Investigation of the molecular mechanisms underlying the retinal degeneration observed in the P347S mutant rhodopsin model of retinitis pigmentosa
A dissertation
Submitted by
Katherine M. Malanson
In partial fulfillment of the requirements for the degree of
Doctor of Philosophy
In
Neuroscience
TUFTS UNIVERSITY Sackler School of Graduate Biomedical Sciences
August 2011
Advisor: Janis Lem
ii
ABSTRACT
Retinitis pigmentosa (RP) is a genetically heterogeneous group of diseases that causes blindness. RP can be inherited as an X-linked, autosomal recessive or autosomal dominant disease. Mutations within the rhodopsin gene account for approximately 25% of autosomal dominantly inherited RP cases. Therefore understanding the mechanisms causing rhodopsin-mediated RP has a significant health impact. To date, results from multiple labs indicate that rhodopsin-mediated RP pathogenesis does not share a common mechanism of degeneration. There is strong evidence that multiple mechanisms are involved, including protein misfolding, mislocalization, release of toxic products and aberrant signaling. This thesis investigates the molecular mechanisms involved in the retinal degeneration of the P347S mutant rhodopsin mouse model of retinitis pigmentosa.
Through the use of transgenic animal models the involvement of persistent photosignaling, aberrant rhodopsin-arrestin complexes, chromophore toxicity and galectin-1 in retinal degeneration are investigated. Additionally, the involvement of glycogenes are investigated with a custom gene microarray.
iii
ACKNOWLEDGEMENTS
I would like to first and foremost thank my advisor, Janis Lem. Without her this work would not have been possible. I am grateful to my committee members, Jim
Schwob, Dan Jay and Noorjahan Panjwani, for their guidance throughout my time at
Tufts. Additionally, I would like to thank my outside examiner, Clint Makino, for carefully reading my thesis and providing valuable feedback.
I am indebted to my fellow lab members, Kibibi Ganz, Jinsong Yang, Fang Yang,
Ed Dudek and Jesse Peterson. They provided helpful advice and daily scientific support.
Additionally, I need to thank Zhiyi Cao and Anna Markowska of the Panjwani Lab for guiding me through the exciting world of glycobiology. I would like to thank the students of the Neuroscience Program at Tufts University Sackler School of Biomedical Sciences.
Having the joy of being their peer kept me motivated even when experiments were not working.
I would also like to express my gratitude to those who offered their expertise scientifically, including Tiansen Li for providing the P347S and K296E mutant rhodopsin mutant mouse lines, and Wolfgang Baehr for providing the VPP mutant mice. Thanks to
Flore Celestin of the Specialized Center of Research in Ischemic Heart Disease Histology
Core and Derek Papalegis of the Division of Animal Laboratory Medicine for assistance with histological sections. Special thanks to Basil Pawlyk and Mike Sandberg of
Massachusetts Eye and Ear Infirmary for their help recording electroretinograms.
iv
Finally, a very special thanks to my family and my husband, Jeffrey, who have instilled in me a passion for learning, a drive for excellence, and a sense of humor and perspective to make this work not only possible, but worthwhile.
This work was financially supported by The Synapse Neurobiology Training Grant,
Program in Neuroscience, Tufts University School of Medicine; CFG grant support
(National Institute of General Medical Sciences Grant GM62116), Foundation Fighting
Blindness, National Eye Institute, Research to Prevent Blindness, and Massachusetts
Lions Eye Research Fund.
v
TABLE OF CONTENTS
Acknowledgements iii
List of Tables vi
List of Figures vi
Chapter 1: An Introduction to Rhodopsin-Mediated Retinitis Pigmentosa 2
Chapter 2: A Novel Form of Transducin-Dependent Retinal Degeneration: Accelerated Retinal Degeneration in the Absence of Rod Transducin 46
Chapter 3: Involvement of Galectin-1 in Rhodopsin-Mediated Retinitis Pigmentosa 69
Chapter 4: Analysis of Glycogenes and their Molecular Pathways in Rhodopsin-Mediated Retinitis Pigmentosa by Microarray 100
Chapter 5: A Discussion of the Molecular Mechanisms Underlying P347S Mutant Rhodopsin Mediated-Retinitis Pigmentosa 123
Figures 144
Bibliography 209
vi
LIST OF TABLES
Table 1: Compiled ERG Data
Table 2: Number of differentially expressed transcripts in P347S mutant rhodopsin retina
compared to control
Table 3: Categories of differentially expressed transcripts in P347S mutant rhodopsin
retina over time
Table 4: Differentially expressed transcripts involved in glycan degradation
Table 5: Growth factor and receptor differentially expressed transcripts
Table 6: Glycosyltransferases differentially expressed transcripts
Table 7: Mouse Housekeeping differentially expressed transcripts
Table 8: Chemokine differentially expressed transcripts
Table 9: Interleukin and receptors differentially expressed transcripts
Table 10: Additional differentially expressed transcripts
Table 11: Trends of transcripts differentially expressed at 1 and 2 Months
vii
LIST OF FIGURES
Figure 1: Schematic of the mammalian eye
Figure 2: Retina structure
Figure 3: Phototransduction cascade
Figure 4: Rhodopsin secondary structure and point mutations associated with ADRP
Figure 5: P23H difference spectra
Figure 6: Cone cell death mechanisms: involvement of rod-derived trophic factor
Figure 7: Gene therapy: The use of ribozymes to treat P23H rats
Figure 8: Comparison of rhodopsin mutant mice on wild-type (Tr ) and -transducin
null (Tr ) genetic backgrounds.
Figure 9: Role of rhodopsin-arrestin complexes in P347S mutant rhodopsin degeneration.
Figure 10: -transducin stabilizes P347S metarhodopsin.
Figure 11: Comparison of A2E and A2E-precursor levels in wild-type and P347S mutant
rhodopsin retinas
Figure 12: P347S mutant rhodopsin protein has decreased stability
Figure 13: Comparison of galectin-1 RNA levels in P347S mutant rhodopsin retinas
compared to control
Figure 14: Galectin-1 protein expression is increased in the P347S mutant rhodopsin
retina compared to control
Figure 15: Galectin-1 sugar-binding column identifies tenascin-R (TN-R) as a galectin-1
binding partner
Figure 16: Mass-spectrophotometry identified peptides within TN-R Sequence
viii
Figure 17: Tenascin-R localizes to synaptic layers in the wild-type and galectin-1
knockout retina
Figure 18: Tenascin-R expression in wild-type and P347S mutant rhodopsin retinas
Figure 19: Morphology of P347S mutant rhodopsin retinas in the presence and absence of
galectin-1 at 2M
Figure 20: Morphology of P347S mutant rhodopsin retinas in the presence and absence of
galectin-1 at 4M
Figure 21: Morphology of P347S mutant rhodopsin retinas in the presence and absence of
galectin-1 at 9M
Figure 22: Involvement of galectin-1 in retinal degeneration measured with
electroretinogram at 4M
Figure 23: Time-to-Peak for 4M ERG
Figure 24: Involvement of galectin-1 in retinal degeneration measured with
electroretinogram at 9M
Figure 25: Time-to-peak for 9M ERG
Figure 26: Galectin-1 localization in wild-type control and galectin-1 knockout mice.
Figure 27: Verification of galectin-1 antibody for immunoprecipitation.
Figure 28: Identification of sugar independent galectin-1 binding partners
Figure 29: Validation of galectin-1 binding with thy-1.
Figure 30: Confirmation of galectin-1 sugar-binding with neuropilin-1 (NP-1) and
tenascin-C (TN-C)
Figure 31: Dendogram showing hierarchical clustering analysis of mutant and control
retinas.
ix
Figure 32: Categories of differentially expressed transcripts in P347S mutant rhodopsin
retina compared to control
Figure 33: Comparison of microarray and quantitative RT-PCR data for CX3 chemokine
receptor (CX3CR1)
Figure 34: Comparison of microarray and quantitative RT-PCR data for colony
stimulating factor 1 (CSF1)
Figure 35: Localization of colony stimulating factor 1 (CSF1) in wild-type and P347S
mutant rhodopsin retinas
Figure 36: Comparison of microarray and quantitative RT-PCR data for colony
stimulating factor 1 receptor (CSF1R)
Figure 37: Localization of colony stimulating factor 1 receptor (CSF1R) in wild-type and
P347S mutant rhodopsin retina
Figure 38: Comparison of microarray and quantitative RT-PCR data for interleukin 6
signal transducer (IL6ST)
1
INVESTIGATION OF THE MOLECULAR MECHANISMS
UNDERLYING THE RETINAL DEGENERATION OBSERVED IN THE
P347S MUTANT RHODOPSIN
MODEL OF RETINITIS PIGMENTOSA
2
CHAPTER 1: AN INTRODUCTION TO RHODOPSIN-MEDIATED
RETINITIS PIGMENTOSA
Portions of this chapter originally appeared as a chapter in Progress in Molecular Biology and Translational Science (Malanson and Lem 2009).
3
ABSTRACT
Retinitis pigmentosa (RP) is a genetically and phenotypically heterogeneous group of diseases that cause blindness. RP is genetically inherited as an X-linked, autosomal recessive or autosomal dominant disease. Mutations within the rhodopsin gene account for approximately 25% of autosomal dominantly inherited RP cases.
Therefore understanding the mechanisms causing rhodopsin-mediated RP has a significant health impact. To date, results from multiple labs indicate that rhodopsin- mediated RP pathogenesis does not share a common mechanism of degeneration. There is strong evidence that multiple mechanisms are involved, including impaired protein folding and localization, release of toxic products and aberrant signaling. Development of effective treatments requires investigation of the mechanisms involved in the degeneration caused by different rhodopsin mutations. This chapter focuses on the mechanisms by which rhodopsin mutations cause retinal degeneration, as well as potential therapeutic strategies to treat the disease.
INTRODUCTION
Retinitis pigmentosa (RP) is a genetically heterogeneous group of diseases that produces phenotypes ranging from relatively mild night blindness (nyctalopia) to severe and complete blindness. Early in the disease process, rod photoreceptors (specialized neuronal cells adapted for exquisite light sensitivity) begin to die, causing night blindness. Patients have difficulty seeing in dim light, such as night driving, and adapting to changes in light intensity, such as entering a darkened movie theater. The disease progresses slowly with the continued loss of photoreceptors, resulting in loss of
4 peripheral vision and producing “tunnel vision” where only central vision is retained.
Additionally, many patients report seeing continued flashes of light (photopsia). In late stages of the disease, central vision is also lost, causing complete blindness.
The key diagnostic test for retinitis pigmentosa is the electroretinogram (ERG), which measures rod and cone function. The ERG detects changes in rod function early in the disease process, often before patients are mentally aware of visual dysfunction.
Because of the remarkable adaptive abilities of the human brain, as much as 90% of rod cells can be lost before a patient is aware of visual changes. Thus, by the time most patients seek medical attention, the disease is in advanced stages. The slow, progressive degeneration of rod photoreceptors inevitably leads to the loss of cone photoreceptors, which are responsible for high acuity, bright light vision. Cone photoreceptors are concentrated in the center of the retina in the “fovea”. The reason for the sequential degeneration is not fully understood, but it appears that cone survival is dependent upon the survival of rod photoreceptors.
The prevalence of RP in the United States is one in 4,000 individuals, affecting approximately 100,000 people. Worldwide, RP affects approximately 1.5 million individuals. RP is genetically inherited as an X-linked, autosomal recessive or autosomal dominant disease. The first mutation associated with human RP was identified in the gene encoding rhodopsin, the G-protein coupled receptor of rod photoreceptor cells.
Since then, mutations in 140 additional genes have been mapped or identified to be involved in RP (for a summary see RetNet at http://www.sph.uth.tmc.edu/retnet/).
Genetic screens of unrelated RP patient populations show that individual affected genes account only for a few percent of RP cases. Upwards of 40% of RP cases are of
5 unknown genetic etiology. However, mutations in the rhodopsin gene account for approximately 25% of autosomal dominant RPs (ADRPs), a significant proportion of RP patients. Thus, understanding the mechanisms by which rhodopsin mutations cause RP has a significant health impact. This chapter focuses on our current knowledge of rhodopsin mutants and their role in retinal degenerative disease.
Because of space limitations, we are unable to discuss many studies and apologize to those investigators whose work is not included here. Our discussions within this chapter will focus on the genetics, biochemistry and physiology of specific rhodopsin mutations that have been studied in greater detail.
RHODOPSIN FUNCTION
RETINA STRUCTURE
The rhodopsin protein is abundantly expressed in the retina, the light sensitive neurosensory tissue that lines the back of the eye. The protein is expressed in a tissue- and cell-specific manner. Shown in Figure 1 is a schematic of a mammalian eye. Vision in mammals begins with the entry of light through the transparent cornea. Light passes through the anterior chamber of the eye, through the pupillary opening and is focused by the lens. A visual image is focused on the retina, the tissue comprised of photoreceptors and other neuronal and glial cells.
The retina is a highly organized, stratified structure (Figure 2). Thus degenerative changes are easily recognized by loss of organization and stratification. The retina is bounded by retinal pigment epithelial (RPE) cells, which provide metabolic support to abutting photoreceptor cells. There are two major classifications of photoreceptor cells –
6 rods and cones. Rods are responsible for dim light vision, whereas cones are responsible for high acuity color vision in bright light. Photoreceptors have a polarized cell structure, with outer segments at one end of the cell and the synapse at the other. Rhodopsin is found within the membranous discs of the outer segments, which stratify as a layer adjacent to the retinal pigment epithelium. The outer segments are joined to the inner segment by a modified cilium. The inner segment contains the mitochondria and endoplasmic reticulum, where proteins are synthesized, then transported through the cilium into the outer segment. The nuclei of rod and cone photoreceptor cells are stratified in a separate “outer nuclear layer” (ONL) and can often be distinguished by the distinctive staining pattern of the nucleus, with rods having densely stained, rounded nuclei and cones having nuclei that stain heterogeneously. Photoreceptor cells form synapses in a single layer referred to as the “outer plexiform layer” (OPL).
