date, it still remains unclear which
rod arrestin residues directly interact with rhodopsin.

Cone Arrestin
Unlike the detailed studies of rod arrestin, there is limited independent
knowledge of cone arrestin (CAR). CAR function and mechanism of action in cones
is thought to be similar to rod arrestin in rods based primarily on sequence homology.
Cones respond much more rapidly than rods, demonstrate unique single photon
responses electrophysiologically, and light activated cone pigments decays 10-100
times faster (Imai et al., 1995). The mechanism of CAR interaction may be to
provide faster binding and dissociation mediating the cell’s demands of its signaling
cascade.

27
More recently, the crystal structure of salamander CAR in its basal state has
been resolved to 2.3Å. Crystallized CAR has the similar canonical arrestin structure
of two concave receptor binding domain (containing 10 β-sheets) connected by a
linker region and a C terminal tail (Sutton et al., 2005). This brings to date a defined
structure for three of the four arrestins.
Receptor specificity of the arrestins is dependent on specific residues within
the binding site. Structurally, CAR appears to be a fusion of the highly specific rod
arrestin and the promiscuous non-visual arrestin-2 (Sutton et al., 2005). Studies
using multiple chimeric rod arrestin and arrestin-2 mutants with in vitro binding
assays to rhodopsin and the m2 muscarinic cholinergic receptor (m2 mAChR)
identifies 2 major determinants of receptor specificity: a 41 residue region in the N
domain and a 32 residue region in the C domain (Vishnivetskiy et al., 2004).
Structural analysis identifies unique H-binding networks in the N domain receptor-
binding segment in CAR and arrestin-2 between Y46 and N159 (residues correspond
to Salamander Ambystoma tigrinum CAR, also known as 1SUJ). Y46 and N14,
along with K16 and D43 are missing from rod arrestin (Sutton et al., 2005). The H-
bonding network is present in mouse cone arrestin (mCAR) except for the initial Y46
and N159 pair (Figure 1.5). The receptor binding C domain exhibits higher
sequence homology between CAR and rod arrestin than arrestin-2 (Sutton et al.,
2005). Sequence alignment and structurally, CAR appears to be a hybrid of the
highly selective rod arrestin and the non-visual arrestin-2. This suggests the affinity
and specificity of cone arrestin is intermediate to that of the highly specific tight rod

28

Figure 1.5: Sequence alignments of homologous salamander cone arrestin (1SUJ), mouse cone
arrestin (mCAR), bovine rod arrestin, and mouse rod arrestin. Based on the sequence analysis, CAR
is thought to function similarly to rod arrestin in binding to activated phosphorylated rhodopsin and
terminating signal transduction by blocking the binding of transducin. Crystal structure analysis of
rod arrestin and CAR display 2 concave receptor binding domains containing 10 β-sheets connected
by a Linker region and a C-terminal tail. The residues in bold represent the H-binding network
identified in salamander CAR that help stabilize the core of the protein. The bold and underlined
residues represent residues that compose the main core of mCAR. There is a high degree of
homology: “*” are identical, “:” conserved substitutions, and “.” semi-conserved residues. Sequences
are from the NCBI Protein database.

29
Figure 1.5 Continued

1SUJ ------madg skvykktcpn aklsiylgkr dfvdhvehve pvdgvvlidp 50
MCAR ---------m stvfkktssn gkfsiylgkr dfvddvdtve pidgvvlvdp
Bovine arrestin -mkankpapn hvifkkisrd ksvtiylgkr dyidhverve pvdgvvlvdp
Mouse arrestin maacgktnks hvifkkvsrd ksvtiylgkr dyvdhvsqve pvdgvvlvdp
::** . : .:****** *::*.*. ** *:*****:**

1SUJ eylkdrkvfv tltcafrygr ddldligmsf rkdlyslatq vyppe---tk 100
MCAR eylegrklfv rltcafrygr ddldvigltf rkdlyvqtkq vapaeptsiq
Bovine arrestin elvkgkrvyv sltcafrygq edidvmglsf rrdlyfsqvq vfp--pvgas
Mouse arrestin elvkgkkvyv tltcafrygq edidvmgltf rrdlyfsrvq vyp--pvgam
* ::.::::* ********: :*:*::*::* *:*** * * *

1SUJ epltplqekl mkklgahayp fcfkmgtnlp csvtlqpgpd dtgkscgvdf 150
MCAR gpltalqerl lhklgvnayp ftlqmvanlp csvtlqpgpe dsgkpcgvdf
Bovine arrestin gattrlqesl ikklgantyp flltfpdylp csvmlqpapq dvgkscgvdf
Mouse arrestin svltqlqesl lkklgdntyp flltfpdylp csvmlqpapq dvgkscgvdf
* *** * ::*** ::** * : : ** *** ***.*: * **.*****

1SUJ evkafcaenl ---eekihkr nsvqlvirkv qfapanlgva pkteitrqfm 200
MCAR evksfcaenl ---eekipks dsvqlvvrkv qfsalepgpg psaqtirsff
Bovine arrestin eikafathst dveedkipkk ssvrllirkv qhaprdmgpq praeaswqff
Mouse arrestin evkafasdit dpeedkipkk ssvrllirkv qhappemgpq psaeaswqff
*:*:*.:. *:** * .**:*::*** *.:. : * * :: .*:

1SUJ lsdrplhlea sldkeiyyhg epinvnvkin nttgkivkki kiiveqvtdv 250
MCAR lssqplqlqa wmdrevhyhg eaisvhvsin nytnkvirri kiavvqttdv
Bovine arrestin msdkplrlav slskeiyyhg epipvtvavt nstektvkki kvlveqvtnv
Mouse arrestin msdkplnlsv slskeiyfhg epipvtvtvt nntdkvvkki kvsveqianv
:*.:**.* . :.:*:::** *.* * * :. * * * :::* *: * * ::*

1SUJ vlfsldkyvk tvcaeetndt vaanstlskt fsvtpmlann rekrglaldg 300
MCAR vlysldkytk tvfvqeftet vaanssfsqt favtpllaan cqkqglaldg
Bovine arrestin vlyssdyyik tvaaeeaqek vppnssltkt ltlvpllann rerrgialdg
Mouse arrestin vlyssdyyvk pvaseetqek vqpnstltkt lvlvpllann rerrgialdg
**:* * * * .* :* :. * .**::::* : :.*:** * :::*:****

1SUJ klkhedtnla sttvirpgmd kevlgilvsy kvkvhlvvar ggilgdltss 350
MCAR klkhedtnla sstilrpgmn kellgilvsy kvrvnlvvsy ggilgglpas
Bovine arrestin kikhedtnla sstiikegid ktvmgilvsy qikvkltvs- -gllgeltss
Mouse arrestin kikhedtnla sstiikegid rtvmgilvsy hikvkltvs- -gflgeltss
*:******** *:*::: *:: : ::****** :::*:*.*: *:** *.:*

1SUJ dvavelpltl mhpkpsddk- ---prseedi iieefarqkl dgekddeeek 400
MCAR dvgvelpvil ihpkpspge- ravatssedi vieefmqhns qtqs------
Bovine arrestin evatevpfrl mhpqpedpdt akesfqdenf vfeefarqnl kdageykeek
Mouse arrestin evatevpfrl mhpqpedp-- akesvqdenl vfeefarqnl kdtgentegk
:*.*:*. * :**:*. . .*:: ::*** ::: .

1SUJ edverees-
MCAR ---------
Bovine arrestin tdqeaamde
Mouse arrestin kdedagqde

30
arrestin and promiscuous lax arrestin-2 binding.
In vitro binding assays were used to test the hypothesis of an intermediate
binding specificity of CAR (Sutton et al., 2005). Binding assays of human cone
arrestin, bovine arrestin-2, and bovine rod arrestin to human green cone opsin,
human m2 mAChR, and bovine rod arrestin were tested. Each arrestin was assayed
with its respective native target and also with the other arrestin proteins. CAR binds
preferentially to the cone opsin protein and m2 mAChR with little rhodopsin binding
(Sutton et al., 2005). In their study, as well as previous studies, rod arrestin binds
preferentially to rhodopsin while arrestin-2 can bind both visual opsins as well as m2
mAChR. Indeed, the absolute level of CAR binding to cone opsin is considerably
less (~1.5 fmol) compared with rod arrestin to rhodopsin (~27 fmol) and the
promiscuous non-visual arrestin-2 (~11 fmol) to m2 mAChR. Using column and
centrifugation assays to separate tightly bound and free arrestin, they also determined
that there was a fairly low affinity of CAR to cone opsin (Sutton et al., 2005). There
is a high rate of dissociation and instability in the CAR-cone opsin complex. Rod
arrestin-rhodopsin complex, on the other hand, is highly stable (Gurevich and
Benovic, 1993). Altering this key protein involved in the phototransduction cascade
helps explain the cone specific photoresponse.
Cones function in high intensity light conditions where the receptor, cone
opsin, is rapidly utilized for signaling and is consumed/ bleached. Rapid use of the
receptor necessitates a high rate of receptor turnover. Arrestin release is necessary
prior to dephosphorylation (Palczewski et al., 1989). The loose CAR-opsin complex

31
may play an instrumental role in accelerating pigment replenishment. In vertebrates,
rod arrestin only needs to bind to rhodopsin while mCAR has 2 cone pigment targets.
Recently emerging data suggests possible rod arrestin expression in cones yet its
functional role remains to be characterized. CAR may need to bind multiple cone
opsin targets so high specificity may be impossible and this may contribute to lower
CAR-opsin complex stability.
Sequence analysis and in vitro binding assays provide a strong foundation to
begin to characterize the mechanism of cone arrestin function in vivo. Sequence
alignments of arrestins are highly conserved suggesting a similar role in termination
of a heterotrimeric G protein signaling cascade. Yet, direct evidence of the in vivo
function of CAR remains lacking. We have taken a genetic approach to analyze the
functional contribution of CAR in photoreceptors. To test the hypothesis that CAR
can terminate a photoresponse and binds less stably with rhodopsin than rod arrestin,
we have generated transgenic mice expressing mCAR in rod photoreceptors lacking
native rod arrestin. The isolation of this specific cone protein in the well studied rod
system allows the functional characterization of CAR with fewer confounding
factors. It also provides an excellent environment in which to compare and contrast
CAR function with the well characterized rod arrestin. The rod phototransduction
signaling cascade has been well defined and permits enough material for biochemical
assays. Rod homologous proteins identified in cone photoreceptors suggest similar
functions within the phototransduction cascade in cones. Many other physiologic
signaling cascades that mediate physiological functions in retinal cells such as loss of

32
axon growth abilities in retinal ganglion cells are not as well defined as the
phototransduction cascade.

Retinal Ganglion Cells
Unlike photoreceptors that initiate the first steps of vision, retinal ganglion
cells mark the final step of information processing by the neural retina and carry this
information to the brain. RGCs reside in the GC layer of the retina (Figure 1.1) and
represent about 3% of retinal cells (Young, 1995). Though mouse RGCs can be
classified in 4 groups and fourteen subtypes based on morphology (Sun et al., 2002),
the putative identification characteristic of RGCs is that these neurons send axons
through the optic nerve to targets in the brain. The estimated total number of RGCs
in a C57/BL/6 wild-type mouse retina based on counting axons in the optic nerve
~300 µm from the posterior pole is about 40,000 cells which means ~54% of the cell
in the GC layer are displaced amacrine cells per retina (Jeon et al. 1998).
Ideally, analysis of internal molecular interactions and signaling cascades
within cells requires the identification and isolation of pure cell types. Though not
abundant, highly pure intact viable RGCs can be isolated using the CD90 (Thy1)
surface marker by immunopanning (Barres et al., 1989). When using a magnetic
assisted cell sort purification of RGC with CD90 technique, retrograde 1,1'-
dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI) tracer injected
into both sides of the superior colliculus prior to purification yields ~92.5% ±20.8%
fluorescently labeled cells with 93.7% ±4% expressing CD90 (Huang et al., 2003).

33
The cell surface molecule CD90 (Thy1) has been used to identify RGCs in
the retina. CD90 is the smallest member of the immunoglobulin superfamily (Morris,
1985), and in the brain, it is expressed exclusively by neurons (Xue et al., 1990;
Barlow and Huntley, 2000). This glycoprotein spans the plasma membrane via a
glycosyl-phosphatidyl inositol linkage domain that can interact with intracellular
signaling pathways, though its role during development and in maintaining
functional neurons remains unclear. Evidence from CD90 null mice suggests it is
not essential for normal neural development, maintenance of axonal pathways or
functional synaptic connections, nor does it play an essential role in mediating
axonal growth, regeneration, nor plasticity in mature CNS neurons (Barlow et al.,
2002). We utilize the CD90 cell surface molecule to identify and purify developing
RGCs in generating transcript expression profiles as a starting point in identifying
transcriptional factors that control the molecular signaling cascade mediating the
developmental loss of axon growth abilities.

Axonal Elongation
As RGCs develop and mature, two complementary processes ensure correct
axonal connections: axonal pathfinding and axonal growth. These two processes
interact complementarily so that neurons correctly locate and synapse to their target.
During axonal pathfinding, growth cones at the tips of extending axons respond to
the attractive or repulsive molecules in the environment. They integrate these signals
and coordinate cytoskeletal reorganizations to allow axonal growth to preferentially

34
occur on one side resulting in axon movement and turning. Briefly, the Rho family
of small GTPases including Rac, cdc42, and Rho has been implicated in mediating
this action. Activation of the Rac and cdc42 pathway with concurrent inhibition of
RhoA leads to actin polymerization. Conversely, RhoA activation and inhibition of
Rac and cdc42 pathway leads to actinomyosin attraction and a repulsive response.
Signaling cues such as netrin, slits, and semaphorins act as both repulsive and
attractive chemical modulators depending on the time and location of the traversing
neuron.
In order for a neuron to be functional it must locate its correct target and be
able to traverse the extraneuronal milieu. Two parallel lines of investigation have
focused on the inhibitory extracellular environment, in particular myelin, and more
recently, the intrinsic loss of axonal growth abilities due to irreversible changes
within the maturing CNS neuron.

Axonal Growth
During neurogenesis, embryonic CNS neurons extend their axons quickly
over long distances to reach their target. Mature CNS neurons lose this ability. CNS
neonatal spinal neurons are able to regrow lesioned axons (Bregman and Goldberger,
1982) even to the extent of recovering function (Bregman et al., 1993). With lesions
to the pyramidal tract in the spinal cord, CNS neurons lose the ability to regenerate
between P4 and P20 (Kalil and Reh, 1982). These in vivo data of young CNS

35
neurons capable of regenerating their axons while mature CNS neurons cannot are
reflected in culture models of other CNS neurons.
The physiological switch during the perinatal period is clearly demonstrated
by co-culture experiments. Retinal tissue containing RGCs were abutted and co-
cultured with midbrain tectal tissue and examined for new axonal projections in
hamsters (Chen et al., 1995). In vitro co-cultures of purified RGCs and dorsal root
ganglion neurons on different ages of optic nerve cryosection

(Shewan et al., 1995),
as well as in vitro cultures of purified rat RGCs (Goldberg et al., 2002), and mice
RGCs (Chen et al., 1997) all reflect the physiologic loss of growth observed in vivo.
Though the specific dates vary due to species variation, in vitro CNS neurons
accurately reflect in vivo developmental loss of axonal growth abilities.
Debate continues regarding contribution of the inhibitory nature of a hostile
CNS environment or an irreversible internal cell signal. Maturation of the CNS
environment including glial cells (astrocytes and oligodendrocytes) and myelin
occurs during this perinatal time period. Inhibitory molecules within the myelin
matrix exert an inhibitory influence on axonal growth, but increasing evidence
support factors within the maturing neuron as playing the more dominate role.

Myelin-Associated Inhibitors of Axon Growth
Historically, the developmental loss of axonal growth was thought to be
primarily due to myelin-associated inhibitors. Three main myelin-associated
inhibitors of axonal growth and regeneration have been identified: Nogo, myelin-

36
associated glycoprotein (Mag), and oligodendrocyte myelin glycoprotein (OMgp).
The presence of any of these inhibitors leads to growth cone collapse and the
termination of axon growth. The convergence of these inhibitors in activating the
Nogo receptor suggests a functional redundancy.
Nogo, a member of the reticulon family that comprises proteins associated
with the endoplasmic reticulum, exists in three isoforms: -A, -B, or -C. The IN-1
monoclonal antibody directed against a fraction of myelin inhibits neurite extension
when added to in vitro (Caroni and Schwab, 1988) and in vivo (Schnell and Schwab,
1990) encourages neurons to grow axons. It remained unclear to what antigen IN-1
targeted until a decade later when Nogo was subsequently cloned by 3 independent
groups (Chen M et al., 2000; Grandpre et al., 2000; Prinjha et al., 2000). A 66
residue sequence common to all 3 isoforms (Nogo-66) is primarily responsible for
inducing neurite growth cone collapse. Interestingly, Nogo-A has been localized in
both the inner and outer myelin membrane of young and mature CNS. The increased
expression of Nogo in the myelin is thought to result in axon growth termination. Its
up regulation at the site of CNS damage inhibits growth across the glial scar (Huber
et al. 2002). It remains unclear if Nogo exerts a similar inhibitory effect on
neighboring axons since Nogo-A and –B are expressed by CNS neurons (Chen et al.,
2000).
Mag, another important myelin inhibitor, is a member of the immunoglobulin
(Ig) superfamily. It contains five extracellular Ig-like domains and is bifunctional.
Of the total CNS myelin proteins, Mag represents only 1% and is found mostly in the

37
inner loop closest to an axon. In neurite outgrowth studies in vitro, Mag inhibits
neurite outgrowth in mature neurons, yet in young neurons Mag promotes axonal
growth (Mukhopadhyay et al., 1994). This suggests Mag is not the initiating signal
for axon growth loss.
Most recently, OMgp, a glycosylphosphatidylinositol (GPI) linked protein,
was identified as another myelin associated inhibitor (Wang et al., 2002). Omgp is
thought to be minor component of myelin that is much lower in abundance than even
Mag. It contains a leucine rich repeat (LRR) domain linked to a C-terminal
intracellular signaling serine/threonine domain. OMgp is not only found on
oligodendrocytes as the name implies, but is also highly expressed on neurons. The
location of Mag and OMgp on neurons, as well as myelin, suggests competitive
neuronal inhibition of axonal growth.

Myelin-Associated Inhibitors Activate the Nogo Receptor
All three myelin-associated inhibitors induce growth cone collapse and
consequently inhibit axonal elongation. They are initiating stimuli that activate a
signaling cascade that leads to loss of axon growth. Using Nogo-66 as a probe,
Nogo receptor (Ngr) was identified as the membrane surface receptor for the Nogo
ligand (Fournier et al., 2001). Ngr is a GPI-linked protein containing a series of
eight LRR and a second region closer to the membrane-associated region of the GPI
terminus domain. Subsequent studies have shown that Mag (Liu, 2002), OMgp
(Wang et al., 2002), and Nogo all use the same receptor complex to signal collapse

38
even though no sequence or domain similarity has been identified between them.
Ngr is a GPI-linked protein without a cytoplasmic signaling domain and is unable to
induce signaling alone. The p75 neurotrophin receptor (p75NTR) is the putative
signal transducer for Ngr. Immunoprecipitation assays for Mag, Nogo, or OMgp
identifies both Ngr and p75NTR supporting a cohesive signaling pathway.
The presence of only one of these myelin-associated inhibitors is capable of
restricting axonal elongation by activating an inhibitory signaling cascade through
the Ngr-p75NTR complex. This complex propagates the inhibition of neurite growth
sequentially via the small GTPase Rho, which activates Rho Kinase (ROCK),
myosin light chain kinase, and myosin light chain that ultimately results in
cytoskeletal rearrangements (Filbin, review 2003).
During the perinatal period when RGCs lose their axonal growth potential,
p75NTR is expressed even at E15.5 (Frade and Barde, 1999). Increased numbers of
RGCs survive axonal lesions in p75NTR knockout mice. It remains to be answered
whether p75NTR knockout CNS neurons retain axonal growth abilities when they
are mature. Such experiments would clarify the contribution of myelin-associated
inhibitors in signaling axon growth loss. In any case, in vitro studies of young CNS
neurons do not exhibit the same limitations that more mature CNS neurons on
myelin (Cai et al., 2001). Wild-type embryonic CNS neurons that express p75NTR
are still capable of extensive axon growth in culture on inhibitory myelin. This
suggests loss of axon growth is controlled by other factors. The delineation of
myelin-associated inhibitors signal transduction pathway via a single Ngr-P75NTR

39
receptor complex and its presence in both young and mature CNS neurons suggests
this biochemical cascade is not the primary mechanism in the developmental loss of
axonal growth capabilities.

Intrinsic Control of Axonal Growth
Even early studies in axon growth suggest a dynamic interplay between the
inhibitory extraneuronal environment and intrinsic CNS neuronal properties.
Replacement of the entire hostile CNS myelin with peripheral nerve grafts into
lesions of the spinal cord demonstrated limited ability for these mature CNS neurons
to grow axons (Richardson et al., 1980). Adult RGCs cannot grow their axons even
if given a permissive environment either. Studies of RGCs traversing a 4 cm
autologous peripheral nerve graft linking one eye to the superior colliculus along an
extracranial course was labeled. Only 20% of surviving RGCs grew new axons
(Vidal-Sanz et al., 1987). In these studies, functional recovery is limited and
proceeds extremely slowly if at all. Surviving RGCs take nearly 2 months to
regenerate through a peripheral graft (Aguayo et al., 1987). Yet, during
neurogenesis, neurons routinely differentiate and reach their targets in days. The
data all indicate that the hostile environment is only part of the cause. More recent
efforts have begun to focus on the intrinsic core molecular changes that refinements
in cell isolation techniques have allowed the purification of individual cell types.
Intrinsic changes in developing CNS neurons play an important role in
restricted axon growth. Co-cultures of progressively maturing RGCs analyzed for

40
neurite penetration into various ages of midbrain tectal tissue determined that young
embryonic neurons extend numerous axons into tectal tissue of all ages, yet postnatal
neurons do not (Chen et al., 1995). The authors demonstrate a precipitous decrease
in postnatal axon growth compared to younger embryonic neurons is primarily due
to the neuron itself and not the CNS environment. This places into new perspective
the studies of nearly a decade earlier with peripheral nerve grafts. Failure to grow
axons given a permissive environment is a result of intrinsic core neuronal changes
that occur in maturing CNS neurons.
Previous reports on the down-regulation of bcl-2 expression (Chen et al.,
1997) and decreases in cAMP levels (Cai et al., 2001) are proposed to be involved in
the biochemical cascade leading to growth loss in maturing neurons. During
expression analysis of neural cell adhesion molecule (NCAM), β-amyloid precursor
protein, and bcl-2 protein (bcl-2) using immunofluorescence staining, bcl-2
expression declines in more mature CNS neurons (Chen et al., 1997). RGCs from
bcl-2 transgenic mice retain the ability to grow axons compared to wild type neurons.
Yet complicating the study, bcl-2 is a proto-oncogene. The observed increases in
axonal growth may be due to increased RGC cell survival. The increase in surviving
neurons can explain the increase in numbers of neurons sprouting axons. Subsequent
studies over-expressing bcl-2 using an adenoviral vector in postnatal RGCs in culture
increases cell survival, but could not induce axon growth in the absence of trophic
factors (Goldberg et al., 2002). Bcl-2 may not be the involved in axon growth
cascade, but may play more of a role in RGC survival than actual neurite growth.

41

cAMP and Neurite Growth
Similarly to bcl-2 expression, the dramatic decrease in cAMP levels during
the perinatal period is thought to play a role in mediating axon growth loss.
Endogenous levels of cAMP decreases with developmental in CNS neurons like
dorsal root ganglion cells (DRG), RGCs, and corticospinal tract neurons as axonal
growth abilities are lost (Cai et al., 2001). Elevating [cAMP] using forskolin (an
inducer of adenylyl cyclase), phosphodiesterase inhibitors (inhibitors of cAMP
degradation), various neurotrophins, and introduction of non-degradable cAMP
analogs (such as chlorophenylthio-cAMP or dibutryl-cAMP) overcomes the
inhibitory effects of Mag and myelin. Neurons with higher cAMP levels extend
more axons compared to controls (Cai et al., 1999). This suggests cAMP is involved
in the biochemical cascade resulting in axon growth loss.
The neuronal response to increases in cAMP consists of a two phase response.
The first phase is an acute and transcription-independent activation of protein kinase
A (PKA). PKA rapidly acts on the cytoskeleton leading to growth cone collapse and
growth inhibition. The second phase is PKA independent but transcription
dependant and encourages axon growth. Conditioning peripheral nerve lesions are
known to increase cAMP levels in DRG cells (Qiu et al., 2002). These peripheral
nerve conditioning lesions prior to lesions of DRG neurons improves cell survival
and axon growth. The initial sprouting of DRG neurites is not as robust in DRG
cells injected with the PKA inhibitor H89 in response to the conditioning lesion.

42
However, the axon growth measured one week later with H89 treated DRG cells is
unaffected compared to DRG cell unexposed to H89 (Qiu et al., 2002). Therefore,
late phase neuronal response to cAMP in axon growth appears to be independent of
PKA. Interestingly, cAMP increases axon growth from mature CNS neurons, but it
does not restore them to the full axon growth abilities of young embryonic neurons.
In vitro neurite growth assays of E20 and P8 rat RGCs in response to the trophic
factor BDNF and increasing levels of non-degradable cAMP analog
chlorophenylthio-cAMP still retains significant growth differences (Goldberg et al.,
2002). Though cAMP cannot fully recover axon growth abilities, increased cAMP
levels increases neurite outgrowth in both E20 and P8 RGCs, but E20 RGCs sends
out significantly longer axons (Goldberg et al., 2002). Together these data suggest
that cAMP does mediate increased axon growth, but it is not directly involved in the
signal transduction cascade for axon growth.

Growth Associated Protein 43
Growth associated protein 43 (GAP-43) has long been used as a marker for
neurite growth, but it may function primarily in axonal pathfinding. It is expressed
at very high levels during development and is predominately localized to the internal
surface of the growth cone membrane (Strittmatter et al., 1995). It is a presynaptic
phosphoprotein regulated by Ca
+2
and other peptides and a major substrate for
protein kinase C (PKC). Overexpression of GAP-43 promotes neurite sprouting in
non-neural cell and PC12 neuronal cell lines (Bomze et al., 2001). Mice generated

43
lacking GAP-43 die in the early postnatal period and exhibit grossly normal CNS,
but retinal axons remain trapped in the chiasm for 6 days before reaching their target
(Strittmatter et al., 1995). In vitro cultures of their DRG do extend axons. PC12 cell
lines lacking GAP-43 also sprout neurites (Baetge, 1991). Both results support
GAP-43’s role in modulating growth cone signaling. GAP-43 is most likely not a
gene that controls axonal growth, but instead increases secondarily during axon
growth.
The discovery that the first 1000 bases 5’ to the protein coding region directs
the expression of GAP-43 in a neuronal-specific manner (Starr et al., 1994) has
prompted many to characterize the promoter. The GAP-43 promoter contains seven
E-boxes, a core sequence of (CANNTG), to which bHLH transcription factors bind
(Chiaramello et al., 1996). The bHLH factors in NeuroD2 family Nex1/Math2 has
been demonstrated to promote GAP-43 activation (Uittenbogaard et al., 2003).
Though GAP-43 may be important in growing axons, its role in primarily
modulating growth cone signaling is rapidly emerging.

Transcriptional Dependence of Axon Growth
cAMP acts as an intracellular messenger molecule and is known to play a
role in modulating transcription. It is unclear which genes cAMP modulates in
RGCs that control axon growth or if cAMP is an effector. Addition of the
transcription inhibitor 5,6-dichloro-1-b-D-ribofuranosyl-benzimidazole (DRB) to
cultures of cerebellar neurons dramatically blocked cAMP and BDNF mediated axon

44
growth through inhibitory myelin (Cai et al., 2002). Axonal growth thus requires
transcriptional changes within the neuron. Inhibition of transcription inhibits axonal
growth. These experiments establish the transcriptional dependence for axonal
growth.

Amacrine cell dependent signal
The developmental loss of axonal growth capabilities to a minor degree can
be attributed to the down regulation of bcl-2 and cAMP. They may play ancillary
roles in modulating the main signaling cascade. The loss of growth was postulated
to involve a response to external signals from surrounding cells. In an effort to
identify the mediating signal, systematic exposure of RGCs to other purified retinal
cells was conducted (Goldberg et al., 2002). The authors show that direct contact
with bipolar, glial cells, not conditioned media from these cells induce axon growth
loss. Not even amacrine cell conditioned media affected RGC neurite growth. Only
amacrine cell membranes induce cultured embryonic RGCs to irreversibly lose their
axonal growth capabilities (Goldberg et al. 2002). Direct amacrine cell membrane
associated factors mediate permanent axon growth ability loss, yet the removal of
these amacrine cell membranes could not recover neonatal RGC growth properties.
It is not known if amacrine membrane contact triggers cAMP and bcl-2 levels to
decrease and thus loss of axonal growth abilities. Though evidence suggests
amacrine cell membrane may initiate the loss of axonal growth, controversy remains.
The loss of axonal growth abilities is developmentally dependent and occurs rapidly

45
during perinatal development. It is also quite clear that axon growth is transcription
dependent (Cai et al., 2002). The transcriptionally regulated molecular factors that
control axonal growth remain to be identified.

Gene Expression and the Retina
Gene expression studies attempting to identify different genes that play a role
in developing mature functional cells have been undertaken predominantly on the
whole retina (Blackshaw et al., 2001; Diaz et al., 2003; Swaroop and Zack, 2002;
Chowers et al., 2003). Newer studies are utilizing high density microarray analysis
instead of the historic method of mRNA subtraction libraries. These studies are
advantageous to screen for genes that may function in retinal development and
physiology when little is known about the molecular factors that are involved and to
identify novel ones. There remains a need to identify and to characterize
transcriptionally dependent molecular factors in the loss of axon growth. To this end,
we characterize the differential gene expression of developing RGCs during the
physiological switch of rapid neonatal axon growth to restricted postnatal growth.

Dissertation Summary
This dissertation examines the biochemical and functional properties of cone
arrestin (CAR) in phototransduction signaling cascades and describes the
transcriptional changes in developing RGCs. We hypothesis that CAR functions in
cone photoreceptors similarly to rod arrestin’s role in terminating rod

46
phototransduction, but that the CAR-opsin complex is less stable than the rod
arrestin-opsin complex. Using transgenic techniques, CAR is expressed in rod
photoreceptors without rod arrestin to allow the direct functional comparison of CAR
photoresponse termination in the well-characterized rod system. Light damage,
translocation, membrane association, and single cell electrophysiology assays show
that CAR functions similarly to rod arrestin in the termination of the rod
photoresponse in rod photoreceptors in vivo although CAR is not as effective in
signal termination as rod arrestin (Chapter II). These data provide direct evidence
that CAR can quench phototransduction signaling in vivo and the CAR-rhodopsin
complex is less stable than a rod arrestin-rhodopsin complex.
In conjunction with characterizing the unique biochemical and functional
properties of CAR, transcription profiles were analyzed to identify the genes that
may play a functional role in the developmental loss of axon growth in RGCs. We
hypothesize that transcription mediated molecular factors control the restriction in
axon growth during perinatal RGCs maturation. The critical period of transcriptional
changes occur when RGC neurite growth ability loss between E16 and E18. High-
density microarray analysis from directly isolated CD90
+
RGCs show 71 out of
26,000 genes as differentially expressed (ANOVA corrected p<0.05 with at least a
2^±1.5 fold difference) in perinatal RGCs from retinas of mice between E16 and P5
before axonal growth loss and immediately following. Semi-quantitative RT-PCR
and in situ hybridization confirms differential temporal and spatial transcript
expression. Semi-quantitative RT-PCR of purified cultured E16 and E18 RGCs

47
identifies similarly differentially transcribed genes (Chapter III). The similar
expression of gene transcripts from directly isolated and in vitro neurite growth
RGCs suggests these factors mediate the loss of axon growth. These data also
provide a database for understanding gene regulation in RGC development and
disease.

48
Chapter I: References
Aguayo AJ, Vidal-Sanz M, Villegas-Perez MP, and Bray GM. (1987) Growth and
connectivity of axotomized retinal neurons in adult rats with optic nerves
substituted by PNS grafts linking the eye and the midbrain. Ann N Y Acad
Sci. 495: 1-9.

Arshavsky V and Bownds MD. (1992) Regulation of deactivation of photoreceptor
G protein by its target enzyme and cGMP. Nature 357: 416-17.

Baetge EE. Hammang JP. (1991) Neurite outgrowth in PC12 cells deficient in GAP-
43. Neuron. 6(1):21-30.

Barlow JZ and Huntley GW. (2000) Developmentally regulated expression of Thy-1
in structures of the mouse sensory-motor system. J. Comp. Neurol. 421: 215-
33.

Barlow JZ, Kelley KA, Bozdagi O, Huntley GW. (2002) Testing the role of the cell-
surface molecule Thy-1 in regeneration and plasticity of connectivity in the
CNS. Neurosci. 111(4): 837-52.

Barres BA, Silverstein BE, Corey DP, Chun LL. (1988) Immunological,
morphological, and electrophysiological variation among retinal ganglion
cells purified by panning. Neuron 1: 791–803.

Blackshaw S, Fraioli RE, Furukawa T, Cepko CL. (2001) Comprehensive analysis of
photoreceptor gene expression and the identification of candidate retinal
disease genes. Cell. 107(5): 579-89

Bomze HM. Bulsara KR. Iskandar BJ. Caroni P. Skene JH. (2001) Spinal axon
regeneration evoked by replacing two growth cone proteins in adult neurons.
Nature Neuroscience 4(1):38-43.

Bregman BS, Goldberger ME. (1982) Anatomical plasticity and sparing of function
after spinal cord damage in neo natal cats. Science. 217(4559): 553-555.

Bregman BS, Kunkel-Bagden E, Reier PJ, Dai HN, McAtee M, Gao D. (1993)
Recovery of function after spinal cord injury: Mechanisms underlying
transplant-mediated recovery of function differ after spinal cord injury in
newborn and adult rats. Experimental Neurology. 123(1): 3-16.

Cai D, Deng K, Mellado W, Lee J, Ratan R and Filbin M. (2002) Arginase I and
Polyamines Act Downstream from Cyclic AMP in Overcoming Inhibition of
Axonal Growth MAG and Myelin In Vitro. Neuron 35 (4): 711-719.

49
Cai D, Qiu J, Cao Z, McAtee M, Bregman BS, and Filbin MT. (2001) Neuronal
cyclic AMP controls the developmental loss in ability of axons to regenerate.
J Neurosci. 21(13): 4731-9.

Cai D, Shen Y, De Bellard M, Tang S, and Filbin MT. (1999) Prior exposure to
neurotrophins blocks inhibition of axonal regeneration by MAG and myelin
via a cAMP-dependent mechanism. Neuron. 22(1): 89-101.

Calvert PD, Krasnoperova NV, Lyubarsky AL, Isayama T, Nicolo M, Kosaras B,
Wong G, Gannon KS, Margolskee RF, Sidman RL, et al. (2000)
Phototransduction in transgenic mice after targeted deletion of the rod
transducin alpha -subunit. Proc Natl Acad Sci USA 97: 13913-13918.

Caroni P and Schwab ME. (1988) Antibody against myelin-associated inhibitor of
neurite growth neutralizes nonpermissive substrate properties of CNS white
matter Neuron. 1(1):85-96.

Carter-Dawson LD and LaVail MM. (1979) Rods and cones in the mouse retina. J
Comp Neurol 188: 245-262.

Chen CK, Inglese J, Lefkowitz RJ, and Hurley JB. (1995) Ca
2+
dependent interaction
of recoverin with rhodopsin kinase. J Biol Chem 270: 18060-18066.

Chen CK, Burns ME, He W, Wensel TG, Baylor DA, Simon MI. (2000) Slowed
recovery of rod photoresponse in mice lacking the GTPase accelerating
protein RGS9-1. Nature 403: 557-60.

Chen CK, Burns ME, Spencer M, Niemi GA, Chen J, Hurley JB, Baylor DA, Simon
MI. (1999) Abnormal photoresponses and light-induced apoptosis in rods
lacking rhodopsin kinase. Proc Natl Acad Sci USA 96: 3718-22.

Chen DF, Jhaveri S, Schneider GE. (1995) Intrinsic changes in developing retinal
neurons result in regenerative failure of their axons. PNAS 92: 7287-7291.

Chen DF, Schneider GE, Martinou J, and Susumu T. (1997) Bcl-2 promotes
regeneration of severed axons in mammalian CNS. Nature 385(6615): 434-
439.

Chen J, Makino CL, Peachey NS, Baylor DA, Simon MI. (1995) Mechanisms of
rhodopsin inactivation in vivo as revealed by a COOH-terminal truncation
mutant. Science 267: 374-77.

50

Chen J, Simon MI, Matthes MT, Yasumura D, LaVail MM. (1999) 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.

Chen MS, Huber AB, van der Haar ME, Frank M, Schnell L, Spillmann AA, Christ
F, and Schwab ME. (2000) Nogo-A is a myelin-associated neurite outgrowth
inhibitor and an antigen for monoclonal antibody IN-1. Nature. 403(6768):
434-439.

Chowers I, Liu D, Farkas RH, Gunatilaka TL, Hackam AS, Bernstein SL,
Campochiaro PA, Parmigiani G, Zack DJ. (2003) Gene expression variation
in the adult human retina. Human Molecular Genetics. 12(22): 2881-93.

Chiaramello A. Neuman T. Peavy DR. Zuber MX. (1996) The GAP-43 gene is a
direct downstream target of the basic helix-loop-helix transcription factors.
Journal of Biological Chemistry 271(36):22035-43.

Craft CM and Whitmore DH. (1995) The arrestin superfamily:cone arrestins are a
fourth family. FEBS Lett. 362: 247-55.

Craft CM, Whitmore DH, and Wiechmann AF. (1994) Cone arrestin identified by
targeting expression of a functional family. J Biol Chem 269(26): 4613- 4619.

Diaz E, Yang YH, Ferreira T, Loh KC, Okazaki Y, Hayashizaki Y, Tessier-Lavigne
M, Speed TP, Ngai J. (2003) Analysis of gene expression in the developing
mouse retina. Proceedings of the National Academy of Sciences of the USA.
100(9): 5491-6.

Dowling JE. (1987) The retina –an approachable part of the brain. Harvard
University Press, Cambridge MA.

Drager UC and Olsen JF. (1980) Origins of crossed and uncrossed retinal projections
in pigmented and albino mice. J. Comp. Neurol. 191: 383-412.

Fain FL and Lisman JE. (1993) Photoreceptor degeneration in vitamin A deprivation
and retinitis pigmentosa: the equivalent light hypothesis. Exp Eye Res. 57:
335-40.

Fesenko EE, Kolesnikov SS, Lyubarsky AL. (1985) Induction by cyclic GMP of
cationic conductance in plasma membrane of retinal rod outer segment.
Nature 313: 310-13.

51

Filbin MT. (2003) Myelin-associated inhibitors of axonal regeneration in the adult
mammalian CNS. Nat Rev Neurosci. 4(9):703-13. Review.