Photoreceptors synapse directly or indirectly to second order cells that localize within another strata called the “inner nuclear layer” (INL). The inner nuclear layer contains bipolar cells, horizontal cells and amacrine cells. In addition, nuclei for the glial support cells are found in the INL. The bipolar, horizontal and amacrine cells form synapses in the “inner plexiform layer” (IPL) with cells of the innermost layer of the retina, the ganglion cells. The ganglion cells are located in the “ganglion cell layer” (GCL), and their axons create the optic nerve through which visual information is sent to the brain.
ROD CELL-SPECIFIC EXPRESSION OF RHODOPSIN
Rhodopsin is the G-protein coupled receptor (GPCR) expressed specifically and exclusively in rod photoreceptor cells. Rhodopsin molecules are densely packed in
7 photoreceptor outer segments, contributing to rod photoreceptors being sensitive to a single photon. Rhodopsin is the prototypical Class A receptor, possessing seven transmembrane domains. Unlike other G-protein coupled receptors in which direct ligand binding activates the receptor, rhodopsin covalently binds its ligand, 11-cis-retinal.
The 11-cis-retinal is an inverse agonist, holding rhodopsin in the inactive state.
A schematic of the phototransduction cascade is presented in Figure 3. Activation by a single photon produces a conformational change, converting 11-cis-retinal to all- trans-retinal. The conformational change activates rhodopsin (R*) so that it can bind the rod-specific heterotrimeric G-protein, transducin. Transducin is activated by the substitution of GDP by GTP on the transducin -subunit. This causes the dissociation of the transducin -complex from the active GTP-bound -subunit. The activated transducin -subunit activates the cyclic guanosine monophosphate phosphodiesterase
(cGMP PDE) by binding the phosphodiesterase -subunits, removing their inhibitory effect on the catalytic cGMP phosphodiesterase - and -subunits. Activated cGMP phosphodiesterase - and -subunits hydrolyze cGMP to 5’GMP, lowering intracellular levels of cGMP, a second messenger of photoreceptor cells (Cobbs, Barkdoll et al. 1985;
Fesenko, Kolesnikov et al. 1985; Haynes, Kay et al. 1986).
Intracellular concentrations of cGMP regulate cyclic nucleotide gated (CNG) channels located within the plasma membrane of photoreceptor cells. In darkness, cGMP intracellular levels are elevated and bind to the CNG channels, holding them in the open configuration. Open CNG channels allow the entry of sodium and calcium into the photoreceptor outer segment. Rhodopsin activation results in the hydrolysis of cGMP, thus decreasing intracellular cGMP concentration. The reduction in cGMP concentration
8 reduces the proportion of cGMP bound channels and increases the proportion of closed
CNG channels. Closure of the CNG channels blocks the flow of sodium and calcium into the cell, effectively hyperpolarizing the plasma membrane. This change in membrane potential is transmitted as a neural signal through the secondary neurons of the retina to the ganglion cells and, ultimately to the brain.
Restoration of light-activated rhodopsin to a light sensitive state is well characterized. Inactivation of rhodopsin is a multi-step process. First, rhodopsin kinase phosphorylates light-activated rhodopsin, uncoupling rhodopsin and transducin. Next, visual arrestin binds phosphorylated rhodopsin, completely quenching rhodopsin signaling activity. All-trans-retinal is restored to its light-sensitive conformation through the visual cycle, which takes place within retinal pigment epithelial cells (Ebrey and
Koutalos 2001; Pepperberg and Crouch 2001; Kono, Goletz et al. 2008).
CLASSIFICATION OF RHODOPSIN MUTATIONS
More than one hundred different rhodopsin mutations have been identified in patients with retinitis pigmentosa. Most of these mutations are point mutations and a small number are deletions. Mutations are located throughout the protein in the intradiscal, transmembrane and cytoplasmic domains (Figure 4). There is poor correlation between disease severity and location of mutations.
The mechanisms by which rhodopsin mutations cause retinitis pigmentosa have been the focus of much scientific research. To date, many different hypotheses have been proposed and tested, including impaired protein folding and mislocalization, release of toxic products, and aberrant signaling. With over one hundred different rhodopsin
9 mutations identified, it is not surprising that a single mechanism that leads to rhodopsin- mediated autosomal dominant retinitis pigmentosa (ADRP) has not been found. The research does suggest that rhodopsin mutations cause ADRP through multiple mechanisms. This research highlights the importance of understanding the underlying mechanism of each mutation to develop effective therapies.
Studies by Nathans and collaborators addressed the mechanism of mutant rhodopsin-mediated retinitis pigmentosa by sorting rhodopsin mutants into two classes
(Sung, Schneider et al. 1991). Mutant rhodopsins were expressed in a heterologous cell system and classified into two groups based on the mutant proteins’ expression levels, ability to regenerate spectrally active rhodopsin by association with 11-cis-retinal, and subcellular localization. Class 1 mutants expressed at levels comparable to that of wild- type rhodopsin, regenerated with 11-cis-retinal to form light-responsive rhodopsin, and localized properly to the plasma membrane. Class 1 mutations tended to be within the C- terminus of rhodopsin. Class 2 mutants were more common than class 1 mutants and were characterized by decreased protein expression compared to expression of wild-type rhodopsin, inability to associate with 11-cis-retinal and mislocalization within heterologous cells. Class 2 mutations tended to occur in the intradiscal, transmembrane and cytoplasmic domains of rhodopsin. The findings of this study suggested that class 2 mutants might cause degeneration through decreased expression of the rhodopsin protein, inability to form spectrally active rhodopsin and mislocalization of protein. However, since the class 1 mutants appeared normal in all tests, the question of mechanism remained for this class of mutations. This classification system remains popular, although other groups have developed other subclasses based on added criteria.
10
A modified classification system taking more recent data into account was proposed in 2005 by Mendes et al. (Mendes, van der Spuy et al. 2005). Expression of some class 1 mutants in transgenic animals revealed that the C-terminal mutations affected trafficking and were not targeted appropriately to the photoreceptor outer segment (Li, Snyder et al. 1996). The new classification system is not mutually exclusive. Mutant rhodopsins can fall within several categories. The system proposed that the new class 1 mutants folded normally but mislocalized within the photoreceptor cells. Class 2 mutants misfolded, were retained within the endoplasmic reticulum, and did not associate with 11-cis-retinal. Class 3 mutants affected endocytosis. Class 4 mutants affected rod opsin stability and post-translational modifications. Class 5 mutants had an increased activation rate for transducin, but no obvious folding defect. Class 6 mutants also had no obvious folding defect and were constitutively active in absence of chromophore and in the dark.
Both classification systems group rhodopsin mutants that share common mechanisms of degeneration and therefore, might share common therapies. While groups of rhodopsin mutations may cause disease through similar mechanisms, no single pathogenic mechanism common to all rhodopsin mutations has been found that would allow a single treatment for all cases of rhodopsin-mediated RP. The diversity of mechanisms causing rhodopsin-mediated RP suggests the requirement for “personalized medicine”.
MECHANISMS OF DEGENERATION
P23H, VPP: A MODEL FOR MISFOLDED AND MISLOCALIZED RHODOPSIN MUTANTS
11
One of the best studied rhodopsin mutations is a point mutation of proline at residue 23 (P23). P23 is located in the N-terminus of rhodopsin, within the intradiscal space (Figure 4). While other rhodopsin mutations are more prevalent in other continents, the P23H rhodopsin mutation is the most prevalent rhodopsin mutation in the
United States, suggesting a founder effect (Dryja, McGee et al. 1990).
In vitro expression of P23H mutant rhodopsin showed mislocalization of misfolded protein, making it a class 2 mutant by the classification system of Sung et al.
Protein misfolding and mislocalization have both been implicated as causes of other neurodegenerative diseases (Muchowski 2002; Soto and Estrada 2008; Williams and
Paulson 2008). Misfolded proteins can cause ER stress, activate the unfolded protein response (UPR), and result in the formation of protein aggregates within cells, all of which can lead to apoptosis. Recently Tam et al. showed that mislocalized rhodopsin killed cells independent of light by activating normally inaccessible signaling pathways
(Alfinito and Townes-Anderson 2002; Galy, Roux et al. 2005; Tam, Xie et al. 2006).
The P23H rhodopsin mutant provides a model system for studying how protein misfolding and mislocalization cause cell death.
Misfolding
Gene expression studies in a heterologous cell system suggested that P23H rhodopsin misfolded and mislocalized (Sung, Schneider et al. 1991). Protein expression levels were reduced and showed altered difference spectra and oligosaccharide profiles.
The reduction in protein expression levels suggested the mutant protein was not stable and was either not formed initially, or was degraded faster than wild-type biosynthesis.
12
To assess whether the P23H mutant produced a functional rhodopsin molecule
(composed of the opsin apo-protein and the 11-cis-retinal chromophore ligand), the difference spectra was analyzed. Normal rhodopsin produces an absorbance spectrum that shows a characteristic shift when exposed to light. Misfolded rhodopsins unable to bind the chromophore ligand do not produce the characteristic light-shifted absorbance spectrum. The difference spectrum is calculated by subtracting the absorbance spectra after photobleaching from the absorbance spectra before photobleaching. The P23H difference spectra was abnormal compared to wild-type rhodopsin, indicating the P23H mutant rhodopsin does not combine with 11-cis-retinal to form spectrally active rhodopsin (Figure 5). Misfolding of P23H rhodopsin was also suggested by the lack of complex oligosaccharides, which are added during protein maturation in the Golgi complex.
P23H mutant rhodopsin misfolding has been confirmed by additional in vitro studies. Liu et al. used circular dichroism (CD) spectral techniques to analyze misfolded and correctly folded rhodopsins (defined by their inability or ability, respectively, to bind retinal). The CD spectra indicated that the misfolded P23H rhodopsin had 50-70% less helical content than wild-type rhodopsin and lacked the characteristic banding pattern typical of wild-type rhodopsin, indicating that the mutant has a different tertiary structure
(Liu, Garriga et al. 1996). Finally, a trypsin digest of misfolded P23H mutant rhodopsin degraded the mutant protein into small fragments whereas properly folded rhodopsin was relatively resistant to trypsin digestion. Both the CD spectra and trypsin digestion further supported the hypothesis that P23H mutant rhodopsin misfolds.
13
As a rule, misfolded proteins are ubiquitinated and targeted for degradation via the proteasome pathway. Illing et al. confirmed the involvement of the ubiquitin pathway in P23H mutant rhodopsin degradation (Illing, Rajan et al. 2002). In the absence of proteasome inhibitors, the mutant rhodopsin was unstable and quickly targeted for degradation via ubiquitination. However, in the presence of proteasome inhibitors and dominant negative ubiquitin, P23H mutant rhodopsin was stabilized. These results strongly suggest that P23H mutant rhodopsin is unstable, ubiquitinated and degraded via the proteasome pathway.
Because rhodopsin molecules can form large oligomeric complexes, it was hypothesized that the presence of mutant rhodopsin could interfere with wild-type rhodopsin function. In fact, it was later confirmed that after expressing both P23H mutant rhodopsin and wild-type rhodopsin in vitro, the presence of P23H mutant rhodopsin impaired delivery of wild-type rhodopsin to the plasma membrane (Rajan and
Kopito 2005). Co-expression of mutant and wild-type rhodopsins lead to an increase in proteasome mediated degradation and steady state ubiquitination of wild-type rhodopsin, demonstrating that P23H mutant rhodopsin has a dominant negative effect. This study is especially critical because it was one of the first to identify how mutant rhodopsin can interfere with wild-type rhodopsin activity.
Mislocalization
Studies by Sung et al. suggested that in addition to misfolding, P23H mutant rhodopsin also did not localize correctly (Sung, Schneider et al. 1991). In the retina, rhodopsin is synthesized on the rough endoplasmic reticulum (ER) and transported via
14 the Golgi apparatus to the plasma membrane. The rhodopsin-laden plasma membrane at the rod cilium invaginates and eventually pinches off to form disc membranes. In the heterologous cell system used in this study, wild-type rhodopsin localized to the plasma membrane. In contrast, the P23H mutant rhodopsin localized intracellularly with a staining pattern indicating that it was trapped within the ER. P23H mutant rhodopsin was noticeably absent from the plasma membrane.
A Drosophila model expressing Rh1P37H, which corresponds to human rhodopsin P23H, was studied to determine in vivo localization of the mutant rhodopsin
(Galy, Roux et al. 2005). Data from these flies indicated that the mutant rhodopsin accumulated primarily in the ER with a small portion located in the rhabdomeres, the fly equivalent to rod outer segments. The study also indicated that P23H mutant rhodopsin was active when localized properly, although with a significantly reduced activity. The mislocalized mutant rhodopsin activated the stress-specific mitogen-activated protein kinases, p38 and JNK. Both p38 and JNK are known to control stress-induced apoptosis.
Activation of these kinases linked mislocalized rhodopsin directly to the induction of apoptosis.
Transgenic frogs have also been used to analyze P23H mutant rhodopsin localization in vivo. Frogs were created that over-expressed frog, bovine, human and murine forms of P23H mutant rhodopsin (Tam and Moritz 2006; Tam and Moritz 2007).
All of the transgenic frogs expressing P23H mutant rhodopsin underwent retinal degeneration. Analysis of the frogs indicated that the mutant rhodopsin was expressed at very low levels and that the mutant protein was retained within the endoplasmic reticulum. Interestingly, it was observed that light sensitivity varied between the mutant
15 rhodopsin species. While frogs expressing bovine P23H mutant rhodopsin had complete protection from degeneration when raised in the dark, frogs expressing human P23H mutant rhodopsin only had a partial rescue, and frogs expressing frog and murine P23H mutant rhodopsin had no protection from degeneration. In the frogs protected against degeneration, the mutant rhodopsin localized properly to the outer segments. It was found that the mutant rhodopsin was only able to localize properly after truncation of the
N-terminus. Truncation of the N-terminus allowed the mutant protein to exit the ER and then transport to the outer segment.
P23H mutant rhodopsin localization in vivo has also been analyzed using transgenic mice. While the initial characterization of the transgenic mice suggested that the mutant rhodopsin predominantly localized correctly to rod outer segments (Olsson,
Gordon et al. 1992; Wu, Ting et al. 1998), later data suggested that the majority of the mutant rhodopsin mislocalized (Roof, Adamian et al. 1994; Frederick, Krasnoperova et al. 2001) and was retained within the inner segment and in the outer plexiform layer early in the degeneration. Interestingly, the buildup of P23H mutant rhodopsin in the outer plexiform layer also correlated with the accumulation of mislocalized phosphodiesterase and transducin to the outer plexiform layer (Roof, Adamian et al.
1994).
Several lines of evidence, including in vitro and in vivo data from Drosophila, frogs and mice, suggest that P23H mutant rhodopsin mislocalizes. The mislocalization of the mutant rhodopsin molecule has been associated with the mislocalization of other signaling molecules, and may lead to degeneration via activation of p38 and JNK kinases.
16
P347S: A MODEL FOR A NON-AUTONOMOUS PATHWAY AND DESTABILIZED RHODOPSIN
MUTANTS
Seven different point mutations of the proline at residue 347 in rhodopsin have been linked to retinitis pigmentosa – P347S, P347L, P347A, P347G, P347C, P347R, and
P347T. P347 is located in the C-terminal cytoplasmic tail of rhodopsin (Figure 4), an area suggested to be critical for proper protein trafficking and sorting. The last five residues of rhodopsin, of which P347 is one, are highly conserved amongst species, indicating their importance.
In vitro data demonstrates that P347S mutant rhodopsin regenerates normally with
11-cis-retinal, has a normal response to light, activates tranducin, is phosphorylated by rhodopsin kinase, can bind arrestin, and localizes properly (Li, Snyder et al. 1996). In summary, all of the in vitro data suggests that P347S mutant rhodopsin behaves the same as wild-type, placing it within Sung et al.’s class 1 rhodopsin mutants.
Non-autonomous degeneration
The creation of transgenic chimeric mice expressing P347S mutant rhodopsin in only a subset of photoreceptor cells demonstrated that degeneration was through a non- autonomous pathway (Huang, Gaitan et al. 1993). The retinas of the chimeric mice had populations of photoreceptor cells expressing only wild-type rhodopsin, and populations of photoreceptor cells expressing both mutant and endogenous rhodopsin. Analysis of three chimeric mice demonstrated that the level of degeneration correlated to the percentage of cells expressing the P347S mutant rhodopsin. Furthermore, mice with a greater number of cells expressing only wild-type rhodopsin had slower rates of retinal
17 degeneration. Yet, photoreceptor cells expressing only wild-type rhodopsin were also dying. The observation that photoreceptor cells expressing only wild-type rhodopsin were degenerating indicated the involvement of a non-autonomous pathway in the degeneration.
After demonstrating the involvement of a non-autonomous pathway, researchers hypothesized that photoreceptor cells expressing the mutant rhodopsin released a toxic factor that killed neighboring photoreceptor cells. If the photoreceptor cells expressing mutant rhodopsin released a toxic factor, then cells adjacent to those expressing the mutant rhodopsin should degenerate first. However, this localized degeneration was not observed. In fact, the photoreceptor cells of chimeric mice degenerated in a uniform fashion regardless of proximity to P347S mutant expressing cells. These results suggested that the release of a diffusible toxic factor from degenerating photoreceptor cells was not the cause of neighboring cell death (Huang, Gaitan et al. 1993). These observations do not rule out the involvement of a trophic factor in the degeneration.
In summary this study not only demonstrated the involvement of a non- autonomous pathway but also suggests that the degeneration was not dependent on the release of a toxic factor from degenerating cells. Later, we will further discuss the involvement of both the release of toxic factor and the necessity of a trophic factor in cone cell degeneration.
Cell sorting
P347 is located within the last five residues of the C-terminal tail of rhodopsin.
These five residues are highly conserved across species, presumably due to their potential
18 involvement in cell sorting and trafficking. Several studies have investigated whether protein mislocalization is involved in the degeneration caused by the P347S mutation in rhodopsin.
As previously mentioned, when P347S mutant rhodopsin is expressed in a heterologous cell system, it localizes at levels similar to wild-type rhodopsin at the plasma membrane, suggesting that P347S rhodopsin is able to localize correctly (Sung,
Schneider et al. 1991). However, mislocalization of mutant protein was predicted when a frog retinal cell-free system was used to examine mutant rhodopsin trafficking (Deretic,
Schmerl et al. 1998). The study found that the last five residues of rhodopsin are critical for proper protein sorting. Synthetic peptides using the last five residues of rhodopsin as competitive inhibitors prevented cell sorting of wild-type rhodopsin. When the last five residues were deleted from the competitive peptide, wild-type rhodopsin sorted to the membrane. Introducing the P347S mutation into the peptide permitted wild-type rhodopsin trafficking, suggesting that residue P347 plays a critical role in proper cell trafficking and sorting.
Two lines of P347S transgenic mice were made in independent labs. The first line of mice created had normal retinal development until the third postnatal week, after which photoreceptor cells started dying (Chang, Hao et al. 1993). By seven weeks of age, these P347S transgenic mice have only half of their outer nuclear layer remaining.
In this mouse model, while some mutant rhodopsin was properly targeted to the rod outer segments, some mutant rhodopsin mislocalized in the inner segment (Chang, Hao et al.
1993).
19
A second independent line of P347S transgenic mice was also made (Li, Snyder et al. 1996). In these mice, there was a direct correlation between the severity of degeneration and the level of P347S mutant rhodopsin mRNA. The new line of P347S transgenic mice had a much slower rate of retinal degeneration than those created by
Chang et al. The P347S mutant rhodopsin localized predominantly in the outer segments, but also accumulated in submicrometer-sized extracellular vesicles near the junction of inner and outer segments. The accumulation of the mutant rhodopsin in these extracellular vesicles suggested that protein trafficking may play a contributory role in the retinal degeneration, but a majority of the mutant rhodopsin does localize properly to the rod outer segments.
Together these studies suggest that protein mislocalization may play a role, but as most of the P347S mutant rhodopsin does localize properly, mislocalization probably does not fully account for the degeneration seen in the P347S model.
Rhodopsin Stability and Signaling
Abnormal signaling has been suggested to be involved in rhodopsin-mediated
ADRP. The involvement of an abnormal signaling pathway has been investigated with
P347S mutant rhodopsin mice. Using the mice created by Chang et al., Weiss et al. demonstrated that there were increased levels of cAMP in P347S retinas compared to wild-type retinas (Weiss, Hao et al. 1995). Since cAMP levels are not normally regulated through the canonical phototransduction pathway, increased levels of cAMP in rhodopsin mutant mice suggested the involvement of a non-canonical pathway. Together with the
P347S mislocalization studies, these data suggest the mutant rhodopsin might be
20 activating a secondary pathway that is not normally utilized when rhodopsin is localized properly to the outer segments.
Previous work has shown that persistent photosignaling can cause retinal degeneration (Hao, Wenzel et al. 2002; Woodruff, Wang et al. 2003). Furthermore, degeneration caused by persistent photosignaling can be rescued by introducing a null mutation for the -subunit of transducin which prevents G-protein mediated signaling.
Since P347S mutant rhodopsin predominantly localizes to rod outer segments, regenerates with 11-cis-retinal, and has a normal light response, it was proposed that
P347S mutant rhodopsin caused degeneration through persistent photosignaling. We tested this hypothesis by placing P347S expressing animals (Li, Snyder et al. 1996) onto an -transducin null background. The results of this experiment are presented in chapter
2 and demonstrated that the presence of -transducin provides protection from degeneration, as removal of the -transducin increased the rate of degeneration (Brill,
Malanson et al. 2007).
The accelerated degeneration in the absence of -transducin also suggested a role for transducin in the stabilization of P347S mutant rhodopsin. Further results, including protection by dark-rearing and increased levels of bis-retinoid fluorophores and lipofuscin granules, suggested that the P347S mutant rhodopsin was unstable and released all-trans-retinal at a faster rate than wild-type rhodopsin (Brill, Malanson et al.
2007). Free all-trans-retinal is toxic to the cells and an increased rate of release could be toxic to the photoreceptor cells. To directly test the role of chromophore toxicity, we placed the P347S expressing mice onto an Rpe65 null genetic background. The Rpe65 enzyme is critical for the recycling of chromophore and, in its absence, cells cannot
21 produce 11-cis-retinal. Without 11-cis-retinal, all-trans-retinal cannot be produced.
Therefore, if the degeneration present in the P347S mutant is caused by increased release of all-trans-retinal, then placing the mice on the Rpe65 null background would provide protection. To our surprise, placing the mice on the Rpe65 null background did not slow the degeneration. We believe that the Rpe65 null mutation did not rescue the degeneration because the P347S mutant rhodopsin may require 11-cis-retinal to properly fold and traffic to the outer segments. Further examination is required to determine the necessity of chromophore in mutant rhodopsin localization.
Taken together, these studies suggest that the mutant P347S rhodopsin has abnormal biochemistry which leads to the degeneration present in the mouse models.
The mutant rhodopsin may activate a non-canonical signaling pathway within the inner segments, resulting in the increased levels of cAMP in the retina. The P347S mutant rhodopsin also is unstable and releases toxic all-trans-retinal faster then wild-type rhodopsin.
K296E: A MODEL FOR PERSISTENT SIGNALING AND ALTERNATIVE SIGNALING PATHWAYS
The lysine to glutamic acid substitution at residue 296 of rhodopsin is associated with human RP. The mutant is well studied at the biochemical level. K296 is located within the seventh transmembrane domain of rhodopsin (Figure 4) and is the site of choromphore attachment. Additionally, K296 forms a salt bridge with G113 to restrict the motion of the protein, keeping it in its inactive conformation (Oprian 1992). The
K296E mutation blocks both the formation of the Schiff base and the salt bridge. Without the formation of the Schiff base, the chromophore cannot attach, which prevents the
22 creation of spectrally active rhodopsin. Additionally, without the formation of the salt bridge, opsin cannot be locked into its inactive conformation. Not surprisingly, K296E is constitutively active in vitro. In a heterologous cell system, Robinson et al. found that
K296E constitutively activated transducin independent of light (Robinson, Cohen et al.
1992). These results suggested that K296E mutant rhodopsin fits into a group of rhodopsin mutations that cause retinal degeneration through persistent photosignaling.
Persistent photosignaling
The creation of a mouse expressing K296E mutant rhodopsin allowed investigators to study the mechanisms involved in the retinal degeneration mediated by this rhodopsin mutation in vivo. Mice expressing K296E rhodopsin undergo progressive retinal degeneration. In vivo, the K296E mutant rhodopsin localized to rod outer segments, was phosphorylated, and bound arrestin (Li, Franson et al. 1995). These results suggested the mutant rhodopsin was in the inactive state and not constitutively active, contrary to in vitro results. Indeed, the degeneration associated with K296E was independent of persistent photosignaling (Li, Franson et al. 1995; Brill, Malanson et al.
2007).
Several lines of evidence suggested that K296E mutant rhodopsin was not causing degeneration through light-dependent signaling in vivo. First, removing all light stimuli by dark-rearing K296E mutant mice did not protect from degeneration. Dark-reared mutant mice had decreased ERG amplitudes (Li, Franson et al. 1995), indicating progressive retinal degeneration despite the lack of light stimulus. Second, blocking persistent photosignaling, by placing the K296E mutant on the -subunit transducin null
23 background, also did not protect from degeneration (Brill, Malanson et al. 2007).
Instead, K296E rhodopsin mutant mice had a faster rate of degeneration in the absence of transducin, suggesting that the presence of -transducin provides some degree of protection against degeneration. In summary, these results suggest that the K296E rhodopsin mutation is not causing degeneration through persistent photosignaling, and that the presence of the -transducin provides protection from degeneration for this rhodopsin mutation. The protective effect of transducin suggests that the K296E mutant rhodopsin may be signaling through a G-protein independent pathway, which may lead to degeneration. Transducin might provide protection by either blocking the ability of
K296E mutant rhodopsin to signal through a G-protein independent pathway or by stabilizing K296E mutant rhodopsin.
Rhodopsin-Arrestin Complex formation
Evidence from Drosophila studies suggested that the formation of stable rhodopsin-arrestin complexes caused degeneration. Preventing the formation of these complexes prevented degeneration (Alloway, Howard et al. 2000; Kiselev, Socolich et al.
2000; Iakhine, Chorna-Ornan et al. 2004). To analyze the involvement of rhodopsin- arrestin complex formation in K296E mediated degeneration, mice were placed onto an arrestin knockout background. The removal of arrestin prevented the formation of rhodopsin-arrestin complexes. If degeneration were dependent on the formation of these complexes, preventing their formation should provide protection from degeneration.
Analysis of K296E, arrestin knockout mice demonstrated that the presence of rhodopsin- arrestin complexes contributed to, but was not the sole mechanism of degeneration
24
(Chen, Shi et al. 2006). In conclusion, these studies demonstrated the potential involvement of abnormal protein complexes in retinal degeneration in rhodopsin- mediated ADRP.
MECHANISMS OF CONE DEGENERATION
Perhaps the most puzzling question about retinitis pigmentosa is why rod cell death is inevitably followed by cone cell death. Although cone cells account for only 5% of photoreceptors, they are essential for high acuity daylight vision and their degeneration has devastating effects. Most of the RP disease gene alleles are rod cell-specific. This makes it all the more perplexing as to why cone photoreceptors progressively die as well.
Yet, the loss of rods is not concomitant with the loss of cones. In human retinitis pigmentosa patients, cone photoreceptor cells can survive for years in the absence of rods before dying. There are no known forms of retinitis pigmentosa where rods die and cones survive. Conversely, cone specific genetic mutations can result only in cone cell death. This has led to the hypothesis that cones depend on rods for their survival. Three hypotheses on the mechanisms underlying cone cell death include: 1) the loss of rod- derived trophic factor(s), 2) release of toxic factor(s) during rod cell death, and (3) metabolic support imbalance following rod cell death. In the following sections we discuss evidence supporting each of these theories.