Fournier AE, GrandPre T, and Strittmatter SM. (2001) Identification of a receptor
mediating Nogo-66 inhibition of axonal regeneration. Nature. 409(6818):
341-6.

Frade JM and Barde YA. (1999) Genetic evidence for cell death mediated by nerve
growth factor and the neurotrophin receptor p75 in the developing mouse
retina and spinal cord. Development. 126(4): 683-90
Gilman AG. (1987) G proteins: transducers of receptor-generated signals. Ann Rev
Biochem 56: 615-649.

Goldberg JL, Klassen MP, Hua Y, and Barres BA. (2002) Amacrine-signaled loss of
intrinsic axon growth ability by retinal ganglion cells. Science 296(5574):
1860-4.

Gorczyca WA, Gray-Keller MP, Detwiler PB, Palczewski K. (1994) Purification and
physiological evaluation of a guanylate cyclase activating protein from retinal
rods. Proc Natl Acad Sci USA 91: 4014-18.

Gorczyca WA, Polans AS, Surgucheva IG, Subbaraya I, Baehr W, Palczewski K.
(1995) Guanylyl cyclase activating protein. A calcium-sensitive regulator of
phototransduction. J Biol Chem 270: 22029-36.

GrandPre T, Nakamura F, Vartanian T, and Strittmatter SM, Stephen M. (2000)
Identification of the Nogo inhibitor of axon regeneration as a Reticulon
protein. Nature. 403(6768): 439-444.

Gurevich VV, and Benovic JL. (1992) Cell-free expression of visual arrestin.
Truncation mutagenesis identifies multiple domains involved in rhodopsin
interaction. J Biol Chem. 267(30): 21919-23.

Gurevich VV, and Benovic JL. (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 VV and Benovic JL. (1997) Mechanism of phosphorylation-recognition by
visual arrestin and the transition of arrestin into a high affinity binding state.
Mol Pharmacol. 51(1): 161-9.

Hamm HE and Gilchrist A. (1996) Heterotrimeric G proteins. Curr Opin Cell Biol 8:
189-96.

52
Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole-Cire P,
Kleyner Y, Taylor A, Simon MI, Grimm C, Reme CE, Lem J. (2002).
Evidence for two apoptotic pathways in light-induced retinal degeneration.
Nat Genet. 32(2): 254-60.

Hirsch JA, Schubert C, Gurevich VV, and Sigler PB. (1999) The 2.8 A crystal
structure of visual arrestin: a model for arrestin's regulation. Cell. 97(2): 257-
69.

Huang X, Wu DY, Chen G, Manji H, Chen DF. (2003) Support of retinal ganglion
cell survival and axon regeneration by lithium through a Bcl-2-dependent
mechanism. Invest Ophthalmol Vis Sci. 44(1): 347-54.

Huber AB, Weinmann O, Brosamle C, Oertle T, and Schwab ME. (2002) Patterns of
Nogo mRNA and protein expression in the developing and adult rat and after
CNS lesions. J Neurosci. 22(9): 3553-67.

Hsu YT and Molday RS. (1993) Modulation of the cGMP-gated channel of rod
photoreceptor cells by calmodulin. Nature 361: 76-9.

Imai H, Imamoto Y, Yoshizawa T, and Shichida Y. (1995) Difference in molecular
properties between chicken green and rhodopsin as related to the functional
difference between cone and rod photoreceptor cells. Biochemistry. 34(33):
10525-31.

Jeon CJ, Strettoi E, Masland RH. (1998) The major cell populations of the mouse
retina. J Neurosci 18(21): 8936-8946.

Kalil K and Reh T. (1982) A light and electron microscopic study of regrowing
pyramidal tract fibers. J Comp. Neurol. 211: 265-275.

Kühn H, Hall SW, and Wilden U. (1984) Light-induced binding of 48-kDa protein to
photoreceptor membranes is highly enhanced by phosphorylation of
rhodopsin. FEBS Lett. 176(2): 473-8.

Krupnick JG, Gurevich VV, and Benovic JL. (1997) Mechanism of quenching of
phototransduction. Binding competition between arrestin and transducin for
phosphorhodopsin. J Biol Chem. 272(29): 18125-31.

Liu BP, Fournier A, GrandPre T, and Strittmatter SM. (2002) Myelin-associated
glycoprotein as a functional ligand for the Nogo-66 receptor. Science.
297(5584): 1190-3.

Lisman J and Fain G. (1995) Support for the equivalent light hypothesis for RP.
Nat Med. 1(12): 1254-5.

53

Morris RJ. (1985) Thy-1 in developing nervous tissue. Dev. Neurosci. 7: 133-60

Mukhopadhyay G, Doherty P, Walsh FS, Crocker PR, and Filbin MT. (1994) A
novel role for myelin-associated glycoprotein as an inhibitor of axonal
regeneration. Neuron. 13(3): 757-67.

Murakami A, Yajima T, Sakuma H, McLaren MJ, Inana G. (1993) X-arrestin: a new
retinal arrestin mapping to the X chromosome. FEBS Lett. 334(2): 203-209

Palczewski K, McDowell JH, Jakes S, Ingebritsen TS, and Hargrave PA. (1989)
Regulation of rhodopsin dephosphorylation by arrestin. J Biol Chem.
264(27): 15770-3.

Palczewski K, Subbaraya I, Gorczyca WA, Helekar BS, Ruiz CC, Ohguro H, Huang
J, Zhao X, Crabb JW, Johnson RS, Walsh KA, Gray-Keller MP, Detwiler PB,
Baehr, W. (1994) Molecular cloning and characterization of retinal
photoreceptor guanylyl cyclase-activating protein. Neuron 13: 395-404.

Pierce KL, Premont RT, Lefkowitz RJ. (2002) Seven-transmembrane receptors.
Nature Reviews 3: 639-50. Review

Prinjha R, Moore SE, Vinson M, Blake S, Morrow R, Christie G, Michalovich D,
Simmons DL, and Walsh FS. (2000) Inhibitor of neurite outgrowth in
humans. Nature. 403(6768): 383-4.

Qiu J, Cai D, Dai H, McAtee M, Hoffman PN, Bregman BS, and Filbin MT. (2002)
Spinal axon regeneration induced by elevation of cyclic AMP. Neuron. 34(6):
895-903.

Richardson PM, McGuinness UM, and Aguayo AJ. (1980) Axons from CNS
neurons regenerate into PNS grafts. Nature. 284(5753): 264-5.

Rodieck RW. (1998) The first steps in seeing. Sinauer Associates Inc, Sunderland,
MA.

Schnell L and Schwab ME (1990) Axonal regeneration in the rat spinal cord
produced by an antibody against myelin-associated neurite growth inhibitors.
Nature. 343(6255): 269-72.

Shewan D, Berry M, Cohen J. (1995) Extensive regeneration in vitro by early
embryonic neurons on immature and adult CNS tissue. J. Neurosci. 15(3 Pt
1): 2057-62.

54

Shinohara T, Dietzschold B, Craft CM, Wistow G, Early JJ, Donoso LA, Horwitz J,
and Tao R. (1987) Primary and Secondary Structure of Bovine Retinal S
Antigen (48-kDa Protein) PNAS 84: 6975-6979.

Smith DS and Skene JH. (1997) A transcription-dependent switch controls
competence of adult neurons for distinct modes of axon growth. J Neurosci
17(2): 646-658.

Sokolov M, Lyubarsky AL, Strissel KJ, Savchenko AB, Govardovskii VI, Pugh Jr
EN, Arshavsky VY. (2002) Massive light-driven translocation of transducin
between the two major compartments of rod cells: a novel mechanism of
light adaptation. Neuron 34: 95-106.

Starr RG. Lu B. Federoff HJ. Functional characterization of the rat GAP-43
promoter. Brain Research 638(1-2):211-20, 1994.

Strittmatter SM. Fankhauser C. Huang PL. Mashimo H. Fishman MC. Neuronal
pathfinding is abnormal in mice lacking the neuronal growth cone protein
GAP-43. Cell 80(3):445-52, 1995.

Sun W, Li N, He S. (2002) Large-scale morphological survey of mouse retinal
ganglion cells. J Comp Neurol 451: 115-126.

Sutton RB, Vishnivetskiy SA, Robert J, Hanson SM, Raman D, Knox BE, Kono M,
Navarro J, and Gurevich VV. (2005) Crystal Structure of Cone Arrestin at
2.3A: Evolution of Receptor Specificity. J Mol Biol. 354(5): 1069-80.

Swaroop A and Zack DJ. (2002) Transcriptome analysis of the retina. Genome
Biology. 3(8): reviews 1022.

Uittenbogaard M. Martinka DL. Chiaramello A. The basic helix-loop-helix
differentiation factor Nex1/MATH-2 functions as a key activator of the GAP-
43 gene. Journal of Neurochemistry 84(4):678-88, 2003.

Vidal-Sanz M, Bray GM, Villegas-Perez MP, Thanos S, and Aguayo AJ. (1987)
Axonal regeneration and synapse formation in the superior colliculus by
retinal ganglion cells in the adult rat. J Neurosci. 7(9): 2894-909.

Vishnivetskiy SA, Paz CL, Schubert C, Hirsch JA, Sigler PB, and Gurevich VV.
(1999) How does arrestin respond to the phosphorylated state of rhodopsin? J
Biol Chem. 274(17):11451-4.

55
Vishnivetskiy SA, Hosey MM, Benovic JL, and Gurevich VV. (2004) Mapping the
arrestin-receptor interface. Structural elements responsible for receptor
specificity of arrestin proteins. J Biol Chem. 279(2): 1262-8.

Wacker WB, Donoso LA, Kalsow CM, Yankeelov JA Jr, and Organisciak DT.
(1977) Experimental allergic uveitis. Isolation, characterization, and
localization of a soluble uveitopathogenic antigen from bovine retina. J
Immunol. 119(6): 1949-58.

Wald G. (1968) The molecular basis of visual excitation. Nature 219: 800-807.

Wang KC, Koprivica V, Kim JA, Sivasankaran R, Guo Y, Neve RL, and He Z.
(2002) Oligodendrocyte-myelin glycoprotein is a Nogo receptor ligand that
inhibits neurite outgrowth. Nature. 417(6892): 941-4.

Wikler KW and Rakic P. (1990) Distribution of photoreceptor subtypes in the retina
of diurnal and nocturnal primates. J Neurosci 10: 3390-3401.

Wilden U and Kühn H. (1982) Light-dependent phosphorylation of rhodopsin:
number of phosphorylation sites. Biochemistry 21: 3014-22.

Wilden U, Hall SW, Kühn H. (1986) Phosphodiesterase activation by photoexcited
rhodopsin is quenched when rhodopsin is phosphorylated and binds the
intrinsic 48-kDa protein of rod outer segments. Proc Natl Acad Sci USA.
83(5): 1174-8.

Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA, Chen J. (1997) Prolonged
photoresponses in transgenic mouse rods lacking arrestin. Nature. 389: 505-
509.

Xue GP, Calvert RA, Morris RJ. (1990) Expression of the neuronal cell surface
glycoprotein Thy-1 is under post-transcriptional control and is spatially
regulated in the developing olfactory system. Development 109: 851-864.

56
Chapter II: Ectopic Cone Arrestin Rescues Photoreceptors from Light Damage
in a Dose Dependant Manner

Chapter II: Abstract
Purpose. To characterize the functional role of cone arrestin and rod arrestin in
terminating light activated rhodopsin signaling in transgenic mice.
Method. Transgenic mice expressing cone arrestin in rod photoreceptors lacking rod
arrestin were generated. Cone arrestin’s ability to protect photoreceptors from light
damage was assessed. Light dependent translocation and membrane association
characterized cone arrestin function. Suction electrode recordings from single rods
were performed to investigate in real time phototransduction termination.
Results. Two independent lines of cone arrestin expressing transgenic mice were
obtained, and each was bred into an arrestin null background. Mouse cone arrestin
protects photoreceptors in a dose dependant manner, but was not as effective as rod
arrestin. Cone arrestin translocates properly and associates to photoreceptor
membranes. Consistent with these data, flash responses show incomplete
phototransduction signal termination by cone arrestin.
Conclusions. Cone arrestin can function similarly to rod arrestin in terminating
rhodopsin signaling and partially protect photoreceptors from light-induced damage,
though it is not as effective as rod arrestin. Although rod arrestin can fully terminate
cone opsin signaling, cone arrestin cannot fully substitute for rod arrestin in vivo and
the cone arrestin-rhodopsin complex is less stable than a rod arrestin-rhodopsin

57
complex. The results suggest that in vivo high affinity specific binding requirements
are distinct for cone versus rod arrestin.

Chapter II: Introduction
The role of the arrestin superfamily of proteins is largely to bind to their
respective G-protein-coupled receptor target and mediate the termination of receptor
signaling. The arrestin superfamily of proteins includes: rod arrestin (S-antigen,
SAG, also referred to as visual arrestin) which interacts with the visual pigment
rhodopsin, β-arrestins (BAR1 and BAR2) which interact with β-adrenergic receptors,
and cone arrestin (CAR). Expression of CAR was initially identified through a
human retinal cDNA library screen using PCR based on the last known conserved
region in homologous sequences of other arrestins with the known vector sequence
(Craft et al., 1994). It was subsequently mapped independently by two groups to the
X-chromosome (Craft et al., 1994; Murakami et al.,1993).
The expression of cone arrestin (CAR) is developmentally regulated both
spatially and temporally. Studies of mouse CAR (mCAR) protein through
immunoblot analysis in sixteen tissue types (adrenal, cerebellum, cortex, heart,
intestine, kidney, liver, lung, olfactory bulb, ovary, pineal, retina, spleen, testis,
thymus, and turbinates) identified mCAR only in the pineal and retina (Zhu et al.,
2002). Further immunohistochemical analysis demonstrated that expression was
restricted to the cone cells of the neural retina (Zhu et al., 2002). During
immunoblot analysis of the developing retina, the full length 381 amino acid long, 44

58
kDa mCAR protein becomes detectable at embryonic day 13.5 and increases with
age until adulthood while the homologous 403 amino acid long, better studied rod
arrestin protein becomes detectable at P5 with peak expression by P9 in wild-type
C57/Bl/6J mice (Zhu et al., 2002). These homologous proteins are expressed at
similar times during development.
The functional role of rod arrestin in terminating the photoresponse has been
well characterized. During the photoresponse in rod cells, a photon of light causes
photoisomerization of a chromophore 11-cis retinal, converting it to all-trans retinal
within the core of the rhodopsin molecule (R), leading to a conformational change
into an activated state (R*), and the subsequent activation of the phototransduction
cascade via the G-protein transducin (Rodieck, 1998). The termination of the
photoresponse involves G-protein-coupled receptor kinase 1 (GRK1)
phosphorylation of R* followed by rod arrestin binding (Wilden et al., 1986).
A mechanism that may be involved in modulating the rod function in
changing light conditions is the light-driven redistribution of phototransduction
proteins into different compartments of the cell (Sokolov et al. 2002). Transducin
and arrestin redistribute in a light dependent manner. Light dependent movement of
transducin was confirmed by combining serial tangential cryosectioning of the retina
with immunoblot analysis (Sokolov et al. 2002). Soluble rod arrestin translocates in
a light dependant manner from the rod inner segment, outer nuclear layer, and outer
plexiform layer to the membranous outer segment, presumably to be physiologically
available to bind to phosphorylated R* (R*-P) and continue the termination of the

59
phototransduction cascade (Broekhuyse et al., 1985). CAR immunostaining in cone
cells also redistributes from a diffuse staining through the entire cell into the cone
outer segment in a light dependent manner (Zhu et al., 2002). The exact mechanism
of light dependent translocation is not known, although R*-P is not critical for
arrestin movement, R* is necessary (Mendez et al., 2003). There is still on going
investigation into diffusion as well as energy dependent mediated light dependent
rod translocation.
In electrophysiology studies, at least one copy of the rod arrestin gene (~30%
of normal protein levels) is required to preserve the normal kinetics of photoresponse
termination in rod photoreceptors (Xu et al., 1997). In homozygous arrestin
knockout mice, light damage experiments result in loss of much of the photoreceptor
outer nuclei layer (Roca et al., 2004; Chen et al., 1999). The thickness of the
photoreceptor outer nuclear layer has classically been used as a quantifiable marker
of photoreceptor cell survival and health (Michon et al., 1991). Without rod arrestin,
the defect in rhodopsin shutoff and prolonged photoresponse signaling leads to
photoreceptor cell death (Fain and Lisman, 1993; Chen et al., 1999). The inability of
rod arrestin knockout retinas to terminate constitutive light activation of the signaling
cascade leads to photoreceptor cell death via a transducin-dependent mechanism
(Hao et al., 2002).
Although the mCAR protein is 70% homologous to arrestin, the functional
role of CAR in protecting photoreceptors from light damage and in termination of
the photoresponse in cone cells remains unclear. The scarcity of cone cells has made

60
studying them difficult since they represent only 3% of the photoreceptors in the
murine retina, while rod cells represent nearly 97% (Dawson and LaVail, 1979, and
Jeon et al. 1998). Thus, elucidating the role of cone specific proteins and their
functional role in the developed mature cell has been difficult due to the challenge of
identifying and isolating enough cone cells. Homologous proteins to those in the rod
phototransduction cascade have been identified in cones, the exact nature of their
interactions is unclear and therefore, the mechanism of cone phototransduction is not
well studied in vertebrates. Rod arrestin, as well as other rod phototransduction
proteins, has homologous proteins in cone photoreceptors. The structural similarity
of rod arrestin and cone arrestin, as well as the presence of other highly related
proteins, suggests that a similar phototransduction cascade exist and each component
functions similarly in both rods and cones.
Like rod arrestin, recent biochemical evidence in the rod degenerative neural
retina leucine zipper knockout mouse model demonstrated that CAR binds in a light-
dependant manner to S and M opsins (Zhu et al., 2004). Competitive binding assays
of salamander CAR and salamander rod arrestin provide evidence that CAR binds
phosphorylated activated frog rhodopsin (Smith et al., 2000). These data suggest
CAR functions similar to rod arrestin in phototransduction signal termination, but
direct evidence of in vivo binding is lacking.
CAR may function in cone phototransduction as rod arrestin does in rods
because of structural homology, cell specificity, and temporal nature of its
expression (Craft and Whitmore, 1995). Recent structural analysis of cone arrestin

61
binding found that it forms weaker arrestin-opsin complexes than rod arrestin-
rhodopsin (Sutton et al., 2005). We hypothesize that CAR plays a role in the
termination of the cone visual phototransduction cascade akin to rod arrestin’s role in
the termination of rod visual phototransduction and that mCAR forms less stable
complexes than rod arrestin. Our goal is to examine the differences in the
biochemical and functional characteristics between rod arrestin and CAR in
photoreceptors of the mouse retina in vivo.
We generated two lines of mouse cone arrestin (mCAR) expressing
transgenic mice in the arrestin knockout background to compare directly the
functional roles of rod arrestin and CAR. We show that ectopic mCAR protects
photoreceptors in a CAR dose dependant manner and defined the threshold of light
intensity for nearly full protection of photoreceptors in mCAR-high transgenic mice.
Proper light dependant translocation of mCAR in these transgenic lines occurred, yet
mCAR did not associate with the photoreceptor membranes to the extent of rod
arrestin. Single cell recordings of single photon responses from mCAR transgenic
photoreceptors showed incomplete termination of the phototransduction cascade.
The inability of mCAR to completely terminate the rod photoresponse may be the
mechanism for dose dependent protection of photoreceptors from light damage. We
provide biochemical and functional evidence that further supports CAR’s
physiological role in the termination of activated opsins.

62
Material and Methods
All procedures in this study concerning animal usage adhered to both the
NIH Health Guide for the Care and Use of Laboratory Animals and both USC
IACUC and UC Davis guidelines.

Transgenic Mice Generation
The well-characterized rod photoreceptor specific 4.4 kb 5’ region of the
mouse rhodopsin promoter (mRhoP) (Lem et al., 1991) was ligated to a 1.3 kb
mCAR cDNA was obtained through reverse transcription coupled to PCR
amplification (RT-PCR) from total retinal RNA using primers designed based on
sequences from Dr. Cheryl Craft. The 0.6 kb mouse protamine 1 polyadenylation
signal was placed 3’ of the cone arrestin cDNA. Linear fragments of the construct
were purified and injected into fertilized single cell embryos. Injected embryos were
transplanted into pseudo-pregnant females and founder pups were analyzed for
construct integration by PCR and Southern blot analysis. PCR analysis of genomic
DNA from the tail biopsy of offspring with the mouse rhodopsin-specific primer
(Rh1.1: GTGCCTGGAGTTGCGCTGTGGG) and the mouse cone arrestin-specific
primer (mCAR: GTCTTGGAAACGGTGGTAGAGGCC) at the following
conditions: 95°C for 3.5 min; and 30 cycles of 94°C for 1 min; 63°C for 1.5 min; and
72°C for 1.5 min were used to identify transgenic positive animals. PCR products
were separated on a 1.5% agarose gel with ethidium bromide and images
photographed by Gel Doc (Bio-RAD). PCR positive founder mice were mated to

63
arrestin knockout mice (arrestin-/-, Lem et al., 1999) to generate mCAR-high and
mCAR low expressing mice.

Retina Immunohistochemistry
Mice were sacrificed by cervical dislocated following CO
2
induced narcosis,
and the eyes were enucleated. The cornea was punctured in fixative buffer
composed of 4% paraformaldehyde (Polysciences Inc, Warrington, PA), 0.5%
glutaraldehyde (Ted Pella Inc, Redding, CA), in 0.1 M cacodylate buffer (Electron
Microscopy Sciences, Forth Worthington, PA) and incubated for 5 min. The cornea
and lens were removed and the eyecup was transferred into new fixative and
incubated for 1.5 hours at room temperature. The eyecups were rinsed three times
with 0.1 M cacodylate buffer, hemisected, embedded, fast frozen in OCT (VWR),
and sectioned with a cryostat (Leica) to 7 µm. To block non-specific antibody
binding, slides were incubated in PBS with 1 mM CaCl
2
, 1 mM MgCl
2
, and 1 mg/ml
BSA (PBS/BSA) for 15 min at room temperature then rinsed with PBS containing 1
mM CaCl
2
and 1 mM MgCl
2
. The slides were incubated with anti-mCAR LUMIJ
polyclonal antibody (1:500 dilution, Zhu et al., 2002) diluted with PBS/BSA in a
humidified chamber for 1 hour and then rinsed three times with PBS for 5 min each.
The sections were incubated with secondary goat anti-rabbit conjugated to FITC
antibody (1:1000 dilution, Vector Laboratory) in PBS/BSA for 30 min, washed three
times in PBS for 5 min each, fixed with 4% paraformaldehyde in PBS for 5 min

64
before mounting with Vectashield (Vector Laboratory). Micrographs were taken
with the Zeiss LSM 510 confocal microscopy.

Quantification of mCAR Over-Expression
Protein samples were prepared by homogenizing retinal tissues in buffer (50
mM Tris, pH 7.2). Total protein concentrations were determined by the Bradford
assay (Bradford, 1976). Immunoblotting analysis was done as described (You et al.,
1999) with minor modifications. Total protein as indicated was denatured and
separated by 12.5% polyacryalmide gel electrophoresis in the presence of sodium
dodecyl sulfate (SDS-PAGE). Proteins were transferred to Hybond-ECL Western
PVDF membranes (Amersham Life Science, Buckinghamshire, UK) at 4°C, and the
transfers were confirmed by Ponceau Red staining. After blocking in 5% nonfat dry
milk, the blots were incubated with mCAR specific primary antibody (1:5000
dilution) followed by the incubation of the secondary antibody (anti-rabbit IgG
coupled to horseradish peroxidase 1:10,000 dilution; Vector Laboratory) and
visualized on film exposed to enhanced chemiluminescence (Hyperfilm-ECL,
Amersham). The immunoblot bands were quantified through analyzing the optical
densities of scanned bands using the Quantity One software (Bio-Rad). Two lines
expressing cone arrestin at low (mCAR-L) or high (mCAR-H) levels were used in
the study.

65
Light Exposure and Morphological Analysis
Dark reared 4 week old C57/BL/6J wild-type control mice (Jackson Labs),
arrestin knockout, mCAR-low, and mCAR-high mice were exposed to white
fluorescent light at ~3000, ~2000, or ~1000 lux for 72 hours while control mice were
kept in the dark. Animals were sacrificed by cervical dislocation following CO
2

induced narcosis, the superior apex of the eye marked, and enucleated. Eyecups
were dissected leaving a flap of cornea marking the superior apex of the eye and
fixed overnight in 2.5% glutaraldehyde, 3.5% paraformaldehyde, and 0.1 M
cacodylate buffer, pH 7.2. They were rinsed 3 times in 0.1 M cacodylate buffer and
embedded in epon resin. The tissue was sectioned at 1 µm along the vertical
meridian from the apex of the superior pole through the optic nerve and stained with
staining solution (1% azure blue mixed with 1% methyl-blue and 1% borax).
ONL width was measured from a single section from each mouse. The total
length was measure from the optic nerve head (“0” was assigned) to the ends of each
hemispheres (negative for the superior and positive for the inferior hemisphere). The
distance for each hemisphere was divided into 11 segments. Three measurements
spaced 25 µm apart were taken at each field and the mean calculated for each
segment for comparisons and plotted. An unpaired independent 2-tail t-test with
α=0.05 was used to test the level of significance for each treatment point.
Calculations were based on the ONL thickness measurements between -2 and -5 and
2 and 5 to quantify the level of photoreceptor rescue as described by LaVail (LaVail
et al., 2000) with a p<0.001 as significant. The level of significance for differential

66
ONL thickness in the superior retina compared to the inferior retina was calculated
using the measurements between -2 and -5 compared to between 2 and 5. An
unpaired independent 2-tail t-test with α=0.05 was used to test for significance with a
p<0.01 as significant.

Translocation of mCar and Arrestin
To examine in vivo mCAR translocation, dark reared mice exposed to 3000
lux of light for 30 min or kept in the dark were scarified following CO
2
induced
narcosis and retinas processed for retinal immunohistochemistry using the same
methods as described above.

Soluble and Membrane-Bound mCAR
Retinas from each wild-type C57/BL/6J or mCAR-high 4 week old mice
reared in the dark were dissected in the dark under infrared conditions. One retina
from a mouse was kept in the dark while the other retina was exposed to 10 min of
light. Each were homogenized in 100 µl of 50 mM Tris, pH 7.4. Samples were
centrifuged at 15000 g at 4
o
C for 10 min and the supernatant removed (90 µl)
immediately and an additional 10 µl buffer added. The pellet was suspended into a
total of 100 µl of buffer. Samples were all treated with SDS sample buffer and
denatured. An equal volume of protein was resolved by SDS-PAGE, blotted as
described above, and quantified for expression. An unpaired independent 2-tail t-test
with α=0.05 was used to test for significance with a p<0.05 as significant. The anti-

67
arrestin antibody C10C10-polyclonal (1:10,000 dilution) was used to detect rod
arrestin in the C57/BL/6J mice.

Single Cell Electrophysiology
Retinas from wild-type, arrestin knockout, and mCAR-high mice dark reared
4-8 weeks were isolated in the dark under infrared conditions in Leibovitz’s L-15
medium (Gibco) with 10 mM glucose and 0.1 mg/ml bovine serum albumin (Sigma)
and stored on ice for electrophysiology as previously described (Krispel et al., 2003).
The retinas were cut into small pieces with a razor blade and loaded into a recording
chamber that was perfused with 37
o
C buffer containing 112.5 mM NaCl, 3.6 mM
KCl, 2.4 mM MgCl
2
, 1.2 mM CaCl
2
, 10 mM HEPES, 20 mM NaHCO
3
, 10 mM
glucose, and 3 mM Na-succinate, pH 7.4 and 290 milliosmolar. A suction electrode
with tips 1-2 µm in diameter gently drew in individual rods and measured the inward
current from a single rod outer segment with a current-to-voltage converter
(Axopatch 1B, Axon Instruments, Inc.). Rod cells were stimulated with 10-ms
flashes or increasing intensities of light at 500 nm. The intensity of light was
measured after each experiment using a silicon photodiode (United Detector
Technology) and controlled using calibrated neutral density filters.

Rhodopsin bleaching
One retina from each wild-type C57/BL/6J or mCAR-high 4 week old mice
reared in the dark were dissect out in the dark under infrared conditions. Retinas

68
were placed in solubilization buffer (10 mM Hepes, pH=7.5, 2 mM MgCl
2
, 2 mM
CaCl
2
, 1% DM, 150 mM NaCl) and incubated at room temperature for 3 hours in the
dark. The absorbance of the dark solubilized retina was taken at 280 nm for total
protein and 500 nm for rhodopsin absorption. Samples were exposed to light for 1.5
min intervals between absorbance readings until absorbance at 500 nm stopped
declining. The change in absorbance between the maximum level of rhodopsin
bleaching and the absorbance of the dark sample was calculated.

Results:
Transgene Expression in Photoreceptors
Transgenic mice were generated to directly compare the functional
characteristics of CAR and rod arrestin in terminating phototransduction signaling.
Two lines of independently generated transgenic mCAR expressing mice were bred
into a homozygous rod arrestin knockout background (Lem et al., 1999) to yield
mCAR-high and mCAR-low lines. The mCAR-high line was created by Dr. Xiao
Liu and a second lower expressing line, mCAR-low, was newly created. Two lines
are needed to compare the dose response effects of CAR. Characterization of protein
expression was undertaken due to the random nature of transgene integration that
affects the level of expression. Examination of serial dilutions of whole retinal
homogenates by western blot revealed a line nearly 120 fold expression over
endogenous mCAR (mCAR-low) expression in wild-type C57/BL/6J mice and the
other line over-expressing mCAR at nearly 1500 fold (mCAR-high) (Figure 2.1b, c).

69
These values represent 1:160 and 1:12.5 molar ratio with rhodopsin. The
physiological ratio of rod arrestin to rhodopsin is 1:1.3. Retinal sections confirmed
that mCAR expression was localized

70

Figure 2.1. Transgene expression of mCAR in the retina.
(a) Transgene construct. The 4.4 kb 5’ region of the mouse rhodopsin promoter (mRhoP) was
ligated to the 1.3 kb mouse cone arrestin cDNA (Accession: NM_133205). The 0.6 kb
mouse protamine 1 polyadenylation signal was placed 3’ to the cone arrestin cDNA.
(b) Western blot analysis of over-expressing mCAR in the retinas of transgenic lines. Serial
dilutions of total amounts of retinal proteins as indicated were resolved by SDS-PAGE. The
expression of mCAR from mCAR-high and mCAR-low lines was compared to the
endogenous level in 50 µg of wild-type C57/BL/6J mice. Two lines of transgenic mCAR
mice were generated to examine the dose dependant effects of mCAR over-expression.
(c) Densitometric analysis of western blots. Immunoblot bands were quantified through
analyzing the optical densities of scanned bands closest to intensity of the wild-type band.
Transgene expression relative to endogenous wild-type C57/BL/6J mice is: ~120 fold
greater in mCAR-low and ~1500 fold greater in mCAR-high lines.
(d) Localization of mCAR expression in the retina. Representative confocal images of cells
stained for the presence of mCAR in green, DAPI labeled nuclei are in blue. The sparse
cone photoreceptor cells in wild-type C57BL/6J retinas are the only cells labeled. The
soluble mCAR is present in the entire length of the photoreceptor layer from the inner
segment in the outer plexiform layer of the retina to the outer segment. mCAR is present
throughout the photoreceptor layer in mCAR-high and mCAR-low. The expression of
mCAR is qualitatively lower in mCAR-low than mCAR-high. The ganglion cell layer (GC),
inner plexiform layer (IPL), inner cellular layer (ICL), outer nuclear layer (ONL), and outer
segment (OS) are as labeled.

71

Figure 2.1. Continued
(a)

(b)

(c)

(d)

GC
IPL
ICL
ONL
OS
mCAR-high mCAR-low Wild-type
20 um
GC
IPL
ICL
ONL
OS
mCAR-high mCAR-low Wild-type
20 um

Relative expression of mCAR to
transgene negative wild-type C57/BL/6J
1
120.6
1474.7
0
200
400
600
800
1000
1200
1400
1600
C57/BL/6 mCAR-low mCAR-high
Relative fold difference

C57/BL/6J
50ug
0.5ug 0.2ug
0.1ug 0.05ug 0.2ug 0.1ug 0.05ug 0.01ug
mCar-Low
mCar-High
C57/BL/6J
50ug
0.5ug 0.2ug
0.1ug 0.05ug 0.2ug 0.1ug 0.05ug 0.01ug
mCar-Low
mCar-High

mRho Promoter
4.4 kb
mCAR
1.3 kb
MP1
Poly A
tail
0.6 kb
mRho Promoter
4.4 kb
mCAR
1.3 kb
MP1
Poly A
tail
0.6 kb

72
appropriately to the photoreceptor layer only and provided qualitative visual
confirmation of levels of mCAR over-expression (Figure 2.2d).

mCAR Protects Photoreceptors from Light Damage in a Dose Dependant Manner
Rod arrestin has been well established as a critical component in protecting
photoreceptors from light damage (Roca et al., 2004; Chen et al., 1999) by
terminating the photoresponse (Xu et al., 1997). By ectopically expressing mCAR in
rod arrestin knockout mice (arrestin -/-), we can directly compare the functional
ability of mCAR to terminate the photoresponse and thereby protect photoreceptors
from light damage thought to be caused by constitutive signaling.
One month old dark reared mice were exposed to ~3000 lux of white
fluorescent light for 72 hours and the photoreceptor outer nuclear layer (ONL)
thickness was measured across the retina to determine CAR’s ability to functionally
substitute for rod arrestin (Figure 2.2). In representative micrographs (Figure 2.2a),
we did not observe differences between the ONL thickness from wild-type retinas
exposed to light nor in any mice kept in the dark. Arrestin1 -/- mice had the most
amount of light induced photoreceptor cell death. Increased ectopic mCAR
expression increasingly protected photoreceptors from light damage. A plot of the
means with standard deviation of the various mice illustrates the ONL thickness
across the retina (Figure 2.2b). As expected light exposed, wild-type mice with
intact endogenous rod arrestin did not display a significant difference to their
littermate controls that were kept in the dark (p>0.001), while light exposed arrestin1

73
knockout mice did show significant levels of photoreceptor cell loss compared to
their littermates that were kept in the dark (p<0.001).
The transgenic mice that ectopically express mCAR showed a dose
dependant protection of photoreceptors lacking rod arrestin across the length of the
retina. The mCAR-high mice had increase photoreceptor survival over mCAR-low
mice, which in turn also protected photoreceptors from light damage compared to
arrestin1 knockout mice (p<0.001 for both). Only mCAR-high and mCAR-low mice
demonstrated a significant (p<0.01) differential effect on photoreceptor cell survival
when comparing the superior and inferior ONL thickness at this intensity level. In
both lines, the inferior retina was more protected, but there was no statistically
significant differential protection in arrestin1 -/- mice. Ectopic expression of mCAR
in arrestin -/- mice did not fully rescue photoreceptors from light damage at 3000 lux.
Photoreceptor were progressively protected from light damage in mice expressing
higher levels of mCAR.

Threshold determination for photoreceptor rescue
Because of the dose dependent relationship with photoreceptor protection,
decreasing incident light should ameliorate the light damage in mCAR transgenic
mice lines. One R*-P binds to one rod arrestin molecule for proper rod
phototransduction termination. It is unclear if the CAR levels are sufficient to
quench every R*-P generated. At lower light levels, fewer photons are incident on a
photoreceptor cell as a result, fewer R* would be generated. Reaching a threshold of

74
Figure 2.2. Dose dependent protection of photoreceptors from light damage.
(a) Representative micrographs of control and light exposed mice. Micrographs are from 1 µm
sections of Epon embedded retinas sectioned along the vertical meridian from the superior
apex through the optic nerve of dark reared 4 week old mice exposed to ~3000 lux for 72
hours (Light as indicated) or control mice kept in the dark (Dark as indicated) at position “-
1” in (b). The retinal pigment epithelial layer (RPE), outer segment (OS), and the outer
nuclear layer (ONL) are labeled.
(b) ONL thickness of control and light exposed mice. ONL thickness at 22 fields (mean and
standard error of the mean, SEM) is as indicated for wild-type C57/BL/6J Dark, N=4;
C57/BL/6J Light, N=5; arrestin knockout (arrestin -/-) Dark, N=5; arrestin -/- Light, N=4;
mCAR-low Dark, N=5, mCAR-low Light, N=4; mCAR-high Dark, N=4; mCAR-high Light,
N=5 with “0” assigned to the optic nerve, negative numbers decline towards the superior
pole, and positive numbers advance toward the inferior pole. The ONL thickness
measurements between -2 and -5 and 2 to 5 were used to quantify levels of photoreceptor
rescue. No mice reared and kept in the dark displayed statistically significant changes in
ONL thickness nor did wild-type C57/BL/6J mice exposed to light. ONLs were thinner in
arrestin -/- (p<0.001), mCAR-low (p<0.001), and mCAR-high (p<0.001) exposed to ~3000
lux of florescence light compared to those kept in the dark. All mice exposed to light
(C57/BL/6J Light, p<0.001; mCAR-low Light, p<0.001; mCAR-high Light p<0.001)
exhibited differences in ONL thickness compared to arrestin-/- mice exposed to light.
Measurements of the ONL thickness of the superior retina between -2 and -5 were used to
compare the ONL thickness of the inferior retina between 2 and 5. Only mCAR-low
(p<0.01) and mCAR-high (p<0.01) mice exposed to light demonstrated a differential
superior to inferior ONL thickness. Even at high levels of expression, ectopic mCAR
expression does not fully rescue photoreceptors from light damage in the absence of rod
arrestin due to exposure to ~3000 lux of diffuse white light for 72 hours.

75

Figure 2.2 Continued.
(c) A representative micrograph of an eyecup sectioned through the vertical meridian used to
measure the ONL. The flap of cornea on the upper left marks the apex of the superior pole
at the “-11” position, the optic nerve head is visible in the middle of the picture at the “0”
position, and the upper right is the base of the inferior retina at position “11” as indicated in
(b).

76
Figure 2.2
(a)

(b)

OS
20 µm
C57/BL/6J Light Arrestin
-/-
Light
C57/BL/6J Dark Arrestin
-/-
Dark
mCAR-low Light mCAR-high Light
mCAR-low Dark mCAR-high Dark
RPE
ONL
OS
20 µm
C57/BL/6J Light Arrestin
-/-
Light
C57/BL/6J Dark Arrestin
-/-
Dark
mCAR-low Light mCAR-high Light
mCAR-low Dark mCAR-high Dark
RPE
ONL
Light Damage 3,100 lux for 72 h
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
C57/BL/6J Dark
mCAR-high Dark
mCAR-low Dark
Arrestin -/- Dark
C57/BL/6J Light
mCAR-high Light
mCAR-low Light
Arrestin -/- Light

77
Figure 2.2 (c)

Corneal flap
marking the
superior
retina
-11
Superior
0
Optic nerve head
11
Inferior
Corneal flap
marking the
superior
retina
-11
Superior
0
Optic nerve head
11
Inferior

78
light generating sufficiently low enough amounts of R*, and therefore R*-P, for
CAR to completely bind to and terminate the photoresponse should result in full
rescue of photoreceptors.
To determine the threshold of light intensity at which mCAR-high can
functionally replace rod arrestin in protecting photoreceptors from light damage, one
month old dark reared mice were exposed to progressively lower intensity of white
fluorescent light at intervals of 1000 lux. The ONL thickness was measured across
the retina for wild-type, mCAR-high, and arrestin -/- mice exposed to ~2000 lux of
light for 72 hrs (Figure 2.3a) and for wild-type, mCAR-low, mCAR-high, and
arrestin -/- mice exposed to ~1000 lux of light (Figure 2.3b). The ONL of all mice
kept in dark remained similar to wild-type C57/BL/6J mice in the dark or light
exposed. The ONL thickness in mCAR-high and arrestin -/- light exposed to 2000
lux was not fully maintained compared to their dark controls (p<0.001) and both
demonstrated a superior regional susceptibility to light damage (p<0.01).
Photoreceptor protection with high levels of mCAR expression did not fully protect
cells at ~2000 lux.
At 1000 lux of light, there was still a statistically significant difference
between the ONL thickness of mCAR-high, mCAR-low, and arrestin -/- compared to
their dark counterparts (p<0.001). Except for the central superior retina (-7 to 0 in
Figure 2.3b), the ONL thickness was rescued in mCAR-high mice exposed to 1000
lux. The superior central region was still more susceptible to light damage than the
inferior retina in transgenic mice and arrestin -/- exposed to the light (p<0.01).

79
Figure 2.3. Progressive increase in protection against light damage at decreasing light intensity. The
threshold for full rescue of photoreceptors by mCAR was identified by decreasing light intensity
levels at intervals of 1000 lux.
(a) ONL thickness of dark reared mice exposed to ~2000 lux for 72 hours (Light) or control
mice kept in the dark (Dark). ONL thickness at 22 fields (mean and SEM) is as indicated for
C57/BL/6J Dark, N=7; C57/BL/6J Light, N=8; arrestin -/- Dark, N=8; arrestin -/- Light,
N=3; mCAR-high Dark, N=8; mCAR-high Light, N=4. No mice reared and kept in the dark
displayed statistically significant changes in ONL thickness nor did wild-type C57/BL/6J
mice exposed to light. ONL thickness were decreased in arrestin -/- (p<0.001) and mCAR-
high (p<0.001) exposed to light compared to those kept in the dark. All mice exposed to
light (C57/BL/6J Light, p<0.001 and mCAR-high Light, p<0.001) exhibited differences in
ONL thickness compared to arrestin -/- mice exposed to light. Measurements of the ONL
thickness of the superior retina between -2 and -5 were used to compare the ONL thickness
of the inferior retina between 2 and 5. Only mCAR-high (p<0.01) and arrestin -/- (p<0.01)
mice exposed to light demonstrated a differential superior to inferior ONL thickness. Even
at high expression levels, ectopic mCAR does not fully rescue photoreceptors from light
damage in the arrestin knockout background from exposure to ~2000 lux of diffuse white
light for 72 hours.
(b) ONL thickness of dark reared 4 weeks old mice exposed to ~1000 lux for 72 hours (Light) or
control mice kept in the dark (Dark). ONL thickness at 22 positions (mean and SEM) are as
indicated for C57/BL/6J Dark, N=4; C57/BL/6J Light, N=4; arrestin -/- Dark, N=3; arrestin -
/- Light, N=3; mCAR-low Dark, N=5, mCAR-low Light, N=6; mCAR-high Dark, N=4;
mCAR-high Light, N=5. No mice reared and kept in the dark display statistically significant
changes in ONL thickness nor did wild-type C57/BL/6J mice exposed to light. Using the
measurements between -2 and -5 and 2 to 5, the ONL thickness were thinner in arrestin -/-
(p<0.001), mCAR-low (p<0.001), and mCAR-high (p<0.001) exposed to light compared to
those kept in the dark. All mice (C57/BL/6J Light, p<0.001; mCAR-low Light, p<0.001;

80
Figure 2.3. Continued
mCAR-high Light p<0.001) exhibited differences in ONL thickness compared to arrestin -/-
mice exposed to light. Measurements of the ONL thickness of the superior retina between -2
and -5 were used to compare the ONL thickness of the inferior retina between 2 and 5. The
arrestin -/- (p<0.001), mCAR-low (p<0.01) and mCAR-high (p<0.01) mice exposed to light
demonstrated a differential superior to inferior ONL thickness. Analysis of mCAR-high
Dark compared to light exposed mice showed incomplete rescue (p<0.001) in the regions
immediately surrounding the optic nerve. Protection from photoreceptor light damage in
mCAR-high appeared to be complete except for portions of the central superior retina.
(c) Arrestin -/- ONL decreased in a light intensity dependent manner. All data for arrestin -/-
was graphed together with light intensity levels as indicated. Under all three light intensity
levels, the arrestin -/- ONL was dramatic and significant thinner than dark reared control
mice (p<0.001). The superior retina was more susceptible to light damage at the ~1000 and
~2000 lux of light (p<0.01). The ONL decreased with increasing light intensity until there
was no significant difference between the superior retina and the inferior retina at ~3000 lux
of light (p>0.01).
(d) Light intensity dependent protection of arrestin -/- photoreceptors from light damage by
ectopic CAR expression in mCAR-low transgenic lines. The data for mCAR-low was
graphed together with light intensity levels as indicated to visualize the increasing protective
effect of CAR at decreasing light levels. The superior retina was thinner than the inferior
retina when exposed to light at ~1000 lux and ~3000 lux (p<0.01). There was no significant
change in the ONL thickness of the superior retina unlike in the inferior retina (p<0.001)
when light exposure was decreased from ~3,000 lux to ~1,000 lux. ONL thickness in the
superior retina was not fully protected in arrestin -/- photoreceptors by mCAR-low even at
our lowest light damage setting of ~1000 lux.

81
Figure 2.3. Continued
(e) Light intensity dependent protection of arrestin -/- photoreceptors from light damage by
ectopic CAR expression in mCAR-high transgenic lines. The data for mCAR-high was
graphed together with light intensity levels as indicated to visualize the increasing protective
effect of CAR at decreasing light levels. The superior retina was thinner than the inferior
retina when exposed to light at ~1000, ~2000, ~3000 lux (p<0.01 for all). ONL thickness in
the superior retina was not fully protected in arrestin -/- photoreceptors by mCAR-high
expression even at our lowest light damage setting of ~1000 lux (p<0.001) even though it
was fully protective of the inferior retina. Outside the central superior region between -2
through -5 used for statistics, the ONL appeared to return to normal levels. The threshold
where CAR fully protects photoreceptors through the entire retina from light damage was
less than 1000 lux.

82
Figure 2.3
(a)
(b)
Light Damage at 2,000 lux for 72 h
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
C57/BL/6J Dark
mCAR-high Dark
Arrestin -/- Dark
C57/BL/6J Light
mCAR-high Light
Arrestin -/- Light
Light Damage at 1,000 lux for 72 h
0
5
10
15
20
25
30
35
40
45
50
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
C57/BL/6J Dark
mCAR-high Dark
mCAR-low Dark
Arrestin -/- Dark
C57/BL/6J Light
mCAR-high Light
mCAR-low Light
Arrestin -/- Light

83
Figure 2.3. Continued
(c)
(d)
Intensity dependent damage in Arrestin -/-
0
5
10
15
20
25
30
35
40
45
50
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
Arrestin -/- Dark
Arrestin -/- Light 1,000 lux
Arrestin -/- Light 2,000 lux
Arrestin -/- Light 3,100 lux
Intensity dependant damage in mCAR-Low
0
5
10
15
20
25
30
35
40
45
50
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
mCAR-low Dark
mCAR-low Light 1,000 lux
mCAR-low Light 3,100 lux

84
Figure 3.3. Continued
(e)

Intensity dependent damage in mCAR-high
0
5
10
15
20
25
30
35
40
45
50
-11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11
Superior Inferior
Outer-Nuclear Layer (um)
mCAR-high Dark
mCAR-high Light 1,000 lux
mCAR-high Light 2,000 lux
mCAR-high Light 3,100 lux

85
The superior retina demonstrates an increased susceptibility to light damage in all
lines. In light exposed control arrestin1 -/- mice, the ONL thickness did not
degenerate as much in conditions of decreasing light intensities (Figure 2.3c). There
is a differential regional susceptibility to light damage in the arrestin1 -/- mice
exposed to 2000 lux for 72 hours (p<0.001) and 1000 lux for 72 hours (p<0.001). In
mCAR-low lines, there was an increased protective effect in ONL thickness in the
inferior retina. No significant change was observed in the superior retina (Figure
2.3d). In mCAR-high lines, decreasing the light intensity increased photoreceptor
ONL thickness to near dark control levels except in the superior retina (Figure 2.3e).
The ectopic expression of mCAR can protect photoreceptors in a dose
dependent manner. There was a light intensity threshold where CAR was able to
replace rod arrestin in protecting photoreceptor cells from constitutive signaling
mediated cell death.

Light Dependant Translocation and Membrane Association of mCAR
Another main characteristic of rod arrestin is its well-documented light
dependant ability to translocate to the outer segment and bind to the disc membrane
(Broekhuyse et al., 1985). Translocation of rod arrestin may play a role in mediating
binding to its substrate, disc membrane embedded phosphorylated light activated
rhodopsin (R*-P). Light dependant translocation of mCAR has been observed in the
cone cells of wild-type mice (Zhu et al., 2002). Physiologically, rod cells translocate
rod arrestin and cone cells translocate CAR in a light dependent manner. However,

86
it is not know if transgenic rods can translocate CAR effectively to be in close
enough proximity to bind R*-P. The unavailability of sufficient quantities of CAR
to bind R*-P and terminate phototransduction would lead to constitutive signaling
and ultimately explain observed photoreceptor death.
To ensure that CAR was available, in the correct location, and at high enough
levels, we examined the light dependant in vivo translocation of mCAR to the outer
segment. Light dependent translocation of mCAR-H and rod arrestin in adult mice
was analyzed by immunohistochemistry in collaboration with Dr. Ana Mendez
(Figure 2.4a) and data for mCAR-L is not shown. In the dark, mCAR in wild-type
cone photoreceptors was diffuse from the synaptic terminal at the base of the ONL
through the IS. In the dark, rod arrestin in the rod photoreceptors of wild-type mice
localized to the IS and ONL. In the dark, mCAR in arrestin -/- rod photoreceptors in
mCAR-high retinas was diffuse throughout the entire cell from the ONL to the OS.
In the light, mCAR in wild-type cone photoreceptors translocates in a light
dependent manner to the OS, though some staining was still present in the IS and
synapse of the ONL. In the light, rod arrestin in wild-type rod photoreceptors
translocates to the OS from the IS and ONL. Similarly to mCAR in cone cells,
immunoreactivity to rod arrestin was still present in rod IS and ONL. In the light,
mCAR in arrestin -/- rod photoreceptors translocated almost entirely to the OS from
the IS and ONL. Light dependent translocation of both mCAR and rod arrestin
occurs in a light dependent manner in both wild-type and transgenic mice. These
data corroborate earlier data of the light dependent translocation of CAR in wild-type

87
Figure 2.4. Comparison of mCAR and arrestin translocation and membrane binding. Light induced
translocation of mCAR and arrestin in photoreceptors. Soluble mCAR and arrestin translocates and
binds to the membrane in a light dependent manner.
(a) Immunofluorescent staining of adult mouse retinal sections for mCAR and arrestin. Dark
reared wild-type C57/BL/6J and mCAR-high mice kept in the dark (Dark, top row) or
exposed to light for 30 min (Light, second row) were sacrificed and processed for
immunohistochemistry. Sections were stained for mCAR (mCAR, lower labels) with the
monoclonal anti-mCAR antibody targeting the C-terminal region, LUMIJ, or rod arrestin
(Arrestin) with the polyclonal anti-arrestin antibody, C10C10. In wild-type mice in the dark,
mCAR staining was diffuse throughout the cone photoreceptor cells in the synaptic terminal
at the base of the outer nuclear layer (ONL), ONL, and inner segment (IS). In wild-type
mice in the dark, rod arrestin staining was also diffuse throughout the rod photoreceptor in
the synaptic terminal at the base of the ONL, ONL, and IS similar to mCAR. In mCAR-high
transgenic mice in the dark, mCAR staining is diffuse throughout the entire cell. In wild-
type mice exposed to light, mCAR translocates in a light dependant manner most visibly
from the ONL and synaptic terminals to the cone cell outer segment. In wild-type mice
exposed to light, rod arrestin translocates to the OS most visibly from the IS and the synaptic
terminals. In mCAR-high transgenic mice exposed to light, mCAR translocates nearly
completely with little staining left in the ONL and IS to the OS.
(b) Light dependent association of soluble rod arrestin (Arrestin) to the membrane fraction.
Retinas from dark reared wild-type C57/BL/6J mice (N=3) were dissected in the dark. One
retina from each animal was kept in the dark while the other one was exposed to light for 10
min. Both were homogenized and centrifuged to separate the soluble (supernatant) from the
insoluble membrane (pellet) fraction. The pellet was resuspended in the same volume of
buffer as the supernatant. Equal volumes of proteins from the supernatant and its
corresponding pellet were resolved by SDS-PAGE followed by western blot analysis of
arrestin levels using a polyclonal anti-arrestin antibody, C10C10. Graph indicates

88
Figure 2.4 Continued
densitometry of the accompanying immunoblots (supernatant (S), and pellet (P); mean of the 3
samples as indicated with the standard deviation, SD). There is a loss of arrestin in the Light S
fraction compared to Dark S (p<0.05) and a concomitant gain of arrestin in the Light P fraction
compared to the Dark P (p<0.01).
(c) Light dependent association of soluble mCAR to the membrane fraction. Retinas from dark
reared mCAR-high were dissected in the dark. One retina from each animal was kept in the
dark while the other one was exposed to light for 10 min. Both were homogenized and
centrifuged to separate the soluble (supernatant) from the insoluble membrane (pellet)
fraction. The pellet was resuspended in the same volume of buffer as the supernatant. Equal
volumes of proteins from the supernatant and its corresponding pellet were resolved by SDS-
PAGE followed by western blot analysis of mCAR levels using a monoclonal anti-mCAR
antibody targeting the c-terminal region, LUMIJ. Graph indicates densitometry of the
accompanying immunoblots (supernatant (S), and pellet (P); mean of the 3 samples as
indicated with the SD). There is a loss of mCAR in the Light S fraction compared to Dark S
(p<0.05) and a concomitant gain of mCAR in the Light P fraction compared to the Dark P
(p<0.05). The binding of mCAR to the membrane in mCAR-high transgenic mice is less
robust than rod arrestin binding in wild-type mice.

89
Figure 2.4. Continued
(a)

(b)

(c)

Dark Supernatant Dark Pellet Light Supernatant Light Pellet Dark Supernatant Dark Pellet Light Supernatant Light Pellet Dark Supernatant Dark Pellet Light Supernatant Light Pellet
Densitometry analysis of Arrestin translocation to
outersegment
0
1000
2000
3000
4000
5000
6000
7000
8000
Dark S Light S Dark P Light P
Intensity
Dark Pellet Light Pellet Dark Supernatant Light Supernatant Dark Pellet Light Pellet Dark Supernatant Light Supernatant Dark Supernatant Light Supernatant
Densitometry analysis of mCar translocation in
mCar-High
0
1000
2000
3000
4000
5000
6000
Dark S Light S Dark P Light P
Intensity
Dark
Light
Wild-type Wild-type mCAR-high
mCAR Arrestin mCAR
Dark
Light
Wild-type Wild-type mCAR-high
mCAR Arrestin mCAR
ONL
OS
IS
ONL
OS
IS
Dark
Light
Wild-type Wild-type mCAR-high
mCAR Arrestin mCAR
Dark
Light
Wild-type Wild-type mCAR-high
mCAR Arrestin mCAR
ONL
OS
IS
ONL
OS
IS
ONL
OS
IS
ONL
OS
IS

90
cone cells (Zhu et al., 2003) and of rod arrestin (Broekhuyse et al., 1985).
Translocation of mCAR in response to light was more dramatic in the mCAR-high
transgenic mice lines. Sufficient amounts of CAR protein for terminating the
photoresponse should be present in the OS during light exposure to bind R*-P and
terminate constitutive signaling and therefore protect photoreceptors from death.
Translocation of CAR to the outer segment was effective in our transgenic
mice, yet CAR binding to its substrate target may be ineffective for
phototransduction termination in our mice. In the dark, rod arrestin and CAR remain
soluble in the cytoplasm. Upon light exposure the trans-membrane receptor
rhodopsin activates the phototransduction cascade, it is then phosphorylated, and rod
arrestin binds to R*-P in the membrane to terminate the photoresponse.
The membrane association assay was used to analyze the ability of CAR to
bind to the disc membrane in a light dependent manner. The retinas of one month
old dark reared mice were dissected in the dark, homogenized, and centrifuged. The
soluble mCAR and rod arrestin in retinal homogenates was released into the
supernatant. The pellet contained membrane bound rhodopsin and the associated
arrestins. Western blot analysis of the pellet and supernatant fractions from
C57/BL/6J wild-type retinas in the dark or in the light showed light inducing arrestin
partitioning from the supernatant into the membrane fraction (Figure 2.4b).
Analysis of mCAR-high transgenic mice also showed light inducing mCAR
partitioning from supernatant into the membrane fraction although membrane
association was not observed to be as great (Figure 2.4c).

91
Electrophysiology
Electrophysiology has been integral in the characterization of the functional
components of the phototransduction cascade. Membrane association assays
suggested that mCAR did not bind to R*-P to terminate rod phototransduction as the
mechanism that led to photoreceptor cell loss in light damage experiments. To
address the issue of constitutive signaling, single cell recordings in response to dim
light flashes were obtained. The current across the membrane of isolated single
photoreceptors was measured in response to the light flashes. Single cell recordings
from our collaborators Drs. Will Ruben and Marie Burns were obtained from
mCAR-high transgenic mice, wild-type, and arrestin knockout mice in response to
dim light flashes. The average of single photon responses from single cell
recordings were plotted (Figure 2.5). Data showed similar initiation phase in
mCAR-high and arrestin -/- compared to wild-type mice. The initial re-polarization
of mCAR-high photoreceptors occurred similarly to arrestin1 -/- mice and has been
attributed to phosphorylation of activated rhodopsin (Xu et al., 1997). Arrestin1
binding follows rhodopsin phosphorylation, sterically blocks the association of
transducin, and therefore terminates transducin dependent constitutive signaling. It
appeared that the recovery mediated by arrestin1 binding was faster in the mCAR-
high rods when compared to the wild-type. The recovery of mCAR rod response
continued to progress past that of arrestin1 -/-, indicative of functional mCAR
binding to phosphorylated rhodopsin to inactivate the phototransduction cascade,
though inactivation is not complete.

92

Figure 2.5

Figure 2.5. Average single photon responses from single cell recording of dim light response. Single
photoreceptors were isolated and the change in the current across the OS cell membrane was
measured in response to light using a suction electrode. The initiation phase is similar in wild-type,
arrestin knockout (Arrestin -/-), and mCAR-high. The recovery phase of mCAR-high appears faster
and proceeds past the level of arrestin -/-, but does not completely shut off the rhodopsin
phototransduction cascade as in wild-type photoreceptors during the time frame observed.
1.0
0.8
0.4
0.2
0
pA
4 2 0
Time (s)
WT
mCAR-High
Arrestin
1.0
0.8
0.4
0.2
0
pA
4 2 0
Time (s)
Wild-type
mCAR-High
Arrestin -/-
1.0
0.8
0.4
0.2
0
pA
4 2 0
Time (s)
WT
mCAR-High
Arrestin
1.0
0.8
0.4
0.2
0
pA
4 2 0
Time (s)
Wild-type
mCAR-High
Arrestin -/-

93

Rhodopsin Content is Unaltered in Arrestin Knockout and mCAR Transgenic Mice
The use of the rhodopsin promoter to drive cell specific expression of a target
gene may have affected the level of endogenous rhodopsin levels. A transgenic
mouse line where blue opsin was expressed under the control of the same rhodopsin
promoter used in these experiments was known to affect rhodopsin levels (Dr. Guang
Shi, unpublished data). Alterations in the level of rhodopsin content will affect the
ability of enough mCAR to bind and terminate light activated rhodopsin signaling.
To ensure that levels of the target substrate, rhodopsin, was unaffected by the
addition of the mCAR transgene, the amount of rhodopsin in the whole retina was
assessed by photo-bleaching. There was no statistically significant difference in the
absorbance of 500 nm light in the arrestin knockout mice compared to mCAR-high
(Table 2.1). Rhodopsin levels in arrestin knockout mice have been shown to be
equivalent to wild-type mice (Xu et al., 1997; Roca et al., 2004). The level of
rhodopsin in our transgenic mice was unaffected by the expression of out transgene.
The inability of CAR to terminate constitutive signaling is not due to altered levels
of it substrate rhodopsin.

94

Table 2.1

Table 2.1: Rhodopsin content is similar in arrestin1 knockout (Arr -/-) and transgenic mCAR-high
mice. The retinas from dark reared one month old mice from arrestin1 knockout mice (N=12) and
mCAR-high mice (N=12) were used to analyze the content of rhodopsin by photo-bleaching. No
statistically significant difference was observed.

Δ Mean 500nm dark-light SD
Arr-/- 0.13 0.02
mCar-high 0.12 0.04

95
Discussion
In the mammalian retina, rod and cone photoreceptors are responsible for
converting light inputs into electric signals. Rods help distinguish between light and
dark under dim conditions while cones are responsible for high-acuity color vision in
the light. Though humans require cone mediated vision in most daily activity, cone
phototransduction remains unclear. Homologous rod phototransduction proteins
have been identified in cone photoreceptors and are thought to function similarly as
their counterparts. By focusing on CAR, we have begun to dissect out the
components thought to be involved in the cone phototransduction cascade and to
compare their functions with their well-characterized homologous rod counterparts.
CAR protects photoreceptors in a dose dependent manner in arrestin -/-
photoreceptors when mice are exposed to light. Degeneration was prevented when
mice were kept in the dark. Immunohistochemistry localization demonstrates proper
light dependent translocation of mCAR to the OS from the IS and ONL in mCAR-
high mice. Membrane association assays showed less CAR binding to the disc
membrane than arrestin. Single cell electrophysiology recordings provide evidence
of ineffective rhodopsin signal termination and a prolonged photoresponse in arrestin
knockout rods that express mCAR. These data suggest that mCAR’s inability to
fully terminate the light mediated excessive signaling leads to cell death. We have
found that CAR has some ability to deactivate R*, but it does not seem to form as
stable a complex as a rod arrestin-rhodopsin complex. This is likely the underlying

96
basis for CAR’s partially protective role in preventing constitutive
phototransduction activation mediated cell death.
Ectopic expression of mCAR in arrestin knockout mice allows us to directly
compare the functional ability of arrestin and mCAR in protecting photoreceptors
from light damage. CAR could not replace rod arrestin and bind to R*-P in an
equivalent manner, this is unlikely a result of differential expression in the amount of
arrestins. In arrestin hemizygous knockout mice approximately 30% of arrestin
protein is retained, yet there was no evidence of photoreceptor light damage and the
kinetics of the photoresponse was unchanged compared to wild-type mice (Xu et al.,
1997). Photoreceptor cell loss, as expected, was not observed in wild-type mice with
endogenous rod arrestin even at 3000 lux of light exposure. Even very low amounts
of a truncated form of arrestin, p44 which expresses at ~10% normal arrestin levels
is able to rescue photoreceptors and restore a normal dim flash response. The
functional difference of CAR and arrestin in the ability to protect photoreceptors is
unlikely due to the concentration of mCAR in transgenic rods, but more likely
secondary to the differences in structure of CAR and rod arrestin.
CAR is capable of binding to R*-P. Competitive binding assays
demonstrated that salamander CAR binds at approximately 50-fold lower affinity for
frog rhodopsin than did salamander rod arrestin (Smith et al., 2000). Competitive
binding assays of mCAR and rod arrestin to rhodopsin will be needed for direct
confirmation, but it remains unlikely that mCAR can functionally substitute for rod
arrestin even at a 50-fold increased level as suggested by observations in salamander

97
binding experiments. While only some of the rod arrestin translocate to the OS,
nearly all of mCAR in our transgenic lines move into the OS. Translocation studies
of rod arrestin and mCAR-high suggest higher mCAR levels in the
microenvironment of the OS with the mCAR-to-rhodopsin molar ratio lower than
1:12.5 calculated in our transgenic photoreceptors and perhaps even lower than the
1:1.3 rod arrestin-to-rhodopsin molar ratio, yet light damage was still observed.
Membrane association experiments separating the membrane containing
pellet and the supernatant soluble mCAR and arrestin in mCAR-high and wild-type
retinas establish biochemical evidence that CAR can associate with the membrane
where the substrate R*-P is located. In wild-type mice, rod arrestin showed a
dramatic movement from the soluble fraction to the membrane in the light.
Although mCAR translocation in vivo is dramatic in mCAR-high, membrane
association assays suggested less binding to the membrane bound target substrate
R*-P. Movement of CAR may be more efficient in rod cells, yet CAR did not
associate with the membrane as effectively as rod arrestin. Consistent with weaker
affinity, this may explain the inability of mCAR to fully protect light damage to
these photoreceptors since membrane bound R*-P can still signal, resulting in a
prolonged photoresponse. Rhodopsin-rod arrestin complexes appeared in greater
amounts than rhodopsin-mCAR complexes with the membrane association assay.
As a corollary in support of a less stable rhodopsin-mCAR complex the
signal terminating rhodopsin-rod arrestin complex, single cell electrophysiology
during the second half never reaches baseline in mCAR transgenic mice. An average

98
flash response occurs either when: 1) mCAR does not fully terminate signaling or 2)
events of high affinity mCAR binding to R*-P that completely quenches signaling
occur simultaneously with events when mCAR binds with very low affinity. Since
we did not observe large fluctuations from trial to trial as would be reflected in an
instance of both binding and non-binding events that yields the observed average
flash response, it is more likely that mCAR binds to all R*-P with low affinity. This
suggests that the CAR-rhodopsin complex is less stable than a physiological rod
arrestin-rhodopsin.
The light damage assay was used to analyze the ability of CAR to
functionally replace rod arrestin. At 1,000 and 2,000 lux of light exposure, the
arrestin -/- control retinas exhibited an increased ONL cell loss in the central superior
retina as observed earlier (Roca et al., 2004). At 3000 lux, this disparity was no
longer apparent in arrestin -/- suggesting a threshold where the factors that mediate
increased resistance to light damage in the inferior retina was no longer effective.
The hemispheric sensitivity in arrestin -/- mice suggests an increased susceptibility to
light damage in the superior retina. This hemispheric sensitivity was also observed
in mCAR transgenic mice. Both mCAR-low and mCAR-high lines protected
photoreceptors from cell death better in the inferior retina with full rescue of the
ONL in the inferior retina of mCAR-high mice exposed to ~1000 lux of light
damage. At progressively lower light intensity, protection from light damage in
photoreceptors was still dose dependant: higher levels of mCAR were better at
maintaining photoreceptors. There is a threshold of complete protection from light

99
damage afforded to rod photoreceptors expressing mCAR most likely at less than
1000 lux.
The mechanism leading to the regional light induced cell death remains
unclear. Greater sensitivity in the superior retina has been found in albino mice
(LaVail et al., 1987) and in normal mice regardless of light direction or mouse
pigmentation (Rapp and William, 1980). Reactive oxygen species are thought to be
the main damaging agent in albino mice (Noell, 1980). This is unlikely in our
experiments, since no mutations were targeted to affect metabolic or oxidative stress
pathways. Our genetic mutations modified molecules in the rod phototransduction
cascade. Differential absorption of photons by rod cells resulting in longer sustained
levels of constitutive light activation may led to increased cell death in the superior
retina. Yet, in studies using an ocular transmission photometer that measured the
scattered light transmitted to the external surface of the sclera after passing through
the normal optics of the eye and then through the back found no difference in the rate
of photon incident on the different regions of the retina (Williams et al., 1998). The
molecular mechanism behind differential rod regional susceptibility to light damage
remains unknown.
Proteins in photoreceptors are compartmentalized and their location may be
critical to their proper function. The proper localization of rhodopsin has been
narrowed to a sequence in the C-terminal region of the protein (Chuang et al. 1998;
Martin et al. 2004). Retinitis pigmentosa is a group of retinal disorders characterized
by progressive dysfunction, cell loss, and atrophy of the retinal tissue that has been

100
associated with improper protein localization. Four C-terminal rhodopsin mutations
that cause retinitis pigmentosa with the following nomenclature describing the one
letter amino acid abbreviation followed by the location and the amino acid
substitution: Q344ter (a nonsense which terminates at 344 and lacks the last 5 amino
acids), V345M, P347S, and P347L abolished normal rhodopsin binding to Tctex-1,
an actin light chain and disrupting proper trafficking (Tai et al., 1999). The improper
localization of rhodopsin leads to cell death. Proper localization of
phototransduction proteins may be mediated by specific sequences.
It has been suggested that rod arrestin translocates in a light dependent
manner to be near the disc membrane where R*-P resides to efficiently bind to and
terminate phototransduction (Lee and Montell, 2004). Similarly, CAR is thought to
translocate to bind to its substrate target in the OS (Zhu et al., 2002). CAR
effectively translocates and weakly associates to the membrane upon exposure to
light and rapid diffusion would not explain such a translocation when compared to
arrestin that causes more of a concentration gradient. In vivo experiments
demonstrate that mCAR in our transgenic lines exhibit the same characteristic of
light dependant translocation from the ONL and IS of photoreceptors to the OS as
native mCAR in cones. Light dependent translocation of rod arrestin depends on
light activation of rhodopsin, but not R*-P in mammals (Mendez et al., 2003), but
the mechanism remains unclear. Drosophila rod arrestin requires phosphoinositides
to interact with NINAC myosin III that promotes light dependent translocation (Lee
and Montell, 2004). The demonstrated ability of CAR to translocate in our

101
transgenic mice provide an additional resource to identify the sequence that govern
localization of the arrestins in the dark as well as the sequence required for proper
translocation. These new lines contain CAR, a homologous protein to rod arrestin,
with many different amino acid residues at multiple sites can be used in conjunction
with full length p48 arrestin and its splice variant p44 to identify the required
signaling sequence for proper protein localization. Arrestin compared to the
truncated p44 moves from the IS to the OS during light exposure, yet p44 usually
resides in the OS due to possible membrane anchoring. By aligning sequences, one
can determine the different amino acid mutations between all three proteins that may
play a role in localization. Chimeras using partial sequences fused with a reporter
construct such as a green fluorescent protein (GFP) constructs can be generated and
introduced into rod photoreceptors via electroporation or viral vectors. Proper
localization and light dependent translocation of chimeras will narrow the specific
region for binding studies to identify associated proteins mediating proper sequence
localization and translocation.
The underlying molecular factors that ultimately lead to light dependent cell
death to date remain unclear. Light exposure to the retina is two magnitudes less in
undilated pigmented mice (Rapp and William, 1980). Therefore, the light reaching
the retina was effectively 30, 20, and 10 lux in our light damage experiments. At
low light intensity, the c-Fos/AP-1 pathway is not involved; instead, downstream
transducin signaling is required for cell death. CAR was capable of rescuing
photoreceptors in a light- and dose-dependant manner. In an effort to explore the

102
mechanism that leads to light damage, single cell recording of flash responses in
these mice were compared to arrestin knockout and wild-type mice. The mCAR
photoreceptors display an inability to completely shutoff phototransduction in the
observed period. Mice kept in the dark with an inactive phototransduction cascade
display normal morphology. Our dim light damage experiments and incomplete
shutoff of dim light flashes implicate constitutive light signaling as the mechanism
leading to cell death.
These transgenic mCAR mice may provide a valuable model in conjunction
with arrestin knockout mice to elucidate the mechanism of dim light damage.
Patients with Oguchi disease, a recessive condition from naturally occurring
mutations that result in nonfunctional arrestin or rhodopsin kinase, retain unaffected
daytime vision (Chen et al, 1999). The inability to terminate rod signaling due the
absence of rod arrestin results in loss of night vision, but the retention of high acuity
color daytime cone mediated vision due to unaffected CAR can be a molecular
mechanism that underlie the clinical presentation of Oguchi.
The inability of CAR to entirely shutoff phototransduction may be due to
CAR’s physiological specificity to S and M opsins as opposed to rhodopsin. Rod
arrestin binds quickly and with high affinity to light activated, phosphorylated
rhodopsin (R*-P). Rod arrestin has been demonstrated to interact with the third and
the first cytoplasmic loop as well as the phosphorylated C-terminal end of rhodopsin
(Krupnick et al., 1994). Mutagenesis studies of individual residues in bovine arrestin
have identified Arg 171, Arg 175, and Lys 176 as critical for phosphate interaction.

103
BLAST results of mCAR sequence show that although the Arg and Lys residues are
near the same vicinity to residues 175 and 176, respectively, in bovine arrestin, there
is no Arg near residue 171 in mCAR. Amino acids between R*-P and arrestin
interact specificity to determine the specificity and kinetics of binding. Difference in
the mCAR sequence and rod arrestin may contribute to differential binding and thus
effective termination of the rod photoresponse. In transgenic mice expressing blue
opsin, a cone visual pigment, rod arrestin works well to terminate the photoresponse
(Dr. Guang Shi, unpublished data) suggesting mCAR has a higher specificity in
recognizing its substrate.
The downstream factors from transducin involved in mediating constitutive
signaling dependent photoreceptor cell death are thought to involve TGF-β (Roca et
al., 2004), reactive oxygen species (Noell, 1980), or prolonged sustained low levels
of Ca
2+
have all been implicated. In darkness cGMP-gated channels are open and
allow the influx of extracellular Ca
2+
while an ion Na
+
/K
+
/Ca
2+
exchanger couples
the influx of 4 Na
+
ions down the electrochemical gradient in exchange for extruding
one K
+
and one Ca
2+
(Makino et al., 2004). In response to light, cGMP gated ion
channels close and Ca
2+
levels fall since extracellular Ca
2+
is no longer entering
while the Na
+
/K
+
/Ca
2+
exchanger is still actively extruding K
+
and Ca
2+
resulting in
hyperpolarization of the cell. Hyperpolarization causes voltage dependent ion
channels to close and the loss of Ca
+2
in the other compartments of the cell. Ca
+2

also act as a second messenger in the signal transduction of many other pathways, so

104
a sustained drop in Ca
+2
levels due to constitutive signaling may have profound
implications on other cellular activities.
Constitutive signaling also leads to sustained low levels of cGMP through
sustained phosphodiesterase activity. cGMP is a second messenger involved in
signal transduction of other molecular pathways. Lower levels of cGMP in the OS
may act as a sink and draw the small cellular signaling cGMP from the IS through
the connecting cilium effectively decreasing photoreceptor signaling cGMP.
Sustained low levels of cGMP or the intermediates needed to regenerate sufficient
quantities of cGMP in the OS of retinas with defective photoresponse shut-off may
mediate a cell death signal. These underlying pathways that lead to light dependent
cell death remain to be clarified.
The functional role of arrestin is to inactivate phototransduction that leads to
the recovery of sensitivity. Electrophysiology data show that mCAR can shutoff
phototransduction although incompletely. The initiation phase of the photoresponse
is similar in the photoreceptors of wild-type, mCAR-high, and arrestin knockout
mice. There appears to be a faster inactivation of the photoresponse in mCAR-high
rods. Cones are known to recover sensitivity much faster with visual pigment
regeneration 2000 fold higher than rod (Zhu et al., 2003) and opsin phosphorylation
occurs at more than 20 times faster in carp cones than rods (Tachibanaki et al., 2001).
Fast flash response recovery have been observed in the rod null phenotype Nrl
knockout (Nikonov et al., 2005), a cone retinal model that expresses enhanced S-
opsin and M-opsin cone proteins. Flash response recovery of cone cells is faster than

105
rod cells (Rodieck, 1998), but the role of mCAR in mediating this needs further
study.
Our results provide biochemical and functional evidence in vivo that CAR
behaves like rod arrestin in response to light and that it binds less stably with
rhodopsin than rod arrestin. CAR can protect photoreceptors from light damage at
proper light levels, translocate and associate to the membrane where R*-P resides,
and partially shutoff phototransduction. We provide strong evidence that CAR plays
a functionally similar role in cone cells as rod arrestin does in rod cells, but it cannot
substitute for rod arrestin. Because R*-P is not the natural target for CAR, we have
begun to characterize transgenic mice that express mouse S opsin and mCAR in rod
photoreceptors lacking arrestin and rhodopsin to compare the interaction of these
cone phototransduction components in the already well characterized rod
environment.

106
Chapter II: References
Bradford MM. (1976) A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal Biochem. 72:248-54.

Broekhuyse RM, Tolhuizen EF, Janssen AP, Winkens HJ. (1985) Light induced
shift and binding of S-antigen in retinal rods. Curr Eye Res. 5:613-8.

Carter-Dawson LD and LaVail MM. (1979) Rods and cones in the mouse retina. I.
Structural analysis using light and electron microscopy. J Comp Neurol.
188(2): 245-62.

Chen CK, Burns ME, Spencer M. (1999) Abnormal photoresponse and light induced
apoptosis in rods lacking rhodopsin. PNAS 96: 3718-22.

Chen J, Simon MI, Matthes MT, Yasumura D, LaVail MM. (1999) 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 JZ and Sung CH. (1998) The cytoplasmic tail of rhodopsin acts as a novel
apical sorting signal in polarized MDCK cells. J Cell Biol. 142(5): 1245-56.

Craft CM, Whitmore DH. (1995) The arrestin superfamily: cone arrestins are a
fourth family. FEBS Lett. 362(2): 247-55.

Craft CM, Whitmore DH, Wiechmann AF. (1994) Cone arrestin identified by
targeting expression of a functional family. J Biol Chem 269: 4613-4619.

Fain GL and Lisman JE. (1993). Photoreceptor degeneration in vitamin A
deprivation and retinitis pigmentosa: the equivalent light hypothesis. Exp Eye
Res. 57(3): 335-40.

Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole-Cire P,
Kleyner Y, Taylor A, Simon MI, Grimm C, Reme CE, Lem J. (2002).
Evidence for two apoptotic pathways in light-induced retinal degeneration.
Nat Genet. 32(2): 254-60.

Jeon CJ, Strettoi E, Masland RH. (1998) The major cell populations of the mouse
retina. J Neurosci. 18(21): 8936-46.

Krispel CM, Chen C, Simon

MI and Burns ME. (2003) Novel Form of Adaptation in
Mouse Retinal Rods Speeds Recovery of Phototransduction. J Gen
Physiology 122(6): 703-712.

107

Krupnick JG, Gurevich VV, Schepers T, Hamm HE, Benovic JL. (1994) Arrestin-
rhodopsin interaction. Multi-site binding delineated by peptide inhibition. J
Biol Chem. 269(5): 3226-32.