TROPHIC FACTORS
This hypothesis suggests that healthy rod photoreceptor cells release trophic factors that cones require for survival. Mohand-Said et al. demonstrated the existence of
25 a diffusible trophic factor that supported cone cell survival in organ co-culture experiments (Mohand-Said, Deudon-Combe et al. 1998). Retinas carrying residual cone cells from retinally degenerate mice were co-cultured with rod-containing retinas. The experiments showed that co-culture with rods increased cone survival in the degenerating retina (Figure 6). Since the co-cultured retinas did not physically touch, the group concluded that the trophic factor was diffusible. The group also concluded that the trophic factor came from rod cells since control experiments with rodless retinas did not increase cone survival in the degenerating retina. Streichert et al. also demonstrated the existence of a rod derived trophic factor which supported the survival of photoreceptor cells in a retinal degeneration mouse model (Streichert, Birnbach et al. 1999). While this group did not specifically analyze the survival of cones, they did find that the rod derived trophic factor extended the life of rod photoreceptor cells in P23H rhodopsin mutant retinas. The survival promoting factor was diffusible, heat labile and absent from RP retinas. The group tested a variety of trophic factors expressed in retinas, including basic fibroblast growth factor (bFGF), brain-derived neurotrophic factor (BDNF) and glial cell derived neurotrophic factor (GDNF), but no one factor provided the protection seen by co-culture with normal rod cells.
Two rod derived trophic factors were recently identified by expression cloning
(Leveillard, Mohand-Said et al. 2004; Chalmel, Leveillard et al. 2007). The groups appropriately named the trophic factors rod-derived cone viability factor (RdCVF) and rod-derived cone viability factor-2 (RdCVF2). Both were truncated thioredoxin-like proteins expressed specifically in photoreceptor cells. They share similar gene and three dimensional protein structures, and both were shown to promote cone survival in vitro.
26
Taken together, these studies suggest that healthy rod cells release a trophic factor that supports the viability of cone cells. Therefore it follows that as rod cells die in retinitis pigmentosa, there is a decrease in this trophic factor and it is this decrease in trophic factor that is responsible for cone photoreceptor cell death. However the fact that cone cells are functional for years after rod cells have died in human RP patients argues against this theory.
TOXIC FACTORS
A second hypothesis on the mechanism of cone cell death in retinitis pigmentosa involves the release of toxic factors in the retina by dying rod cells or activated microglia.
The toxic factors could be reactive oxygen species, excess glutamate or products of apoptosis that accumulate in retinas lacking rod photoreceptor cells.
Ripps proposed a mechanism whereby toxic factors released from degenerating rods cells travel through gap junctions and cause cone cell death (Ripps 2002). Possible toxic factors include glutamate and products of apoptosis. While the field generally agrees that the release of excess glutamate is unlikely to be the cause of cone cell death, the release of apoptotic byproducts may play a role in cone cell death in RP.
In the central nervous system, activated microglia migrate to damaged areas and phagocytose cellular debris. Activated microglia also secrete molecules that can kill surrounding healthy neurons. These molecules include nitric oxide, reactive oxygen species, excitatory amino acids, proteases and proinflammatory cytokines (Suzumura,
Takeuchi et al. 2006). Gupta et al. demonstrated the presence of activated microglia in degenerating human retinas using immunocytochemistry (Gupta, Brown et al. 2003). In
27 normal human retina, quiescent microglia were small, stellate cells associated with the inner retina blood vessels. However, in degenerating retinas, numerous activated microglia were present in the outer nuclear layer in degenerating areas. The activated microglia were enlarged amoeboid cells that contained rhodopsin-positive cytoplasm.
While the group did not demonstrate that the microglia were secreting toxic substances that might kill neighboring photoreceptor cells, previous evidence from the central nervous system suggests that the release of toxic substances from microglia might play a role in retinitis pigmentosa.
Reactive oxygen species may also increase following rod photoreceptor cell death. Rod photoreceptor cells are responsible for most of the oxygen consumption in the outer retina. Oxygen levels in the outer retina vary inversely with oxygen consumption. As the number of rod photoreceptors decreases so does the oxygen consumption, which increases the levels of oxygen in the outer retina (Yu, Cringle et al.
2000; Yu, Cringle et al. 2004). The increase in oxygen in the outer retina may lead to the creation of toxic reactive oxygen species which may be responsible for cone cell death.
Shen et al. demonstrated the involvement of reactive oxygen species in cone photoreceptor cell loss (Shen, Yang et al. 2005). Using P347L mutant rhodopsin transgenic pig retinas, biomarkers of oxidative damage to lipids, proteins and DNA were found within cone inner segments after the loss of rod photoreceptors. This study strongly suggests the involvement of oxidative stress in cone cell loss.
The involvement of oxidative stress in retinal degeneration is also supported by pharmacological studies in which administration of antioxidants increased photoreceptor survival (Komeima, Rogers et al. 2006; Komeima, Rogers et al. 2007). Treatment with
28 antioxidants promoted cone survival in three separate RP mouse models. Antioxidant administration increased cone cell density and rhodopsin mRNA levels, slowed the decline of function as measured by ERG, and slowed the thinning of the outer nuclear layer. These results strongly support the involvement of oxidative damage in cone photoreceptor cell loss.
Taken together these results support the idea that cone photoreceptor cells are lost following rod cell death due to the increase of toxic substances in the retina. These toxic substances may be released by dying rod photoreceptor cells or by activated microglia.
The toxic substances may be a byproduct of rod cell death, as is the case for increased oxygen levels, resulting in oxidative damage.
METABOLIC SUPPORT
A third hypothesis for the mechanism of cone cell death following rod photoreceptor cell loss involves the requirement of rod photoreceptor cells for cone cell metabolic support (Punzo, Kornacker et al. 2009). This hypothesis hinges on the idea that photoreceptor outer segments receive most of their metabolic support from the retinal pigment epithelium (RPE). There are approximately 20-30 photoreceptor outer segments in contact with each RPE cell. Rods account for 95% of photoreceptors in the retina.
Therefore, a single RPE cell makes contact with only 1 or 2 cone photoreceptor outer segments, which might not be enough to maintain a functional connection. This ratio of rod to cone photoreceptors is not true in the macula of the human retina, an area that is exclusively cones. However, rodents lack a macula, which limits our abilities to completely analyze this third hypothesis.
29
A study comparing four mouse models of retinitis pigmentosa, including the
P23H/VPP mutant rhodopsin mouse, suggests cone cells die due to lack of metabolic support (Punzo, Kornacker et al. 2009). Gene expression arrays were used to identify common changes that occurred in all four models at the onset of cone death. The study identified a large number of genes that were involved in cellular metabolism, suggesting cones were suffering from a shortage of nutrients. The study also found that cones showed signs of autophagy, a process of self-digestion. One group of genes identified included several components of the mammalian target of rapamycin (mTOR) pathway, which regulates cellular metabolism. To bypass this pathway, mice were treated with insulin. Mice receiving insulin had transiently prolonged cone survival compared to controls. The results of this study suggested that as rod photoreceptor cells died, the flow of nutrients, including glucose, to the cone photoreceptor outer segments is disrupted.
While this study is very convincing, there are two pieces of evidence that suggest changes in metabolic support are not the only mechanism of cone cell death in RP. First, if lack of metabolic support is the sole mechanism causing cone photoreceptors to die, the macula, an area of the human retina containing only cone cells, should never be affected.
While the cone photoreceptors in the macula live the longest in human RP, they too eventually die, suggesting that the disruption of outer segment RPE interaction cannot be the sole mechanism of cone cell death. Second, in a rat model of RP, photoreceptors could be rescued from degeneration by the application of various growth factors even when a thick layer of debris was present that separated the photoreceptors from their source of nourishment (Steinberg 1994). Therefore, while metabolic support does seem
30 to play a role in the degeneration of cone photoreceptors, it seems unlikely that it is the sole mechanism of cone cell loss in retinitis pigmentosa.
TREATMENTS
There is currently no universally effective treatment for retinitis pigmentosa.
While different mechanisms have been found to play a role in the degeneration associated with individual or groups of rhodopsin mutations, no common mechanism has been found that provides a therapeutic target for most or all cases of rhodopsin-mediated
ADRP. In the absence of a shared mechanism, many different therapies have been suggested and tested within labs, including gene therapy, transplantation with stem cells or retinal tissue, and pharmacological treatment with supplements to protect photoreceptor cells. This section will focus on some of the work that has been done with each of these potential therapies.
GENE THERAPY
Gene therapy is theoretically an ideal treatment for rhodopsin-mediated ADRP.
As previously discussed, mutations within rhodopsin cause degeneration through a toxic gain-of-function as opposed to a loss-of-function. Therefore the goal of gene therapy for rhodopsin-mediated ADRP is to decrease the toxic gain-of-function of the mutant rhodopsin. Decreasing the gain-of-function of the mutant rhodopsin has typically been achieved through inhibiting the expression of the mutant protein or increasing the ratio of wild-type to mutant rhodopsin, as the rate of degeneration slows when the ratio of wild-
31 type to mutant rhodopsin is high. Decrease of the toxic gain-of function has also involved targeting the apoptotic pathway to extend the life of photoreceptor cells.
LaVail and colleagues demonstrated the potential of gene therapy for the treatment for rhodopsin-mediated ADRP using recombinant adeno-associated virus expressing hairpin and hammerhead ribozymes (LaVail, Yasumura et al. 2000).
Ribozymes are small RNAs that cleave mutant transcripts in an allele specific manner, leaving the wild-type transcript intact. Rats expressing P23H mutant rhodopsin were injected with ribozymes at P15, before the onset of degeneration, and at P30 or P45, when approximately 40-45% of the photoreceptors had degenerated. Rats injected at both the early and late time points with ribozymes targeting the mutant rhodopsin had increased ONL thickness compared to controls, indicating that the treatment increased the lifespan of photoreceptor cells (Figure 7). Rats injected at P15 also had significantly increased ERG amplitudes compared to controls six months after treatment, indicating that the therapy not only increased the lifespan of photoreceptor cells, but also preserved their function. It is important to note that the treatment had the same protective effect at all time points studied, suggesting that this therapy can be initiated in patients after the onset of degeneration. This study demonstrated that decreasing the expression of mutant rhodopsin could protect photoreceptors from degeneration and preserve function. Yet, many questions still remain on the feasibility for treatment in humans.
One such question is the ability to design viruses that would specifically target mutant rhodopsins. With over 100 different mutations identified within rhodopsin that cause ADRP, having a specific therapy for each mutant is a large hurdle to overcome.
32
One way to bypass the need for mutant-specific therapy is to target a more generalized mechanism, such as apoptosis, the end stage found within rhodopsin-mediated ADRP.
Using an adenovirus expressing an apoptosis inhibitor, Leonard et al. demonstrated the potential use of a more generalized therapy (Leonard, Petrin et al.
2007). The study used an adenovirus that expressed the X-linked inhibitor of apoptosis
(XIAP) which inhibits caspase-3, -7 and -9. The virus was injected into rats expressing either a class 2 mutant rhodopsin, P23H, or a class 1 mutant rhodopsin, S334ter at P14-
P17 before degeneration was evident. At 4 months, injection with virus preserved ONL thickness for both rhodopsin mutants. However the ERG results suggested there was only functional protection for the class 2, P23H mutant rhodopsin rats. ERG results at all time points tested demonstrate that the S334ter mutant rhodopsin rats had no change in
ERG amplitudes compared to uninjected mutant controls. However, the injected P23H mutant rhodopsin rats had increased ERG amplitudes compared to uninjected mutant controls at all time points studied, with the greatest change over time. The ERG findings may downplay the positive results since the injection only treats approximately 20% of the retina, and the ERG measures a total retinal response, not isolated to the injected area.
By expressing an inhibitor for apoptosis, this study demonstrated that gene therapy does not need to only focus on specific rhodopsin mutations, but could focus on a shared end pathway and still protect against degeneration. The difference in response to adenovirus injection of S334ter and P23H rhodopsin mutants suggests that the apoptotic pathways utilized may be different between the two mutants.
Another group has tested the idea that suppression of both mutant and wild-type rhodopsin using short hairpin RNA, followed by replacement with wild-type rhodopsin
33 that is immune to RNA degradation could rescue degeneration (O'Reilly, Palfi et al.
2007; O'Reilly, Millington-Ward et al. 2008; Chadderton, Millington-Ward et al. 2009).
Suppression of both mutant and wild-type rhodopsin also gets around the hurdle of needing to personalize therapy for each of the 100 different rhodopsin mutations. The group used a modified recombinant adeno-associated virus that preferentially infected photoreceptor cells. While injection with the virus rescued the degeneration present in
P347S mutant rhodopsin mice, the rescue was localized to areas receiving the injection and was transient, suggesting the need for multiple injections in a single patient. The investigators suggested replacement with wild-type rhodopsin that is immune to degeneration by the short hairpin RNA, but they have yet to demonstrate the feasibility of doing so after onset of degeneration. Still, the approach is promising for the development of generalized therapy for rhodopsin-mediated ADRP.
Many studies have illustrated the potential benefits of gene therapy in treatment of rhodopsin-mediated ADRP. However, several questions still remain, including the timing of therapy and viral toxicity. One critical issue that needs to be determined with gene therapy is the timing of treatment. As each rhodopsin mutation has a different rate of photoreceptor cell death, the field will need to determine if and when there is a time point when rescue of photoreceptor degeneration is no longer possible. Patients typically do not visit the clinic until they have significant vision impairment, when a majority of their photoreceptor cells have died. Thus, it will be critical to determine if gene therapies are therapeutically effective after a majority of photoreceptor cells have died.
Another complication of gene therapy is the adverse side effects, including fatal forms of severe combined immunodeficiency (SCID) syndrome, leukemia, and fatal
34 inflammatory response that have been associated with the use of adenoviruses.
Investigators are making great strides at creating viruses that will only infect photoreceptor cells and thus minimize the off target effects. Only once the viruses are proven safe will gene therapy be a potential treatment for human retinitis pigmentosa patients.