LaVail MM, Gorrin GM, Repaci MA, Yahumura D. (1987) Light-induced retinal
degeneration in albino mice and rats: strain and species differences. In:
Hollyfield JG, Anderson RE, LaVail MM, eds. Degenerative retinal
disorders: clinical and laboratory investigations. New York: Alan R. Liss.

LaVail, MM, Yasumura, D, Matthes, MT, Drenser, KA, Flannery, JG, Lewin, AS,
Hauswirth, WW. (2000) Ribozyme rescue of photoreceptor cells in P23H
transgenic rats: long-term survival and late-stage therapy. Proc. Natl. Acad.
Sci., USA. 97: 11488-93.

Lee SJ and Montell C. (2004) Light-Dependent Translocation of Visual Arrestin
Regulated by the NINAC Myosin III. Neuron 43: 95-103

Martin RE, Ranchon-Cole I, Brush RS, Williamson CR, Hopkins SA, Li F,
Anderson RE. (2004) P23H and S334ter opsin mutations: Increasing
photoreceptor outer segment n-3 fatty acid content does not affect the course
of retinal degeneration. Mol Vis. 10: 199-207.

Lem J, Applebury ML, Falk JD, Flannery JG, Simon MI. (1991) Tissue-specific and
developmental regulation of rod opsin chimeric genes in transgenic mice.
Neuron 6: 201-10

Makino C, Dodd R, Chen J, Burns ME, Roca A, Simon M, and Baylor D. (2004).
Recoverin regulates light-dependent phosphodiesterase activity in retinal rods.
J Gen Physiol 123: 729-741.

Mears AJ, Kondo M, Swain PK, Takada Y, Bush RA, Saunders TL, Sieving PA,
Swaroop A. (2001) Nrl is required for rod photoreceptor development. Nat
Genet 29: 447-452.

Mendez A, Lem J, Simon M, Chen J. (2003) Light-dependent translocation of
arrestin in the absence of rhodopsin phosphorylation and transducin signaling.
J Neurosci. 23(8): 3124-9.

Michon JJ, Li ZL, Shioura N, Anderson RJ, Tso MO. (1991) A comparative study
of methods of photoreceptor morphometry. Invest Ophthalmol Vis Sci.
32(2): 280-4.

108
Nikonov S, Danielle L, Zhu X, Craft C, Swaroop A, Pugh E. (2005) Photoreceptors
of Nrl -/- mice coexpress functional S- and M-cone opsins having distinct
inactivation mechanisms. J General Physiology 125: 287-303.

Noell WK. (1980). Possible mechanisms of photoreceptor damage by light in
mammalian eyes. Vis Res 20: 1163-1171.

Rapp LM and William TP. (1980) A parametric study of retinal light damage in
albino and pigmented rats. In: Williams TP, Baker BN, eds. The effect of
constant light in visual process: NY Plenum. 137-159.

Roca A, Shin KJ, Liu X, Simon MI, Chen J. (2004). Comparative analysis of
transcriptional profiles between two apoptotic pathways of light-induced
retinal degeneration. Neuroscience. 129(3): 779-90.

Rodieck RW. (1998) The first steps in seeing. Sinauer Associates Inc, Sunderland,
MA.

Smith WC, Gurevich EV, Dugger DR, Vishnivetskiy SA, Shelamer CL, McDowell
JH, Gurevich VV. (2000) Cloning and functional characterization of
salamander rod and cone arrestins. Investigative Ophthalmology & Visual
Science. 41(9): 2445-55.

Sokolov M, Lyubarsky AL, Strissel KJ, Savchenko AB, Govardovskii VI, Pugh Jr
EN, Arshavsky VY. (2002) Massive light-driven translocation of transducin
between the two major compartments of rod cells: a novel mechanism of
light adaptation. Neuron 34: 95-106.

Tachibanaki S, Tsushima S, Kawamura S. (2001) Low amplification and fast visual
pigment phosphorylation as mechanisms characterizing cone photoresponses.
Proc Natl Acad Sci USA. 98(24): 14044-9

Tai AW, Chuang J-Z, Bode C, Wolfrum U, and Sung C-H. (1999) Rhodopsin’s
Carboxy-Terminal Cytoplasmic Tail Acts as a Membrane Receptor for
Cytoplasmic Dynein by Binding to the Dynein Light Chain Tctex-1 97: 877-
887.

Wilden U, Hall SW, Kuhn H. (1986) Phosphodiesterase activation by photoexcited
rhodopsin is quenched when rhodopsin is phosphorylated and binds the
intrinsic 48-kDa protein of rod outer segments. Proc Natl Acad Sci USA.
83(5): 1174-8.

Williams TP, Webbers JPP, Giordano L, Henderson RP. (1998) Distribution of
photon absorption rates across the rat retina. J of Physiology 508(2): 515-522.

109
Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA, Chen J. (1997) Prolonged
photoresponses in transgenic mouse rods lacking arrestin. Nature. 389: 505-9.

You L, Chan SK, Bruce JM, Archibeque-Engle S, Casanova M, Corton JC, Heck H.
(1999) Modulation of testosterone-metabolizing hepatic cytochrome P-450
enzymes in developing Sprague-Dawley rats following in utero exposure to
p,p'-DDE. Toxicol Appl Pharmacol. 158(2): 197-205.

Zhu X, Li A, Brown B, Weiss ER, Osawa S, Craft CM. (2002) Mouse cone arrestin
expression pattern: light induced translocation in cone photoreceptors.
Molecular Vision. 8: 462-71.

Zhu X, Brown B, Li A, Mears AJ, Swaroop A, Craft CM. (2003) GRK1-dependent
phosphorylation of S and M opsins and their binding to cone arrestin during
cone phototransduction in the mouse retina. J Neurosci. 23(14): 6152-60.

110
Chapter III: Gene Expression Analysis During Perinatal
Retinal Ganglion Cell Development

Chapter III: Abstract
Purpose. To identify the molecular events underlying the dramatic changes between
embryonic retinal ganglion cells (RGCs) that undergoing rapid axonal elongation
and postnatal RGCs with much reduced axonal growth.
Methods. Gene expression profiles of purified RGCs were systemically analyzed
from embryonic day 16 through postnatal day 5, before and after RGCs lose their
ability to elongate axons, using Affymetrix
TM
cDNA microarray. The temporal
expression profiles of select transcripts were confirmed using real-time reverse
transcriptase-polymerase chain reaction (RT-PCR) and in situ hybridization (ISH).
Moreover, expression of the selected candidates was further investigated in purified
RGC cultures of an in vitro model of neurite growth.
Results. Analysis of over 22,600 transcripts identified 71 genes that were
differentially expressed between the embryonic and postnatal RGCs. These results
were conformed by RT-PCR and in situ hybridization. A subset of these transcripts
also exhibited similar patterns of expression in vivo as they did in vitro in purified
RGC cultures, suggesting that isolated RGCs preserve the transcriptional mediation
of gene expression in culture.
Conclusions. Differential expression of gene transcripts in perinatal RGCs may
contribute to the loss of axon growth capacity in these neurons. These results provide

111
an invaluable database for our understanding of gene regulation on RGC
development and disease, including glaucoma.

Chapter III: Introduction
Retinal ganglion cells (RGCs) are a critical component in the visual cycle as
the final step of transmitting information from the neural retina to the brain. These
cells are preferentially affected in glaucoma, one of the most devastating diseases of
the developed world. RGCs are glutaminergic neurons widely used as a model
central nervous system (CNS) neuron in studying development, injury, degeneration,
and regeneration. Identifying and characterizing the molecular pathways dictating
RGC function is important to understanding fundamental neuronal processes.
Cellular differentiation of RGCs is controlled by key transcriptional factors. Unlike
neurons of the peripheral nervous system and even immature CNS neurons that
readily regenerate after damage, mature CNS neurons exhibit a significant loss in
their axonal regenerative potential.
Regulation of transcription is commonly used to control the levels of gene
expression and thus cellular functions. Transcriptional factors have played an
integral role in coordinating developmental events in RGCs. A study of developing
retinal cells using pulse labeling of progenitors exiting the cell cycle with [
3
H]
thymidine demonstrate cell-types in the differentiating neuroretina emerges from the
precursor population in an invariant temporal sequence (Cepko et al., 1996).
Transcriptional factors expressed in a temporal specific manner are the main

112
regulatory factors in controlling proper gene specific expression mediating the
correct number and functionality of RGCs.
In RGC development, transcriptional factors are critical in dictating
transcriptional events that generate a unique signaling pathway to respond to
particular sets of cues. Progression of RGC differentiation involves the properly
timed sequential expression of transcription factors including: Pax6, Math5, and
Brn3b (Ashery-Paden et al., 2001; Brown et al., 1998). Pax6 is member of the Pax
family with two DNA binding motifs whose expression is maintained in proliferating
cells. Pax6 conditional mutants resulted in reduced retinal progenitor cell
proliferation and the exclusive generation of amacrine interneurons (Marquardt,
2001). The basic-loop-helix (bHLH) transcription factor Math5 is expressed at E11
in retinal progenitors and promotes RGC cell fate (Brown et al., 1998). Math5 loss
of function mutants results in the loss of most RGCs and a concomitant increase in
other retinal cell types (Bertrand et al., 2002). Brn-3b, a member of the POU domain
family of transcription factors, expression peaks at E12-E15 during which the
greatest number of ganglion cells undergo their final mitosis (Gan et al., 1996).
Homozygous Brn-3b deletion resulted in a selective loss of 70% of RGCs (Gan et al.,
1996). Therefore, transcriptional regulators play critical roles in the differentiation
of RGCs. Though RGC genesis has readily been studied, not much is known about
the molecular factors involved in the later stages of RGC development.
After cell commitment, immature embryonic CNS neurons such as RGCs
extend their axons to reach their targets in a coordinated manner. Nearly 90% of

113
embryonic RGCs project axons to contralateral targets (Zhu et al., 1999) by
extending their axons across the midline optic chiasm at about E14, and have fibers
in place by E16 (Zhu et al., 1999). RGCs with ipsilateral targets comprise the
remaining 10% of the cell type and project axons to their targets from E18 to birth
(Goldberg et al., 2002; Godement et al., 2002). There is a high degree of uniformity
in the actions of RGCs during perinatal development. Most of these neurons are
functioning in the same manner, to send axons to reach their targets, and are thought
to express a similar set of genes encoding for this process.
The molecular mechanisms underlying why mature CNS neurons display
little axonal growth in contrast with neonatal CNS neurons that readily extend axons
remain unclear. Axonal restriction is not due to the inhibitory myelin since neonatal
neurons extend long axons when grown on a total myelin substrate (Cai et al., 2001)
which supports the idea that internal cell signals govern axon growth restriction. The
physiological level of cAMP declines during perinatal development when axonal
growth capability is restricted. Elevated levels of cAMP and activation of protein
kinase A (PKA) promotes neurite outgrowth in mature CNS neurons (Cai et al.,
2002), yet it is not known whether direct downstream effectors are activated or
indirectly via activation of transcription factors.
Intrinsic control of axon growth is transcription dependant. Neurite
outgrowth is blocked when the transcription inhibitor 5,6-dichloro-1-b-D-
ribofuranosyl-benzimidazole (DRB) is introduced (Cai et al., 2001). Transcription
factors are key components that govern the expression of inherent molecular

114
programs therefore, some genes and proteins must be made to facilitate axon growth.
We hypothesis that transcription factors control transcription dependant molecular
factors mediating axon growth are differentially expressed in neonatal and mature
RGCs.
Defining the mechanism of how neurons control extension of their axons and
termination of this process is critical to understanding the development of the
nervous system, yet the molecular pathway underlying this process remains unknown.
The ability to control and modulate CNS axonal re-growth has profound clinical
implications. The loss of neurite growth in RGCs marks the transition from
immature to mature functional neurons and is transcriptional dependant marks a
developmental milestone.
To identify the differentially expressed transcriptional factors that regulate
the dramatic change in axon growth between immature and functional neurons, we
use purified RGCs. In vitro cultures of purified RGCs defined the loss of axonal
growth as between E16 and E18. Data from Affymetrix high-density microarrays
allows the simultaneous systematic analysis of multiple physiologically relevant
genes from directly isolated developing RGCs before and after axonal restriction.
The microarray data was used as a guide to identify the most developmentally
dynamic transcripts and to identify general trends in their expression and major
categories of gene expression. Components in molecular pathways interact with one
another and function together to mediate cellular responses. A simultaneous change
in the transcripts of a molecular pathway is highly suggestive of a physiologically

115
relevant role. Since transcription factors are known to control multiple biochemical
cascades make and exert widespread cellular changes, the rapid fluctuations of
transcription factors during this period is highly suggestive of a functional role in the
loss of axonal growth ability. A subset of the identified physiologically differentially
expressed transcripts changed similarly in the RGC neurite growth culture
suggesting that these molecular factors may mediate transcription dependent control
of axonal growth.

Materials and Methods
Animals
All procedures in this study concerning animal usage adhered to both the
NIH Health Guide for the Care and Use of Laboratory Animals and USC IACUC
guidelines and conformed to the standards of the Association for Research in Vision
and Ophthalmology.

Purification of RGCs
To coordinate the timing of pregnancy, female C57BL/6J mice (Jackson
Labs) were mated for 24 hrs and then housed separately. The first day post coitus
was termed embryonic day (E0). Pups were born on post-natal day 0 (P0). RGCs
from E16, E18, P0, or P5 C57BL/6J mice were isolated by a magnetic-bead
separation system (Miltenyi Biotec, Auburn, CA). The retinas were dissected in
Ca
2+
and Mg
2+
-free Hanks’ buffered salt solution (HBSS) (Invitrogen, Carlsbad, CA)

116
and dissociated in Ca
2+
and Mg
2+
-free HBSS with 5% papain (Worthington
Biochemicals, Lakewood, NJ) for 5 min at 37
o
C. To stop the digestion, 1%
ovoidmucoid (Worthington Biochemicals, Lakewood, NJ) was added, and cells were
triturated. The single cell suspension was incubated with rabbit anti-mouse CD90
(Thy1.2) antibody conjugated to micrometal beads (Miltenyi Biotec Inc, Auburn,
CA) for 15 min at 4
o
C in elution buffer (Ca
2+
and Mg
2+
-free HBSS with 0.5% bovine
serum albumin and 2 mM EDTA; Sigma-Aldrich). The cells were loaded onto a
metal column in the presence of a magnetic field and washed with elution buffer.
Cells were then eluted in the absence of the magnetic field with elution buffer. To
ensure a consistent level of purity the cells were reloaded onto another column,
washed with elution buffer and then eluted off the column.

Analysis of Cell Purity
Purified RGCs were incubated with CD90 (Thy1.2) antibodies conjugated to
FITC (Pharmingen) for 10 min at 4
o
C or an IgG2b isotype control conjugated to
FITC (Pharmingen). Fluorescent Activated Cell analysis of 10,000 cells was
performed with the final purified cell population using Beckman Coulter Epics-XL-
MCL.

Culture of Purified RGCs
Purified RGCs were cultured in 24-well plates (VWR, San Dimas, CA) that
had been precoated with 10 µg/ml of poly-D-lysine (Sigma, St. Louis, MO)

117
overnight followed by 2 µg/mL merosin (Peprotech, Rocky Hill, NJ) for 3 hrs. Cells
were cultured in Neurobasal A (Gibco) supplemented with Sato supplement: B27
(Gibco), penicillin-streptomycin (Gibco), L-glutamine (Sigma), IST (Gibco), CNTF
(0.02 µg/µl, Peprotech), bFGF (0.01 µg/ml, Invitrogen), forskolin (5 mM, A.G.
Scientific), BNDF (0.01 µg/µl, Invitrogen) and pyruvate (100 mM, Sigma).

Immunohistochemistry
Cultured cells were fixed in a solution containing 4% paraformaldehyde
(PFA) and 15% sucrose in phosphate buffered saline (PBS) for 15 min at 37
o
C and
rinsed three times in PBS-0.2% Triton 100 (PBST). Non-specific binding was
blocked with 20% horse serum (Hyclone) in PBST for 30 min at room temperature.
Primary antibodies (anti-neurofilament high, medium and low molecular weight
(Chemicon) and β-actin (Chemicon)) were diluted in PBST (1:200) and incubated
for 1 hr at room temperature. The cells were rinsed with PBST three times for 5 min
each and then incubated with corresponding secondary antibody, including anti-
rabbit FITC (Vector Research) and anti-mouse (Vector Labs) for 1 hr at room
temperature, and rinsed with PBS. Slides were mounted with Vectashield (Vector
Labs) and photographed with Zeiss LSM 510 confocal microscopy.

Quantification of RGC neurite growth
Equal numbers of purified RGCs were plated and cultured for 48 hrs. Cells
were loaded with 2 µM calcein-AM (Molecular Probes) for 1 hr. Eight confocal

118
pictures were taken per well from at least 3 wells per sample. Total number of
labeled cells and cells with neurites ≥ 3 times their body width were counted.
Percentage of cells bearing a neurite longer than 3 times of their body length was
calculated.

RNA Preparation
RNA was isolated from sorted RGCs by the reagent, Trizol (Invitrogen), and
stored as a precipitate in 70% EtOH at –80
o
C. RNA was dried, dissolved in
0.01mol/1 Tris-HC1, 0.1mmol/1. EDTA pH 7.4 (TE
-4
), and treated with DNase
(Ambion). RNA was quantified using Ribogreen (Molecular Probes) by
spectroscopy against dilutions of a known standard. RNA integrity was analyzed
using a Bioanalyzer (Agilent) for 28S:18S ratio>1.5.

Microarray Hybridization:
Two-hundred nanograms of total RNA was biotin labeled during the second
round of in vitro transcription using the MessageAMP system (Ambion). Biotin
labeled RNA was stored at –80
o
C before hybridization onto MOE 430A microarray
chips (Affymetrix Inc.).
These in vitro transcription products were cleaned with Rneasy Mini kit and
15 µg were fragmented to sizes from 35-200 bases by incubating at 94°C for 35 min.
The biotinylated and fragmented cRNA probes were hybridized for 16 hrs at 40°C to
MOE430A oligonucleotide arrays in the GeneChip Fluidics Station 400 (Affymetrix

119
Inc.), washed, and stained with streptavidin-conjugated phycoerythrin. Probe arrays
were scanned with the Hewlett-Packard Scanner (HP).

Oligonucleotide Data Collection and Analysis
The intensity of individual spot was obtained with Array Suite 5.0
(Affymetrix). The raw data for all the chips were normalized using the Robust
Multi-Chip Averaging (RMA) based algorithm and analyzed further with Avadis 2.0
(Strand Genomics). The gene expression scores were in log base 2. Scores below
7.0 in all chips were set to 7.0 in order to minimize noise associated with less robust
measurements of rare transcripts. These represented the lower 50th percentile of all
expression values in all the data sets. Significance of differential expression levels
(after grouping samples into E16, E18, P0, and P5 time points) were examined using
one-way ANOVA on all groups, all-at-a-time, parametric, with asymptotic p-values
and Westfall-Young testing correction for the set. Fold changes were based on the
mean gene expression score between each group to yield log-ratios. Differentially
expressed genes were identified if they satisfied the following criteria: ANOVA for
the difference of mean expression between at least two groups with a significance
level of p<0.001, Westfall-Young corrected p-value<0.05 for the set, and fold change
between at least two groups that lie above 2.8 or below –2.8 (which is represented by
log-ratios of 1.5 or -1.5 because all scores were based in log 2).
Gene clustering profiles were generated using hierarchical and k-means
algorithms. For Hierarchical clustering profile, each gene was subtracted from the

120
mean of E16 (baseline) and clustered based on the average Euclidean correlation
(dChip 3.0).

Semi-quantitative real-time RT-PCR
Semi-quantitative RT-PCR was performed to confirm the data obtained from
oligonucleotide microarray as described (Chan et al., 2004) with minor
modifications. Two hundred nanograms of RNA from sorted cells treated for DNA
contamination was reverse transcribed with iScript cDNA synthesis kit (Bio-Rad) at
reaction conditions as follows: 25
o
C for 5 min; 42
o
C for 30 min; 85
o
C for 5 min.
Primers were designed based on NCBI sequences using Primer Express software
(Applied Biosystems) and confirmed for uniqueness by NCBI BLAST. Real-time
PCR reactions were performed with cDNA template, iQSuperMix with SYBR green
(Bio-Rad), and 165 nM of gene specific primers at reaction conditions as follows:
50
o
C for 2 min, 95
o
C for 10 min, and 40 cycles at 95
o
C for 15 sec with annealing and
extension at 60
o
C for 1 min. Ribosomal L-32 was used as reference (housekeeping)
gene and levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were
monitored as secondary confirmation. Transcripts were analyzed in triplicates per
sample and 3 samples per transcript. Fluorescent levels of PCR products were
monitored by iCycler (Bio-Rad).
Fourteen transcripts defined as differentially expressed by microarray
analysis were selected for semi-quantitative PCR confirmation along with L32 and
GAPDH as 2 housekeeping genes. The primers for each genes are as listed 5’ to 3’:

121
GGGTGCGGAGAAGGTTCAA and TGGTTTTCTTGTTGCTCCCTA for L32;
TGCCAAGTATGATGACATCAAGAA and GCCCAAGATGCCCTTCAGT for
GAPDH; AGAGGCGGGTCCCTGACT and TCTTGCACACGCAGGGTAAC for
Ly6/neurotoxin 1; CACTGGCCTTTGAGCACATCT and
TGGCCTTGGCTTTGAGACTAG for sortilin-related receptor, LDLR class A
repeats-containing; GCGAGTGGCAGTGTGATGAG and
CTGCACGCTTCCTTTCACTTC for synaptotagmin-like 2;
GCCAGGGCAGGAAACCA and ATAGGCTGTCATCCAGCACATG for guanine
nucleotide binding protein (G protein), gamma 4 subunit;
GACCAGATCAAGGAGGCAATG and TTGGGTCTTCGGGCGATT for
tachykinin 1; CCATGACTGATAACGTCCAATGAC and
TCTGCCTCTCTGGATGCAAA for synaptoporin; CGGAGTCAGGCCAACATGA
and TCTCCTGACGTCTTGTCTTTTCC for Purkinje cell protein 4;
ACCCAGCAGGTGGACAGTATCT and ACCAAGCGAGCGGCTTTT for
nephronectin; GCACGTGTCCAGTGTTGCAT and
GCAAATCAGCCTCAGAATCATCT for cyclin-dependent kinase-like 2 (CDC2-
related kinase); GCTCCTCGTCGTCATCTTCCT and TCTGGCCTGGCCCCTTAG
for bromodomain containing 2; GGTGGAAGAAATCCAGGACTAACA and
AGTGCTGGTGGGACTGCAA for axotrophin; GGCCCTTGCTCCCTTCA and
TGAGGGCCCAAACCCATT for early growth response1;
AACCAGCGGTACTACCTTAAGCA and GGCCACATCCACCTCCAA for
special AT rich; and AGTGCCCCCCCTTTGGT and

122
GCACCTCATAGTGGGAATTTCAT for Kruppel-like factor 9. The cycle when
the fluorescent product reaches the set threshold, (C
T
) was determined. The
efficiency of primer pairs for each gene was compared to L-32 at a dilution range of
the same pool of cDNA template. The ∆C
T
on the y-axis was plotted against log
(cDNA) to generate a line with a slope <0.01 to verify the efficiency of amplification
of the gene of interest to be similar to that of L-32. Relative multiples of change in
expression was determined by calculating ∆∆C
T
. An ANOVA was used to test for
statistically significance differences between the time points using α=0.05 with a p<
0.05 as significant.

In situ Hybridization
In situ hybridization analysis was performed in E16, E18, P0, and P5 retinal
sections. Probe templates were cloned from reverse transcribed cDNA from purified
P5 RGC into pBluescript II (Stratagene, Santa Clara). Following primer sequences
include the restriction endonuclease sequence for BamHI and EcoRI sites (in small
letters) that were used for cloning, followed by sequences from the genes of interest:
cgggatccAGGAAGCTGGGTCAGCG and ggaattcACGTGTCCGTCCATGC for
Ly6/neurotoxin 1; cgggatccCGGAAAACGACCACGTTCT and
ggaattcGCGAAGCCGACACCCA for sortilin-related receptor, LDLR class A
repeats-containing; cgggatccTGGCGTGATGCCGACA and
cccaagcttGGTTACAGGAGTCCGGGT for synaptotagmin-like 2;
cgggatccCCCAATGGAGGCGGTAAC and

123
ggaattcATCTCGTGCCCAGTAGTGAA for guanine nucleotide binding protein (G
protein), gamma 4 subunit; cgggatccCAATGCCGGAGCCCTT and
ggaattcTGGGGAGGTCACCACATTAG for tachykinin1;
cgggatccACTGGCTGGCATTCCTC and ggaattCCCAGGCCTTTGTGTCT for
synaptoporin; cgggatccACAAAGTGCCGGAGCG and
ggaattcAGGGGGCATAAATACTATGGGTT for Purkinje cell Protein 4;
cgggatccTTCCACGGCAGCCGAC and ggaattcGTGTGCCCGAGTGCAG for
nephronectin; cgggatccTCGGGCGGGTTAGACG and
ggaattcGTGGCAACTGGCCTAGC for cyclin-dependent kinase like 2;
cgggatccAAGCTGGGTCGAGTAGTACA and
ggaattcCTAGATGCACTGAGACGGG for Bromodomain containing 2;
cgggatccGTGAGGGCAGGGCAGC and ggaattcGCCCCAAGGGAGTCACC for
Axotrophin; cgggatccCACGCCTTGCCGATGG and
ggaattcGCAATAGAGCGCATTCAATGTGTT for early growth response1; and
cgggatccAAGACGCCACAATGAATGACT and ggaattcAGGTGGGAACCCCGAG
for Kruppel-like factor 9. Cloned plasmid sequences were confirmed against the
NCBI gene sequence database. The cloned Tachykinin 1 fragment was missing a 10
nucleotide sequence between nucleotide 320 and 364 as compared to the NCBI
sequence. RNA probes incorporating digoxigenin-UTP (DIG-UTP, Roche Applied
Biosciences) were generated with using the T7 (sense-RNA) or T3 promoter (anti-
sense RNA) maxiScript (Ambion).
Mice at indicated times were sacrificed, the eyes enucleated, and fixed in 4%

124
paraformaldehyde in phosphate buffered solution (PBS) for 1 hr. Sections for all 4
time-points were processed simultaneously for each gene. The eyecups were rinsed
three times with PBS, hemisected, embedded, and fast frozen in OCT (VWR).
Eyecups were sectioned with a cryostat (Leica) to 7 µm. Slides were warmed to
55
o
C for 30 min and dried at 65
o
C for 10 min. Sections were fix in 4% PFA for 20
min, rinsed in DEPC-PBS twice for 5 min, and incubated in protease K solution (50
µg/ml of 50 mM Tris-HCl pH 7.5, 5 mM EDTA) for 3 min. Sections were washed
in DEPC-PBS for 5 min, fix in 4% PFA for 15 min, rinsed in DEPC-water,
acetylated in 0.25% acetic anhydride in 0.1M triethanolamine (TEA), pH=8.0 for 10
min, and washed in DEPC-PBS for 5min. Sections were incubated in hybridization
solution (50% formamide, 5X SSC, 0.3 mg/ml yeast tRNA, 100 ug/ml heparin, 1X
Denhardt’s, 0.1% Tween 20, 5 mM EDTA, and 0.1% CHAPS) at 65
o
C for 2 hrs and
then hybridized with 1 µg/ml of probe overnight. Sections were washed in 0.2X
SSC at 65
o
C for 15 min then 30 min twice.
Antibody binding mediated visualization of ISH probes. The anti-DIG
conjugated to alkaline phosphatase (AP) antibody was purified prior to use. Five
mg/ml mouse embryo powder was dissolved in PBT solution (PBS, 2 mg/ml BSA,
0.1% Triton X-100) at 70
o
C for 30 min, vortexed for 10 min, cooled on ice prior to
the addition of 12.5 µl lamb serum and 2.5 µl anti-DIG Fab fragment (Roche
Applied Research) and mix at 4
o
C for 1 hr. The solution was centrifuged at 13,000 g
for 10 min at 4
o
C, and the supernatant was transferred to a fresh tube. PBT and lamb
serum was added to a final volume of 5 ml with 20% lamb serum. Sections were

125
washed in PBT for 20 min twice and non-specific binding was blocked with 20%
lamb serum in PBT for 30 min. Slides were placed in a humidified chamber with
350 µl of purified anti-DIG-AP antibody at room temperature for 1 hr. Sections
were washed in PBT for 20 min three times. Sections were equilibrated with AP
buffer (100 mM Tris pH 9.5, 50 mM MgCl
2
, 100 mM NaCl, 0.1% Tween 20 and 5
mM levamisole) for 5 min twice. Sections were developed with 1 µl NitroBlue
Tetrazolium (NBT) and 3.5 µl Bromo-Chloro-Indolyl Phosphate (BCIP) per ml of
AP buffer, fresh reagents were added 8-12 hr later as needed. Sections were washed
in PBS twice and fixed in 4% PFA for 30 min and mounted.

Transcript Analysis of Cultured RGCs
RNA was isolated from equal numbers of purified E16 and E18 RGCs that
had been cultured for 48 hrs as described, using Trizol. Transcript levels from 3-5
separate samples were analyzed using semi-quantitative RT-PCR. Relative multiples
of change in expression was determined by calculating ∆∆C
T
as described above. A
student T-test was used to test for statistically significance differences between E16
and E18 using α=0.05 with a p< 0.05 as significant.

Transfection of Cultured RGCs
Linear fragments of cesium chloride purified CMV promoter sequence linked
to DsRed and poly A signaling tail were introduced into 24 hr old cultures of purified
E16 RGCs with Lipofectamine 2000 (Invitrogen) following their recommendations.

126
Optimization of transfection protocol was attempted with a DNA (μg):
Lipofectamine 2000 (µl) ratio of 1:0, 1:0.1, 1:0.5, 1:0.8, 1:1, and 1:2. To visualize
the RGCs, cells were loaded with 2 µM of the vital dye calcein-AM (Molecular
Probes) for 1 hr. Confocal photomicrographs were taken at 24 hrs and 48 hrs post
transfection.

Results
Purification of RGCs
To precisely define the molecular changes that occur within retinal ganglion
cells (RGCs) during development, it is necessary to quickly obtain pure and viable
cells. To purify RGCs from neuroretinas at various developmental stages, we
adopted a method of micrometal bead-conjugated antibody cell separation system,
using anti-CD90 antibody that recognizes a RGC specific marker known as Thy1.2
(Wu et al., 2003) Purity of RGCs was then analyzed by flow cytometry after cells
were immunolabeled with FITC-conjugated anti-CD90. The method consistently
yielded RGCs of high purity, with an average of 96.5 ± 3.1% (mean ± standard
deviation, n=3) CD90
+
cells (Figure 3.1), while the IgG2b isotype control used in
place of the FITC conjugated antibody exhibited only 1.0% autofluorescence
(Figure 3.2).
Purified CD90
+
cells were then cultured, fixed, and processed for
immunohistochemistry to confirm cell viability and verify cell identity. Virtually, all
of the purified cells were co-immunostained by the RGC marker, anti-neurofilament

127

Figure 3.1. RGCs are consistently obtained at high purity. Representative flow cytometry of purified
cells co-labeled with FITC-conjugated anti-CD90 and micrometal beads-conjugated anti-CD90
antibodies isolated by magnetic assisted cell sort. All settings were preserved from the negative
isotype control (Figure 2) for direct comparison. (a) Forward scatter (FS on the y-axis)
corresponding to cell size and side scatter (SS on the x-axis) corresponding to cell granularity with
gates on all viable cells for analysis in (b) and (c). (b) Graph of two-color analysis with
predominantly FITC
+
cells. PE (y-axis) is not being analyzed. (c) A histogram of data in (b) with
98.6% positive cell. An average 96.5%± 3.1% (mean ± standard deviation, n=3) cell are CD90
positive.

Figure 3.2. A very low level of auto-fluorescence in isotype negative control. Representative flow
cytometry of purified cells labeled with micrometal beads-conjugated with CD90 and stained with
IgG2 isotype negative control in place of FITC-conjugated CD90 antibody. Samples were use to
define the baseline of sensitivity for analysis.
(c)
(a)
(b)
Figure 3.1 Figure 3.2
(c)
(a)
(b)

128

Figure 3.3

Figure 3.3. Co-localization of cells stained with antibodies recognizing neurofilament (an RGC
marker) and β-actin in nearly all cells. Representative confocal images of highly purified CD90
+
cells
cultured for 48 hours were processed for immunohistochemistry. (a) Antibody to β-actin, a structural
protein in all cells, stains red. (b) Antibody to neurofilament stains most cells green. (c) Bright-field
showing cells. (d) Merge image co-localization of all fields.
(a)
(b)
(c) (d)
(a)
(b)
(c) (d)

129
Defining the Timing When RGCs Lose Axon Growth Capacity
Immature RGCs, like most CNS neurons, have the capacity to elongate and
regenerate their axons, while this ability is lost following maturation. To define
when in development RGC axons lose their intrinsic capacity to grow, RGCs of
various ages (E16 - P5) were purified and cultured for 48 hrs (Figure 3.4a). With
the use of a vital dye, we were able to analyze axon growth in living RGCs only.
Cells visually assessed for health and survival indicated similar survival rates at E16
and E18, ~90% at P0, and ~60% at P5. Cells with neurites that extended ≥ 3 times
their body length were counted with a total of 3,400 to 6,000 live cells from multiple
litters per time point (Figure 3.4b). We found that from E16 to E18, there was a
drastic decline in numbers of RGCs that extended neurites over 3 times of body
length, dropping from 33.0 ± 4.5% of cells counted at E16 to 9.6 ± 3.4% at E18
(Figure 3.4c). Thus, RGCs undergo a dramatic loss of neurite growth capacity
between E16 and E18.

Assessment of Small Sample Amplification
The use of highly purified viable cells restricts the number of RGCs available
per animal and therefore limits the amount of initial mRNA. To test the feasibility of
our proposal, we quantified the amount of total RNA recovered from purified cells.
A litter of five mice can consistently yield around 6E4 cells and 260 ng of total RNA
can be recovered via Trizol (Invitrogen) extraction based on absorbance at 260nm.
Conventional methods of generating biotin-labeled probes for oligonucleotide

130
Figure 3.4

Age
N
litters
Total Cells
Counted
E16 4 5967
E18 3 6019
P0 3 5714
P5 4 3469

P5
E16
E18
P0
P5
E16
E18
P0
(a)
P5
E16
E18
P0
P5
E16
E18
P0
(a)
(b)

131
Figure 3.4. Continued

Figure 3.4. Developmental stage progression marked by dramatic decrease in RGC neurite growth.
(a) Representative confocal images of purified RGCs cultured in vitro for 48 hours after being loaded
with the fluorescence vital dye calcein-AM at E16, E18, P0, and P5. A narrowly defined time
window between E16 and E18 when RGCs lose their neurite growth ability marks a critical switch in
developmental. (b) Graph of mean percent and standard deviation of RGCs with neurites that extend
≥3 times their body length of total cell counted. (c) Number of samples and total number of RGCs
counted at each time point. A total of between 3,400 and 6,000 cells were counted at each time point.
Growth Curve of Ganglion Cells
0
5
10
15
20
25
30
35
40
E16 E18 P0 P5
Age
% Cell with Growing Axons
Wild Type
(c)

132
arrays require 5 ug of total mRNA. Based on these numbers, 97 mice would be
needed for one sample. Thus, for triplicate replicates, we would need 288 mice for
each time point.
A new method of generating probes from small samples was assessed to
circumvent this challenge of obtaining enough starting material. By incorporating an
additional amplification step to the conventional method of generating
oligonucleotide probes, we limit the number of animals needed. The use of 2 cycles
of in vitro transcription (IVT) was assessed for both reliability and reproducibility.
To compare the effects of an additional IVT step as well as to evaluate the
effectiveness of different kits, biotin labeled probes were generated from
MessageAMP (Ambion) and RiboAMP (Arcturus) at different amounts of initial
total retinal RNA as indicated using the conventional 1 or 2 IVT and then hybridized
to oligonucleotide microarray chips (Affymetrix Inc.). Pearson correlation
coefficients represent the linear relationships (r-values) between 2 samples were
calculated: r=1.0 is a perfect positive correlation, r=-1.0 as a perfect negative
correlation, and r=0 as no linear relation. The r-value is a numerical value to
globally assess how one experimental condition related to another. A scatter plot of
the intensity of each gene was generated (Figure 3.5).
The MessageAMP method represented more consistent generation of
probes within 2 cycles of IVT: 200 ng 2 IVT: 60 ng 2 IVT (r=0.98); 60 ng 2 IVT: 20
ng 2 IVT (r=0.99); 200 ng 2 IVT: 20 ng 2 IVT (r=0.98) while the RiboAMP method
varied significantly within 2 IVT (r=0.74-0.96). Analysis of 2ug 1 IVT:2 IVT

133

Figure 3.5

Figure 3.5. Scatter plot of gene intensity and correlation coefficients (r-value in parenthesis
in the upper left corner of each quadrant). The x-axis corresponds to a particular transcript’s
intensity from the experiment described on the top of the column. The y-axis represents the
same transcript’s intensity from the experiment described at the end of the row to the right.
All 12,234 transcripts are represented. The top 4 experiments (black) used the MessageAMP
system while the bottom 3 (orange) used the RiboAMP system. 2 IVT MessageAMP gives
consistently reproducible results (0.99-0.98) as does RiboAMP (0.96). MessageAMP 1 IVT
results a higher fidelity with 2 IVT (0.85-0.89) than RiboAMP (0.74-0.75). 1 IVT labeling
between MessageAMP and RiboAMP is consistent (0.92). Because of the higher fidelity rate,
we used the MessageAMP method.
(0.86)
(0.92) (0.89) (0.88) (0.90)
(0.75)
(0.74)
(0.96)
(0.85)
(0.67)
(0.99) (0.98)
(0.99)
(0.86)
(0.92) (0.89) (0.88) (0.90)
(0.75)
(0.74)
(0.96)
(0.85)
(0.67)
(0.99) (0.98)
(0.99)

134
(r=0.85-0.89) indicates that samples undergoing 2 IVT labeling should not be
directly compared with data obtained from 1 IVT. The comparisons will be distorted.
The lack of perfect correlation between 2 IVT and 1 IVT underscores the need to
verify differential gene expression data by another method, in our case, semi-
quantitative real-time PCR. Though the conventional 1 IVT provides a better
representation of the original pool of transcripts, there is a reasonable correlation
between 1 and 2 IVT. Two-step amplification allows us to use a reasonable number
of animals (one litter) per sample given the limited amount of total RNA obtained
from the purified RGCs. The need to reverse transcribe twice for the 2 IVT method
most likely contributes to some of the error because probes generated with 2IVT are
shorter than probes from 1 IVT. 2 IVT amplification using RiboAMP did not
correlate well with 1 IVT (0.74 and 0.75) using the same kit. There is a better
correlation with the MessageAMP method (0.86- 0.89) for 1 IVT to 2 IVT. We
determined that the MessageAMP system would be better for our requirements than
RiboAMP.
To see whether a better correlation between 1 IVT and 2 IVT can be achieved,
the length of IVT probe generation was varied from 6 h to 13 h (Table 3.1). 5 ug of
initial total retinal RNA was labeled by one cycle of IVT for 6 h or 13 h, while 150
ng of initial total RNA was transcribed for 6 h or 13 h for the first cycle of IVT
followed by labeling during the 13 h second cycle of IVT. The length of IVT did
not seem to skew the generation of labeled probes when 5 ug of initial RNA was
used (r=0.99) or when 150 ng of RNA was used for 2 IVT (0.98). There is a

135

Table 3.1

Table 3.1: Comparison of the effects of alternatives in the methods of generating IVT labeled probes.
The same pool of RNA was used in a series of 1 or 2 cycles of IVT and the Pearson correlation
coefficient was calculated (r-value) based on all of the data from a high density microarray. The
amounts of initial starting total mRNA is listed followed by the number of hours IVT. If 2 cycles of
IVT was used, a second number follows.

Amount of total mRNA
and length of IVT times r-value
5ug 6h vs 5ug 13h 0.993
5ug 6h vs150ng 6,13h 0.886
5ug 6h vs 150ng 13,13h 0.884

5ug 13h vs 150ng 6, 13h 0.884
5ug 13h vs 150ng 13,13h 0.887

150ng 6, 13h vs 150ng 13,
13h 0.981

136
discrepancy between 1 cycle of IVT and 2 (r=0.88) and therefore analysis must be
done with samples undergoing the same labeling technique. Finally, we determined
that using as little as 150 ng of total RNA, we can produce reliable biotin labeled
probes to study differential gene expression by microarray analysis.

Intact RNA is Consistently Isolated
To ensure the reliability of our data, the technique of cell purification coupled
with RNA isolation was assessed. Cells undergoing significant stress degrade their
mRNA as they prepare a compensatory response. Trizol isolation of total RNA
immediately followed cell separation was assessed for amount using Ribogreen
(Molecular Probes) and integrity with Bioanalyzer (Agilent) made available to us
through our collaboration with Dr. Mel Simon’s lab at Caltech. Picograms of total
RNA analyzed for 28S and 18S presence and stability consistently yielded a 28:18S
ratio>1.5 (Figure 3.6) of intact RNA samples from this method of cell purification.

Characterizing Molecular Changes with cDNA Microarray
The fact that RGCs cultured under a similar condition display drastic change
in their ability to extend axons suggests a cell autonomous mechanism in neurite
growth regulation. However, molecular signals mediating the transcriptional control
of axon growth during development are not well understood. To identify
differentially expressed genes that could account for the differences in axon growth,
we used an unbiased differential gene expression analysis of Affymetrix cDNA

137

Figure 3.6

Figure 3.6: Bioanalyzer analysis of the integrity of total RNA isolated from purified RGCs
prior to labeling. The top graph show fluorescence staining binding to nucleic acids (y-axis)
as it flows through the capillary system (time on x-axis). A represent gel of the same data is
to the right. Samples with 28S:18S ratio above 1.5 from the output data were selected for
labeling. This represented intact RNA.
18S
28S
Fluorescence
Time (seconds)
0
5
10
15
20
25
30
19 24 29 34 39 44 49 54 59 64 69
Total-RNA-Pico_01699_2003-07-07 Sample 1
Fragment Name Start_Time(secs) End_Time(secs) Area %_of_total_Area
1 18S 40.50 42.40 29.55 16.67
2 28S 46.00 48.80 51.33 28.96
RNA Area 177.20
RNA Concentration(pg/ul) 1,233.00
rRNA Ratio [28S / 18S] 1.74

138
expression array with over 22,600 transcripts in developing RGCs before and after
the loss of their axon growth capacity were compared. Filtering out the noise and
non-expressed genes resulted in 7,945 detected genes. An ANOVO test for
differential mean expression with a p<0.05 resulted in 1,408 genes. Genes
classically defined as biologically significant change by >2 or <0.5-fold by
conventional microarray analysis yielded 224 genes. Linear amplification of small
samples slightly skews expression ratios (Luzzi et al., 2001) a >2.83 or <0.38 fold
change resulted in the 71 most differentially expressed genes.
The analysis yielded 71 transcripts (>2.83 or <0.38 fold change, Figure 3.7).
A large portion of these genes are associated with growth factors/receptors/signaling
molecules/cell cycle factors (15.5%), transcriptional regulators and other nuclear
factors (15.5%), cell surface/adhesion/matrix/cytoskeletal proteins/membrane
proteins (22.5%), hormones/proteases/secretory apparatus/extracellular matrix
components (8%), and metabolism/maintenance (15.5%). The majority of the
differentially expressed transcripts (53%) increased as RGCs matured. Many
transcripts (19.7%) identified as differentially expressed were novel genes that have
unknown functions. The transcripts identified as differentially expressed are listed in
Table 3.2.
Transcripts with similar trends in expression may be indicative of an
induction or a suppression of particular unique molecular pathways. Hierarchical
and K-means algorithms were used to group transcripts that exhibited similar
patterns of expression to identify novel gene relationships from E16 to P5. For

139

Figure 3.7

Figure 3.7. Functional categorization of transcripts defined as differentially expressed between E16
and P5 (ANOVA corrected p≤0.05 with at least 2^±1.5 fold difference). Seventy-one transcripts of
26,000 genes screened were identified as differentially expressed.
Categories of Differentially Expressed Genes
11
11
16 8
11
14
Growth factors, receptors and signaling molecules and Cell cycle
Transcriptional regulators and other nuclear factors
Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
Hormones and proteases and secretory apparatus (Extracellular matrix)
Metabolism/maintenance
Others

140
Table 3.2
Accession
number Gene name
E16 mean-
E16 mean
E18 mean-
E16 mean
P0 mean-
E16 mean
P5 mean-
E16 mean
Corrected
P-Value
K-means
cluster

Growth factors, receptors and signaling molecules and Cell cycle
1417283_at Ly6/neurotoxin 1 0 0.1 0.4 1.7 0 Cluster 1
1420872_at guanylate cyclase 1, soluble, beta 3 0 0.7 0.3 1.7 0.05 Cluster 1
1425518_at
cAMP-regulated guanine nucleotide exchange
factor II 0 0.7 0.7 1.9 0 Cluster 1
1426258_at
sortilin-related receptor, LDLR class A repeats-
containing 0 0.5 0.6 2.1 0 Cluster 1
1455462_at adenylate cyclase 2 0 0.6 0.5 1.8 0 Cluster 1
1424638_at cyclin-dependent kinase inhibitor 1A (P21) 0 0 0.3 1.1 1.51 Cluster 1
1417943_at
guanine nucleotide binding protein (G protein),
gamma 4 subunit 0 0.8 0.6 2.4 0 Cluster 2
1449229_a_at
cyclin-dependent kinase-like 2 (CDC2-related
kinase) 0 1.1 1.5 2.5 0.05 Cluster 2
1416783_at tachykinin 1 0 1.5 1.8 2.2 0 Cluster 3
1422474_at phosphodiesterase 4B, cAMP specific 0 2.0 1.9 1.7 0 Cluster 3

Table 3.2: Differentially expressed transcripts grouped into functional categories. The mean levels of induction (positive number) and reduction
(negative numbers) compared to the mean E16 levels of that gene are in log base 2. The Westfall-Young corrected p-values are as indicated as well as
the K-means expression profile.

141

1416034_at CD24a antigen 0 -0.3 -0.2 -2.7 0 Cluster 4
Transcriptional regulators and other nuclear factors
1448954_at nuclear receptor interacting protein 3 0 0.3 0.3 2.1 0 Cluster 1
1456341_a_at Kruppel-like factor 9 0 2.1 2.3 4.0 0 Cluster 3
1417679_at growth factor independent 1 0 -2.1 -2.6 -2.6 0 Cluster 4
1422610_s_at insulin-like growth factor 2, binding protein 3 0 -0.0 -1.1 -2.3 0 Cluster 4
1416723_at transcription factor 4 0 2.0 1.3 0.7 0 Cluster 5
1417065_at early growth response 1 0 1.5 1.6 0.9 0.03 Cluster 5
1420390_s_at zinc finger protein 354A 0 1.6 0.8 -0.0 0.01 Cluster 6
1434336_s_at REST co-repressor 0 1.6 0.1 -0.2 0 Cluster 6
1436182_at special AT-rich sequence binding protein 1 0 1.83 0.90 0.4 0 Cluster 6
1437210_a_at bromodomain containing 2 0 -2.54 -1.24 -0.2 0.05 Cluster 7
1456194_a_at RNA binding protein regulatory subunit 0 -2.23 -1.22 0.0 0 Cluster 8

Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
1417373_a_at tubulin, alpha 4 0 0.3 0.8 2.0 0 Cluster 1
1452298_a_at myosin Vb 0 0.5 0.7 1.7 0 Cluster 1
1452107_s_at nephronectin 0 0.2 1.0 2.7 0 Cluster 2
1416361_a_at dynein, cytoplasmic, intermediate chain 1 0 1.3 1.1 2.7 0 Cluster 2
1417672_at
solute carrier family 4, sodium bicarbonate
cotransporter-like, member 10 0 0.8 1.4 3.0 0 Cluster 2

Table3.2. Continued

142

1421594_a_at synaptotagmin-like 2 0 0.9 1.2 2.3 0 Cluster 2
1435495_at myosin binding protein H 0 0.2 0. 5 3.6 0 Cluster 2
1450121_at
Mus musculus sodium channel 27 mRNA
fragment. 0 0.6 1.5 2.7 0 Cluster 2
1452981_at contactin 1 0 1.0 1.5 2.0 0 Cluster 2
1415904_at lipoprotein lipase 0 2.3 2.7 2.4 0 Cluster 3
1423640_at synaptoporin 0 1.1 1.9 2.5 0 Cluster 3
1436836_x_at calponin 3, acidic 0 -0.5 -1.2 -1.9 0 Cluster 4
1455883_a_at leucine rich repeat transmembrane neuronal 1 0 2.3 1.1 1.6 0 Cluster 5
1424721_at
similar to Microfibril-associated glycoprotein 3
precursor 0 -1.8 -0.9 -0.1 0.02 Cluster 7
1448113_at stathmin 1 0 -1.7 -0.8 -0.5 0.03 Cluster 7
1419734_at actin, beta, cytoplasmic 0 -1.8 -2.0 -0.0 0.03 Cluster 8

Hormones and proteases and secretory apparatus(Extracellular matrix)
1416114_at SPARC-like 1 (mast9, hevin) 0 0.6 0.6 1.7 0 Cluster 1
1417702_a_at histamine N-methyltransferase 0 0.5 1.1 1.7 0 Cluster 1
1426785_a monoglyceride lipase 0 0.3 0.8 2.1 0 Cluster 1
1450708_at secretogranin II 0 1.2 1.1 2.3 0 Cluster 2
1427256_at chondroitin sulfate proteoglycan 2 0 -0.7 -0.9 -2.6 0 Cluster 4

Table3.2. Continued

143

1423567_a_at
proteasome (prosome, macropain) subunit, alpha
type 7 0 -2.1 -0.8 -0.1 0.02 Cluster 7
1452035_at procollagen, type IV, alpha 1 0 -1.5 -0.9 2.0 0.05 Cluster 9
1455439_a_at lectin, galactose binding, soluble 1 0 -1.1 -0.8 2.1 0 Cluster 9

Metabolism/maintenance
1422456_at N-ethylmaleimide sensitive fusion protein 0 0.6 0.7 1.7 0 Cluster 1
1424394_at selenoprotein M 0 -0.4 0.3 1.8 0 Cluster 1
1418829_a_at enolase 2, gamma neuronal 0 0.1 1.3 2.3 0 Cluster 2
1439036_a_at ATPase, Na+/K+ transporting, beta 1 polypeptide 0 0.9 1.2 2.8 0 Cluster 2
1417799_at ATPase, H+ transporting, V1 subunit G isoform 2 0 1.2 1.3 1.7 0 Cluster 5
1435148_at ATPase, Na+/K+ transporting, beta 2 polypeptide 0 1.6 1.0 1.1 0 Cluster 5
1419499_at
glycerol-3-phosphate acyltransferase,
mitochondrial 0 1.8 1.0 0.5 0.02 Cluster 6
1422126_a_at
Nudt13 nudix (nucleoside diphosphate linked
moiety X)-type motif 13 0 1.7 0.6 0.6 0 Cluster 6
1415744_at H2-K region expressed gene 2 0 -2.1 -1.3 -0.0 0.05 Cluster 7
1456642_x_at S100 calcium binding protein A10 (calpactin) 0 -1.8 -1.3 -0.9 0 Cluster 7
1437729_at ribosomal protein L27a 0 -2.1 -1.9 -0.5 0.01 Cluster 8

Table3.2. Continued

144

Others
1420822_s_at
Mus musculus adult male stomach cDNA, RIKEN
full-length enriched library, clone:2210016L03
product:sphingosine-1-phosphate phosphatase 1,
full insert sequence.
0 0.6 0.9 1.6 0.05 Cluster 1
1423135_at thymus cell antigen 1, theta 0 0.1 0.5 1.8 0.03 Cluster 1
1451674_at Mus musculus, clone IMAGE:5368459, mRNA 0 0.7 0.6 2.0 0 Cluster 1
1454897_at expressed sequence AI450287 0 0.7 0.5 1.6 0 Cluster 1
1460214_at Purkinje cell protein 4 0 1.0 1.6 3.6 0 Cluster 2
1420955_at visinin-like 1 0 1.8 2.0 2.3 0 Cluster 3
1420191_s_at hypothetical protein MNCb-4137 0 1.3 1.6 2.3 0.03 Cluster 3
1449947_s_at RIKEN cDNA A230102L03 gene 0 -1.5 -1.2 -1.5 0.02 Cluster 4
1433737_at expressed sequence AA673513 0 1.7 1.2 1.3 0.02 Cluster 5
1452227_at RIKEN cDNA 2310045A20 gene 0 1.6 0.2 0.4 0 Cluster 6
1436362_x_at 0 -2.0 -0.8 -0.5 0 Cluster 7
1436893_a_at axotrophin 0 -2.3 -0.7 -0.2 0.05 Cluster 7
1448685_at RIKEN cDNA 2900010M23 gene 0 -2.0 -1.3 0.0 0.05 Cluster 8
1427280_at RIKEN cDNA A230052E19 gene 0 -2.1 -0.4 1.4 0.03 Cluster 9

Table3.2. Continued

145
Hierarchical clustering profile, each gene was subtracted from the mean of E16
(baseline) and clustered based on the average Euclidean correlation (dChip 3.0, Li et
al., 2001, Figure 3.8). The top of the dendrogram show that at each developmental
time, samples share an overall similar expression profile based on all of the genes
indicating a high degree of consistency. It is also an unbiased internal control to
validate our hypothesis that there is a high degree of uniformity in transcript
expression since most RGCs are doing the same thing. Genes were then grouped
into 4 groups consisting of 9 clusters with K-means algorithm based on the similar
relative levels of induction and the general pattern of expression to help identify
novel molecular pathways that may be involved in perinatal RGC development
(Avadis, Figure 3.9). The majority 80.3% of genes increased in expression.
Identified molecular factors that are implicated in axonal growth were
specifically focused on during our analysis, but they were not differentially
expressed (Table 3.3).

Confirmation of Microarray Data by Real Time RT-PCR
To validate the data obtained from cDNA microarray, new RNAs were
isolated from samples of purified RGCs for semi-quantitative real time RT-PCR.
Fourteen differentially expressed genes were selected for confirmation as well as the
housekeeping gene ribosomal protein L32 and glyceraldehyde-3-phosphate
dehydrogenase (GAPDH) at E16, 18, P0, and P5. They were, for the most part,

146

Figure 3.8. Hierarchical cluster profile of gene transcripts and samples. Clustering based on average
Euclidean correlation (dChip) with data presented in log based 2 compared to the mean of each gene
transcript at E16 (baseline). All 71 differentially expressed transcripts are represented from all
individual samples at each time point (E16, n=6; E18, n=4; P0, n=4; P5 n=4). The dendrogram on the
top groups samples that share similar transcript expression profiles. Samples within time points group
together and E16, E18, P0, and P5 are separate from each other into distinct clusters. P0 and E18 are
more similar to each other, than E16, and P5. The dendrogram on the left groups transcripts that
exhibit similar expression profiles. Expression level colors are represented as follows: red, higher
than the baseline; black, no change from baseline, and green, below baseline. Each colors in the Gene
Ontology, Protein Domain, Pathway and Chromosome columns represent different functional
descriptions based on NCBI LocusLink database
Figure 3.8
P5-2-baseline
P5-4-baseline
P5-1-baseline
P5-3-baseline
E16-1-baseline
E16-2-baseline
E16-3-baseline
E16-4-baseline
E16-5-baseline
E16-6-baseline
P0-3-baseline
P0-4-baseline
P0-1-baseline
P0-2-baseline
E18-1-baseline
E18-2-baseline
E18-3-baseline
E18-4-baseline
Gene Ontology
Protein Domain
Pathway
Chromosome
1456341_a_at, Kruppel-like factor 9
1415904_at, lipoprotein lipase
1434336_s_at, hypothetical protein 5730409O 11
1452227_at, RIKEN cDNA 2310045A20 gene
1436182_at, special AT-rich sequence binding protein 1
1420390_s_at, zinc finger protein 354A
1419499_at, glycerol-3-phosphate acyltransferase, mitochondrial
1422126_a_at, RIKEN cDNA 4933433B15 gene
1455883_a_at, RIKEN cDNA 4632401D06 gene
1416723_at, transcription factor 4
1433737_at, expressed sequence AA673513
1435148_at, ATPase, Na+ /K+ transporting, beta 2 polypeptide
1417065_at, early growth response 1
1435495_at, myosin binding protein H
1460214_at, Purkinje cell protein 4
1424394_at, selenoprotein M
1418829_a_at, enolase 2, gamma neuronal
1452107_s_at, gb:AA223007 /DB_XREF= gi:1843285 /DB_XREF= mw01g12.r1 /CLO NE= IMAG E:663334 /FE A=FLmRNA /CNT=87 /TID=Mm.205021.4 /TIER=ConsEnd /S TK=0 /UG = Mm.205021 /LL= 1142...
1417943_at, guanine nucleotide binding protein (G protein), gamma 4 subunit
1424638_at, cyclin-dependent kinase inhibitor 1A (P21)
1417702_a_at, histamine N-methyltransferase
1417283_at, Ly6/neurotoxin 1
1423135_at, thymus cell antigen 1, theta
1454897_at, expressed sequence AI450287
1420822_s_at, Mus musculus adult male stomach cDNA, RIKEN full-length enriched library, clone:2210016L03 product:sphingosine-1-phosphate phosphatase 1, full insert sequence.
1452298_a_at, myosin Vb
1455462_at, adenylate cyclase 2
1451674_at, Mus musculus, clone IMAG E:5368459, mRNA
1420872_at, guanylate cyclase 1, soluble, beta 3
1416114_at, SPARC-like 1 (mast9, hevin)
1425518_at, cAMP-regulated guanine nucleotide exchange factor II
1422456_at, N-ethylmaleimide sensitive fusion protein
1426785_s_at, monoglyceride lipase
1417373_a_at, tubulin, alpha 4
1448954_at, RIKEN cDNA A330103B05 gene
1426258_at, sortilin-related receptor, LDLR class A repeats-containing
1422474_at, phosphodiesterase 4B, cAMP specific
1420191_s_at, hypothetical protein MNCb-4137
1416783_at, tachykinin 1
1420955_at, visinin-like 1
1439036_a_at, ATP ase, Na+ /K+ transporting, beta 1 polypeptide
1450121_at, Mus musculus sodium channel 27 mRNA fragment.
1417672_at, solute carrier family 4, sodium bicarbonate cotransporter-like, member 10
1449229_a_at, cyclin-dependent kinase-like 2 (CDC2-related kinase)
1423640_at, synaptoporin
1417799_at, ATPase, H+ transporting, V 1 subunit G isoform 2
1452981_at, contactin 1
1416361_a_at, dynein, cytoplasmic, intermediate chain 1
1450708_at, secretogranin II
1421594_a_at, synaptotagmin-like 2
1419734_at, actin, beta, cytoplasmic
1437729_at, ribosomal protein L27a
1449947_s_at, RIKEN cDNA A230102L03 gene
1456642_x_at, S100 calcium binding protein A10 (calpactin)
1437210_a_at, bromodomain containing 2
1448685_at, RIKEN cDNA 2900010M23 gene
1415744_at, H2-K region expressed gene 2
1456194_a_at, RNA binding protein regulatory subunit
1436893_a_at, axotrophin
1423567_a_at, proteasome (prosome, macropain) subunit, alpha type 7
1436362_x_at, gb:AA798895 /DB_XREF= gi:2861850 /DB_XREF= vv94e07.r1 /CLO NE= IMAG E:1230084 /FE A=EST /CNT= 14 /TID= Mm.220953.2 /TIER= Stack /STK=9 /UG = Mm.220953 /UG _TITLE= Mus ...
1424721_at, similar to Microfibril-associated glycoprotein 3 precursor
1448113_at, stathmin 1
1427280_at, RIKEN cDNA A230052E19 gene
1455439_a_at, lectin, galactose binding, soluble 1
1452035_at, procollagen, type IV , alpha 1
1417679_at, growth factor independent 1
1436836_x_at, calponin 3, acidic
1422610_s_at, insulin-like growth factor 2, binding protein 3
1416034_at, CD24a antigen
1427256_at, chondroitin sulfate proteoglycan 2
P5-2-baseline
P5-4-baseline
P5-1-baseline
P5-3-baseline
E16-1-baseline
E16-2-baseline
E16-3-baseline
E16-4-baseline
E16-5-baseline
E16-6-baseline
P0-3-baseline
P0-4-baseline
P0-1-baseline
P0-2-baseline
E18-1-baseline
E18-2-baseline
E18-3-baseline
E18-4-baseline
Gene Ontology
Protein Domain
Pathway
Chromosome
1456341_a_at, Kruppel-like factor 9
1415904_at, lipoprotein lipase
1434336_s_at, hypothetical protein 5730409O 11
1452227_at, RIKEN cDNA 2310045A20 gene
1436182_at, special AT-rich sequence binding protein 1
1420390_s_at, zinc finger protein 354A
1419499_at, glycerol-3-phosphate acyltransferase, mitochondrial
1422126_a_at, RIKEN cDNA 4933433B15 gene
1455883_a_at, RIKEN cDNA 4632401D06 gene
1416723_at, transcription factor 4
1433737_at, expressed sequence AA673513
1435148_at, ATPase, Na+ /K+ transporting, beta 2 polypeptide
1417065_at, early growth response 1
1435495_at, myosin binding protein H
1460214_at, Purkinje cell protein 4
1424394_at, selenoprotein M
1418829_a_at, enolase 2, gamma neuronal
1452107_s_at, gb:AA223007 /DB_XREF= gi:1843285 /DB_XREF= mw01g12.r1 /CLO NE= IMAG E:663334 /FE A=FLmRNA /CNT=87 /TID=Mm.205021.4 /TIER=ConsEnd /S TK=0 /UG = Mm.205021 /LL= 1142...
1417943_at, guanine nucleotide binding protein (G protein), gamma 4 subunit
1424638_at, cyclin-dependent kinase inhibitor 1A (P21)
1417702_a_at, histamine N-methyltransferase
1417283_at, Ly6/neurotoxin 1
1423135_at, thymus cell antigen 1, theta
1454897_at, expressed sequence AI450287
1420822_s_at, Mus musculus adult male stomach cDNA, RIKEN full-length enriched library, clone:2210016L03 product:sphingosine-1-phosphate phosphatase 1, full insert sequence.
1452298_a_at, myosin Vb
1455462_at, adenylate cyclase 2
1451674_at, Mus musculus, clone IMAG E:5368459, mRNA
1420872_at, guanylate cyclase 1, soluble, beta 3
1416114_at, SPARC-like 1 (mast9, hevin)
1425518_at, cAMP-regulated guanine nucleotide exchange factor II
1422456_at, N-ethylmaleimide sensitive fusion protein
1426785_s_at, monoglyceride lipase
1417373_a_at, tubulin, alpha 4
1448954_at, RIKEN cDNA A330103B05 gene
1426258_at, sortilin-related receptor, LDLR class A repeats-containing
1422474_at, phosphodiesterase 4B, cAMP specific
1420191_s_at, hypothetical protein MNCb-4137
1416783_at, tachykinin 1
1420955_at, visinin-like 1
1439036_a_at, ATP ase, Na+ /K+ transporting, beta 1 polypeptide
1450121_at, Mus musculus sodium channel 27 mRNA fragment.
1417672_at, solute carrier family 4, sodium bicarbonate cotransporter-like, member 10
1449229_a_at, cyclin-dependent kinase-like 2 (CDC2-related kinase)
1423640_at, synaptoporin
1417799_at, ATPase, H+ transporting, V 1 subunit G isoform 2
1452981_at, contactin 1
1416361_a_at, dynein, cytoplasmic, intermediate chain 1
1450708_at, secretogranin II
1421594_a_at, synaptotagmin-like 2
1419734_at, actin, beta, cytoplasmic
1437729_at, ribosomal protein L27a
1449947_s_at, RIKEN cDNA A230102L03 gene
1456642_x_at, S100 calcium binding protein A10 (calpactin)
1437210_a_at, bromodomain containing 2
1448685_at, RIKEN cDNA 2900010M23 gene
1415744_at, H2-K region expressed gene 2
1456194_a_at, RNA binding protein regulatory subunit
1436893_a_at, axotrophin
1423567_a_at, proteasome (prosome, macropain) subunit, alpha type 7
1436362_x_at, gb:AA798895 /DB_XREF= gi:2861850 /DB_XREF= vv94e07.r1 /CLO NE= IMAG E:1230084 /FE A=EST /CNT= 14 /TID= Mm.220953.2 /TIER= Stack /STK=9 /UG = Mm.220953 /UG _TITLE= Mus ...
1424721_at, similar to Microfibril-associated glycoprotein 3 precursor
1448113_at, stathmin 1
1427280_at, RIKEN cDNA A230052E19 gene
1455439_a_at, lectin, galactose binding, soluble 1
1452035_at, procollagen, type IV , alpha 1
1417679_at, growth factor independent 1
1436836_x_at, calponin 3, acidic
1422610_s_at, insulin-like growth factor 2, binding protein 3
1416034_at, CD24a antigen
1427256_at, chondroitin sulfate proteoglycan 2
P5-2-baseline
P5-4-baseline
P5-1-baseline
P5-3-baseline
E16-1-baseline
E16-2-baseline
E16-3-baseline
E16-4-baseline
E16-5-baseline
E16-6-baseline
P0-3-baseline
P0-4-baseline
P0-1-baseline
P0-2-baseline
E18-1-baseline
E18-2-baseline
E18-3-baseline
E18-4-baseline
Gene Ontology
Protein Domain
Pathway
Chromosome
1456341_a_at, Kruppel-like factor 9
1415904_at, lipoprotein lipase
1434336_s_at, hypothetical protein 5730409O 11
1452227_at, RIKEN cDNA 2310045A20 gene
1436182_at, special AT-rich sequence binding protein 1
1420390_s_at, zinc finger protein 354A
1419499_at, glycerol-3-phosphate acyltransferase, mitochondrial
1422126_a_at, RIKEN cDNA 4933433B15 gene
1455883_a_at, RIKEN cDNA 4632401D06 gene
1416723_at, transcription factor 4
1433737_at, expressed sequence AA673513
1435148_at, ATPase, Na+ /K+ transporting, beta 2 polypeptide
1417065_at, early growth response 1
1435495_at, myosin binding protein H
1460214_at, Purkinje cell protein 4
1424394_at, selenoprotein M
1418829_a_at, enolase 2, gamma neuronal
1452107_s_at, gb:AA223007 /DB_XREF= gi:1843285 /DB_XREF= mw01g12.r1 /CLO NE= IMAG E:663334 /FE A=FLmRNA /CNT=87 /TID=Mm.205021.4 /TIER=ConsEnd /S TK=0 /UG = Mm.205021 /LL= 1142...
1417943_at, guanine nucleotide binding protein (G protein), gamma 4 subunit
1424638_at, cyclin-dependent kinase inhibitor 1A (P21)
1417702_a_at, histamine N-methyltransferase
1417283_at, Ly6/neurotoxin 1
1423135_at, thymus cell antigen 1, theta
1454897_at, expressed sequence AI450287
1420822_s_at, Mus musculus adult male stomach cDNA, RIKEN full-length enriched library, clone:2210016L03 product:sphingosine-1-phosphate phosphatase 1, full insert sequence.
1452298_a_at, myosin Vb
1455462_at, adenylate cyclase 2
1451674_at, Mus musculus, clone IMAG E:5368459, mRNA
1420872_at, guanylate cyclase 1, soluble, beta 3
1416114_at, SPARC-like 1 (mast9, hevin)
1425518_at, cAMP-regulated guanine nucleotide exchange factor II
1422456_at, N-ethylmaleimide sensitive fusion protein
1426785_s_at, monoglyceride lipase
1417373_a_at, tubulin, alpha 4
1448954_at, RIKEN cDNA A330103B05 gene
1426258_at, sortilin-related receptor, LDLR class A repeats-containing
1422474_at, phosphodiesterase 4B, cAMP specific
1420191_s_at, hypothetical protein MNCb-4137
1416783_at, tachykinin 1
1420955_at, visinin-like 1
1439036_a_at, ATP ase, Na+ /K+ transporting, beta 1 polypeptide
1450121_at, Mus musculus sodium channel 27 mRNA fragment.
1417672_at, solute carrier family 4, sodium bicarbonate cotransporter-like, member 10
1449229_a_at, cyclin-dependent kinase-like 2 (CDC2-related kinase)
1423640_at, synaptoporin
1417799_at, ATPase, H+ transporting, V 1 subunit G isoform 2
1452981_at, contactin 1
1416361_a_at, dynein, cytoplasmic, intermediate chain 1
1450708_at, secretogranin II
1421594_a_at, synaptotagmin-like 2
1419734_at, actin, beta, cytoplasmic
1437729_at, ribosomal protein L27a
1449947_s_at, RIKEN cDNA A230102L03 gene
1456642_x_at, S100 calcium binding protein A10 (calpactin)
1437210_a_at, bromodomain containing 2
1448685_at, RIKEN cDNA 2900010M23 gene
1415744_at, H2-K region expressed gene 2
1456194_a_at, RNA binding protein regulatory subunit
1436893_a_at, axotrophin
1423567_a_at, proteasome (prosome, macropain) subunit, alpha type 7
1436362_x_at, gb:AA798895 /DB_XREF= gi:2861850 /DB_XREF= vv94e07.r1 /CLO NE= IMAG E:1230084 /FE A=EST /CNT= 14 /TID= Mm.220953.2 /TIER= Stack /STK=9 /UG = Mm.220953 /UG _TITLE= Mus ...
1424721_at, similar to Microfibril-associated glycoprotein 3 precursor
1448113_at, stathmin 1
1427280_at, RIKEN cDNA A230052E19 gene
1455439_a_at, lectin, galactose binding, soluble 1
1452035_at, procollagen, type IV , alpha 1
1417679_at, growth factor independent 1
1436836_x_at, calponin 3, acidic
1422610_s_at, insulin-like growth factor 2, binding protein 3
1416034_at, CD24a antigen
1427256_at, chondroitin sulfate proteoglycan 2

147

Figure 3.9

Gene ID Gene Name
Growth factors, receptors and signaling molecules and Cell cycle
1417283_at Ly6/neurotoxin 1
1420872_at guanylate cyclase 1, soluble, beta 3
1425518_at cAMP-regulated guanine nucleotide exchange factor
1426258_at sortilin-related receptor, LDLR class A repeats-
1455462_at adenylate cyclase 2
1424638_at cyclin-dependent kinase inhibitor 1A (P21)
1417943_at guanine nucleotide binding protein (G protein),
1449229_a_at cyclin-dependent kinase-like 2 (CDC2-related
1416783_at tachykinin 1
1422474_at phosphodiesterase 4B, cAMP specific
Transcriptional regulators and other nuclear factors
1448954_at nuclear receptor interacting protein 3
1456341_a_at Kruppel-like factor 9
Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
1417373_a_at tubulin, alpha 4
1452298_a_at myosin Vb
1452107_s_at nephronectin
1416361_a_at dynein, cytoplasmic, intermediate chain 1
1417672_at solute carrier family 4, sodium bicarbonate
1421594_a_at synaptotagmin-like 2
1435495_at myosin binding protein H
1450121_at Mus musculus sodium channel 27 mRNA fragment.
1452981_at contactin 1
1415904_at lipoprotein lipase
1423640_at synaptoporin
Hormones and proteases and secretory apparatus(Extracellular matrix)
1416114_at SPARC-like 1 (mast9, hevin)
1417702_a_at histamine N-methyltransferase
1426785_a monoglyceride lipase
1450708_at secretogranin II
Metabolism/maintanence
1422456_at N-ethylmaleimide sensitive fusion protein
1424394_at selenoprotein M
1418829_a_at enolase 2, gamma neuronal
1439036_a_at ATPase, Na+/K+ transporting, beta 1 polypeptide
Others
1420822_s_at Mus musculus adult male stomach cDNA, RIKEN full-
1423135_at thymus cell antigen 1, theta
1451674_at Mus musculus, clone IMAGE:5368459, mRNA
1454897_at expressed sequence AI450287
1460214_at Purkinje cell protein 4
1420955_at visinin-like 1
1420191_s_at hypothetical protein MNCb-4137
Cluster 1,2, and 3- 38 items
-3
-1
1
3
E16 E18 P0 P5
Time
Log fold difference
(base 2)
Group 1

148
Figure 3.9. Continued

Gene ID Gene Name
Transcriptional regulators and other nuclear factors
1416723_at transcription factor 4
1417065_at early growth response 1
1420390_s_at zinc finger protein 354A
1434336_s_at REST co-repressor
1436182_at special AT-rich sequence binding protein 1
Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
1455883_a_at leucine rich repeat transmembrane neuronal 1
Metabolism/maintanence
1417799_at ATPase, H+ transporting, V1 subunit G isoform 2
1435148_at ATPase, Na+/K+ transporting, beta 2 polypeptide
1419499_at
glycerol-3-phosphate acyltransferase, mitochondrial
1422126_a_at
Nudt13 nudix (nucleoside diphosphate linked
moiety X)-type motif 13
Others
1433737_at expressed sequence AA673513
1452227_at RIKEN cDNA 2310045A20 gene
Cluster 5 and 6- 13 items
-3
-1
1
3
E16 E18 P0 P5
Time
Log fold difference
(base 2)
Gene ID Gene Name
Growth factors, receptors and signaling molecules and Cell cycle
1416034_at CD24a antigen
Transcriptional regulators and other nuclear factors
1417679_at growth factor independent 1
1422610_s_at insulin-like growth factor 2, binding protein 3
Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
1436836_x_at calponin 3, acidic
Hormones and proteases and secretory apparatus(Extracellular matrix)
1427256_at chondroitin sulfate proteoglycan 2
Others
1449947_s_at RIKEN cDNA A230102L03 gene
Cluster 4- 6 items
-3
-1
1
3
E16 E18 P0 P5
Time
Log fold difference
(base 2)
Group 2

149
Figure3. 9. Continued

Figure 3.9. K-mean cluster analysis of defined differentially expressed transcripts. Seventy-one
transcripts were placed into 4 groups based on relative change in expression of each gene at E16, E18,
P0, and P5. The mean differential expression levels of the genes are represented in the graphs. The
gene transcripts are listed below in functional categories.
Gene ID Gene Name
Transcriptional regulators and other nuclear factors
1437210_a_at bromodomain containing 2
1456194_a_at RNA binding protein regulatory subunit
Cell surface/adhesion/matrix/cytoskeletal proteins/membrane proteins
1424721_at similar to Microfibril-associated glycoprotein 3
1448113_at stathmin 1
1419734_at actin, beta, cytoplasmic
Hormones and proteases and secretory apparatus(Extracellular matrix)
1423567_a_at proteasome (prosome, macropain) subunit, alpha
1452035_at procollagen, type IV, alpha 1
1455439_a_at lectin, galactose binding, soluble 1
Metabolism/maintanence
1415744_at H2-K region expressed gene 2
1456642_x_at S100 calcium binding protein A10 (calpactin)
1437729_at ribosomal protein L27a
Others
1436362_x_at
1436893_a_at axotrophin
1448685_at RIKEN cDNA 2900010M23 gene
1427280_at RIKEN cDNA A230052E19 gene
Cluster 7, 8, and 9- 15 items
-3
-1
1
3
E16 E18 P0 P5
Time
Log fold difference
(base 2)
Group 4

150

Table 3.3

Gene name
Plexin A3
Plexin B2
Plexin C1
Complexin 1
Complexin 2
Growth associated protein 43
Semaphorin 3A
Semaphorin 4A
Semaphorin 4G
Semaphorin 6C

Table 3.3. Neurite growth related genes analyzed by microarray. The mean level of these transcripts
at E16, E18, P0, and P5 were not differentially expressed.

151
chosen based on literature searches on their hypothesized functional significance in
neural development and control of axon growth.
Transcriptional factors regulate inherent molecular signaling cascades
mediating core cellular processes. To identify novel roles for transcription factors
central in controlling axon growth, four were selected for further validation.
Kruppel-like factor 9 (Klf9), a transcription factor whose family members are
involved in cell proliferation, differentiation and apoptosis (Liu et al., 1996)
progressively increasing to 50.0-fold higher (±34.1) at P5 (Figure 3.10). Of the
genes analyzed, transcript levels for Klf9 were the lowest at E16, by comparison
GAPDH expression was near 188,967- fold greater than Klf9.
Special AT-rich sequence binding protein 1 (SATB1) is a nuclear protein
involved in recruiting chromatin-remodeling factors by selectively tethering
specialized DNA sequences including myc and a brain-specific gene (Cai et al., 2003,
Figure 3.10) and recruits histone deacetylase to modify the tertiary structure of DNA
opening genes for transcription. Though SATB1 was statistically differentially
expressed by real-time RT-PCR, it did not change over 2-fold. This data suggest that
SATB1 may not play a role in the control of axon growth.
The early growth response 1 (egr1) transcriptional factor, an immediate early
gene, implicated in playing a role in memory (Jones et al., 2001) increased to 4.2-
fold (±0.7) at P5 (Figure 3.10). Axons must grow to the correct location for them to
form appropriate synapses and the storage of memory. Egr1 may also be involved in
neural plasticity since memory is more closely related to cellular plasticity.