Another question for therapies focused on decreasing mutant rhodopsin expression is the level of suppression required to maintain long term effects. Most human patients with rhodopsin-mediated RP have a mutation in one allele. The one allele mutation results in approximately 50% of the total rhodopsin being mutant protein. In comparison, animal models currently being studied express mutant rhodopsin in addition to endogenous wild-type rhodopsin. In many animal models, the mutant rhodopsin only accounts for 1%, at most, of the total rhodopsin. Yet, this 1% can lead to total blindness.
Degeneration in animal models with only 1% mutant rhodopsin suggests that therapies focusing on decreasing expression of mutant rhodopsin will need to completely, or near completely, eliminate the expression of the mutant rhodopsin to have a lasting therapeutic effect. At best, reducing levels of mutant rhodopsin will delay, but not arrest photoreceptor degeneration. For the time being, delaying degeneration and prolonging sight could improve many patients’ quality of life.
VITAMIN A AND PHARMACOLOGICAL SUPPLEMENTATION
Ligand binding has long been shown to increase the stability of the apo-protein opsin (Hubbard 1958). Therefore, it follows that supplementation with vitamin A, which
35 is converted into the 11-cis-retinal, may provide a therapeutic effect by stabilizing the mutant rhodopsin.
Berson et al. demonstrated the beneficial effects of vitamin A supplementation for patients with retinitis pigmentosa (Berson, Rosner et al. 1993; Berson, Rosner et al.
1993). Vitamin A supplementation slowed the rate of degeneration, measured by ERG function, in patients with common forms of RP. Since the results of this study were averages of groups irrespective of genotype, the effectiveness of vitamin A supplementation for rhodopsin-mediated ADRP could not be determined.
Follow up studies specifically addressed the effect of vitamin A supplementation on rhodopsin-mediated RP. Vitamin A supplementation was shown to protect mice expressing a class 2 rhodopsin mutation (Li, Sandberg et al. 1998), which are classically thought to cause degeneration through protein misfolding. Mice expressing a class 2 mutant rhodopsin that received vitamin A supplementation had preserved ERG amplitudes, longer inner and outer segments, as well as thicker outer nuclear layers compared to control. As a control for the experiment, mice expressing P347S mutant rhodopsin, a class 1 mutation, which is not thought to misfold, were also analyzed. Mice expressing the class 1 mutant rhodopsin did not receive the same beneficial effects following vitamin A supplementation, but instead showed a more rapid degeneration.
These results suggested that vitamin A supplementation may have beneficial effects for degenerations caused by class 2 rhodopsin mutants. However, supplementation given for a different class of mutants can be detrimental, highlighting the importance of understanding the mechanism of degeneration.
36
In vitro work has also suggested that pharmacological supplementation to aid protein folding can protect against degeneration for class 2 rhodopsin mutants.
Pharmacological chaperones applied to cells expressing P23H mutant rhodopsin, rescued some of the misfolding effects (Noorwez, Kuksa et al. 2003). The mutant rhodopsin in the presence of the chaperone had mature glycosylation patterns, was light sensitive, and localized to the plasma membrane (Noorwez, Malhotra et al. 2004). Retinal chromophore applied during opsin synthesis also increased P23H protein levels five- or six-fold, suggesting that in the absence of retinal, the opsin was targeted for protein degradation. In summary, in vitro and in vivo data supports the beneficial therapeutic effect of vitamin A or a similar chaperone to aid in proper folding of mutant rhodopsin.
TRANSPLANTATION OF STEM CELLS OR RETINAL SHEETS
Retinal transplantation has had a long history, and it comes as no surprise that transplantation of either retina tissue or retinal stem cells has been investigated as a potential therapy for retinitis pigmentosa.
Therapeutic studies originally focused on the use of retina tissue for transplantation. Later, it was found that integration with the host retina was vastly improved when retinal progenitor cells were transplanted as opposed to whole retina sheets. Retinal progenitor cells, also called retinal stem cells, are responsible for producing all the cells of the retina during development and are derived from fetal or newborn mice and rats, or human fetuses. Studies using these progenitor cells transplanted into either healthy or degenerating rat and mice retinas found that the cells migrate into retinal layers, develop morphological characteristics of retinal cell types, but
37 do not fully integrate into the outer nuclear layer and do not express photoreceptor cell specific genes (Chacko, Rogers et al. 2000; Coles, Angenieux et al. 2004; Klassen, Ng et al. 2004; Canola, Angenieux et al. 2007). However, MacLaren et al. recently demonstrated that transplantation of newly born rod cells, harvested at the peak of rod genesis, integrated morphologically and made functional connections (MacLaren,
Pearson et al. 2006). Transplanted cells restored response to dim light in mice that were previously blind. However, using newly born rod cells for therapy in humans has an added complication because in humans these cells would need to be harvested from fetuses in the 2nd trimester. There is hope that cultured cells with the same multipotency can be developed.
Ghosh et al. grafted E48 pig retinas into a 6 month old P347L transgenic pig when the outer nuclear layer had thinned to less than half its original thickness (Ghosh,
Engelsberg et al. 2007). While the graft and host retina did not make neuronal connections, and the function of the host retina was initially reduced following transplantation, the graft did rescue rods and cones in the host retina from degeneration.
The survival of host photoreceptor cells was not a local phenomenon, as host rod survival was more pronounced at a distance from the graft. These results suggest that grafting a developing retina can provide protection from degeneration potentially through the release of trophic factors or other molecules that protect from degeneration.
Together these results suggest transplantation of either retinal stem cells or retinal sheets may provide protection from degeneration. However, several hurdles still exist before this therapy can be adequately used in humans, including the source of the stem cells and retinal tissue.
38
CONCLUSIONS
Results of studies from multiple independent labs indicate that rhodopsin- mediated ADRP pathogenesis does not utilize a common mechanism of degeneration.
Molecular genetic studies and therapeutic testing reveal response differences for each of the rhodopsin mutants. There is strong evidence for multiple pathologic mechanisms, including protein misfolding, mislocalization, release of toxic products and aberrant signaling. It is also clear that the mechanisms are not mutually exclusive, with each mechanism contributing differentially to disease pathogenesis. Yet, it is likely that there are other mechanisms that have not yet been discovered and have a major contribution to the degenerative process. In order to be highly effective, future therapies will need to take into account the various mechanisms that apply to different mutations. It is our hope that, with new information, we can group similar mutation mechanisms together that will share effective therapies for the successful treatment of inherited retinitis pigmentosa.
39
APPENDIX: AN INTRODUCTION TO GLYCOGENES AND THEIR
INVOLVEMENT IN NEURONAL DEGENERATION
INTRODUCTION
Carbohydrates, often called glycans, are one of the four major biomolecules along with proteins, nucleic acids and lipids. Through the process of glycosylation a vast, diverse and highly regulated set of glycans are attached to proteins and lipids to form glycoproteins, proteoglycans and glycolipids. Due to their structural diversity, glycans are more complex than either nucleic acids or proteins. Instead of being linked by identical bonds to form linear molecules, the subunits of glycans are linked in a variety of fashions to form branched molecules. The spectrum of glycan structures has been predicted to be exponentially more immense than the number of proteins in a cell (Freeze
2006).
It has been predicted that more than half of all proteins are glycosylated
(Apweiler, Hermjakob et al. 1999). The addition of carbohydrates is a post-translational modification that can widely alter protein function (Lowe and Marth 2003). Glycans have been shown to have altered expression in response to physiological and pathological stimuli. The assortment of glycans is regulated by glycosyltransferases and glycosidases and other enzymes that synthesize and remodel glycan chains, as well as accessory enzymes involved in synthesis and transport of nucleotide sugars. The genes encoding these enzymes are collectively called glycogenes. Glycogenes have tissue specific expression (Comelli, Head et al. 2006), and have been shown to be regulated through
40 developmental and in pathological settings (Diskin, Kumar et al. 2006; Kroes, He et al.
2010; Saravanan, Cao et al. 2010).
Glycans are involved in many cellular processes. They mediate host-pathogen interactions and innate immunity. The addition of specific carbohydrate chains target proteins and lipids to different cellular compartments. One example of carbohydrate- mediated targeting is the addition of mannose-6-phosphate which targets proteins to the lysosome. Additionally, glycans serve as ligands for glycan-binding proteins that mediate cell trafficking, cell adhesion, and cell signaling. These glycan-binding proteins include C-type lectins, I-type lectins, and galectins. Together they decode the information stored in the glycan structure. Finally, the addition of glycans has also been shown to aid in protein folding and conformational stability. Several studies have suggested that rhodopsin requires glycosylation to fold and localize properly to rod outer segments (Fliesler and Basinger 1985; Kaushal, Ridge et al. 1994). Additionally, when rhodopsin is not glycosylated properly, retinal degeneration results (Zhu, Jang et al. 2004;
White, Fritz et al. 2007).
GLYCOSYLATION
Glycans are attached to proteins and lipids through the process of glycosylation.
Mammals utilize nine monosaccharides. Biosynthetic pathways produce all nine monosaccharides from sugars and precursors present in the diet. Glycosylation occurs mostly within the secretory pathway of the cell, and produces three main classes of glycans. N-linked glycans are attached to a nitrogen of asparagine amino acids of proteins. O-linked glycans are attached to the hydroxyl oxygen of serine or threonine.
41
Lipid glycosylation creates glycolipids. N-linked glycosylation begins within the endoplasmic reticulum with the transfer of a preformed precursor oligosaccharide composed of N-acetylglucosamine, mannose, and glucose. The original precursor oligosaccharide is further modified within the endoplasmic reticulum and the Golgi apparatus. This further modification creates the diversity of N-linked glycans. O-linked glycans are formed within the Golgi apparatus through pathways that are not yet understood.
While most glycosylation occurs within the endoplasmic reticulum and Golgi apparatus, GlcNAcylation, linking N-acetylglucosamine to serine or threonine, happens outside the secretory pathway. Like phosphates added through phosphorylation, O-
GlcNAc added in GlcNAcylation, has a shorter half-life than the protein to which it is attached. Its removal can allow for phosphorylation or an additional round of
GlcNAcylation. GlcNAcylation is highly regulated and required for viability of many mammalian cell types, perhaps because the process can act as a regulator for phosphorylation or protein turnover (Zhang, Su et al. 2003; O'Donnell, Zachara et al.
2004; Zachara and Hart 2004).
GLYCAN BINDING PARTNERS
The process of decoding the biological information stored in glycans is tasked to glycan-binding proteins called lectins. Lectins are ubiquitously expressed, found in animals, plants, and microorganisms. The interactions between glycans and lectins are weaker than protein-protein interactions. However, since most lectins bind in a multivalent fashion, glycan-lectin interactions do have significant avidity. Lectins are
42 highly selective for specific glycan structures. In addition to glycan structures, protein sequence and conformation can also contribute to lectin binding specificity (Collins, Blixt et al. 2004; Han, Collins et al. 2005; Ohtsubo, Takamatsu et al. 2005).
Lectins are classified into groups based on sequence motifs and biological properties. The lectin groups include C-type lectins, I-type lectins and galectins. C-type lectins get their name because they require calcium to form bonds. The C-type lectins have a conserved 120 amino acid carbohydrate recognition domain. Calcium acts as a bridge by directly interacting with sugar hydroxyl groups. In addition, two glutamate residues within the lectin bind to both the calcium ion and the sugar, and several other protein side chains form hydrogen bonds with additional glycan hydroxyl groups.
Selectins are members of the C-type lectin family. They bind immune system cells to site of injury in inflammatory response. I-type lectins get their name because they mediate glycan interactions via an immunoglobulin (Ig)-like domain. The major subfamily of the
I-type lectins are the siglecs or sialic-acid-binding lectins.
Galectins, once called S-type lectins, are a family of lectins with affinity for - galactosides due to a conserved 130 amino acid carbohydrate recognition domain. To date, fifteen mammalian galectins have been identified. The galectins are classified into three groups. Prototype galectins include galectin-1, -2, -5, -7, -10, -11, -13, -14, and -15.
The prototype galectins contain one carbohydrate recognition domain and can form dimers through noncovalent interactions. The tandem repeat galectins include galectin-4,
-6, -8, -9, and -12. Tandem repeat galectins contain two carbohydrate recognition domains and are constitutively bivalent. The chimera type galectins, which currently only include galectin-3, have a proline and glycine rich domain adjacent to the
43 carbohydrate recognition domain. Galectin-3 can form pentamers with multivalent ligands. The carbohydrate recognition domain of all galectins recognizes a minimum saccharide ligand of N-acetyllactosamine. Individual galectins can recognize modifications to this saccharide, allowing for tissue and developmental specificity.
Galectins do not have a protein secretion signal, but are found within the extracellular matrix. Galectins are also found on the cell surface as well as in the cytoplasm and nucleus. Generally, extracellular functions of galectins are dependent on their sugar-binding abilities, whereas their intracellular functions can be sugar independent. Galectins have been shown to play a role in the cell-cell adhesion, cell- matrix adhesion, cell growth and viability. Many galectins are expressed within tumors, leading to the hypothesis that they are involved in tumorigenesis and metastasis.
GALECTIN-1
Galectin-1, a prototypical galectin, was the first galectin identified. It can exist as a monomer (of approximately 14.4 kDa) or as a homodimer (of approximately 29 kDa).
In oxidizing conditions, three disulfide bonds form to create the oxidized form of galectin-1, which lacks the ability to bind sugars (Cho and Cummings 1996). It is believed that galectin-1 is secreted in a pathway similar to fibroblast growth factor-2 via inside-out transportation involving direct translocation across the plasma membrane.
Galectin-1’s carbohydrate recognition domain contains two antiparallel -sheets forming a sandwich-like structure.
Several galectin-1 binding partners have been identified. Galectin-1 binding partners can be both sugar dependent and sugar independent. Galectin-1 has several
44 sugar dependent binding partners within the extracellular matrix, including laminin, cellular fibronectin, thrombospondin, plasma fibronectin, vitronectin and osteopontin
(Camby, Le Mercier et al. 2006). Galectin-1 also has several sugar dependent binding partners on the cell surface, including integrins, GM1 ganglioside and CD2, 3, 7, 43, and
45. Recently, galectin-1 has been shown to bind gemin4 (Park, Voss et al. 2001), ras
(Paz, Haklai et al. 2001; Rotblat, Belanis et al. 2010) and Fos B (Nishioka, Sakumi et al.