152

Figure 3.10

Figure 3.10. A subset of differentially expressed transcripts of transcriptional regulators and other
nuclear factors were examined by semi-quantitative RT-PCR. The genes were confirmed to be
differentially expressed (p<0.05) except for bromodomain containing 2 (Brd2). Real-time assessment
of developmentally dynamic expression of Brd2 and special AT-rich sequence binding protein 1 did
not change by more than 2-fold as suggest by microarray data.
Relative expression of transcriptional
regulators and other nuclear factors
-5.00
5.00
15.00
25.00
35.00
45.00
55.00
Kruppel-like
factor 9
special AT-
rich sequence
binding protein
1
early growth
response 1
bromodomain
containing 2
Gene
Relative expression
Mean E16
Mean E18
Mean P0
Mean P5

153
The transcriptional factor bromodomain containing 2 (brd2), implicated in as
a possible cause of neural developmental abnormalities associated with idiopathic
epilepsy (Pal et al., 2003), while human brd2 associates with cell cycle-driving
transcription factors E2F-1 and E2F-2 (Crowley et al., 2004) was found to declined
in expression at E18 followed by a return to E16 levels in microarray analysis
(Figure 3.9, Cluster 7). Real-time RT-PCR did not detect significant differential
expression of Brd2 suggesting that it may not play a significant role in axon growth
nor RGC development.
Since the transcript dependant factors mediating axonal growth are unclear, it
is possible that as yet undefined function of novel genes may control axon growth.
Of the numerous novel developmentally differentially expressed RGC transcripts
identified during our screen, three were selected for real-time PCR for their possible
role in axon growth.
Purkinje cell protein 4 (pcp-4) is a neuron specific protein expressed in the
CNS implicated in cerebellar hypoplasia in Down Syndrome (Chen et al., 1996) and
structurally shares a weak match to a 17 amino acid motif with growth associated
protein 43 (GAP-43). Pcf-4 increased 3.4-fold (±1.1) at E18 to 6.2-fold (±6.6) by P5
(Figure 3.11). This protein may mediate axon growth or pathfinding like GAP-43.
Axotrophin plays a role in corpus callosum formation and GDNF dependent
sensory neuron survival in the CNS (Baker et al., 1997). Transcript levels by real-
time RT-PCR analysis did not show a greater than 2-fold change although the trend
matches microarray data (Figure 3.11).

154
Figure 3.11

Figure 3.11. Novel transcripts identified as differentially expressed in RGC were confirmed by semi-
quantitative real-time RT-PCR. Purkinje cell protein 4 and axotrophin was differentially expressed
(p<0.05) except for the housekeeping GAPDH. As expected, the housekeeping GAPDH transcript
levels were not significantly different.
Relative expression of others
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
Purkinje cell
protein 4
axotrophin GAPDH
Gene
Relative expression
Mean E16
Mean E18
Mean P0
Mean P5

155
Genes that encode for growth factors, receptors, signaling molecules, and cell
cycle proteins are generic programs for cell growth. To identify genes that mediate
axon growth within these large molecular programs for generalized growth, six were
selected for semi-quantitative RT-PCR validation (Figure 3.12). Guanine nucleotide
binding protein (G-protein) gamma 4 subunit, was selected because G-protein
cascades are involved in signal transduction of external stimuli. The gamma subunit
is more divergent than the other components of the G-protein receptor cascade and is
what contributes to the specificity of the signal propagated (Gilman, 1987).
Specifically, the gamma 4 subunit is expressed primarily in the CNS (Kalyanaraman
et al., 1995) and may be important in axon growth and neuron functionality. Real-
time RT-PCR showed a 2.5-fold (±0.3) at E18 and continues to increase as RGCs
mature.
Sortilin-related receptor, LDLR class A repeats-containing (Sort in Figure
3.12) resembles the low-density lipoprotein receptors and may play a role in
neuronal differentiation and outgrowth (Hirayama et al., 2000). This receptor may
be necessary for the uptake of compounds used to make the lipid membranes of long
axons. Sortilin-related receptor increases 4.2-fold (±2.0) at E18 and remains
elevated.
Tackykinin 1 encodes for many proteins secondary to differential splicing
and post-translational modification including: neurokinin A, neuropeptide K,
neuropeptide gamma, and substance P. Tachykinin can modulate hippocampal
excitability (Zimmer et al., 1998) and is indicative of a more mature functioning

156
Figure 3.12

Figure 3.12. Transcripts that function in growth, signaling, and cell cycle were examined by semi-
quantitative RT-PCR and ISH. The genes examined were confirmed to be differentially expressed
(p<0.05) and >2-fold change at any time point.

Relative expression of growth factors, receptors, and
signaling molecules, and cell cycle
0.00
5.00
10.00
15.00
20.00
25.00
guanine
nucleotide
binding protein (G
protein), gamma 4
subunit
sortilin- related
receptor, LDLR
class A repeats-
containing
tachykinin 1 Ly6/neurotoxin 1 cyclin- dependent
kinase- like 2
(CDC2- related
kinase)
Gene
Relative expression
Mean E16
Mean E18
Mean P0
Mean P5

157
RGC. It was selected as an internal control for monitoring RGC development.
Tackykinin-1 expression increases 6.4-fold (±3.8) at E18 and remains elevated
(Figure 3.12).
Ly6/neurotoxin 1 modulates nicotinic acetylcholine receptors (nAChR, Miwa
et al., 1999) and was selected for confirmation since its role is unclear in
glutaminergic RGCs. Ly6 expression also increased dramatically to 3.7-fold (±0.7)
at E18 and continued to progress to 16.4-fold (±5.7) by P5 (Figure 3.12). It may be
expressed at levels capable of silencing all nAChR induced signaling thereby
yielding glutaminergic neurons.
The growth inhibitory cyclin-dependent kinase-like 2 (cdkl2) functions
mainly in mature neurons and is distinct from those of immediate-early genes (Sassa
et al., 2004) was selected for further confirmation as a control to ensure the validity
of our experimental design and the data generated since maturing RGCs should
transcribe more cdkl2. Cdkl2 transcription progressively increased 10.2-fold (±1.8)
by P5 (Figure 3.12).
Because of their roles in synaptogenesis, synaptoporin (SPO) and
synaptotagmin-like 2 transcript expressions were validated. As axon growth declines
physiologically, RGCs have reached their targets and begun to form stable synapses
to ensure cell survival. Synaptotagmin-like 2, a member of the synaptotagmin-like
protein (Slp) family, is involved in the GTPase Rab27A-mediated membrane
transport and regulates vesicle trafficking and fusion (Nagashima et al., 2002). It
increased in transcription 2.5-fold (±0.7) at E18 and remained elevated to 4.2-fold

158

Figure 3.13

Figure 3.13. Transcripts of structural and synapse associated genes were examined by semi-
quantitative RT-PCR. The genes examined were confirmed to be differentially expressed (p<0.05)
and >2-fold change at any time point.
Relative expression of cell
surface/adhesion/matrix/cytoskeletal
proteins/membrane proteins
0.00
2.00
4.00
6.00
8.00
10.00
12.00
nephronectin synaptoporin synaptotagmin-
like 2
Gene
Relative
Expression
Mean E16
Mean E18
Mean P0
Mean P5

159
(±1.8) at P5 (Figure 3.13). SPO, an integral membrane protein in synaptic vesicles
is enriched in the granule cell axons (Singec et al., 2002). SPO increased 5.0-fold
(±0.4) at E18 and remained at that level (Figure 3.13). Of the cell and membrane
structural transcripts that were differentially expressed, synaptoporin and
synaptotagmin-like 2 were confirmed to have an increasing level of transcription
during perinatal development similar to microarray data. The increases in
transcription of vesicle associated proteins suggest maturing RGCs are becoming
more functional by generating vesicles for neurotransmitters.
To also investigate the possible mechanism of how RGCs respond to the
inhibitory myelin, the nephronectin adhesion molecules that bind to collagen was
selected for further validation. Nephronectin progressively increased 2.2-fold (±0.9)
at E18 to 13.9-fold (±5.6) by P5. Adhesion molecules and their interactions with
extracellular matrix proteins like collagen are beginning to emerge as important
mediators of neuronal targeting in synaptic connections (Yamagata et al., 2004).
When new samples were assessed by real-time RT-PCR, most of the genes
characterized displayed similar expression pattern to microarray analysis although
exact fold change varied. Only bromodomain containing 2 which was not
statistically significant and early growth response 1 did not exhibit an expression
trend similar to microarray data. Real-time assessment of transcript levels of
axotrophin, special AT-rich sequence binding protein 1, and early growth response 1

159

Mean E16 relative gene expression
520
1
10124
49
4475
2137
269
944
342
1889
339
8005
203
7762
0
2000
4000
6000
8000
10000
12000
sortilin-related
receptor, LDLR
class A repeats-
containing
Kruppel-like factor 9 special AT-rich
sequence binding
protein 1
Ly6/neurotoxin 1 early growth
response 1
guanine nucleotide
binding protein (G
protein), gamma 4
subunit
synaptotagmin-like
2
synaptoporin tachykinin 1 Purkinje cell protein
4
nephronectin bromodomain
containing 2
cyclin-dependent
kinase-like 2
(CDC2-related
kinase)
axotrophin
Gene
Relative expression levels
Figure 3.14. The relative expression levels of transcripts analyzed by semi-quantitative RT-PCR were compared to Kruppel-like factor 9 (the lowest
expressed mean transcript) at E16. GAPDH was expressed at 18,8967-fold greater than Kruppel-like factor 9.

Figure 3.14

160

were not at least 2-fold differentially expressed as suggested by microarray data.
The relative transcript levels were graphed relative to the mean level of the
klf9 expression at E16, the lowest expressing transcript (Figure 3.14). All of the
genes analyzed were expressed at levels much lower than GAPDH.

Confirmation of Microarray Data by In Situ Hybridizations (ISH)
In situ hybridization (ISH) was used to localize gene expression. All 12
genes previously found to be differentially expressed by semi-quantitative RT-PCR
were in developing RGC. A representative ISH time course of Pcp-4 was similar at
all times as compared to E16 (Figure 3.15). The other genes characterized by ISH
also did not exhibit significant changes during development, only representative
micrographs at PO are presented for the rest. The transcription factors Klf9 and egr1
were preferentially expressed in RGCs (Figure 3.16). Cell signaling and
proliferation transcripts showed preferential expression in the RGC layer but were
not as distinct (Figure 3.17). Transcripts involved in cellular adhesion and
synaptogenesis localized nephronectin and synaptotagmin-like 2 to the RGC layer
(Figure 3.18).

Confirmation of these molecular changes in purified RGCs.
To ensure that transcript levels characterized earlier exhibit the same
expression profile in our in vitro model of neurite growth and to investigate their
possible involvement in mediating axonal elongation, transcripts were analyzed

161
by real-time RT-PCR from purified cultured E16 and E18 RGCs. Physiologically
relevant differentially expressed genes provide a basis to characterize the role of
similarly changing gene expression patterns in cultured RGCs that mediate
transcription dependent axon growth.
Of the candidate 14 genes selected previously, all were analyzed for in vitro
differential transcript expression. Five of these genes that had statistically significant
changes and all of them increased in expression from E16 with axotrophin to 3.9-
fold (±1.6), Kruppel-like factor 9 to 7.4-fold (±3.1), Purkinje cell protein 4 to 3.5-
fold (±1.6), synaptoporin to 3.6-fold (±0.9), and tachykinin 1 to 6.6-fold (±1.2) over
the E16 baseline by E18 (Figure 3.19). The transcriptional profile in four of these
five genes was similar to that observed in directly isolated RGCs except for
axotrophin, which RT-PCR from isolated RGCs suggested no change.

Transfection of RGCs
To investigate more the functional contribution of these differentially
expressed genes in axonal growth, we have established an in vitro model of neurite
growth. Introduction of individual genes into cultured RGCs will facilitate the
molecular dissection of the pathway involved in axonal growth loss. Transfection of
inhibitory genes will limit E16 RGC neurons from extending axons. As a corollary,
if the transfected gene is responsible for mediating axonal growth, then more E18
RGC neurons will extend axons when introduced to them.

162

Figure 3.15

Figure 3.15. A time course in situ hybridization study on Purkinje cell protein 4 (Pcp4). Qualitative
differences in expression are similar. A low and high power micrograph from PO is presented for
axotrophin and the other genes. The lower most dark pigmented layer is the RPE while the innermost
layer of cells are RGC. Pcf4 is preferentially expressed in RGCs.
Pcp4 E16 Pcp4 E18 Pcp4 P0 Pcp4 P5

Pcp4 E16 Pcp4 E18 Pcp4 P0 Pcp4 P5
AxotrophinP0
AxotrophinP0
AxotrophinP0
AxotrophinP0

163

Figure 3.16

Figure 3.16. Low and high power views of ISH for transcription factors Kruppel-like factor 9 (kfl9)
and early growth response-1 (egr1). Klf9 and egr1 are preferentially expressed in the RGC layer.
Representative sections at P0 are shown.
Klf9 P0
egr1 P0
Klf9 P0 egr1 P0

164

Figure 3.17

Figure 3.17. Low and high power views of ISH from P0 retinas of signaling and cell proliferation.
The transcripts for these genes are expressed in the RGC layer and throughout the retina. Guanine
nucleotide binding protein gamma 4 subunit (G-prot); sortilin- related receptor (Sort), and cyclin-
dependent kinase like 2 (cdkl2).

G prot P0 Sort P0
Ly6 P0
cdkl2 P0
G-prot P0 Sort P0
Ly6 P0
cdkl2 P0

165

Figure 3.18

Figure 3.18. Low and high power views of ISH for transcript involved in adhesion and
synaptogenesis. All sections were from P0 retinas as indicated. The transcripts for nephronectin and
synG are preferentially expressed in the RGC layer and throughout the retina. Synaptotagmin-like 2
(SynG like 2).
SynG like 2 P0
Synaptoporin P0
SynG like 2 P0 Synaptoporin P0

NephronectinP0
NephronectinP0
NephronectinP0
NephronectinP0

166

Figure 3.19. Differentially expressed transcripts from cultured RGCs at E16 and E18 were examined
by semi-quantitative PCR. (a) The genes examined were confirmed to be differentially expressed
(p<0.05) are represented in the graph and in the accompanying table. The transcriptional profile
described was similar to microarray data from directly isolated RGCs except for axotrophin. (b)
Relative expression of transcripts analyzed by semi-quantitative PCR. The means of all transcript
levels at E16 scaled to L-32 (housekeeping gene) were compared to the expression level of Kruppel-
like factor 9.

167
Figure 3.19. Continued

(a)
Transcriptional changes in cultured RGCs
0
2
4
6
8
10
12
Mean E16 Mean E18
Time
Fold Change
axotrophin
Kruppel-like factor 9
Purkinje cell protein 4
synaptoporin
tachykinin 1

Mean E16
Std Dev
E16 Mean E18
Std Dev
E18 P-value
axotrophin 1.0 0.2 3.9 1.6 0.00
Kruppel-like factor 9 1.1 0.2 7.4 3.1 0.01
Purkinje cell protein
4 1.0 0.2 3.5 1.6 0.01
synaptoporin 1.1 0.5 3.5 0.9 0.00
tachykinin 1 1.1 0.4 6.6 1.2 0.03

(b)
Relative gene expression levels oc cultured E16 RGCs
177
1
98
30
51
0
50
100
150
200
axotrophin Kruppel-like
factor 9
Purkinje cell
protein 4
synaptoporin tachykinin 1
Gene
Relative fold difference

168

Lipofectamine 2000 (Ambion), though evaluated as a method to transfect
genes into RGCs, is not optimal. A plasmid with a strong ubiquitous CMV promoter
linked to a DsRed and a poly A sequence was transfected into E16 RGCs. Various
ratios of DNA and Lipofectamine 2000 were tested to optimize transfection and
survival rates. RGC cell survival was severely compromised in all experimental
conditions (Figure 3.16). Primary cultures of CNS neurons are particularly sensitive
and challenging. Alternative technique utilizing viral vectors may be more effective
since lentiviral based vector are less traumatic.

169

Figure 3.20

Figure 3.20. Co-localization of transfected viable E16 RGCs expressing DsRed. Representative
confocal images of cultured purified E16 CD90
+
RGCs transfected with a plasmid containing a CMV
promoter , DsRed, and poly A sequence. (a) Successfully transfected RGCs expressing DsRed, stains
red. (b) Calcein-AM labeled viable RGCs stains cells green. (c) Bright-field. (d) Merge image co-
localization of all fields. Low RGC survival and transfection rates limit the effectiveness of
Lipofectamine 2000 as a viable option.
(a)
(c)
(b)
(d)

170
Discussion

In the present study we used gene microarray analysis, combined with in situ
hybridization and real time PCR, to identify genes expressed in mouse RGC during
visual system development, and contrast the relative gene expression level between
the rapid RGC axon growth (E16) with the period axon growth are recessing (P5).
We observed both changes in the types of genes and statistically significant
differences in the levels of the genes expressed in the two time periods. Many of the
housekeeping genes, such as beta actin, glyceraldehyde-3-phosphate dehydrogenase,
and 18S RNA are maintained throughout this neonatal period; however, we
identified 71 differentially expressed transcripts out of 22,623. The genes expressed
at E16 tended to be growth and differentiation associated, while genes elevated at P5
tend to be more functionally associated. Noticeably, we observed more complex
gene expression repertoire in the later developmental stage, suggesting that the loss
of retinal axon growth at the postnatal period results from additional gene regulation,
instead of merely shutting down genes associated with growth.

Gene Expression Complexity Increases as RGCs Mature
Predominately, the transcription of most genes (~54%) demonstrated a
progressive and sustained increase. The coordinated increased number of genes
expressed suggests an increased accumulation of functional molecular pathways
mediating the role of RGCs in conduction of visual information to the brain.
Transcripts encoding for components implicated in functional neurons such

171
as receptors and cell surface proteins were increased in developing RGCs. Contactin
1, a glycosyl phosphatidylinositol (GPI) membrane anchored cell adhesion and
recognition glycoprotein protein increased in developing RGCs. Contactin knockout
mice display severe ataxia and do not survive past P18 (Berglund et al., 1999).
Premature expression of contactin 1 in granule cells affects the proliferation and
development of these neurons as well as the differentiation of Purkinje cells, their
synaptic partners (Bizzoca et al., 2003) suggesting a similar role in RGCs.
The increase transcription of cell surface ion channels like Mus musculus
sodium channel 27 mRNA fragment and solute carrier family 4, sodium bicarbonate
cotransporter-like, member 10 indicated that RGC are gaining more functional
components as they mature. A sustained increase in the ATPase, Na
+
/K
+

transporting, beta 1 polypeptide encoding transcript may play a role in establishing
chemoelectrical gradient for functional neurons to be able to re-polarize/hyper-
polarize. The role in RGC of the transcript encoding for a stilbene-sensitive Na
+
-
HCO
-3
transporter involved in cerebral spinal fluid (CSF) production in choroids
plexus epithelial cells (Praetorius et al., 2004) is unclear since neurons are not known
as a producers of CSF.
The thymus cell antigen 1, theta transcript encode for the CD90 (Thy1) cell
surface molecule used to purify RGC and has been demonstrated to increase in
developing RGCs (Morris and Barber, 1983) and the promoter is used to identify
RGCs (Feng et al., 2000). CD90 is also a marker for Purkinje cells (Messer et al.,
1984) and suggests a high degree of similarity in CNS neurons (Farkas et al., 2004).

172
The identification of CD90 as differentially expressed was expected and is an
important indicator corroborating our work with what is already known about RGC
development.
There are increases in the transcripts of synapse related genes in developing
RGCs. The N-ethylmaleimide sensitive fusion protein (NSF) transcript encodes a
protein that is involved in membrane fusion and is an integral part of a SNARE
complex. The release of neurotransmitters at the proper synaptic junction is
mediated primarily by the ability of the SNARE complex to localize vesicle fusion to
the correct membrane location (Rothman et al., 1994; Matveeva et al., 2003).
Increased expression of NSF suggests RGCs are gaining neuronal proteins that
mediate neuronal functions. Secretogranin II, a member of the granin protein family
selectively expressed in neurosecretory neurons, remained increased in the time
observed. The secretogranin II promoter region contains a cAMP response element
and may be a marker of neuronal differentiation (Scammell et al., 2000).
The increase in ATPase, H
+
transporting, V1 subunit G isoform 2, which
plays a role in acidification of synaptic vesicles (Murata et al., 2002), indicates
maturing RGCs are gaining functional properties. This gene involved in the
acidification of vesicles was induced early then declines unlike the expression
profiles of other synaptic vesicle related genes found in Clusters 1, 2, and 3 that
tends to steadily increase.
The increase in control of secondary messengers suggests maturing RGCs are
becoming more complex with increasing signaling pathways. Many transcripts for

173
proteins that control other signaling components increase as RGCs mature such as:
adenylate cyclase 2; cAMP-regulated guanine nucleotide exchange factor II;
guanylate cyclase 1, soluble, beta 3; guanine nucleotide binding protein (G protein),
gamma 4 subunit; phosphodiesterase 4B. The levels of cAMP and cGMP are both
involved in integrating responses to external stimuli. An increase in cAMP is known
to mediate cell survival and axon growth (Meyer-Franke et al., 1998). Molecular
factors that control cAMP levels suggest RGCs are gaining functional properties. It
may also be playing a role with cell survival factors in RGCs that have established
their correct connections. In addition, increased levels of cAMP may help explain
developmental induction of secretogranin II transcript expression since it has a
cAMP responsive element in its promoter region. The Mus musculus adult male
stomach cDNA, RIKEN full-length enriched library, clone: 2210016L03 product:
sphingosine-1-phosphate phosphatase 1, full insert sequence encodes for a
phosphatase which can act intracellularly as a second messenger and extracellular by
binding to a G-protein coupled receptor.
Only six genes exhibit a sustained progressive decrease in expression. Two
transcriptional factors: growth factor independent 1 and insulin-like growth factor 2
declined as RGCs effectively progressed through maturation under their influence
and are down-regulated when their function is no longer needed. Growth factor
independent 1 is similar to the "senseless (sens)" protein involved in activating
proneural genes in Drosophila (Nolo et al., 2000). Insulin-like growth factor 2,
binding protein 3 (Igf2bp3) associates specifically with the 5’ untranslated region of

174
the human insulin-like growth factor II (IGF2). IGF2 is a mitogen for many cell
types and is important for muscle growth and differentiation suggesting Igf2bp3 is
involve in RGC development. Igf2bp3, a zipcode binding protein-1 (ZBP-1) family
member, is expressed in undifferentiated neuroepithelial cells and some postmitotic
neurons at early embryonic stages (E10.5--E12.5). Igf2bp3 declines from E12.5 to
birth (Mori et al., 2001) and also in developing RGCs. Insulin is a supplement added
into the cultures to ensure RGC survival and for proper glucose metabolism. The
mere addition of growth factors to stimulate properties of immature neurons like
axonal growth may be unrealistic if major components of the inherent molecular
machinery are no longer available.

Identification of Neurite Growth Related Transcripts
Our transcript screen identified genes that correlate well with the data
currently available regarding neurite growth. There are emerging parallels between
neurite growth and angiogenesis. In the body, nerves travel with major blood vessels.
Histamine N-methyltransferase is involved in histamine metabolism and guanylate
cyclase 1 controls nitric oxide synthesis, both play a role in vasodilation. The
possible link between vasodilation and neurite growth in developing RGCs is
beginning to emerge (Wang et al., 2003; Serini et al., 2003). Our data also suggests
a similar relationship.
Chondroitin sulfate proteoglycan 2 (Cspg2 also known as NG2) was down-
regulated as RGCs developed perinatally. Extracellular Cspg2 blocks neurite growth

175
in vitro and forms at the glial scar (Rezajooi et al., 2004). The regulation of Cspg2
expression is complex, it can be transactivated by p53 (Yoon et al., 2002) and thus
indirectly regulated by the increases in p21 transcription that suppresses p53 activity.
NG2
+
cells display multipotent stem cell-like properties in vitro and in vivo
(Belachew et al., 2003) and declines in expression as cells mature. As expected, this
gene is down-regulated in developing RGCs. A clearer understanding of the
interplay between extracellular Cspg2 and cell surface Cspg2 is necessary.
Along with Cspg2, the NOGO-66 receptor (NgR) is known to mediate axonal
growth inhibition. The GPI anchored protein, NgR is responsible for interacting
with the extracellular domain of Nogo-A (Nogo-66), oligodendrocyte myelin
glycoprotein (OMgp), and myelin-associated glycoprotein (MAG), which have been
implicated in creating a non-permissive environment for axonal regeneration (Wang
et al., 2002). NgR in complex with the low affinity neurotrophin receptor (p75
NTR) is the only receptor known to mediate this axonal inhibitory signal in neurons
(Lauren et al., 2003). We have characterized the decline of the transcription of the
surface antigen CD24a. CD24a is a short peptide with a GPI membrane anchor and
may be involved in cell communication. CD24a and contactin as described earlier
may be novel additional inhibitory signal transducers. CD24a is also involved in B
cell development (Kay et al., 1991) and may play a role in neuronal development.
The transcript of the cell surface protein, leucine rich repeat transmembrane
neuronal 1 (LRRN6A), involved in cell adhesion in Drosophila, exhibits a temporal
spike in expression followed by a decline. LRRN6A contains the leucine-rich repeat

176
region similar to the Nogo receptor NgR which binds to myelin associated axonal
growth inhibitors. Because of the structural similarity and expression in the CNS, it
may be involved in axonal guidance, migration, and nervous system development
and regeneration (Carim-Todd et al., 2003) in other CNS neurons as well as RGCs.
The identification of genes that function in controlling axonal growth can
then be used to identify developmental regulators. The dissection of regulatory
elements that control the expression of these neurite growth control genes will
identify molecular factors that govern RGC development.

Markers and Regulators of Perinatal RGC Development
Although we have provided a framework to test candidate genes for their
functional effect by assessing neurite re-growth, we understand that RGC physiology
is changing during this period. The use of neurite outgrowth as the sole determinate
of a genes involvement in RGC development would not be prudent. Many
phenotypic changes occur in maturing RGCs during this period include: altered in
vitro survival to the neurotrophins CNTF, LIF, bFGF, TGFα, RGC paracrine factors,
and oligodendrocyte factors (Meyer-Franke et al., 1995); shift in neurite outgrowth
reliance from laminin to merosin (Cohen et al., 1991); distinctive in vivo growth
cone morphology (Bovolenta et al., 1987); and loss of axonal growth ability in more
mature RGCs (Vidal-Sanz et al. 1987; Chen et al., 1995; Shewan et al. 1995;
Goldberg et al., 2002). Many of the candidate genes identified may also play a role
in these other neuronal processes and so, depending on the gene being studied,

177
appropriate functional assays should be used.
We have created a transcriptional profile database of developing perinatal
RGCs from immature to more functional neurons. The genes identified in this study
will facilitate the functional analysis of transcription factors associated with the
phenotypic changes in these CNS neurons and well as establish potential RGC
specific markers. Nonetheless, a combination of these transcripts will help identify
the developmental stage of maturing RGCs. We have uncovered many novel
candidates that may be involved in RGC development. Though there are differences
in activity, protein stability, post-transcriptional and translational modification,
nuclear localization, and different DNA binding characteristics, we have provided a
broad and concerted profile of the transcriptional change in developing CNS neurons
for further functional analysis.
In an effort to characterize possible molecular factors involved in loss of
axon growth abilities, we have analyzed a subset of transcripts identified as
differentially expressed. Cultured E18 RGCs have significantly less axon growth
abilities than E16 RGCs. Axotrophin, Kruppel-like factor 9, Purkinje cell protein 4,
synaptoporin, and tachykinin 1 are statistically significant and are expressed with >2
fold difference. This suggests that of the transcripts identified as differentially
expressed during normal RGC maturation, these five genes are more likely to be
involved in a signaling cascade that limits axon growth. Further evaluation of the
biochemical and functional role these genes posses will be needed. Initial
transfection assays with Lipofectamine 2000 underscore the technical difficulties

178
pervasive in studies with CNS neurons. Alternative methods will need to be
optimized so that introduction of these genes into RGCs can be completed to realize
their function.

179
Appendix to Chapter III:
Delineation of the other identified transcripts
The differentially expressed transcripts were grouped into nine different
clusters based on their expression level (Figure 9). The general trend of their
expression is stated and each gene is described briefly below. Transcripts discussed
earlier are not repeated.

Cluster 1, 2, and 3: progressive increase in expression
The transcriptional profiles of 38 differentially expressed transcripts in these
3 clusters progressively increased as RGCs continued to mature. They are in three
separate clusters based on their mean expression levels at each time point.
The role of cell cycle genes in neuronal development is unclear since RGCs
during this period are post-mitotic and suggests a secondary role. Transcription of
cyclin-dependent kinase inhibitor 1A (P21) increased. Smad proteins activated by
TGF-beta form a complex with FoxO proteins to turn on p21 (Seoane et al., 2004).
p21 regulates p53 and serves as a control in maintaining cells in the G0 phase.
The transcriptional factors that increase between E16 and E18 and remain
upregulated are of significant interest since they coordinate a broad range of
transcriptional events. The role of the novel RIKEN cDNA A330103B05 gene
(renamed: Nrip3 nuclear receptor interacting protein 3) is unknown.
There are increases in transcription of cell structure genes. Myosin Vb, a
molecular motor, may also play a role in identifying the correct neuronal target for

180
synaptogenesis (Fan et al., 2004). Dynein, cytoplasmic, intermediate chain 1,
encodes for a dynein based multisubunit microtubule-based molecular motor
(Crackower et al., 1999). Increasing number of molecular motors are perhaps
required as RGCs develop their functional properties in order to transport vesicles
and other components up and down long axon. Tubulin, alpha 4 is critical to the
microtubule architecture of the cell (Oakley et al., 1987). Increased transcription of
myosin binding protein H, a protein containing an 'RKPS' sequence which is thought
to be a cAMP- and cGMP-dependent protein kinase A phosphorylation target site,
though characterized in skeletal muscle, its role in neurons is not well understood
(Jung et al., 1998).
Among the differentially expressed transcripts that demonstrated this
temporal profile, many encoded genes involved in maintenance of homeostasis and
synthesis of cellular components. Selenoprotein M transcription is increased. The
addition of selenium, an antioxidant, into cell culture for increases RGC survival and
plays a physiological activation of focal adhesions kinase Akt pathway (Lee et al.,
2003). Increased transcription of enolase 2, gamma neuronal, a glycolytic enzyme,
was expected since the neuron specific enolase promoter has long been used to
specifically label neurons. Increased transcription of monoglyceride lipase and
lipoprotein lipase indicate increased lipid metabolism and the synthesis of the lipid
bilayers. In adipocytes, monoglyceride lipase catalyzes the last step in the hydrolysis
of stored triglycerides and presumably complements lipoprotein lipase in degrading
triglycerides from chylomicrons and very low density lipoproteins (Panzenboeck et

181
al., 1997).
Increased expression of Mus musculus, clone IMAGE: 5368459 and mRNA
hypothetical protein MNCb-4137 was observed, though these genes have not been
well characterized.

Cluster 4
The transcriptional profile of the genes in this cluster decreased progressively
from E16 to P5. Of these six transcripts, two for surface proteins, two for
transcriptional regulators, one encoded for structural proteins, and one was an
unknown RIKEN cDNA.
There was a decrease in the transcription of the not well characterized acidic
calponin 3, a structural component that is associated with the actin cytoskeleton, but
it is not involved in contraction (Maguchi et al., 1995), as well as RIKEN cDNA
A230052E19.

Cluster 5
The expression of six transcripts increased from baseline E16 levels at E18
and gradually decreased. The transcripts with this profile included two ATPases,
two transcriptional factors, a surface protein, and a kinase mediating cell structure.
The ATPase, Na
+
/K
+
transporting, beta 2 polypeptide may mediate adhesion
(Gloor et al., 1990; Antonicek and Schachner, 1988).
The transcripts for two transcriptional factors were observed to increase at

182
E18 and gradually declined. Transcriptional factor 4 (Tcf4) increased and then
declines but still remains elevated and may play a critical role in RGC development.
Tcf4 controls myc expression, a cell survival gene (Hu and Rosenblum, 2005), and
may be the mechanism that mediates survival of RGCs that correctly reached their
target. It may also act independent of the myc pathway in regulating pituitary
growth and development (Brinkmeier et al., 2004). Early growth response 1 was
described previously.
The exact nature of the cell cycle gene expressed sequence AA673513 (also
known as U2AF homology motif (UHM) kinase 1 (Uhmk1 or kinase-interacting
stathmin (KIS)) is unknown. It interacts with p27kip and increases in Uhmk1
overcomes growth arrest as a result of p27 regulation (Boehm et al., 2002). This
kinase that contains an RNA binding motif (Maucuer et al., 1997) that interacts with
stathmin, a cytosolic protein that interacts with tubulin and destabilizes microtubule
organization. Aged stathmin knockout mice developed reduction of motor nerve
conduction, axonal degeneration, dysmylination and glial reactivity (Liedtke et al.,
2002). Stathmin 1 transcription decreases from E16 to E18, similarly to KIS, but it
does not return to E16 levels. Regulation of stathmin may be important in the
development of a functional CNS.

Cluster 6
The six transcripts in this cluster increased at E18 and returned to near E16
levels further on in RGC development. Three encoded for transcriptional regulators:

183
REST corepressor 1 (Rcor1), special AT-rich binding protein 1 (STAB1), and zinc
finger protein 354A. Rcor1 acts in concert with the zinc finger NRSE/REST to
recruit proteins involved in DNA methylation mediated gene silencing of neuronal-
specific genes (Lunyak et al., 2002). Since Rcor1 functions in non-neuronal cells
have been characterized, the rapid spike in Rcor1 transcript expression in neuronal
RGCs suggests a secondary role for Rcor1 in regulating neuronal maturation.
Another transcriptional regulator, zinc finger protein 354A is involved in renal cell
differentiation and may play a regulatory role in developing RGCs (Tekki-Kessaris
et al., 1999).
Two transcripts involved in metabolisms and biosynthesis also exhibited a
similar transcript expression pattern. Nudix (nucleoside diphosphate linked moiety
X)-type motif 13 is involved in isoprenoid biosynthesis, and also NADH
diphosphatase (Abdelraheim et al., 2003). Mitochondrial glycerol-3-phosphate
acyltransferase, plays a role in glycerolipid biosynthesis and is the key regulator of
cellular triacylglycerol (TAG) and phospholipid levels (Hammond et al., 2002). The
phospholipids are key components of membranes including synaptic vesicles.
The transcript of RIKEN cDNA 2310045A20 also demonstrated a similar
temporal expression profile, yet its function is unknown.

Cluster 7 and 8
Twelve of the 71 differentially expressed transcripts were grouped in these
clusters. The temporal expression pattern decreased from E16 levels at E18,

184
followed by a rise in transcript levels back to near E16 levels by P5.
The functional properties of the remaining transcript in this cluster remain
unclear. The cell surface S100 calcium binding protein A10 (calpactin) also known
as annexin II, is a member of the EF-hand superfamily like SPARC adhesion
molecule that share a high degree of sequence similarity. The transcript expression
of stathmin 1 is converse to KIS as mentioned earlier, and may be indicative of KIS
exerting a regulatory influence on the expression of stathmin 1 for proper RGC
development. Microfibril-associated glycoprotein 3 precursor (Mfap3), is part of the
extracellular matrix. Proteasome (prosome, macropain) subunit, alpha type 7 is
involved in protein degradation. The functional significance of H2-K region
expressed gene 2 and the RIKEN cDNA 2900010M23 are not well characterized.

Cluster 9
Transcriptional profiles of genes in this cluster dramatically decreased at E18
and rise to levels higher than the E16 baseline by P5. Of those, a cell cycle
regulatory soluble 1 galactose binding lectin, an autocrine regulator of cell
proliferation with role in maintenance of G0, may play a role in CNS neurite growth
since recombinant galectin-1 has recently been shown to promote the rate of
peripheral nerve regeneration (McGraw et al., 2004) as well as a signaling molecule
(Gauthier et al., 2002).
The role of the transcriptional factor RIKEN cDNA A230102L03 gene,
renamed: (Atbf1) AT motif binding factor 1 is unclear nor the implication of

185
interactions with the transcriptional factor Special AT binding mentioned in cluster 6.
Procollagen, type IV, alpha 1 is part of the extracellular matrix. There is an
emerging role of laminin directed synapse formation (Yamagata et al., 2002). The
differential expression suggests that more mature and functional RGCs also help
direct the correct synaptic formation with other retinal cell types through secretory
signals.

186
Chapter III: References
Abdelraheim SR, Spiller DG, McLennan AG. (2003) Mammalian NADH
diphosphatases of the Nudix family: cloning and characterization of the
human peroxisomal NUDT12 protein. Biochem J. 374(Pt 2): 329-35.

Antonicek H and Schachner M. (1988) The adhesion molecule on glia (AMOG)
incorporated into lipid vesicles binds to subpopulations of neurons. J
Neurosci. 8: 2961-6.

Ashery-Padan R and Gruss P. (2001) Pax6 lights-up the way for eye development.
Current Opinion in Cell Biology 13(6): 706-14.

Metcalfe SM, Muthukumarana PA, Thompson HL, Haendel MA, Lyons GE. (2005)
Leukemia inhibitory factor (LIF) is functionally linked to axotrophin and
both LIF and axotrophin are linked to regulatory immune tolerance. FEBS
Lett. 579(3): 609-14.

Barres BA, Silverstein BE, Corey DP, Chun LL. (1988) Immunological,
morphological, and electrophysiological variation among retinal ganglion
cells purified by panning. Neuron 1: 791–803.

Belachew S, Chittajallu R, Aguirre AA, Yuan X, Kirby M, Anderson S, Gallo V.
(2003) Postnatal NG2 proteoglycan-expressing progenitor cells are
intrinsically multipotent and generate functional neurons. J Cell Biol. 161(1):
169-86.
Berglund EO, Murai KK, Fredette B, Sekerkova G, Marturano B, Weber L,
Mugnaini E, Ranscht B. (1999) Ataxia and abnormal cerebellar
microorganization in mice with ablated contactin gene expression. Neuron.
24(3): 739-50.

Bertrand N, Castro DS, Guillemot F. (2002) Proneural genes and the specification of
neural cell types. Nature Reviews Neuroscience 3(7): 517-30.

Bizzoca A, Virgintino D, Lorusso L, Buttiglione M, Yoshida L, Polizzi A, Tattoli M,
Cagiano R, Rossi F, Kozlov S, Furley A, Gennarini G. (2003) Transgenic
mice expressing F3/contactin from the TAG-1 promoter exhibit
developmentally regulated changes in the differentiation of cerebellar
neurons. Development. 130(1): 29-43.

Boehm M, Yoshimoto T, Crook MF, Nallamshetty S, True A, Nabel GJ, Nabel EG.
(2002) A growth factor-dependent nuclear kinase phosphorylates p27 (Kip1)
and regulates cell cycle progression. EMBO J. 21: 3390-401.

187
Brinkmeier ML, Potok MA, Cha KB, Gridley T, Stifani S, Meeldijk J, Clevers H,
Camper SA. (2003) TCF and Groucho-related genes influence pituitary
growth and development. Mol Endocrinol. 17(11): 2152-61.

Brown NL, Kanekar S, Vetter ML, Tucker PK, Gemza DL, Glaser T. (1998) Math5
encodes a murine basic helix-loop-helix transcription factor expressed during
early stages of retinal neurogenesis. Development 125(23):4821-33.

Cai D, Deng K, Mellado W, Lee J, Ratan R and Filbin M. (2002) Arginase I and
Polyamines Act Downstream from Cyclic AMP in Overcoming Inhibition of
Axonal Growth MAG and Myelin In Vitro. Neuron 35 (4): 711-719.

Cai D, Qiu J, Cao Z, McAtee M, Bregman BS, and Filbin MT. (2001) Neuronal
cyclic AMP controls the developmental loss in ability of axons to regenerate.
J Neurosci. 21(13): 4731-9.

Cai S, Han HJ, Kohwi-Shigematsu T. (2003) Tissue-specific nuclear architecture and
gene expression regulated by SATB1. Nat Genet. 34(1):42-51.

Carim-Todd L, Escarceller M, Estivill X, Sumoy L. (2003) LRRN6A/LERN1
(leucine-rich repeat neuronal protein 1), a novel gene with enriched
expression in limbic system and neocortex. Eur J Neurosci. 18(12):3167-82.

Chan CK, Pham LN, Chinn C, Spee C, Ryan SJ, Akhurst RJ, Hinton DR. (2004)
Mouse strain-dependent heterogeneity of resting limbal vasculature. Invest
Ophthalmol Vis Sci 45(2):441-7

Chen DF, Jhaveri S, Schneider GE. (1995) Intrinsic changes in developing retinal
neurons result in regenerative failure of their axons. PNAS 92: 7287-91.

Chen DF, Schneider GE, Martinou J, Susumu T. (1997) Bcl-2 promotes regeneration
of severed axons in mammalian CNS. Nature 385(6615): 434-39.

Chen H, Bouras C, Antonarakis SE. (1996) Cloning of the cDNA for a human
homolog of the rat PEP-19 gene and mapping to chromosome 21q22.2-q22.3.
Hum Genet. 98(6): 672-77.

Chen ZF, Paquette AJ, Anderson DJ. (1998) NRSF/REST is required in vivo for
repression of multiple neuronal target genes during embryogenesis. Nat
Genet. 20(2): 136-42.

Cepko CL, Austin CP, Yang X, Alexiades M, Ezzeddine D. (1996) Cell fate
determination in the vertebrate retina. PNAS 93(2):589-95.

188
Cohen J and Johnson AR. (1991) Differential effects of laminin and merosin on
neurite outgrowth by developing retinal ganglion cells. Journal of Cell
Science - Supplement 15:1-7.

Crackower MA, Sinasac DS, Xia J, Motoyama J, Prochazka M, Rommens JM,
Scherer SW, Tsui LC. (1999) Cloning and characterization of two
cytoplasmic dynein intermediate chain genes in mouse and human.
Genomics. 55(3): 257-67.

Crowley T, Brunori M, Rhee K, Wang X, Wolgemuth DJ. (2004) Change in nuclear-
cytoplasmic localization of a double-bromodomain protein during
proliferation and differentiation of mouse spinal cord and dorsal root ganglia.
Brain Res Dev Brain Res. 149(2): 93-101.

Fan GH, Lapierre LA, Goldenring JR, Sai J, Richmond A. (2004) Rab11-family
interacting protein 2 and myosin Vb are required for CXCR2 recycling and
receptor-mediated chemotaxis. Mol Biol Cell. 15(5): 2456-2469.

Farkas RH, Qian J, Goldberg JL, Quigley HA, Zack DJ. (2004) Gene expression
profiling of purified rat retinal ganglion cells. Invest Ophthalmol Vis Sci.
45(8): 2503-13.

Feng G, Hood R, Bernstein M, Keller-Peck C, Nguyen Q, Wallace M, Nerbonne JM,
Lichtman JW, Sanes JR. (2000) Imaging neuronal subsets in transgenic mice
expressing multiple spectral variants of GFP. Neuron 28: 41-51.

Gauthier L, Rossi B, Roux F, Termine E, Schiff C. (2002) Galectin-1 is a stromal
cell ligand of the pre-B cell receptor (BCR) implicated in synapse formation
between pre-B and stromal cells and in pre-BCR triggering. Proc Natl Acad
Sci USA. 99(20): 13014-19.

Gan L. Xiang M. Zhou L. Wagner DS. Klein WH. Nathans J. (1996) POU domain
factor Brn-3b is required for the development of a large set of retinal
ganglion cells. Proc Natl Acad Sci USA 93(9): 3920-25.

Gilman AG. (1987) G proteins: transducers of receptor-generated signals. Annu
Rev Biochem 56: 615-649.

Gloor S, Antonicek H, Sweadner KJ, Pagliusi S, Frank R, Moos M, Schachner M.
(1990) The adhesion molecule on glia (AMOG) is a homologue of the beta
subunit of the Na,K-ATPase. J Cell Biol. 110(1): 165-74.

Greenwald RJ, Tumang JR, Sinha A, Currier N, Cardiff RD, Rothstein TL, Faller
DV, Denis GV. (2004) E mu-BRD2 transgenic mice develop B-cell
lymphoma and leukemia. Blood. 103(4): 1475-84.

189

Godement P, Salaun J, Imbert M. (1984) Prenatal and postnatal development of
retinogeniculate and retinocollicular projections in the mouse. Journal of
Comparative Neurology 230(4): 552-75.

Goldberg JL, Klassen MP, Hua Y, Barres BA. (2002) Amacrine-signaled loss of
intrinsic axon growth ability by retinal ganglion cells. Science 296(5574):
1860-4.

Hammond LE, Gallagher PA, Wang S, Hiller S, Kluckman KD, Posey-Marcos EL,
Maeda N, Coleman RA. (2002) Mitochondrial glycerol-3-phosphate
acyltransferase-deficient mice have reduced weight and liver triacylglycerol
content and altered glycerolipid fatty acid composition. Mol Cell Biol.
22(23): 8204-14.

Hirayama S, Bujo H, Yamazaki H, Kanaki T, Takahashi K, Kobayashi J, Schneider
WJ, Saito Y. (2000) Differential expression of LR11 during proliferation and
differentiation of cultured neuroblastoma cells. Biochem Biophys Res
Commun. 275(2): 365-73.

Hu MC and Rosenblum ND. (2005) Smad1, beta-catenin and Tcf4 associate in a
molecular complex with the Myc promoter in dysplastic renal tissue and
cooperate to control Myc transcription. Development. 132(1): 215-25.

Huang X, Wu DY, Chen G, Manji H, Chen DF. (2003) Support of retinal ganglion
cell survival and axon regeneration by lithium through a Bcl-2-dependent
mechanism. Invest Ophthalmol Vis Sci. 44(1): 347-54.

Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U. (2003)
Speed TP. Exploration, normalization, and summaries of high density
oligonucleotide array probe level data. Biostatistics. 4(2): 249-64.

Jones MW, Errington ML, French PJ, Fine A, Bliss TV, Garel S, Charnay P, Bozon
B, Laroche S, Davis S. (2001) A requirement for the immediate early gene
Zif268 in the expression of late LTP and long-term memories. Nat Neurosci.
4(3): 289-96.

Jung J, Oh J, Lee K. (1998) Nucleotide and deduced amino acid sequences of rat
myosin binding protein H (MyBP-H). Arch Pharm Res. 21(6): 712-7.

Kalyanaraman S, Kalyanaraman V, and Gautam N. (1995) A brain-specific G
protein gamma subunit. Biochem. Biophys. Res. Commun. 216: 126–132.

190
Kay R, Rosten PM, Humphries RK. (1991) CD24, a signal transducer modulating B
cell activation responses, is a very short peptide with a glycosyl
phosphatidylinositol membrane anchor. J Immunol. 147(4): 1412-6.

Lauren J, Airaksinen MS, Saarma M, Timmusk T. (2003) Two novel mammalian
Nogo receptor homologs differentially expressed in the central and peripheral
nervous systems. Mol Cell Neurosci. 24(3): 581-94.

Lee YC, Tang YC, Chen YH, Wong CM, Tsou AP. (2003) Selenite-induced survival
of HuH7 hepatoma cells involves activation of focal adhesion kinase-
phosphatidylinositol 3-kinase-Akt pathway and Rac1. J Biol Chem. 278(41):
39615-24.

Li C and Wong WH. (2001) Model-based analysis of oligonucleotide arrays:
expression index computation and outlier detection. Proc Natl Acad Sci USA.
98(1): 31-36.

Liao Y, Shikapwashya ON, Shteyer E, Dieckgraefe BK, Hruz PW, Rudnick DA.
(2004) Delayed hepatocellular mitotic progression and impaired liver
regeneration in early growth response-1-deficient mice. J. Biol. Chem. 279:
43107-16.

Liedtke W, Leman E, Fyffe R, Raine C, Schubart U. (2002) Stathmin-deficient mice
develop an age-dependent axonopathy of the central and peripheral nervous
system Am J Path 160 (2): 469-80.

Liu C, Calogero A, Ragona G, Adamson E, Mercola D. (1996) EGR-1, the reluctant
suppression factor: EGR-1 is known to function in the regulation of growth,
differentiation, and also has significant tumor suppressor activity and a
mechanism involving the induction of TGF-beta1 is postulated to account for
this suppressor activity. Crit Rev Oncog. 7(1-2): 101-25. Review

Lunyak V, Burgess R, Prefontaine G, Nelson C, Sze S, Chenoweth J, Schwartz P,
Pevzner P, Glass C, Mandel G, Rosenfeld M. (2002) Corepressor-dependent
silencing of chromosomal regions encoding neuronal genes Science
298:1747-52.

Luzzi V, Holtschlag V, Watson MA. (2001) Expression profiling of ductal
carcinoma in situ by laser capture microdissection and high-density
oligonucleotide arrays. Am. J. Pathol., 158: 2005-2010.

Maguchi M, Nishida W, Kohara K, Kuwano A, Kondo I, Hiwada K. (1995)
Molecular cloning and gene mapping of human basic and acidic calponins.
Biochem Biophys Res Commun. 217(1):238-44.

191
Maucuer A, Ozon S, Manceau V, Gavet O, Lawler S, Curmi P, Sobel A. (1997) KIS
is a protein kinase with an RNA recognition motif. J Biol Chem. 272(37):
23151-6.

Marquardt T, Ashery-Paden R, Andrejewski N, Scardigil R, Guillemot F, Gruss P.
(2001) Pax6 is required for the multipotent state of retinal progenitor cells.
Cell 105: 43-55.

Matveeva EA, Whiteheart SW, Slevin JT. (2003) Accumulation of 7S SNARE
complexes in hippocampal synaptosomes from chronically kindled rats. J
Neurochem. 84(3): 621-4.

McGraw J, McPhail LT, Oschipok LW, Horie H, Poirier F, Steeves JD, Ramer MS,
Tetzlaff W. (2004) Galectin-1 in regenerating motoneurons. Eur J Neurosci.
20(11): 2872-80.

Messer A, Snodgrass GL, Maskin P. (1984) Enhanced survival of cultured cerebellar
Purkinje cells by plating on antibody to Thy-1. Cell Mol Neurobiol. 4(3):
285-90.

Meyer-Franke A, Kaplan MR. Pfrieger FW. Barres BA. (1995) Characterization of
the signaling interactions that promote the survival and growth of developing
retinal ganglion cells in culture. Neuron 15(4):805-19.

Meyer-Franke A, Wilkinson GA, Kruttgen A, Hu M, Munro E, Hanson MG Jr,
Reichardt LF, Barres BA. (1998) Depolarization and cAMP elevation rapidly
recruit TrkB to the plasma membrane of CNS neurons. Neuron. 21(4): 681-
93.

Miwa JM, Ibanez-Tallon I, Crabtree GW, Sanchez R, Sali A, Role LW, Heintz N.
(1999) Lynx1, an endogenous toxin-like modulator of nicotinic acetylcholine
receptors in the mammalian CNS. Neuron. 23(1): 105-14.

Mori H, Sakakibara S, Imai T, Nakamura Y, Iijima T, Suzuki A, Yuasa Y, Takeda M,
Okano H. (2001) Expression of mouse igf2 mRNA-binding protein 3 and its
implications for the developing central nervous system. J Neurosci Res.
64(2): 132-43.

Morris RJ and Barber PC. (1983) Fixation of Thy-1 nervous tissue for
immunohistochemistry. J. Histochem. Cytochem. 31: 263-274.

Murata Y, Sun-Wada GH, Yoshimizu T, Yamamoto A, Wada Y, Futai M. (2002)
Differential localization of the vacuolar H+ pump with G subunit isoforms
(G1 and G2) in mouse neurons. J Biol Chem. 277(39): 36296-303.

192
Nagashima K, Torii S, Yi Z, Igarashi M, Okamoto K, Takeuchi T, Izumi T. (2002)
Melanophilin directly links Rab27a and myosin Va through its distinct
coiled-coil regions. FEBS Lett. 517(1-3): 233-8.

Nolo R, Abbott LA, Bellen HJ. (2000) Senseless, a Zn finger transcription factor, is
necessary and sufficient for sensory organ development in Drosophila. Cell.
102(3): 349-62.

Oakley BR, Oakley CE, Rinehart JE. (1987) Conditionally lethal tubA alpha-tubulin
mutations in Aspergillus nidulans. Mol Gen Genet. 208(1-2): 135-44.

Pal DK, Evgrafov OV, Tabares P, Zhang F, Durner M, Greenberg DA. (2003) BRD2
(RING3) is a probable major susceptibility gene for common juvenile
myoclonic epilepsy. Am J Hum Genet. 73(2): 261-70.

Panzenboeck U, Wintersberger A, Levak-Frank S, Zimmermann R, Zechner R,
Kostner GM, Malle E, Sattler W. (1997) Implications of endogenous and
exogenous lipoprotein lipase for the selective uptake of HDL3-associated
cholesteryl esters by mouse peritoneal macrophages. J Lipid Res. 38(2): 239-
53.

Praetorius J, Nejsum LN, Nielsen S. (2004) A SCL4A10 gene product maps
selectively to the basolateral plasma membrane of choroid plexus epithelial
cells. Am J Physiol Cell Physiol. 286(3): C601-10.

Rezajooi K, Pavlides M, Winterbottom J, Stallcup WB, Hamlyn PJ, Lieberman AR,
Anderson PN. (2004) NG2 proteoglycan expression in the peripheral nervous
system: upregulation following injury and comparison with CNS lesions. Mol
Cell Neurosci. 25(4): 572-84.

Rothman JE. (1994) Mechanisms of intracellular protein transport. Nature.
372(6501): 55-63. Review

Sassa T, Gomi H, Itohara S. (2004) Postnatal expression of Cdkl2 in mouse brain
revealed by LacZ inserted into the Cdkl2 locus. Cell Tissue Res. 315(2): 147-
56.

Scammell JG, Reddy S, Valentine DL, Coker TN, Nikolopoulos SN, Ross RA.
(2000) Isolation and characterization of the human secretogranin II gene
promoter. Brain Res Mol Brain Res. 75(1): 8-15.

Scherzer CR, Offe K, Gearing M, Rees HD, Fang G, Heilman CJ, Schaller C, Bujo
H, Levey AI, Lah JJ. (2004) Loss of apolipoprotein E receptor LR11 in
Alzheimer disease. Arch Neurol. 61(8): 1200-5.

193
Schoenherr CJ and Anderson DJ. (1995) The neuron-restrictive silencer factor
(NRSF): a coordinate repressor of multiple neuron-specific genes. Science.
267(5202): 1360-3.

Seoane J, Le HV, Shen L, Anderson SA, Massague J. (2004) Integration of Smad
and forkhead pathways in the control of neuroepithelial and glioblastoma cell
proliferation. Cell. 117(2): 211-23.

Serini G, Valdembri D, Zanivan S, Morterra G, Burkhardt C, Caccavari F,
Zammataro L, Primo L, Tamagnone L, Logan M, Tessier-Lavigne M,
Taniguchi M, Puschel AW, Bussolino F. (2003) Class 3 semaphorins control
vascular morphogenesis by inhibiting integrin function. Nature 424: 391–397.

Shewan D, Berry M, Cohen J. (1995) Extensive regeneration in vitro by early
embryonic neurons on immature and adult CNS tissue. Journal of
Neuroscience 15(3 Pt 1): 2057-62.

Singec I, Knoth R, Ditter M, Hagemeyer CE, Rosenbrock H, Frotscher M, Volk B.
(2002) Synaptic vesicle protein synaptoporin is differently expressed by
subpopulations of mouse hippocampal neurons. J Comp Neurol. 452(2): 139-
53.

Sun W, Li N, He S. (2002) Large-scale morphological survey of mouse retinal
ganglion cells. J Comp Neurol 451: 115-126.

Tekki-Kessaris N, Bonventre JV, Boulter CA. (1999) Characterization of the mouse
Kid1 gene and identification of a highly related gene, Kid2. Gene. 240(1):
13-22.

Vidal-Sanz M, Bray GM, Villegas-Perez MP, Thanos S, Aguayo AJ. (1987) Axonal
regeneration and synapse formation in the superior colliculus by retinal
ganglion cells in the adult rat. Journal of Neuroscience 7(9): 2894-909.

Wang B, Xiao Y, Ding BB, Zhang N, Yuan X, Gui L, Qian KX, Duan S, Chen Z,
Rao Y, Geng JG. (2003) Induction of tumor angiogenesis by Slit-Robo
signaling and inhibition of cancer growth by blocking Robo activity. Cancer
Cell 4: 19–29.

Wang KC, Kim JA, Sivasankaran R, Segal R, He Z. (2002) P75 interacts with the
Nogo receptor as a co-receptor for Nogo, MAG and OMgp. Nature.
420(6911): 74-8.

Wu DY, Zheng JQ, McDonald MA, Chang B, and Twiss JL. (2003) PKC isozymes
in the enhanced regrowth of retinal neurites after optic nerve injury. Invest.
Ophthalmol. Vis. Sci. 44: 2783-90.

194

Yamagata M, Weiner JA, Sanes JR. (2002) Sidekicks: synaptic adhesion molecules
that promote lamina-specific connectivity in the retina. Cell. 110(5): 649-60.

Yoon H, Liyanarachchi S, Wright FA, Davuluri R, Lockman JC, de la Chapelle A,
Pellegata NS. (2002) Gene expression profiling of isogenic cells with
different TP53 gene dosage reveals numerous genes that are affected by TP53
dosage and identifies CSPG2 as a direct target of p53. Proc Natl Acad Sci
USA. 99(24): 15632-7

Zhu Q and Julien JP. (1999) A key role for GAP-43 in the retinotectal topographic
organization. Experimental Neurology 155(2):228-42.

Zimmer A, Zimmer AM, Baffi J, Usdin T, Reynolds K, Konig M, Palkovits M,
Mezey E. (1998) Hypoalgesia in mice with a targeted deletion of the
tachykinin 1 gene. Proc Natl Acad Sci USA. 95(5): 2630-5.

195
Chapter IV: Conclusions and Implications

Unique patterns of gene expression determine biochemical signaling cascades
and thus the ability of a cell to respond to changing environmental stimuli. The
phototransduction cascade of molecular events has been characterized in rod
photoreceptors. Based on these studies and the sequence homology with proteins
expressed in cone photoreceptors, cone phototransduction is thought to be quite
similar. Many signaling pathways critical for vision are not as well understood as
phototransduction. RGCs are critical in conveying neuronal impulses from the retina
to the brain, but little is known about the molecular events involved with the rapid
loss of axonal growth ability in developing retinal ganglion cells.
To identify the contributions of specific genes to cellular signaling pathways,
the functional significance of the CAR protein in photoreceptors was characterized
and the transcripts of developing RGCs were screen for differential expression. We
provide evidence that CAR is involved in phototransduction termination and thus is
effective in protecting photoreceptors from constitutive phototransduction cascade
activation induced cell death (Chapter II). CAR in vivo forms a less stable complex
with rhodopsin, suggesting that it may also form less stable complexes with
physiologic cone opsin proteins. These characteristics of CAR may contribute to the
functional differences between rods and cones. Through a transcript screening of
purified RGCs, we provide a starting point for functional studies of developmentally
dependent differentially transcribed genes in RGC development and physiology

196
(Chapter III). Differential gene expression has long been used as the initial step to
identify molecular factors that are involved in molecular cascades when little is
known.
Direct evidence of the cone phototransduction mechanism mediating high
acuity color vision is lacking. The genetic approach of expressing CAR in rods
allows the direct functional comparison of CAR with rod arrestin. Using a similar
experimental scheme as outlined (Chapter II), the introduction of presumptive cone
phototransduction proteins into the well-characterized rod system enables the
evaluation of their biochemical and functional characteristics in phototransduction.
Null mutants inactivating many of the key rod phototransduction cascade elements
including rhodopsin (Lem et al., 1991), transducin (Calvert et al., 2000), cyclase-
activating proteins (GCAPs) (Mendez et al., 2001), rhodopsin kinase (Chen CK et al.,
1999), arrestin (Xu et al., 1997), and recoverin (Makino et al., 2004) are available.
The homologous cone proteins are already identified.
Ectopic introduction of individual cone proteins into rod photoreceptors using
the cell specific rhodopsin promoter will generate enough material for effective in
vivo biochemical and physiological analysis. This method is advantageous in
providing a similar and controlled environment to compare the function of
individually introduced cone homologous phototransduction proteins with their rod
counterpart. Since the photoresponse of monkey rods are 50-100 times more
sensitive, but several times slower than monkey cones (Baylor et al., 1984; Schnapf
et al. 1990), the different isoforms of phototransduction proteins, their interactions

197
with each other, and a unique cone OS structure may all contribute to rod and cone
functional differences. By isolating the individual component, we can determine the
functional contribution of a cone protein to the cone photoresponse.
The ultimate goal is to understand the functional contributions of all
interacting phototransduction proteins to the cone photoresponse. The data presented
suggest that CAR does function and is important in the termination of a rod
photoresponse. Further confirmation of endogenous CAR’s physiological role
requires a cone specific environment. The transcriptional factor neural retina leucine
zipper (nrl) is preferentially expressed in rod photoreceptors. When genetically
ablated (nrl
-/-
), these retinas lack rod function and rod specific gene expression with a
concomitant increase in S-cone function and gene expression (Mears et al., 2001).
The abnormal morphology suggesting a rod-cone intermediate may allow for
biochemical studies of S-cones. More recent studies of the structural, ultrastructural,
histochemical, molecular, and electrophysiological characteristics of these nrl
-/-

photoreceptors suggest they are significantly cone-like, yet it is unclear the
functional impact may contribute since their outer segments are shorter than those of
wide-type cones, exhibit disorder, and appear to deteriorate (Daniele et al., 2005).
Interestingly, these cone-like photoreceptors functionally deteriorate when exposed
to light flashes (Nikonov et al., 2005), but cones function predominately in bright
light conditions. The nrl
-/-
photoreceptors have cone-like properties, but a better
model for physiological cone function needs to be developed.
We have worked to generate transgenic mice lines with cone cells labeled

198
with enhance green fluorescence protein (eGFP) under the control of the
characterized human blue opsin promoter (Chen et al., 1994) to quickly and easily
identify cone photoreceptors. Though to date, we have been unsuccessful. The
creation of human red/green opsin promoter controlling green fluorescence protein
(GFP) in mouse cone photoreceptors (Fei, 2003) will be useful in characterizing the
physiological function of specific genes within the native cone environment. The
living cell marker GFP allows the identification and isolation of cone cells for
analysis.
The introduction of cone phototransduction proteins in rod photoreceptors is
currently the best methodology of comparing the in vivo biochemical and functional
characteristics of homologous cone proteins with their rod counterpart. Our lab has
transgenic mice expressing S-cone opsin in the rod photoreceptor (Dr. Shi et al.,
unpublished data). We have generated double transgenic mice with CAR and cone
opsin in an effort to characterize their molecular interactions and the effect on the
photoresponse. Preliminary data suggests that mCAR deactivates S-opsin poorly,
but rod arrestin (arrestin1) is effective. Yet, the CAR-cone opsin complex may not
be stable as a rod arrestin-rhodopsin complex. It should be stable enough to initiate
complete electrophysiological signal termination. The inability of CAR to terminate
cone opsin initiated signaling in wild-type cone cells would lead to constitutive
cascade activation, but this does not occur in wild type cones. The identification of
arrestin1 in cones (Zhu et al, 2005, 46: ARVO Abstract 1179) in conjunction to CAR
as well as CAR inability to terminate signaling alone suggests a possible cooperative

199
relationship. Continuously active signally is known to cause cell death, but wild-
type cones are obviously viable, and thus signal termination definitely occurs. It is
possible that CAR binds light activated phosphorylated opsins transiently prior to
arrestin1. This mechanism may explain the faster initial termination of the cone
photoresponse compared to rods.
The ability to identify and purify specific cell types is critical to dissect the
functional significance of specifically expressed genes. Coupled to information on
homologous protein function and structural analysis of common motifs, cell specific
gene contributions can be assessed in defined systems as we are accomplishing in
rod photoreceptors. This is exceedingly difficult when the components in a
molecular pathway have yet to be identified.
Through a developmental approach, we provided evidence of differential
gene regulation in developing RGCs during the loss of axonal growth abilities.
Unlike immature CNS neurons, mature CNS neurons fail to regenerate correct
axonal connections once lesioned. As a result, lesions from trauma, stroke, and
degenerative diseases lead to permanent paralysis, blindness, hearing loss, and other
implications. It is known that when mRNA synthesis is permanently disrupted with
α-amanitin or transiently with 5,6-dichlorobenzimidazol riboside (DRB), axon
growth was inhibited (Smith and Skene, 1997). Transcription dependent loss of
axon growth suggests that differential transcript profiles will identify molecular
factors involved in the loss of axon growth cascade.
We provide evidence of possible candidates that may mediate this loss of

200
axonal growth. Magnetic assisted cell sorting technique was utilized to effectively
and consistently isolate highly pure RGCs for in vitro cultures and biochemical
analysis. We have identified differentially expressed transcripts from in vivo
developing RGCs, established an in vitro assay to assess neurite growth, and further
determined the differential gene expression for candidate genes that may be involved
in axonal growth in our in vitro cultures. Attempts to transfect purified RGCs and
express ectopic genes have been unsuccessful to date. Specific genes can be
delivered via adenovirus into in vitro cultures of RGCs (Goldberg et al., 2002) that
would facilitate the functional characterization of the identified differentially
expressed genes.
Our transcript screen of immature RGCs that can extend neurites to their
target compared to mature RGCs that no longer extend their neurites is very likely to
identify novel genes involved in neurite growth and development. The transfection
of the identified developmentally differentially expressed genes coupled with the
quantifiable neurite growth assay will be used to characterize molecular factors
involved in RGC development, axonal growth, and survival. Using the same
technique to isolate immature and mature RGCs provides exceptional controls for
gain and loss of axonal growth function in different media conditions and when
different genes are introduced.
Unbiased comprehensive high throughput microarray profiling and ISH
analysis facilitates the identification of developmentally dynamic genes mediating
the molecular events responsible for phenotypic changes in developing RGCs. The

201
additional characterization of in vitro cultured RGCs during the loss of axonal
growth periods lends more information on physiologically relevant candidates. The
exact functional nature of the identified differentially transcribed genes is unclear in
many cases and needs further analysis.
Our profiling of directly isolated RNA from developing RGCs coupled with
in vitro profiling of these same expressed genes is advantageous in assessing
potentially real in vivo relevant genes that may play a role in developing functional
RGCs. When both are similarly expressed, there is a higher likelihood that they are
functionally relevant and not just experimental artifacts. The speed and quantitative
nature of the neurite growth assays from in vitro cultures can also be used as a
quantifiable characteristic of maturation. These data serve as the first steps to
identify the molecular factors involved in axonal growth in the CNS as well as
potentially identifying genes that define mature RGCs. Understanding the factors
controlling the switch from immature CNS neurons capable of sending out extensive
axons to a more mature CNS neuron with limited growth abilities is important in
developing therapies for individuals to recover from CNS injury.
These data, and perhaps more importantly, the methodology presented to
evaluate presumptive cone phototransduction components in a well defined rod
photoreceptor system and to identify and characterize candidate molecular factors
involved in neurite growth are important in determining the contributions of
individual genes to a cell’s functional response to external stimuli. The identification
and study of expressed genes involved in specific retinal cells are essential not only

202
in understanding and preserving sight, characterizing other similar signaling
cascades, and preserving neurons, but also in ultimately harnessing the power of cell
based and cell targeted therapies.

203
Chapter IV: References
Baylor DA, Nunn BJ, and Schnapf JL. (1984) The photocurrent, noise and spectral
sensitivity of rods of the monkey Macaca fascicularis. J Physiol 357: 575-607

Calvert PD, Krasnoperova NV, Lyubarsky AL, Isayama T, Nicolo M, Kosaras B,
Wong G, Gannon KS, Margolskee RF, Sidman RL, et al.. (2000)
Phototransduction in transgenic mice after targeted deletion of the rod
transducin alpha -subunit. Proc Natl Acad Sci USA 97: 13913-13918.

Chen CK, Burns ME, Spencer M, Niemi GA, Chen J, Hurley JB, Baylor DA, and
Simon MI (1999) Abnormal photoresponses and light-induced apoptosis in
rods lacking rhodopsin kinase. Proc Natl Acad Sci USA 96:3718–3722.

Chen J, Tucker CL, Woodford B, Szel A, Lem J, Gianella-Borradori A, Simon MI,
Bogenmann E. (1994) The human blue opsin promoter directs transgene
expression in short-wave cones and biopolar catl Acad Sci USA
USAxpression in short-wave cone 91: 2611-2615.

Daniele LL, Lillo C, Lyubarsky AL, Nikonov SS, Philp N, Mears AJ, Swaroop A,
Williams DS, and Pugh EN. (2005) Cone-like Morphological, Molecular, and
Electrophysiological Features of the Photoreceptors of the Nrl Knockout
Mouse. IOVS 46: 2156-2167.)

Farkas RH, Qian J, Goldberg JL, Quigley HA, Zack DJ. (2004) Gene expression
profiling of purified rat retinal ganglion cells. Invest Ophthalmol Vis Sci.
45(8): 2503-13.

Fei Y. (2003) Development of the cone photoreceptor mosaic in the mouse retina
revealed by fluorescent cones in transgenic mice. Mol Vis 9: 31-42.

Goldberg JL, Klassen MP, Hua Y, Barres BA. (2002) Amacrine-signaled loss of
intrinsic axon growth ability by retinal ganglion cells. Science 296: 1860-
1864.

Lem J, Krasnoperova NV, Calvert PD, Kosaras B, Cameron DA, Nicolo M, Makino
CL, and Sidman RL. (1999) Morphological, physiological, and biochemical
changes in rhodopsin knockout mice. Proc Natl Acad Sci USA 96: 736-741.

Makino CL, Dodd RL, Chen J, Burns ME, Roca A, Simon MI, and Baylor DA.
(2004) Recoverin regulates light-dependent phosphodiesterase activity in
retinal rods. J Gen Physiol 123: 729-741.

Mears AJ, Kondo M, Swain PK, Takada Y, Bush RA, Saunders TL, Sieving PA,
Swaroop A. (2001) Nrl is required for rod photoreceptor development. Nat
Genet 29: 447-452.

204

Mendez A, Burns ME, Sokal I, Dizhoor AM, Baehr W, Palczewski K, Baylor DA,
and Chen J. (2001) Role of guanylate cyclase-activating proteins (GCAPs) in
setting the flash sensitivity of rod photoreceptors. Proc Natl Acad Sci USA
98: 9948-9953.

Nikonov SS, Daniele LL, Zhu X, Craft CM, and Swaroop A. (2005) Photoreceptors
of Nrl-/-Mice Coexpress Functional S-and M-cone Opsins Having Distinct
Inactivation. J Gen Physiology, 125: 287-304

Schnapf J L, Nunn B J, Meister M, and Baylor DA. (1990) Visual transduction in
cones of the monkey Macaca fascicularis. J Physiol 427: 681-713.

Smith DS and Skene JH. (1997) A transcription-dependent switch controls
competence of adult neurons for distinct modes of axon growth. J Neurosci
17(2): 646-658.

Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA, Chen J. (1997) Prolonged
photoresponses in transgenic mouse rods lacking arrestin. Nature. 389: 505-9.

205
Bibliography

Abdelraheim SR, Spiller DG, McLennan AG. (2003) Mammalian NADH
diphosphatases of the Nudix family: cloning and characterization of the
human peroxisomal NUDT12 protein. Biochem J. 374(Pt 2): 329-35.

Aguayo AJ, Vidal-Sanz M, Villegas-Perez MP, and Bray GM. (1987) Growth and
connectivity of axotomized retinal neurons in adult rats with optic nerves
substituted by PNS grafts linking the eye and the midbrain. Ann N Y Acad
Sci. 495: 1-9.

Antonicek H and Schachner M. (1988) The adhesion molecule on glia (AMOG)
incorporated into lipid vesicles binds to subpopulations of neurons. J
Neurosci. 8: 2961-6.

Arshavsky V and Bownds MD. (1992) Regulation of deactivation of photoreceptor
G protein by its target enzyme and cGMP. Nature 357: 416-17.

Ashery-Padan R and Gruss P. (2001) Pax6 lights-up the way for eye development.
Current Opinion in Cell Biology 13(6): 706-14.

Baetge EE. Hammang JP. Neurite outgrowth in PC12 cells deficient in GAP-43.
Neuron. 6(1):21-30, 1991.

Baker RK, Haendel MA, Swanson BJ, Shambaugh JC, Micales BK and Lyons GE.
(1997) In vitro preselection of gene-trapped embryonic stem cells for
characterizing novel developmentally regulated genes in the mouse. Dev.
Biol. 185: 201–214

Barlow JZ and Huntley GW. (2000) Developmentally regulated expression of Thy-1
in structures of the mouse sensory-motor system. J. Comp. Neurol. 421: 215-
33.

Barlow JZ, Kelley KA, Bozdagi O, Huntley GW. (2002) Testing the role of the cell-
surface molecule Thy-1 in regeneration and plasticity of connectivity in the
CNS. Neurosci. 111(4): 837-52.

Barres BA, Silverstein BE, Corey DP, Chun LL. (1988) Immunological,
morphological, and electrophysiological variation among retinal ganglion
cells purified by panning. Neuron 1: 791–803.

Baylor DA, Nunn BJ, and Schnapf JL. (1984) The photocurrent, noise and spectral
sensitivity of rods of the monkey Macaca fascicularis. J Physiol 357: 575-607

206
Belachew S, Chittajallu R, Aguirre AA, Yuan X, Kirby M, Anderson S, Gallo V.
(2003) Postnatal NG2 proteoglycan-expressing progenitor cells are
intrinsically multipotent and generate functional neurons. J Cell Biol. 161(1):
169-86.

Berglund EO, Murai KK, Fredette B, Sekerkova G, Marturano B, Weber L,
Mugnaini E, Ranscht B. (1999) Ataxia and abnormal cerebellar
microorganization in mice with ablated contactin gene expression. Neuron.
24(3): 739-50.

Bertrand N, Castro DS, Guillemot F. (2002) Proneural genes and the specification of
neural cell types. Nature Reviews Neuroscience 3(7): 517-30.

Bizzoca A, Virgintino D, Lorusso L, Buttiglione M, Yoshida L, Polizzi A, Tattoli M,
Cagiano R, Rossi F, Kozlov S, Furley A, Gennarini G. (2003) Transgenic
mice expressing F3/contactin from the TAG-1 promoter exhibit
developmentally regulated changes in the differentiation of cerebellar
neurons. Development. 130(1): 29-43.

Blackshaw S, Fraioli RE, Furukawa T, Cepko CL. (2001) Comprehensive analysis of
photoreceptor gene expression and the identification of candidate retinal
disease genes. Cell. 107(5): 579-89

Boehm M, Yoshimoto T, Crook MF, Nallamshetty S, True A, Nabel GJ, Nabel EG.
(2002) A growth factor-dependent nuclear kinase phosphorylates p27(Kip1)
and regulates cell cycle progression. EMBO J. 21: 3390-401,

Bomze HM. Bulsara KR. Iskandar BJ. Caroni P. Skene JH. Spinal axon regeneration
evoked by replacing two growth cone proteins in adult neurons. Nature
Neuroscience 4(1):38-43, 2001.

Bradford MM. (1976) A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal Biochem. 72:248-54.

Bregman BS, Goldberger ME. (1982) Anatomical plasticity and sparing of function
after spinal cord damage in neo natal cats. Science. 217(4559): 553-555.

Bregman BS, Kunkel-Bagden E, Reier PJ, Dai HN, McAtee M, Gao D. (1993)
Recovery of function after spinal cord injury: Mechanisms underlying
transplant-mediated recovery of function differ after spinal cord injury in
newborn and adult rats. Experimental Neurology. 123(1): 3-16.

207
Brinkmeier ML, Potok MA, Cha KB, Gridley T, Stifani S, Meeldijk J, Clevers H,
Camper SA. (2003) TCF and Groucho-related genes influence pituitary
growth and development. Mol Endocrinol. 17(11): 2152-61.

Broekhuyse RM, Tolhuizen EF, Janssen AP, Winkens HJ. (1985) Light induced
shift and binding of S-antigen in retinal rods. Curr Eye Res. 5:613-8.

Brown NL, Kanekar S, Vetter ML, Tucker PK, Gemza DL, Glaser T. (1998) Math5
encodes a murine basic helix-loop-helix transcription factor expressed during
early stages of retinal neurogenesis. Development 125(23):4821-33.

Cai D, Deng K, Mellado W, Lee J, Ratan R and Filbin M. (2002) Arginase I and
Polyamines Act Downstream from Cyclic AMP in Overcoming Inhibition of
Axonal Growth MAG and Myelin In Vitro. Neuron 35 (4): 711-719.

Cai D, Qiu J, Cao Z, McAtee M, Bregman BS, and Filbin MT. (2001) Neuronal
cyclic AMP controls the developmental loss in ability of axons to regenerate.
J Neurosci. 21(13): 4731-9.

Cai D, Shen Y, De Bellard M, Tang S, and Filbin MT. (1999) Prior exposure to
neurotrophins blocks inhibition of axonal regeneration by MAG and myelin
via a cAMP-dependent mechanism. Neuron. 22(1): 89-101.

Cai S, Han HJ, Kohwi-Shigematsu T. (2003) Tissue-specific nuclear architecture and
gene expression regulated by SATB1. Nat Genet. 34(1):42-51.

Calvert PD, Krasnoperova NV, Lyubarsky AL, Isayama T, Nicolo M, Kosaras B,
Wong G, Gannon KS, Margolskee RF, Sidman RL, et al.. (2000)
Phototransduction in transgenic mice after targeted deletion of the rod
transducin alpha -subunit. Proc Natl Acad Sci USA 97: 13913-13918.

Carim-Todd L, Escarceller M, Estivill X, Sumoy L. (2003) LRRN6A/LERN1
(leucine-rich repeat neuronal protein 1), a novel gene with enriched
expression in limbic system and neocortex. Eur J Neurosci. 18(12):3167-82.

Caroni P and Schwab ME (1988) Antibody against myelin-associated inhibitor of
neurite growth neutralizes nonpermissive substrate properties of CNS white
matter Neuron. 1(1):85-96.

Carter-Dawson LD and LaVail MM. (1979) Rods and cones in the mouse retina. I.
Structural analysis using light and electron microscopy. J Comp Neurol.
188(2): 245-62.

208
Chan CK, Pham LN, Chinn C, Spee C, Ryan SJ, Akhurst RJ, Hinton DR. (2004)
Mouse strain-dependent heterogeneity of resting limbal vasculature. Invest
Ophthalmol Vis Sci 45(2):441-7

Chen CK, Burns ME, He W, Wensel TG, Baylor DA, Simon MI. (2000) Slowed
recovery of rod photoresponse in mice lacking the GTPase accelerating
protein RGS9-1. Nature 403: 557-60.