2002; Scott and Zhang 2002) in a sugar independent fashion. Galectin-1 interaction with ras strengthens ras’s membrane association.
Galectin-1 has been shown to be involved in many signaling pathways including regulation of cell growth and cell migration by altering adhesion, motility and invasion.
Galectin-1 has also been shown to be involved in pathology of the nervous system.
Autoantibodies to galectin-1 are increased in patients with neurological disorders (Kato,
Ren et al. 2005). In ischemia models galectin-1 expression increased in neurons that were resistant to injury, implying that galectin-1 provides protection (Ishibashi, Kuroiwa et al. 2007; Qu, Wang et al. 2011). The oxidized form of galectin-1 promotes neurite outgrowth and increases axon regeneration in both the peripheral and central nervous systems even at low (picoM) concentrations (Horie, Inagaki et al. 1999). In a transgenic model of amyotrophic lateral sclerosis, treatment with galectin-1 provided protection
(Chang-Hong, Wada et al. 2005). Finally, galectin-1 has been shown to induce astrocyte differentiation which in turn induces astrocyte secretion of brain derived neurotropic factor (BDNF) which has been shown to increase neuronal survival (Qu, Wang et al.
2010).
45
CONCLUSION
Through the process of glycosylation a vast, diverse and highly regulated set of glycans are attached to protein and lipids. The expression of glycans is altered in response to physiological and pathological stimuli. Glycan assortment is regulated through a collection of enzymes encoded by glycogenes, which have tissue specific expression. Previously, glycogenes have been shown to have altered expression during pathology. Glycan-binding proteins decode the information stored in glycans by binding and altering signal transduction pathways. One glycan-binding protein, galectin-1, will be further discussed in chapter 3 due to its previously characterized involvement in neurodegeneration. We will also further examine glycogenes and the role they may play in retinitis pigmentosa in chapter 4.
46
CHAPTER 2: A NOVEL FORM OF TRANSDUCIN-DEPENDENT
RETINAL DEGENERATION: ACCELERATED RETINAL
DEGENERATION IN THE ABSENCE OF ROD TRANSDUCIN
Portions of this chapter originally appeared as an article in IOVS (Brill, Malanson et al.
2007).
47
ABSTRACT
Rhodopsin mutations account for approximately 25% of human autosomal dominantly inherited cases of retinitis pigmentosa. However, the molecular mechanisms by which rhodopsin mutations cause photoreceptor cell death are unclear. Mutations in genes involved in termination of rhodopsin signaling activity have previously been shown to cause degeneration. These genes cause degeneration by persistent activation of the phototransduction cascade. This chapter examines whether the disease-associated rhodopsin mutations P347S, K296E and the triple mutant V20G, P23H, P27L (VPP) caused degeneration by persistent transducin-mediated signaling activity. Transgenic mice expressing each of the rhodopsin mutants were crossed onto a transducin -subunit
-/- null (Tr ) background, and the rates of photoreceptor degeneration were compared to transgenics on a wild-type background.
Mice expressing VPP mutant rhodopsin showed the same severity of degeneration in the presence or absence of -transducin. Unexpectedly, mice expressing P347S or
K296E mutant rhodopsins exhibited faster degeneration on a -transducin null (Tr ) background. To test if absence of -transducin contributed to degeneration by favoring
formation of stable rhodopsin-arrestin complexes, mutant P347S, Tr mice lacking arrestin (Arr) were analyzed. Rhodopsin-arrestin complexes were found not to contribute to degeneration. We hypothesize that the decay of metarhodopsin to apo-opsin and free all-trans-retinaldehyde is faster with P347S mutant versus wild-type rhodopsin.
Consistent with this, the lipofuscin fluorophores A2PE, A2E and A2PE-H2 which form from retinaldehyde, were elevated in P347S transgenic mice.
48
INTRODUCTION
Autosomal dominant retinitis pigmentosa (ADRP) is a genetically heterogeneous group of inherited retinal degenerations causing blindness in man. Mutations in several genes encoding proteins in the phototransduction cascade have been causatively associated with ADRP (Daiger 1996-2007). Rhodopsin mutations collectively account for the most common known cause of ADRP, with more than 100 different mutations identified (Daiger 1996-2007). Thus, elucidating the molecular mechanisms underlying cell death in this class of mutations is of importance.
Previous research from our lab has described transgenic mouse mutants that cause degeneration by prolonged activation of the phototransduction cascade (Lem and Fain
2004). Null mutations in the rhodopsin kinase (Chen, Burns et al. 1999) and arrestin (Xu,
Dodd et al. 1997) genes, both of which play a role in terminating rhodopsin activity, caused a light-dependent retinal degeneration. Protection from retinal degeneration was
observed when either mutation was crossed onto a Tr background (Hao, Wenzel et al.
2002). Retinal degeneration in Rpe65 null (Rpe65) mutant mice (Redmond, Yu et al.
1998) was blocked when placed on a Tr background (Woodruff, Wang et al. 2003).
Degeneration in Rpe65 mice results from persistent signaling by apo-opsin caused by impaired synthesis of 11-cis-retinaldehyde (11-cis-RAL) chromophore (Lem and Fain
2004), which functions as an inverse agonist.
Constitutively active rhodopsin mutants that activate transducin in a light- independent manner have previously been described under in vitro conditions
(Zhukovsky, Robinson et al. 1991; Cohen, Oprian et al. 1992; Robinson, Cohen et al.
1992; Rao, Cohen et al. 1994; Robinson, Buczylko et al. 1994; Rim and Oprian 1995;
49
Sieving, Richards et al. 1995; Rao and Oprian 1996; Zvyaga, Fahmy et al. 1996; Han,
Lou et al. 1997; Kim, Altenbach et al. 1997; Han, Smith et al. 1998; Sieving, Fowler et al. 2001; Ramon, del Valle et al. 2003). Constitutive signaling activity in Drosophila is also associated with retinal degeneration (Iakhine, Chorna-Ornan et al. 2004; Wang, Xu et al. 2005). Three activated rhodopsin mutants associated with congenital night blindness have been reported in humans (Rao, Cohen et al. 1994; Rao and Oprian 1996;
Gross, Rao et al. 2003; Jin, Cornwall et al. 2003). In this study, we investigated whether persistent photosignaling activity by rhodopsin mutants was also a cause of retinal degeneration. The severity of retinal degeneration was compared in transgenic mouse lines carrying one of three mutant rhodopsin transgenes placed either on a wild type
(Tr ) or a Tr genetic background. If abnormal rhodopsin signaling caused retinal
degeneration, we predicted that rhodopsin mutants on the Tr background would be
protected from degeneration. Importantly, retinas of Tr mice did not degenerate except at advanced ages (6+ months), where less than 10% of photoreceptors were lost
(Calvert, Krasnoperova et al. 2000).
We studied transgenic mice expressing three forms of mutant rhodopsin that cause autosomal dominant retinitis pigmentosa in humans. VPP transgenic mice express the disease associated P23H plus two substitutions, V20G and P27L. VPP-rhodopsin mRNA is expressed at levels equivalent to wild-type (Naash, Hollyfield et al. 1993), although relative levels of mutant to wild-type protein are lower (Wu, Ting et al. 1998).
Histological analysis of VPP transgenic mice showed abnormal disc morphogenesis at the base of rod outer segments (Liu, Wu et al. 1997). Immunohistochemical methods
50 localized the majority of VPP mutant rhodopsin to rod outer segment disks (Wu, Ting et al. 1998).
The second transgenic mouse line we studied expressed K296E mutant opsin
(K296E) (Li, Franson et al. 1995). The K296E rhodopsin mutation is also associated with retinitis pigmentosa in humans (Owens, Fitzke et al. 1994). The K296 residue is the
Schiff base attachment site for the 11-cis-retinal chromophore. Substitutions at this residue prevent association of apo-opsin with 11-cis-retinal to form rhodopsin. In vitro,
K296E mutant opsin constitutively activated -transducin independent of light. It remains a point of controversy whether K296E mutant opsin is phosphorylated by rhodopsin kinase and binds arrestin (Li, Franson et al. 1995; Rim and Oprian 1995) or not
(Robinson, Buczylko et al. 1994). In contrast to results from in vitro studies, the mutant opsin in K296E transgenic mice localized to rod outer segments, was constitutively phosphorylated and bound to arrestin (Li, Franson et al. 1995).
The third transgenic mouse line that we studied expressed disease-associated
P347S mutant rhodopsin (Li, Snyder et al. 1996). In vitro studies showed this mutant rhodopsin regenerated normally with 11-cis-retinal and, upon light exposure, exhibited a spectral absorbance shift comparable to wild-type rhodopsin (Sung, Davenport et al.
1993). P347S mutant rhodopsin also activated -transducin, was phosphorylated by rhodopsin kinase and subsequently bound arrestin (Weiss, Hao et al. 1995). Although
P347S mutant rhodopsin was present in outer segments of these transgenic mice, mutant rhodopsin was also present in inner segments and accumulated in submicrometer-sized, extracellular vesicles near the junction between outer and inner segments (Li, Snyder et al. 1996).
51
In the current studies, we examined whether VPP, K296E or P347S mutant rhodopsins triggered -transducin-mediated cell death by producing transgenic mice
expressing these mutant rhodopsins on a Tr genetic background. We show that the absence of -transducin had no effect on the rate of photoreceptor degeneration in VPP transgenic mice. Unexpectedly, we observed accelerated retinal degeneration in mice
expressing K296E and P347S mutant rhodopsin mutants on a Tr background. This contrasts with the protection conferred by the absence of -transducin in rhodopsin kinase (Hao, Wenzel et al. 2002), arrestin (Hao, Wenzel et al. 2002) and Rpe65
(Woodruff, Wang et al. 2003) null mutant mice. Our results show that persistent photosignaling is not a mechanism of retinal degeneration for these three rhodopsin mutations. Further, the accelerated retinal degeneration in K296E and P347S transgenic mice suggests a mechanism whereby -transducin confers a protective effect. Possible explanations for the accelerated degeneration are preferential formation of rhodopsin- arrestin complexes (Alloway and Dolph 1999; Alloway, Howard et al. 2000; Kiselev,
Socolich et al. 2000; Lee, Xu et al. 2003; Orem, Xia et al. 2006; Wang and Montell 2007) or the destabilization of light activated mutant rhodopsin in the absence of -transducin.
METHODS
Animals. All procedures were carried out in accordance with the Association for
Research in Vision and Ophthalmology statement concerning the use of animals in ophthalmic and vision research and within the guidelines of the Tufts Medical Center
Institutional Animal Care and Use Committee. Rhodopsin mutant mice on a Tr and
Tr genetic background were maintained as independent lines under normal cyclic light
52
(5 to 100 lux, in-cage readings). The mutant rhodopsin transgenes were maintained and studied in the heterozygous state. K296E, line A (Li, Franson et al. 1995) and P347S, line
C1 (Li, Snyder et al. 1996) rhodopsin transgenic mice were used in these studies.
Genotyping. The α-transducin genotype was determined by Southern blot or PCR analysis. For Southern blots, purified genomic DNA was digested with XbaI and probed with a radiolabeled 1.6 kb XhoI/EcoRI fragment encompassing exons 3 to 6 of Tr. For
PCR analysis, DNA primers used to amplify the Tr gene were Tr1: 5’-TAT CCA CCA
GGA CGG GTA TTC-3’ (forward primer); Tr2: 5’-GGG AAC TTC CTG ACT AGG
GGA GG-3’ (reverse primer); and Tr3: 5’-GCG GAG TCA TTG AGC TGG TAT-3’
(reverse primer). The amplification yielded a 387 and 273 bp gene product for the wild- type and null mutant alleles, respectively. Primers used to amplify the Arr gene were:
Neo8AR 5’CCATCTTGTTCAATGGCCGATCCC3’; Tail2
5’GACAATGGGACTGAGATGGTGGG3’; E2B
5’GGACAGACAGCATGGCAGCCTG3’. The amplification yielded a 280 bp wild type gene product and a 349 bp null mutant gene product.
Rhodopsin transgene positive animals were identified by PCR analysis. Both the
P347S and K296E rhodopsin mutants expressed subcloned human rhodopsin genes and were identified using the same PCR amplification strategy. The primer pairs specific to human rhodopsin were R6: 5’-CGT TCC AAG TCT CCT GGT GT-3’ and R7: 5’-GAC
CTA GGC TCT TGT TGC TG-3’, producing a ~200 bp PCR product. Annealing for one minute was ramped from 65 to 58C for 7 cycles, dropping at a rate of one degree per cycle, then denatured at 94 C for 35 seconds, annealed at 56 C for one minute and
53 extended at 72C for 40 seconds for an additional 23 cycles. VPP mutant rhodopsin was amplified using primer pairs Rh3: 5’-AA CCA TGG CAG TTC TCC ATG CT-3’ and
OP2B: 5’-GTC CTT GGC CTC TCT GAA C-3’ specific to mouse rhodopsin. After denaturation, the annealing temperature was ramped from 70 C to 60 C for one minute, minus one degree per cycle for ten cycles. This was followed by 22 cycles of denaturation at 94 C for 30 seconds, annealing at 59C for one minute and extension at
72C for one minute. To distinguish the mutant from the endogenous wild type rhodopsin gene, the PCR product was digested with NcoI. Wild type rhodopsin was cleaved to produce a 200 and 300 bp fragment while the VPP mutant opsin, which lacked the NcoI restriction site, yielded a 500 bp fragment.
All mice studied were homozygous for the Leu450 allele of the Rpe65 gene.
Histology. Mice were anesthetized with 0.017 cc avertin/g body weight prior to cardiac perfusion of 100 cc fixative (1% paraformaldehyde, 2% glutaraldehyde in 0.1 M phosphate buffer) at a rate of 200 cc fixative/hour. Eyes were oriented at the superior most point of the eye with a cauterizing pen at the ora serrata. Eyes were excised and rotated in fixative for two hours at room temperature, then the anterior segment was removed and eyes were fixed overnight at 4C. Eyecups were cut into four quadrants marked as the superior nasal, superior temporal, inferior nasal and inferior temporal.