Chen CK, Burns ME, Spencer M, Niemi GA, Chen J, Hurley JB, Baylor DA, Simon
MI. (1999) Abnormal photoresponses and light-induced apoptosis in rods
lacking rhodopsin kinase. Proc Natl Acad Sci USA 96: 3718-22.

Chen DF, Jhaveri S, Schneider GE. (1995) Intrinsic changes in developing retinal
neurons result in regenerative failure of their axons. PNAS 92: 7287-91.

Chen DF, Schneider GE, Martinou J, Susumu T. (1997) Bcl-2 promotes regeneration
of severed axons in mammalian CNS. Nature 385(6615): 434-39.

Chen H, Bouras C, Antonarakis SE. (1996) Cloning of the cDNA for a human
homolog of the rat PEP-19 gene and mapping to chromosome 21q22.2-q22.3.
Hum Genet. 98(6): 672-77.

Chen J, Makino CL, Peachey NS, Baylor DA, Simon MI. (1995) Mechanisms of
rhodopsin inactivation in vivo as revealed by a COOH-terminal truncation
mutant. Science 267: 374-77

Chen J, Simon MI, Matthes MT, Yasumura D, LaVail MM. (1999) 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.

Chen J, Tucker CL, Woodford B, Szel A, Lem J, Gianella-Borradori A, Simon MI,
Bogenmann E. (1994) The human blue opsin promoter directs transgene
expression in short-wave cones and biopolar cells in the mouse retina. Proc
Natl Acad Sci USA 91: 2611-2615.

Chen MS, Huber AB, van der Haar ME, Frank M, Schnell L, Spillmann AA, Christ
F, and Schwab ME. (2000) Nogo-A is a myelin-associated neurite outgrowth
inhibitor and an antigen for monoclonal antibody IN-1. Nature. 403(6768):
434-439.

Chen ZF, Paquette AJ, Anderson DJ. (1998) NRSF/REST is required in vivo for
repression of multiple neuronal target genes during embryogenesis. Nat
Genet. 20(2): 136-42.

209
Chiaramello A. Neuman T. Peavy DR. Zuber MX. The GAP-43 gene is a direct
downstream target of the basic helix-loop-helix transcription factors. Journal
of Biological Chemistry 271(36):22035-43, 1996.

Chowers I, Liu D, Farkas RH, Gunatilaka TL, Hackam AS, Bernstein SL,
Campochiaro PA, Parmigiani G, Zack DJ. (2003) Gene expression variation
in the adult human retina. Human Molecular Genetics. 12(22): 2881-93.

Chuang JZ and Sung CH. (1998) The cytoplasmic tail of rhodopsin acts as a novel
apical sorting signal in polarized MDCK cells. J Cell Biol. 142(5): 1245-56.

Cepko CL, Austin CP, Yang X, Alexiades M, Ezzeddine D. (1996) Cell fate
determination in the vertebrate retina. PNAS 93(2):589-95.

Cohen J and Johnson AR. (1991) Differential effects of laminin and merosin on
neurite outgrowth by developing retinal ganglion cells. Journal of Cell
Science - Supplement 15:1-7.

Crackower MA, Sinasac DS, Xia J, Motoyama J, Prochazka M, Rommens JM,
Scherer SW, Tsui LC. (1999) Cloning and characterization of two
cytoplasmic dynein intermediate chain genes in mouse and human.
Genomics. 55(3): 257-67.

Craft CM, Whitmore DH. (1995) The arrestin superfamily: cone arrestins are a
fourth family. FEBS Lett. 362(2): 247-55.

Craft CM, Whitmore DH, Wiechmann AF. (1994) Cone arrestin identified by
targeting expression of a functional family. J Biol Chem 269: 4613-4619.

Crowley T, Brunori M, Rhee K, Wang X, Wolgemuth DJ. (2004) Change in nuclear-
cytoplasmic localization of a double-bromodomain protein during
proliferation and differentiation of mouse spinal cord and dorsal root ganglia.
Brain Res Dev Brain Res. 149(2): 93-101.

Daniele LL, Lillo C, Lyubarsky AL, Nikonov SS, Philp N, Mears AJ, Swaroop A,
Williams DS, and Pugh EN. (2005) Cone-like Morphological, Molecular, and
Electrophysiological Features of the Photoreceptors of the Nrl Knockout
Mouse. IOVS 46: 2156-2167.)

David S and Aguayo AJ. (1981) Axonal elongation into peripheral nervous system
"bridges" after central nervous system injury in adult rats. Science 214(4523):
931-3.

210
Diaz E, Yang YH, Ferreira T, Loh KC, Okazaki Y, Hayashizaki Y, Tessier-Lavigne
M, Speed TP, Ngai J. (2003) Analysis of gene expression in the developing
mouse retina. Proceedings of the National Academy of Sciences of the USA.
100(9): 5491-6.

Dowling JE. (1987) The retina –an approachable part of the brain. Harvard
University Press, Cambridge MA.

Drager UC and Olsen JF. (1980) Origins of crossed and uncrossed retinal projections
in pigmented and albino mice. J. Comp. Neurol. 191: 383-412.

Fain GL and Lisman JE. (1993). Photoreceptor degeneration in vitamin A
deprivation and retinitis pigmentosa: the equivalent light hypothesis. Exp Eye
Res. 57(3): 335-40.

Fan GH, Lapierre LA, Goldenring JR, Sai J, Richmond A. (2004) Rab11-family
interacting protein 2 and myosin Vb are required for CXCR2 recycling and
receptor-mediated chemotaxis. Mol Biol Cell. 15(5): 2456-2469.

Farkas RH, Qian J, Goldberg JL, Quigley HA, Zack DJ. (2004) Gene expression
profiling of purified rat retinal ganglion cells. Invest Ophthalmol Vis Sci.
45(8): 2503-13.

Fei Y. (2003) Development of the cone photoreceptor mosaic in the mouse retina
revealed by fluorescent cones in transgenic mice. Mol Vis 9: 31-42.

Feng G, Hood R, Bernstein M, Keller-Peck C, Nguyen Q, Wallace M, Nerbonne JM,
Lichtman JW, Sanes JR. (2000) Imaging neuronal subsets in transgenic mice
expressing multiple spectral variants of GFP. Neuron 28: 41-51

Fesenko EE, Kolesnikov SS, Lyubarsky AL. (1985) Induction by cyclic GMP of
cationic conductance in plasma membrane of retinal rod outer segment.
Nature 313: 310-13.

Filbin MT. (2003) Myelin-associated inhibitors of axonal regeneration in the adult
mammalian CNS. Nat Rev Neurosci. 4(9):703-13. Review.

Fournier AE, GrandPre T, and Strittmatter SM. (2001) Identification of a receptor
mediating Nogo-66 inhibition of axonal regeneration. Nature. 409(6818):
341-6.

Frade JM and Barde YA. (1999) Genetic evidence for cell death mediated by nerve
growth factor and the neurotrophin receptor p75 in the developing mouse
retina and spinal cord. Development. 126(4): 683-90

211
Gauthier L, Rossi B, Roux F, Termine E, Schiff C. (2002) Galectin-1 is a stromal
cell ligand of the pre-B cell receptor (BCR) implicated in synapse formation
between pre-B and stromal cells and in pre-BCR triggering. Proc Natl Acad
Sci USA. 99(20): 13014-19.

Gan L. Xiang M. Zhou L. Wagner DS. Klein WH. Nathans J. (1996) POU domain
factor Brn-3b is required for the development of a large set of retinal
ganglion cells. Proc Natl Acad Sci USA 93(9): 3920-25.

Gilman AG. (1987) G proteins: transducers of receptor-generated signals. Annu
Rev Biochem 56: 615-649.

Gloor S, Antonicek H, Sweadner KJ, Pagliusi S, Frank R, Moos M, Schachner M.
(1990) The adhesion molecule on glia (AMOG) is a homologue of the beta
subunit of the Na,K-ATPase. J Cell Biol. 110(1): 165-74.

Greenwald RJ, Tumang JR, Sinha A, Currier N, Cardiff RD, Rothstein TL, Faller
DV, Denis GV. (2004) E mu-BRD2 transgenic mice develop B-cell
lymphoma and leukemia. Blood. 103(4): 1475-84.

Godement P, Salaun J, Imbert M. (1984) Prenatal and postnatal development of
retinogeniculate and retinocollicular projections in the mouse. Journal of
Comparative Neurology 230(4): 552-75.

Goldberg JL, Klassen MP, Hua Y, Barres BA. (2002) Amacrine-signaled loss of
intrinsic axon growth ability by retinal ganglion cells. Science 296(5574):
1860-4.

Gorczyca WA, Gray-Keller MP, Detwiler PB, Palczewski K. (1994) Purification and
physiological evaluation of a guanylate cyclase activating protein from retinal
rods. Proc Natl Acad Sci USA 91: 4014-18.

Gorczyca WA, Polans AS, Surgucheva IG, Subbaraya I, Baehr W, Palczewski K.
(1995) Guanylyl cyclase activating protein. A calcium-sensitive regulator of
phototransduction. J Biol Chem 270: 22029-36.

GrandPre T, Nakamura F, Vartanian T, and Strittmatter SM, Stephen M. (2000)
Identification of the Nogo inhibitor of axon regeneration as a Reticulon
protein. Nature. 403(6768): 439-444.

Gurevich VV, and Benovic JL. (1992) Cell-free expression of visual arrestin.
Truncation mutagenesis identifies multiple domains involved in rhodopsin
interaction. J Biol Chem. 267(30): 21919-23.

212
Gurevich VV, and Benovic JL. (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 VV and Benovic JL. (1997) Mechanism of phosphorylation-recognition by
visual arrestin and the transition of arrestin into a high affinity binding state.
Mol Pharmacol. 51(1): 161-9.

Hamm HE and Gilchrist A. (1996) Heterotrimeric G proteins. Curr Opin Cell Biol 8:
189-96.

Hammond LE, Gallagher PA, Wang S, Hiller S, Kluckman KD, Posey-Marcos EL,
Maeda N, Coleman RA. (2002) Mitochondrial glycerol-3-phosphate
acyltransferase-deficient mice have reduced weight and liver triacylglycerol
content and altered glycerolipid fatty acid composition. Mol Cell Biol.
22(23): 8204-14

Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole-Cire P,
Kleyner Y, Taylor A, Simon MI, Grimm C, Reme CE, Lem J. (2002).
Evidence for two apoptotic pathways in light-induced retinal degeneration.
Nat Genet. 32(2): 254-60.

Hirayama S, Bujo H, Yamazaki H, Kanaki T, Takahashi K, Kobayashi J, Schneider
WJ, Saito Y. (2000) Differential expression of LR11 during proliferation and
differentiation of cultured neuroblastoma cells. Biochem Biophys Res
Commun. 275(2): 365-73.

Hirsch JA, Schubert C, Gurevich VV, and Sigler PB. (1999) The 2.8 A crystal
structure of visual arrestin: a model for arrestin's regulation. Cell. 97(2): 257-
69.

Hu MC and Rosenblum ND. (2005) Smad1, beta-catenin and Tcf4 associate in a
molecular complex with the Myc promoter in dysplastic renal tissue and
cooperate to control Myc transcription. Development. 132(1): 215-25

Huang X, Wu DY, Chen G, Manji H, Chen DF. (2003) Support of retinal ganglion
cell survival and axon regeneration by lithium through a Bcl-2-dependent
mechanism. Invest Ophthalmol Vis Sci. 44(1): 347-54.

Huber AB, Weinmann O, Brosamle C, Oertle T, and Schwab ME. (2002) Patterns of
Nogo mRNA and protein expression in the developing and adult rat and after
CNS lesions. J Neurosci. 22(9): 3553-67.

Hsu YT and Molday RS. (1993) Modulation of the cGMP-gated channel of rod
photoreceptor cells by calmodulin. Nature 361: 76-9.

213

Imai H, Imamoto Y, Yoshizawa T, and Shichida Y. (1995) Difference in molecular
properties between chicken green and rhodopsin as related to the functional
difference between cone and rod photoreceptor cells. Biochemistry. 34(33):
10525-31.

Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U. (2003)
Speed TP. Exploration, normalization, and summaries of high density
oligonucleotide array probe level data. Biostatistics. 4(2): 249-64.

Jeon CJ, Strettoi E, Masland RH. (1998) The major cell populations of the mouse
retina. J Neurosci. 18(21): 8936-46.

Jones MW, Errington ML, French PJ, Fine A, Bliss TV, Garel S, Charnay P, Bozon
B, Laroche S, Davis S. (2001) A requirement for the immediate early gene
Zif268 in the expression of late LTP and long-term memories. Nat Neurosci.
4(3): 289-96.

Jung J, Oh J, Lee K. (1998) Nucleotide and deduced amino acid sequences of rat
myosin binding protein H (MyBP-H). Arch Pharm Res. 21(6): 712-7.

Kay R, Rosten PM, Humphries RK. (1991) CD24, a signal transducer modulating B
cell activation responses, is a very short peptide with a glycosyl
phosphatidylinositol membrane anchor. J Immunol. 147(4): 1412-6.

Kalil K and Reh T. (1982) A light and electron microscopic study of regrowing
pyramidal tract fibers. J Comp. Neurol. 211: 265-275.

Krispel CM, Chen C, Simon

MI and Burns ME. (2003) Novel Form of Adaptation in
Mouse Retinal Rods Speeds Recovery of Phototransduction. J Gen
Physiology 122(6): 703-712.

Krupnick JG, Gurevich VV, and Benovic JL. (1997) Mechanism of quenching of
phototransduction. Binding competition between arrestin and transducin for
phosphorhodopsin. J Biol Chem. 272(29): 18125-31.

Krupnick JG, Gurevich VV, Schepers T, Hamm HE, Benovic JL. (1994) Arrestin-
rhodopsin interaction. Multi-site binding delineated by peptide inhibition. J
Biol Chem. 269(5): 3226-32.

Kuhn H, Hall SW, and Wilden U. (1984) Light-induced binding of 48-kDa protein to
photoreceptor membranes is highly enhanced by phosphorylation of
rhodopsin. FEBS Lett. 176(2): 473-8.

214
Lauren J, Airaksinen MS, Saarma M, Timmusk T. (2003) Two novel mammalian
Nogo receptor homologs differentially expressed in the central and peripheral
nervous systems. Mol Cell Neurosci. 24(3): 581-94.

LaVail MM, Gorrin GM, Repaci MA, Yahumura D. (1987) Light-induced retinal
degeneration in albino mice and rats: strain and species differences. In:
Hollyfield JG, Anderson RE, LaVail MM, eds. Degenerative retinal
disorders: clinical and laboratory investigations. New York: Alan R. Liss.

LaVail, MM, Yasumura, D, Matthes, MT, Drenser, KA, Flannery, JG, Lewin, AS,
Hauswirth, WW. (2000) Ribozyme rescue of photoreceptor cells in P23H
transgenic rats: long-term survival and late-stage therapy. Proc. Natl. Acad.
Sci., USA. 97: 11488-93.

Lee SJ and Montell C. (2004) Light-Dependent Translocation of Visual Arrestin
Regulated by the NINAC Myosin III. Neuron 43: 95-103

Lee YC, Tang YC, Chen YH, Wong CM, Tsou AP. (2003) Selenite-induced survival
of HuH7 hepatoma cells involves activation of focal adhesion kinase-
phosphatidylinositol 3-kinase-Akt pathway and Rac1. J Biol Chem. 278(41):
39615-24.

Lem J, Applebury ML, Falk JD, Flannery JG, Simon MI. (1991) Tissue-specific and
developmental regulation of rod opsin chimeric genes in transgenic mice.
Neuron 6: 201-10

Lem J, Krasnoperova NV, Calvert PD, Kosaras B, Cameron DA, Nicolo M, Makino
CL, and Sidman RL. (1999) Morphological, physiological, and biochemical
changes in rhodopsin knockout mice. Proc Natl Acad Sci USA 96: 736-741.

Li C and Wong WH. (2001) Model-based analysis of oligonucleotide arrays:
expression index computation and outlier detection. Proc Natl Acad Sci USA.
98(1): 31-36.

Liao Y, Shikapwashya ON, Shteyer E, Dieckgraefe BK, Hruz PW, Rudnick DA.
(2004) Delayed hepatocellular mitotic progression and impaired liver
regeneration in early growth response-1-deficient mice. J. Biol. Chem. 279:
43107-16.

Liedtke W, Leman E, Fyffe R, Raine C, Schubart U. (2002) Stathmin-deficient mice
develop an age-dependent axonopathy of the central and peripheral nervous
system Am J Path 160 (2): 469-80

215
Liu BP, Fournier A, GrandPre T, and Strittmatter SM. (2002) Myelin-associated
glycoprotein as a functional ligand for the Nogo-66 receptor.
Science. 297(5584): 1190-3.

Liu C, Calogero A, Ragona G, Adamson E, Mercola D. (1996) EGR-1, the reluctant
suppression factor: EGR-1 is known to function in the regulation of growth,
differentiation, and also has significant tumor suppressor activity and a
mechanism involving the induction of TGF-beta1 is postulated to account for
this suppressor activity. Crit Rev Oncog. 7(1-2): 101-25. Review

Lisman J and Fain G. (1995) Support for the equivalent light hypothesis for RP.
Nat Med. 1(12): 1254-5.

Lunyak V, Burgess R, Prefontaine G, Nelson C, Sze S, Chenoweth J, Schwartz P,
Pevzner P, Glass C, Mandel G, Rosenfeld M. (2002) Corepressor-dependent
silencing of chromosomal regions encoding neuronal genes Science
298:1747-52

Makino CL, Dodd RL, Chen J, Burns ME, Roca A, Simon MI, and Baylor DA.
(2004) Recoverin regulates light-dependent phosphodiesterase activity in
retinal rods. J Gen Physiol 123: 729-741.

Maguchi M, Nishida W, Kohara K, Kuwano A, Kondo I, Hiwada K. (1995)
Molecular cloning and gene mapping of human basic and acidic calponins.
Biochem Biophys Res Commun. 217(1):238-44.

Maucuer A, Ozon S, Manceau V, Gavet O, Lawler S, Curmi P, Sobel A. (1997) KIS
is a protein kinase with an RNA recognition motif. J Biol Chem. 272(37):
23151-6

Marquardt T, Ashery-Paden R, Andrejewski N, Scardigil R, Guillemot F, Gruss P.
(2001) Pax6 is required for the multipotent state of retinal progenitor cells.
Cell 105: 43- Martin RE, Ranchon-Cole I, Brush RS, Williamson CR,
Hopkins SA, Li F, Anderson RE. (2004) P23H and S334ter opsin mutations:
Increasing photoreceptor outer segment n-3 fatty acid content does not affect
the course of retinal degeneration. Mol Vis. 10: 199-207.

Matveeva EA, Whiteheart SW, Slevin JT. (2003) Accumulation of 7S SNARE
complexes in hippocampal synaptosomes from chronically kindled rats. J
Neurochem. 84(3): 621-4.

McGraw J, McPhail LT, Oschipok LW, Horie H, Poirier F, Steeves JD, Ramer MS,
Tetzlaff W. (2004) Galectin-1 in regenerating motoneurons. Eur J Neurosci.
20(11): 2872-80

216
Mears AJ, Kondo M, Swain PK, Takada Y, Bush RA, Saunders TL, Sieving PA,
Swaroop A. (2001) Nrl is required for rod photoreceptor development. Nat
Genet 29: 447-452.

Mendez A, Burns ME, Sokal I, Dizhoor AM, Baehr W, Palczewski K, Baylor DA,
and Chen J. (2001) Role of guanylate cyclase-activating proteins (GCAPs) in
setting the flash sensitivity of rod photoreceptors. Proc Natl Acad Sci USA
98: 9948-9953.

Mendez A, Lem J, Simon M, Chen J. (2003) Light-dependent translocation of
arrestin in the absence of rhodopsin phosphorylation and transducin signaling.
J Neurosci. 23(8): 3124-9.

Messer A, Snodgrass GL, Maskin P. (1984) Enhanced survival of cultured cerebellar
Purkinje cells by plating on antibody to Thy-1. Cell Mol Neurobiol. 4(3):
285-90.

Meyer-Franke A, Kaplan MR. Pfrieger FW. Barres BA. (1995) Characterization of
the signaling interactions that promote the survival and growth of developing
retinal ganglion cells in culture. Neuron 15(4):805-19.

Meyer-Franke A, Wilkinn MG Jr, ÿ lture. Neuron 15(4):805-19.

Meyer- Reichardt LF, Barres BA. (1998) Depolarization and cAMP elevation
rapidly recruit TrkB to the plasma membrane of CNS neurons. Neuron.
21(4): 681-93.

Michon JJ, Li ZL, Shioura N, Anderson RJ, Tso MO. (1991) A comparative study
of methods of photoreceptor morphometry. Invest Ophthalmol Vis Sci.
32(2): 280-4.

Miwa JM, Ibanez-Tallon I, Crabtree GW, Sanchez R, Sali A, Role LW, Heintz N.
(1999) Lynx1, an endogenous toxin-like modulator of nicotinic acetylcholine
receptors in the mammalian CNS. Neuron. 23(1): 105-14.

Mori H, Sakakibara S, Imai T, Nakamura Y, Iijima T, Suzuki A, Yuasa Y, Takeda M,
Okano H. (2001) Expression of mouse igf2 mRNA-binding protein 3 and its
implications for the developing central nervous system. J Neurosci Res.
64(2): 132-43.

Morris RJ. (1985) Thy-1 in developing nervous tissue. Dev. Neurosci. 7: 133-60

Morris RJ and Barber PC. (1983) Fixation of Thy-1 nervous tissue for
immunohistochemistry. J. Histochem. Cytochem. 31: 263-274.

217
Mukhopadhyay G, Doherty P, Walsh FS, Crocker PR, and Filbin MT. (1994) A
novel role for myelin-associated glycoprotein as an inhibitor of axonal
regeneration. Neuron. 13(3): 757-67.

Murakami A, Yajima T, Sakuma H, McLaren MJ, Inana G. (1993) X-arrestin: a new
retinal arrestin mapping to the X chromosome. FEBS Lett. 334(2): 203-209

Murata Y, Sun-Wada GH, Yoshimizu T, Yamamoto A, Wada Y, Futai M. (2002)
Differential localization of the vacuolar H+ pump with G subunit isoforms
(G1 and G2) in mouse neurons. J Biol Chem. 277(39): 36296-303.

Nagashima K, Torii S, Yi Z, Igarashi M, Okamoto K, Takeuchi T, Izumi T. (2002)
Melanophilin directly links Rab27a and myosin Va through its distinct
coiled-coil regions. FEBS Lett. 517(1-3): 233-8.

Nikonov S, Danielle L, Zhu X, Craft C, Swaroop A, Pugh E. (2005) Photoreceptors
of Nrl -/- mice coexpress functional S- and M-cone opsins having distinct
inactivation mechanisms. J General Physiology 125: 287-303.

Noell WK. (1980). Possible mechanisms of photoreceptor damage by light in
mammalian eyes. Vis Res 20: 1163-1171.

Nolo R, Abbott LA, Bellen HJ. (2000) Senseless, a Zn finger transcription factor, is
necessary and sufficient for sensory organ development in Drosophila. Cell.
102(3): 349-62.

Oakley BR, Oakley CE, Rinehart JE. (1987) Conditionally lethal tubA alpha-tubulin
mutations in Aspergillus nidulans. Mol Gen Genet. 208(1-2): 135-44.

Pal DK, Evgrafov OV, Tabares P, Zhang F, Durner M, Greenberg DA. (2003) BRD2
(RING3) is a probable major susceptibility gene for common juvenile
myoclonic epilepsy. Am J Hum Genet. 73(2): 261-70.

Palczewski K, McDowell JH, Jakes S, Ingebritsen TS, and Hargrave PA. (1989)
Regulation of rhodopsin dephosphorylation by arrestin. J Biol Chem.
264(27): 15770-3.

Palczewski K, Subbaraya I, Gorczyca WA, Helekar BS, Ruiz CC, Ohguro H, Huang
J, Zhao X, Crabb JW, Johnson RS, Walsh KA, Gray-Keller MP, Detwiler PB,
Baehr, W. (1994) Molecular cloning and characterization of retinal
photoreceptor guanylyl cyclase-activating protein. Neuron 13: 395-404

218
Panzenboeck U, Wintersberger A, Levak-Frank S, Zimmermann R, Zechner R,
Kostner GM, Malle E, Sattler W. (1997) Implications of endogenous and
exogenous lipoprotein lipase for the selective uptake of HDL3-associated
cholesteryl esters by mouse peritoneal macrophages. J Lipid Res. 38(2): 239-
53.

Pierce KL, Premont RT, Lefkowitz RJ. (2002) Seven-transmembrane receptors.
Nature Reviews 3: 639-50. Review

Praetorius J, Nejsum LN, Nielsen S. (2004) A SCL4A10 gene product maps
selectively to the basolateral plasma membrane of choroid plexus epithelial
cells. Am J Physiol Cell Physiol. 286(3): C601-10.

Prinjha R, Moore SE, Vinson M, Blake S, Morrow R, Christie G, Michalovich D,
Simmons DL, and Walsh FS. (2000) Inhibitor of neurite outgrowth in
humans. Nature. 403(6768): 383-4.

Qiu J, Cai D, Dai H, McAtee M, Hoffman PN, Bregman BS, and Filbin MT. (2002)
Spinal axon regeneration induced by elevation of cyclic AMP. Neuron. 34(6):
895-903.

Rapp LM and William TP. (1980) A parametric study of retinal light damage in
albino and pigmented rats. In: Williams TP, Baker BN, eds. The effect of
constant light in visual process: NY Plenum. 137-159.

Rezajooi K, Pavlides M, Winterbottom J, Stallcup WB, Hamlyn PJ, Lieberman AR,
Anderson PN. (2004) NG2 proteoglycan expression in the peripheral nervous
system: upregulation following injury and comparison with CNS lesions. Mol
Cell Neurosci. 25(4): 572-84.

Richardson PM, McGuinness UM, and Aguayo AJ. (1980) Axons from CNS
neurons regenerate into PNS grafts. Nature. 284(5753): 264-5.

Roca A, Shin KJ, Liu X, Simon MI, Chen J. (2004). Comparative analysis of
transcriptional profiles between two apoptotic pathways of light-induced
retinal degeneration. Neuroscience. 129(3): 779-90.

Rodieck RW. (1998) The first steps in seeing. Sinauer Associates Inc, Sunderland,
MA.

Rothman JE. (1994) Mechanisms of intracellular protein transport. Nature.
372(6501): 55-63. Review

219
Sassa T, Gomi H, Itohara S. (2004) Postnatal expression of Cdkl2 in mouse brain
revealed by LacZ inserted into the Cdkl2 locus. Cell Tissue Res. 315(2): 147-
56

Scammell JG, Reddy S, Valentine DL, Coker TN, Nikolopoulos SN, Ross RA.
(2000) Isolation and characterization of the human secretogranin II gene
promoter. Brain Res Mol Brain Res. 75(1): 8-15.

Scherzer CR, Offe K, Gearing M, Rees HD, Fang G, Heilman CJ, Schaller C, Bujo
H, Levey AI, Lah JJ. (2004) Loss of apolipoprotein E receptor LR11 in
Alzheimer disease. Arch Neurol. 61(8): 1200-5.

Schnell L and Schwab ME (1990) Axonal regeneration in the rat spinal cord
produced by an antibody against myelin-associated neurite growth inhibitors.
Nature. 343(6255):269-72.

Schoenherr CJ and Anderson DJ. (1995) The neuron-restrictive silencer factor
(NRSF): a coordinate repressor of multiple neuron-specific genes. Science.
267(5202): 1360-3.

Schnapf J L, Nunn B J, Meister M, and Baylor DA. (1990) Visual transduction in
cones of the monkey Macaca fascicularis. J Physiol 427: 681-713.

Seoane J, Le HV, Shen L, Anderson SA, Massague J. (2004) Integration of Smad
and forkhead pathways in the control of neuroepithelial and glioblastoma cell
proliferation. Cell. 117(2): 211-23.

Serini G, Valdembri D, Zanivan S, Morterra G, Burkhardt C, Caccavari F,
Zammataro L, Primo L, Tamagnone L, Logan M, Tessier-Lavigne M,
Taniguchi M, Puschel AW, Bussolino F. (2003) Class 3 semaphorins control
vascular morphogenesis by inhibiting integrin function. Nature 424: 391–397.

Shewan D, Berry M, Cohen J. (1995) Extensive regeneration in vitro by early
embryonic neurons on immature and adult CNS tissue. Journal of
Neuroscience 15(3 Pt 1): 2057-62.

Shinohara T, Dietzschold B, Craft CM, Wistow G, Early JJ, Donoso LA, Horwitz J,
and Tao R. (1987) Primary and Secondary Structure of Bovine Retinal S
Antigen (48-kDa Protein) PNAS 84: 6975-6979.

Singec I, Knoth R, Ditter M, Hagemeyer CE, Rosenbrock H, Frotscher M, Volk B.
(2002) Synaptic vesicle protein synaptoporin is differently expressed by
subpopulations of mouse hippocampal neurons. J Comp Neurol. 452(2): 139-
53.

220
Smith DS and Skene JH. (1997) A transcription-dependent switch controls
competence of adult neurons for distinct modes of axon growth. J Neurosci
17(2): 646-658.

Smith WC, Gurevich EV, Dugger DR, Vishnivetskiy SA, Shelamer CL, McDowell
JH, Gurevich VV. (2000) Cloning and functional characterization of
salamander rod and cone arrestins. Investigative Ophthalmology & Visual
Science. 41(9): 2445-55.

Sokolov M, Lyubarsky AL, Strissel KJ, Savchenko AB, Govardovskii VI, Pugh Jr
EN, Arshavsky VY. (2002) Massive light-driven translocation of transducin
between the two major compartments of rod cells: a novel mechanism of
light adaptation. Neuron 34: 95-106.

Starr RG. Lu B. Federoff HJ. Functional characterization of the rat GAP-43 promoter.
Brain Research 638(1-2):211-20, 1994.

Strittmatter SM. Fankhauser C. Huang PL. Mashimo H. Fishman MC. Neuronal
pathfinding is abnormal in mice lacking the neuronal growth cone protein
GAP-43. Cell 80(3):445-52, 1995.

Sun W, Li N, He S. (2002) Large-scale morphological survey of mouse retinal
ganglion cells. J Comp Neurol 451: 115-126.

Sutton RB, Vishnivetskiy SA, Robert J, Hanson SM, Raman D, Knox BE, Kono M,
Navarro J, and Gurevich VV. (2005) Crystal Structure of Cone Arrestin at
2.3A: Evolution of Receptor Specificity. J Mol Biol. 354(5): 1069-80.

Swaroop A and Zack DJ. (2002) Transcriptome analysis of the retina. Genome
Biology. 3(8):REVIEWS1022.

Tachibanaki S, Tsushima S, Kawamura S. (2001) Low amplification and fast visual
pigment phosphorylation as mechanisms characterizing cone photoresponses.
Proc Natl Acad Sci USA. 98(24): 14044-9

Tai AW, Chuang J-Z, Bode C, Wolfrum U, and Sung C-H. (1999) Rhodopsin’s
Carboxy-Terminal Cytoplasmic Tail Acts as a Membrane Receptor for
Cytoplasmic Dynein by Binding to the Dynein Light Chain Tctex-1 97: 877-
887.

Tekki-Kessaris N, Bonventre JV, Boulter CA. (1999) Characterization of the mouse
Kid1 gene and identification of a highly related gene, Kid2. Gene. 240(1):
13-22.

221
Uittenbogaard M. Martinka DL. Chiaramello A. The basic helix-loop-helix
differentiation factor Nex1/MATH-2 functions as a key activator of the GAP-
43 gene. Journal of Neurochemistry 84(4):678-88, 2003.

Vidal-Sanz M, Bray GM, Villegas-Perez MP, Thanos S, Aguayo AJ. (1987) Axonal
regeneration and synapse formation in the superior colliculus by retinal
ganglion cells in the adult rat. Journal of Neuroscience 7(9): 2894-909.

Vishnivetskiy SA, Paz CL, Schubert C, Hirsch JA, Sigler PB, and Gurevich VV.
(1999) How does arrestin respond to the phosphorylated state of rhodopsin? J
Biol Chem. 274(17):11451-4.

Vishnivetskiy SA, Hosey MM, Benovic JL, and Gurevich VV. (2004) Mapping the
arrestin-receptor interface. Structural elements responsible for receptor
specificity of arrestin proteins. J Biol Chem. 279(2): 1262-8.

Wacker WB, Donoso LA, Kalsow CM, Yankeelov JA Jr, and Organisciak DT.
(1977) Experimental allergic uveitis. Isolation, characterization, and
localization of a soluble uveitopathogenic antigen from bovine retina. J
Immunol. 119(6): 1949-58.

Wald G. (1968) The molecular basis of visual excitation. Nature 219: 800-807.

Wang B, Xiao Y, Ding BB, Zhang N, Yuan X, Gui L, Qian KX, Duan S, Chen Z,
Rao Y, Geng JG. (2003) Induction of tumor angiogenesis by Slit-Robo
signaling and inhibition of cancer growth by blocking Robo activity. Cancer
Cell 4: 19–29.

Wang KC, Kim JA, Sivasankaran R, Segal R, He Z. (2002) P75 interacts with the
Nogo receptor as a co-receptor for Nogo, MAG and OMgp. Nature.
420(6911): 74-8

Wang KC, Koprivica V, Kim JA, Sivasankaran R, Guo Y, Neve RL, and He Z.
(2002) Oligodendrocyte-myelin glycoprotein is a Nogo receptor ligand that
inhibits neurite outgrowth. Nature. 417(6892): 941-4.

Wikler KW and Rakic P. (1990) Distribution of photoreceptor subtypes in the retina
of diurnal and nocturnal primates. J Neurosci 10: 3390-3401.

Wilden U and Kuhn H. (1982) Light-dependent phosphorylation of rhodopsin:
number of phosphorylation sites. Biochemistry 21: 3014-22.

222

Wilden U, Hall SW, Kuhn H. (1986) Phosphodiesterase activation by photoexcited
rhodopsin is quenched when rhodopsin is phosphorylated and binds the
intrinsic 48-kDa protein of rod outer segments. Proc Natl Acad Sci USA.
83(5): 1174-8.

Williams TP, Webbers JPP, Giordano L, Henderson RP. (1998) Distribution of
photon absorption rates across the rat retina. J of Physiology 508(2): 515-522.

Wu DY, Zheng JQ, McDonald MA, Chang B, and Twiss JL. (2003) PKC isozymes
in the enhanced regrowth of retinal neurites after optic nerve injury. Invest.
Ophthalmol. Vis. Sci. 44: 2783-90.

Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA, Chen J. (1997) Prolonged
photoresponses in transgenic mouse rods lacking arrestin. Nature. 389: 505-9.

Xue GP, Calvert RA, Morris RJ. (1990) Expression of the neuronal cell surface
glycoprotein Thy-1 is under post-transcriptional control and is spatially
regulated in the developing olfactory system. Development 109: 851-864.

Yamagata M, Weiner JA, Sanes JR. (2002) Sidekicks: synaptic adhesion molecules
that promote lamina-specific connectivity in the retina. Cell. 110(5): 649-60.

Yoon H, Liyanarachchi S, Wright FA, Davuluri R, Lockman JC, de la Chapelle A,
Pellegata NS. (2002) Gene expression profiling of isogenic cells with
different TP53 gene dosage reveals numerous genes that are affected by TP53
dosage and identifies CSPG2 as a direct target of p53. Proc Natl Acad Sci
USA. 99(24): 15632-7

You L, Chan SK, Bruce JM, Archibeque-Engle S, Casanova M, Corton JC, Heck H.
(1999) Modulation of testosterone-metabolizing hepatic cytochrome P-450
enzymes in developing Sprague-Dawley rats following in utero exposure to
p,p'-DDE. Toxicol Appl Pharmacol. 158(2): 197-205.

Young RW. (1985) Cell proliferation during postnatal development of the retina in
the mouse. Dev Brain Res 21: 229-239.

Zhu X, Li A, Brown B, Weiss ER, Osawa S, Craft CM. (2002) Mouse cone arrestin
expression pattern: light induced translocation in cone photoreceptors.
Molecular Vision. 8: 462-71.

Zhu X, Brown B, Li A, Mears AJ, Swaroop A, Craft CM. (2003) GRK1-dependent
phosphorylation of S and M opsins and their binding to cone arrestin during
cone phototransduction in the mouse retina. J Neurosci. 9;23(14): 6152-60.

223
Zhu Q and Julien JP. (1999) A key role for GAP-43 in the retinotectal topographic
organization. Experimental Neurology 155(2):228-42.

Zimmer A, Zimmer AM, Baffi J, Usdin T, Reynolds K, Konig M, Palkovits M,
Mezey E. (1998) Hypoalgesia in mice with a targeted deletion of the
tachykinin 1 gene. Proc Natl Acad Sci USA. 95(5): 2630-5.

Abstract (if available)

Abstract The mature retina is a specialized tissue critical for vision. Individual cell-types that comprise the retina have specialized functional roles mediated by differential gene expression. To identify the contribution of specific genes in defining a cell-type, the function of the cone arrestin protein was analyzed in vivo in photoreceptors and a differential transcript expression profile generated for perinatal developing retinal ganglion cells.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Cone arrestin 4 contributes to vision, cone health, and desensitization of the dopamine receptor D4

PDF

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

PDF

Transcriptomic maturation of developing human cone precursors in fetal and 3D hESC-derived tissues

PDF

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

Elements of photoreceptor homeostasis: investigating phenotypic manifestations and susceptibility to photoreceptor degeneration in genetic knockout models for retinal disease

PDF

Characterization of human embryonic stem cell derived retinal pigment epithelial cells for age-related macular degeneration

PDF

Towards a high resolution retinal implant

PDF

Multimodal investigation of rod photoreceptor structure and function in healthy and diseased states

PDF

Morphological and functional evaluation of hESC-RPE cell transplantation in RCS rats

PDF

Differential role of two coactivators, CCAR1 and CARM1, for dysregulated beta-catenin activity in colorectal cancer cell growth and gene expression

PDF

Differential expression of genes between nickel compounds/3-methylcholanthrene-transformed, and non-transformed, C3H/10T1/2 cl 8 mouse embryo cell lines, and thiol regulation of VEGF-A and VEGF-A...

Asset Metadata

Creator Chan, Sanny Kai-Wai (author)

Core Title Signaling cascades: a functional characterization of cone arrestin and a differential gene expression analysis of developing retinal ganglion cells

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Cell and Neurobiology

Publication Date 10/02/2007

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

Tag cone arrestin,OAI-PMH Harvest,phototransduction,retinal ganglion cells

Language English

Advisor Chen, Jeannie (committee chair), Craft, Cheryl (committee member), Garner, Judy A. (committee member), Hinton, David R. (committee member)

Creator Email sannycha@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m841

Unique identifier UC1469631

Identifier etd-Chan-20071002 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-583837 (legacy record id),usctheses-m841 (legacy record id)

Legacy Identifier etd-Chan-20071002.pdf

Dmrecord 583837

Document Type Dissertation

Rights Chan, Sanny Kai-Wai

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