Tissue quadrants were rinsed several times in 0.1 M phosphate buffer then fixed in 1% osmium tetroxide in 0.1 M phosphate buffer for one hour. Following fixation, tissue quadrants were rinsed several times in PBS, then passed through a dehydrating series of ethanol, with alcohol concentrations increasing from 30 to 90% by increments of 10% for
54 five-minute intervals. The tissue quadrants were immersed in 100% EtOH for 10 minutes three times, then propylene oxide for ten minutes three times. Retina quadrants were soaked for 30 minutes with 33% araldite in propylene oxide, then for 90 minutes in 66% araldite in propylene oxide. Finally, eyes were infiltrated in a solution of 100% araldite containing dodecenyl succinic anhydride (DDSA) and 2,4,6-Tris [dimethyl-aminomethyl] phenol (DMP-30) and the resin was hardened for 48 hours in a 60C oven. Half-micron sections were cut on a MT6000 DuPont Sorvall Ultramicrotome. Sections were stained with 1% toluidine blue, 1% sodium borate in 0.1 M phosphate buffer.
Quantification of retinal degeneration. Retinal sections cut along the vertical meridian of the eye at the optic nerve head were analyzed. For each animal, several sections were examined. Rows of nuclei in the outer nuclear layer were counted at the central and superior and inferior mid-peripheral retina. Values were averaged for each animal and then the average value from several animals at the same age and of the same genotype were averaged to obtain a mean and standard error of the mean. Statistical analysis was performed using one-way ANOVA.
Analysis of A2E and A2E-precursors. All manipulations were done on ice under dim red light (Kodak Wratten 1A). One mouse eyecup containing retina plus RPE was homogenized in 1.0 ml PBS, pH 7.2. Samples were homogenized further by adding 4.0 ml of chloroform/methanol (2:1, vol/vol), extracted by addition of 4.0 ml chloroform and
3.0 ml of dH2O, and centrifuged at 1,000 x g for 10 min. The chloroform extracts were dried under a stream of argon and the residues were dissolved in 100 µL of 2-propanol
55 for analysis by HPLC. Phospholipid extracts were analyzed by normal phase HPLC on a silica column (Zorbax Rx-Sil 5 µm, 250 x 4.6 mm, Agilent, Palo Alto, CA) using an
Agilent model 1100 liquid chromatograph equipped with photodiode-array detector
(Agilent Technologies, Wilmington, DE). The mobile phase (hexane: 2-propanol: ethanol: 25 mM potassium phosphate: glacial acetic acid, 485:376:100:45:0.275, v/v) was filtered, then pumped through the system at 0.5 – 1.4 ml/min. Column and solvent temperatures were maintained at 35°C.
Lipofuscin granule quantitation. Mice were fixed by vascular perfusion with 2% formaldehyde and 2.5% glutaraldehyde in 100 mM sodium phosphate buffer, pH 7.2.
Secondary fixation was in 1.0% osmium tetroxide. The eyes were dissected into quadrants, dehydrated in ethanol, and embedded in Araldite 502. Ultrathin sections for electron microscope viewing were cut on a Leica Ultracut UCT and picked up on 200 mesh uncoated copper grids. The sections were stained with uranium and lead salts and viewed with a Zeiss 910 electron microscope.
The pigment epithelial fields were imaged using a KeenviewTM digital camera.
Eleven fields were collected from 1 control mouse 2 months old and 10 fields were collected from 1 control mouse 3 months old. Thirty-four fields were collected from 3 experimental mice 2-months old and twenty-three fields were collected from 2 experimental mice 3-months old. Measurements were made at a constant magnification of 16,000X. Using analySISTM software, the pigment epithelial cytoplasm area was outlined and each lipofuscin body was outlined. For each field, total lipofuscin area was compared to total pigment epithelial area. Each field was considered as n=1. Results
56 were presented as means and standard deviation. Statistical analysis was performed using the Student t-test.
RESULTS
The absence of -transducin in mice with normal rhodopsin does not cause retinal degeneration.
To determine whether rhodopsin mutants caused degeneration by aberrant photosignaling, we compared the degeneration rates of rhodopsin mutant mice reared in
cyclic light on a Tr or Tr background. For each of the three rhodopsin mutant lines studied, degeneration was examined at several time points using an endpoint of >50% photoreceptor cell loss. Degeneration severity was assessed by counting rows of
photoreceptor cell nuclei in the outer nuclear layer (ONL). Contribution of the Tr
phenotype to degeneration was minimal. ONL thicknesses of Tr mice at one, two, three, four and six months of age were comparable to age-matched wild type retinas (Fig
8A), demonstrating that the Tr phenotype did not contribute to degeneration. This is
consistent with our initial published characterization of Tr mice (Calvert,
Krasnoperova et al. 2000).
VPP mutant rhodopsin mouse retinas degenerate in a signal-independent fashion.
+ Retinal morphologies of VPP mutant mice on Tr or Tr backgrounds were compared at one, three and six months of age (Fig. 8B). Degeneration severity increased with age. However there was no difference in the rate of degeneration on the two
57 different genetic backgrounds. Since the time course of retinal degeneration was similar
on both the Tr and Tr backgrounds, these results indicate that VPP mutant rhodopsin does not cause photoreceptor cell death by persistent activation of the visual transduction cascade.
K296E mutant rhodopsin mouse retinas show accelerated retinal degeneration in the absence of -transducin.
To determine whether the K296E rhodopsin mutation caused retinal degeneration by inappropriate consistent photosignaling, we examined the retinal morphologies of
mice expressing K296E mutant opsin at three and six months of age on a Tr or Tr background (Fig 8C). At both three and six months of age, more rapid degeneration was
observed in K296E rhodopsin mutants on the Tr genetic background versus the Tr genetic background. These results demonstrate that transducin-mediated signaling does not cause degeneration for this rhodopsin mutation.
Absence of -transducin accelerates degeneration in P347S mutant rhodopsin mice.
Retinal morphologies of P347S mutant mice (Li, Snyder et al. 1996) on a Tr or
Tr background were compared at one, two, four and six months of age (Fig. 8D). As with the K296E mutant rhodopsin, the severity of degeneration was significantly
+/+ accelerated with the P347S mutant on the Tr genetic background versus Tr genetic background. These results demonstrated that for the P347S mutant rhodopsin, aberrant transducin-mediated signaling was not a cause of degeneration.
58
Our observations refute transducin-mediated signaling as a mechanism of degeneration in the three rhodopsin mutations studied. However, the accelerated degeneration observed for the K296E and P347S mutant rhodopsins was not expected.
There are several possible explanations for the observed results. We tested these possibilities in the P347S mutant rhodopsin mouse line since it degenerates faster than the K296E mutant mouse.
One possibility is that elevated levels of total rhodopsin may cause retinal degenerations, as previously observed with transgenic over-expression of wild-type rhodopsin (Olsson, Gordon et al. 1992). The combined levels of endogenous and transgene-encoded rhodopsin may contribute to degeneration. This seems unlikely, however, as this line of P347S transgenic mice has a low level of expression of mutant rhodopsin, a level at which additional wild-type rhodopsin does not cause degeneration
(Li, Snyder et al. 1996). Additionally, we previously showed that the absence of Tr does not alter rhodopsin levels (Calvert, Krasnoperova et al. 2000). We also expect that the gene expression level of the P347S mutant rhodopsin transgene would be the same
+/+ whether expressed on a Tr or Tr genetic backgrounds, since the transgene integration site is the same. Direct measurement of rhodopsin levels by difference
spectroscopy showed one-month-old P347S mutant mice on a Tr background retained
190 20 pmol rhodopsin per retina (n = 3), compared to 320 30 pmol rhodopsin per
retina in P347S mutants on a Tr genetic background (n = 3; p < 0.002). The decrease
in rhodopsin concentration in P347S, Tr mice is likely attributable to the more rapid degeneration.
59
Rhodopsin-arrestin complex formation does not contribute to retinal degeneration in
P347S mutant mice.
The formation of stable rhodopsin-arrestin complexes is reported to cause retinal degeneration in Drosophila (Alloway, Howard et al. 2000; Kiselev, Socolich et al. 2000) and mice (Chen, Shi et al. 2006). Since arrestin and -transducin bind competitively to phosphorylated rhodopsin (Krupnick, Gurevich et al. 1997), it is possible that P347S mutant rhodopsin preferentially binds arrestin in the absence of -transducin.
Furthermore, P347S mutant rhodopsin peptide shows increased phosphorylation kinetics
(Ohguro 1997) and can bind arrestin (Weiss, Hao et al. 1995). This may favor formation of stable rhodopsin-arrestin complexes. To test whether rhodopsin-arrestin complex formation contributes to degeneration, we produced P347S rhodopsin mutant mice on a
double null -transducin/arrestin (P347S, Tr , Arr genetic background.
We compared the severity of degeneration in four-month-old P347S, Tr , Arr
and P347S, Tr , Arr mice reared in cyclic light (Fig. 9). We showed previously that
Arr mice placed on a Tr genetic background were protected from light-induced degeneration (Hao, Wenzel et al. 2002). Evaluation of an independent set of animals again revealed accelerated degeneration in P347S rhodopsin mutant mice in the absence
of -transducin. P347S rhodopsin mutants on Tr , Arr background retained 2.7
0.7 (n=6) rows of nuclei compared to P347S mutants on Tr , Arr background, which retained 1.9 0.8 (n=9) rows of nuclei. We failed to observe protection from degeneration in the absence of arrestin protein, indicating that rhodopsin-arrestin
60 complex formation did not contribute to retinal degeneration in the P347S mutant rhodopsin mouse model.
Dark rearing slows degeneration in P347S, Trα mice.
Another possible explanation for the accelerated degeneration in mice on the Tr genetic background is -transducin is needed to stabilize light-activated P347S mutant metarhodopsin by binding of -transducin. Since transducin binding is initiated by light
activation of rhodopsin, we assessed whether dark-rearing P347S, Tr mice provided protection from degeneration.
We compared the severity of degeneration in four-month-old P347S, Tr and
P347S, Tr mice reared in cyclic light or complete darkness (Fig. 10). Compared to
light-reared P347S, Tr mice (3.0 ± 0.9, n = 12, p<0.01), dark-reared P347S, Tr mice were protected from retinal degeneration (4.8 ± 0.8, n = 4), supporting our hypothesis that light activation of the mutant P347S mutant rhodopsin destabilized the
mutant rhodopsin. In contrast, dark-reared P347S, Tr control mice conferred no additional protection (5.7 ± 1.0 rows of nuclei; n = 6) compared to cyclic light-reared
P347S, Tr mice (6.4 ± 0.6 rows of nuclei, n = 8). This observation is consistent with our hypothesis that -transducin binding stabilizes the mutant metarhodopsin, providing a protective effect.
Lipofuscin fluorophores are elevated in P347S mutant rhodopsin mice.
61
The accelerated degeneration observed in P347S transgenic mice may result from destabilization of light-activated P347S metarhodopsin. Wild-type metarhodopsin decays slowly to yield apo-opsin and free all-trans-retinal. Free all-trans-retinal is highly cytotoxic. It is eliminated by reduction to all-trans-retinol. The reduction of all-trans- retinal in rods is a rate-limiting step in the visual cycle (Saari, Garwin et al. 1998).
Possibly, P347S mutant metarhodopsin decays faster than all-trans-retinal clearance from the cell.
In photoreceptors, free all-trans-retinal condenses spontaneously with phosphatidylethanolamine in outer segment disks to form N-retinylidene- phosphatidylethanolamine (N-ret-PE). N-ret-PE can react with another all-trans-retinal to form a family of toxic bis-retinoid fluorophores that include A2PE-H2, A2PE, A2E and iso-A2E (Parish, Hashimoto et al. 1998; Mata, Weng et al. 2000; Bui, Han et al. 2006).
Both A2PE and A2E are photosensitizers that are subject to photooxidation, producing reactive moieties that can modify DNA and proteins, resulting in cell death. These fluorophores are formed in photoreceptor outer segments and accumulate in the retinal pigment epithelium (RPE), where they might modify RPE proteins and DNA.
To test whether bis-retinoid fluorophores accumulate in the retinal pigment epithelium of P347S mutant transgenic mice, we measured levels of A2E and its
precursors in dark-adapted two-month-old wild-type and P347S transgenic mice on Tr
and Tr backgrounds (Fig. 11A). No significant difference was seen for N-ret-PE in
P347S, Tr versus wild-type eyecups. This was expected since all-trans-retinal is not produced in dark-adapted mice and N-ret-PE forms in rapid equilibrium with all-trans-
retinal. However, both A2PE and A2E were elevated two- to three-fold in P347S, Tr
62 mutant eyecups compared to wild-type controls. Similarly, iso-A2E was elevated 5-fold and A2PE-H2 was elevated 27-fold relative to wild-type littermate control eyecups.
Eight-week old P347S, Tr transgenic and non-transgenic littermate Tr mice were
+/+ also assessed for accumulation of phospholipids. Similar to P347S, Tr mice, P347S,
Tr mice showed significantly elevated levels of A2PE, iso-A2E and A2PE-H2
compared to Tr controls (data not shown). These results reveal an accumulation of bis-retinoid fluorophores in eyecups of P347S mutant rhodopsin transgenic mice, supporting our hypothesis that P347S mutant rhodopsin is destabilized.
P347S rhodopsin mutant mice have higher lipofuscin granule density.
Excess A2E fluorophore accumulation in the RPE is associated with a corresponding increase in lipofuscin granule density. To determine whether the increase in A2PE, A2E, iso-A2E and A2PE-H2 correlated with an increase in lipofuscin granules, lipofuscin granule density was compared in two- and three-month-old P347S transgenic and littermate control mice by measuring square microns of lipofuscin granule per square micron of RPE cytoplasm (Fig. 11B). Ten to twelve fields were counted per mouse.
Two-month old P347S transgenic mice (34 fields, n = 3) showed greater than 2-fold increased lipofuscin granule density over littermate wild-type controls (p < 8 x 10-6).
Three-month old animals also showed increased lipofuscin granule density (p < 0.04).
Both data sets show a corresponding increase in lipofuscin granule density associated with elevated levels of A2E and its precursors.
63
Chromophore Toxicity alone does not account for the retinal degeneration in the P347S mutant rhodopsin mice.
Increased levels of bis-retinoid fluorophores and lipofuscin granule density suggest that the decay of metarhodopsin to apo-opsin and free all-trans-retinal may be faster in the P347S model. Free all-trans-retinal is toxic to cells and an increased rate of release could be toxic to the photoreceptor cells. To directly test the role of chromophore toxicity, P347S expressing mice were crossed onto an Rpe65 null genetic background.
The Rpe65 enzyme is critical for the recycling of chromophore and, in its absence, cells cannot produce 11-cis-retinal (Redmond, Yu et al. 1998). Without 11-cis-retinal, all- trans-retinal cannot be produced. Therefore, if the degeneration present in the P347S mutant is caused by increased release of all-trans-retinal, then placing the mice on the
Rpe65 null background would provide protection.
Previously, our lab had shown that Rpe65 mice placed on a Tr genetic background were protected from light-induced degeneration (Woodruff, Wang et al.
2003). We compared the degeneration in four-month-old P347S, Tr , Rpe65 and
P347S, Tr , Rpe65 mice reared in cyclic light (Fig. 12). P347S rhodopsin mutants
on Tr , Rpe65 background retained 2.7 0.7 (n=6) rows of nuclei, compared to
P347S mutants on Tr , Rpe65 background, which retained 1.7 0.6 (n=6) rows of nuclei. Therefore, we failed to observe protection from degeneration in the absence of the Rpe65 protein, indicating that chromophore toxicity alone does not account for the degeneration in the P347S rhodopsin mutant mouse model.
64
DISCUSSION
We have evaluated the contribution of -transducin signaling to retinal degenerative disease in three mouse models of rhodopsin-mediated ADRP. Previous studies showed that null mutations in the rhodopsin kinase and arrestin genes, required for termination of the photoresponse, caused degeneration because of persistent photosignaling (Hao, Wenzel et al. 2002; Woodruff, Wang et al. 2003). We tested whether persistent signaling by mutant rhodopsins also caused degeneration. Of three rhodopsin mutants examined, none were protected from retinal degeneration when placed on a -transducin null background. These results show that persistent photosignaling is not a mechanism of retinal degeneration in transgenic mice expressing VPP, K296E or
P347S mutant rhodopsins. VPP rhodopsin mutant mice, a model for the human P23H
+/+ mutation, showed no change in the rate of degeneration on Tr versus Tr backgrounds. The absence of -transducin, however, accelerated degeneration in K296E and P347S transgenic mice.
The similar rates of degeneration in VPP transgenic mice in the absence or presence of α-transducin indicates that photoreceptor cell apoptosis is unrelated to activation of the visual transduction cascade. Our results contrast with those of
Samardzija et al. (Samardzija, Wenzel et al. 2006) who reported that VPP, Tr double mutant mice showed protection from photoreceptor degeneration. The reasons for the discrepancy are not clear, but likely stem from the effect of genetic modifiers. The
Rpe65 genotype is a known genetic modifier of sensitivity to light damage. Mapping studies indicate the existence of additional modifiers of light damage sensitivity
(Danciger, Lyon et al. 2004). The Rpe65 genotypes were assessed in both studies and did
65 not account for the differing result. Published data on the VPP mutant mouse is also consistent with the presence of genetic modifiers. The VPP mouse line shows both increased susceptibility to light damage (Wang, Lam et al. 1997), and partial, but incomplete protection from retinal degeneration when dark-reared (Naash, Peachey et al.
1996). This suggests the role of both light-dependent and signal-independent degenerative mechanisms. The light-independent component may relate to the faster degeneration observed in VPP mutant albino mice that is unrelated to increased retinal illumination since dark-reared albinos still showed faster degeneration (Naash, Ripps et al. 1996). Furthermore, the VPP mouse line used by Samardzija et al. degenerated nearly twice as fast as the VPP subline used in our studies and the initially characterized mouse line (Naash, Hollyfield et al. 1993). Identifying genetic modifiers that regulate disease susceptibility will be important for understanding disease mechanisms. Other reports describe concurrently operating degenerative mechanisms in other retinal degeneration models (Hao, Wenzel et al. 2002; Iakhine, Chorna-Ornan et al. 2004; Wang, Xu et al.
2005).
K296E transgenic mice on a Tr genetic background did not show protection from degeneration. Our results agree with the conclusions of Li et al. (Li, Franson et al.
1995), that constitutive activation of the visual transduction cascade does not cause retinal degeneration in this animal model. Additionally, we observed accelerated
degeneration in 6-month old K296E, Tr mice which was less pronounced at 3 months of age. Chen et al. (Chen, Shi et al. 2006) do not report accelerated degeneration in
K296E, Tr mice. They examined animals only up to 2.5 months of age, which may explain the apparent discrepancy. Chen et al. also report that K296E transgenic mice
66 undergo degeneration by formation of stable rhodopsin-arrestin complexes (Chen, Shi et al. 2006). It is possible that the accelerated degeneration of K296E transgenic mice that we observe in the absence of -transducin is caused by the favored formation of rhodopsin-arrestin complexes, since -transducin and arrestin competitively bind phosphorylated rhodopsin (Krupnick, Gurevich et al. 1997).
The P347S transgenic mouse line also showed accelerated degeneration in the absence of -transducin. Photoreceptor degeneration was significantly slower in dark-
-/- versus cyclic light-reared P347S, Tr mice, indicating that light activation of the mutant
rhodopsin is an initiating event. However dark-reared P347S, Tr mice were not protected from degeneration compared to cyclic light-reared mice of the same genotype, suggesting that the presence of -transducin provided protection. Rhodopsin-arrestin complex formation was not a major contributor to degeneration, since placing the P347S mutation on a double -transducin and arrestin null mutant background did not provide protection from retinal degeneration.
The P347 residue is part of the highly conserved VAPA C-terminal end of rhodopsin known to play an important role in trafficking rhodopsin to rod outer segments
(for review, see (Deretic 2006)). Mutations in this region result in mis-localization of rhodopsin to inner segments. The mislocalized opsin has been suggested to cause degeneration by aberrant activation of signaling pathways in the inner segment (Alfinito and Townes-Anderson 2002). Our results do not support transducin activation by mislocalized opsin, although it does not exclude signaling mediated by an -transducin independent pathway. Tam et al. also report that mislocalized C-terminal rhodopsin mutants do not signal via transducin activation (Tam, Xie et al. 2006).
67
Our results are consistent with stabilization of light-activated P347S metarhodopsin by -transducin binding. In its absence, the mutant P347S metarhodopsin may decay more rapidly to its component parts, apo-opsin and all-trans-retinal. The all-trans-retinal condenses with phosphatidylethanolamine to form N-ret-PE, which can react with a second molecule of all-trans-retinal to form toxic bis-retinoid fluorophores (Parish,
Hashimoto et al. 1998; Mata, Weng et al. 2000; Bui, Han et al. 2006). The elevated levels of the fluorophores A2E, iso-A2E and A2E-precursors in P347S transgenic mouse eyecups and the correlative increase in lipofuscin granules support our hypothesis. Both
A2PE and A2E are subject to photo-oxidation and can cause cell death (Fishkin, Sparrow et al. 2005; Jang, Matsuda et al. 2005; Kim, Nakanishi et al. 2006). Bis-retinoids are associated with normal aging of the human eye, but accumulate at higher levels in association with some forms of human retinal degenerations including Stargardt macular dystrophy, retinitis pigmentosa and cone-rod dystrophy (Sparrow and Boulton 2005).
Elevation of A2E and other lipofuscin fluorophores in RPE cells have also been reported in animal models of retinal and macular degeneration (Katz, Eldred et al. 1987; Weng,
Mata et al. 1999; Liu, Itagaki et al. 2000; Mata, Weng et al. 2000; Rakoczy, Zhang et al.
2002; Ambati, Anand et al. 2003; Karan, Lillo et al. 2005; Maeda, Maeda et al. 2005;
Zhang, Brankov et al. 2005). Vitamin A supplementation can ameliorate disease severity for some retinal degenerations (Berson, Rosner et al. 1993; Li, Sandberg et al. 1998), but it may be detrimental to patients with destabilized rhodopsin mutations that decay rapidly to release all-trans-retinal.
To further test our hypothesis that the mutant P347S metarhodopsin decays more rapidly to its component parts, apo-opsin and free all-trans-retinal, we utilized the Rpe65
68 null background which eliminates the creation of all-trans-retinal. However, the results obtained using the Rpe65 null mice are inconsistent with the results measuring bis- retinoid fluorophores and lipofuscin granule density. We believe that the Rpe65 null mutation did not rescue the degeneration, because the P347S mutant rhodopsin may require 11-cis-retinal to properly fold and traffic to the outer segments. We did attempt to determine the localization of the P347S mutant rhodopsin on the Rpe65 null background, but were unable to find an antibody that recognized only the mutant rhodopsin and not also the wild-type rhodopsin. Further experiments, including the generation of an antibody specific for P347S mutant rhodopsin, are required to determine if 11-cis-retinal is required for P347S mutant rhodopsin to fold and traffic properly to rod outer segments. Additionally, one could test the ability of P347S mutant rhodopsin to localize in vitro in the presence and absence of chromophore, or generate a tagged P347S mutant rhodopsin and examine its localization on the Rpe65 null background.
Only one disease-associated mutation in the transducin -subunit has been described, which is associated with Nougaret’s night blindness (Dryja, Hahn et al. 1996;
Muradov and Artemyev 2000). Therefore, it is unlikely that loss of transducin function comprises a prevalent mechanism of retinal degeneration. However, it is very likely that a significant subset of rhodopsin mutations impairs transducin binding and this may represent a highly disease-relevant mechanism of degeneration that is worthy of further exploration.
69
CHAPTER 3: INVOLVEMENT OF GALECTIN-1 IN RHODOPSIN-
MEDIATED RETINITIS PIGMENTOSA
70
ABSTRACT
Previous studies reveal a role for galectin-1 in the degeneration of neuronal processes (Chang-Hong, Wada et al. 2005; Sakaguchi, Shingo et al. 2006; Ishibashi,
Kuroiwa et al. 2007; Jin, Na et al. 2007; Plachta, Annaheim et al. 2007; Kajitani, Nomaru et al. 2009). This study examines the involvement galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse model of retinitis pigmentosa (RP). To characterize galectin-1 expression in the wild-type versus P347S mutant rhodopsin retina, immunohistochemistry, quantitative PCR, and immunoblots were performed. Galectin-1 binding partners were identified in the retina through the use of a sugar-binding column followed by mass spectrometry. The role of galectin-1 in retinal degeneration was determined by crossing the P347S mutant rhodopsin onto a galectin-1 knockout background, and comparing the rate of retinal degeneration to the P347S mutant rhodopsin control.
Our study revealed that galectin-1 is upregulated in the P347S mutant retina compared to control at 3 and 4 months at the RNA level, and at 2, 3 and 4 months at the protein level. This study also identifies tenascin-R (TN-R), an extracellular matrix glycoprotein, as a galectin-1 binding partner in the central nervous system. Our studies on the involvement of galectin-1 in retinal suggest that galectin-1 plays a minor role in modulating the extent of retinal degeneration. Our results may underestimate the extent that galectin-1 can modulate degeneration since the experiments utilized the galectin-1 knockout mouse and other galectins may compensate for the removal of galectin-1 in this model.
71
INTRODUCTION
Retinitis pigmentosa (RP) is a genetically heterogeneous group of diseases that leads to blindness. Initially, patients experience night blindness due to the loss of rod photoreceptor cells. The loss of rod photoreceptors is slowly followed by the loss of cone photoreceptors. As these cells are lost, total blindness occurs. Approximately 25% of autosomal dominant retinitis pigmentosa cases have been linked to mutations within rhodopsin, the G-protein couple receptor of rod photoreceptor cells. Over 100 different mutations in rhodopsin have been identified. Seven different point mutations of the proline at residue 347 in rhodopsin have been linked to human RP. This study utilizes one such mutation, the P347S mutant rhodopsin as a model for retinitis pigmentosa.
Located in the c-terminal cytoplasmic tail, the P347S mutation in rhodopsin has been well studied, yet the mechanism causing retinal degeneration is still under debate.
In vitro studies suggest the P347S mutant rhodopsin protein functions normally. In vitro
P347S mutant rhodopsin was found to regenerate with 11-cis retinal, have a normal spectral shift in response to light, activate -transducin, be phosphorylated by rhodopsin kinase, and bind arrestin (Sung, Schneider et al. 1991). In vivo evidence suggests that the majority of P347S mutant rhodopsin localizes properly to the rod outer segments.
However, some of the mutant rhodopsin also accumulates in extracellular vesicles near the junction of inner and outer segments (Li, Snyder et al. 1996). Work from our own lab suggests that transducin activity provides protection from degeneration for the P347S mutant rhodopsin (Brill, Malanson et al. 2007). Our work also indicates that the release of all-trans-retinal may be faster from P347S mutant rhodopsin compared to wild-type
72 rhodopsin (Brill, Malanson et al. 2007). While P347S mutant rhodopsin has been well studied, the mechanism causing degeneration is still unknown.
The P347S mutant rhodopsin mouse is an attractive model for our studies because the time course of degeneration has been carefully studied. In the P347S mouse, degeneration is barely detectable at 1 month of age, yet biochemical evidence suggests that the rod photoreceptor cells have already begun to die. At 4 months of age, approximately half of the rod photoreceptor cells have degenerated (Li, Snyder et al.
1996; Brill, Malanson et al. 2007). We have selected four time points (1, 2, 3, and 4 months of age) to analyze that target early, mid- and late stages of retinal degeneration in order to distinguish involvement of galectin-1 at the various stages of retinal degeneration.
Galectins are a family of carbohydrate binding proteins with affinity for - galactoside, due to a conserved carbohydrate recognition domain. The conserved carbohydrate recognition domain recognizes a minimum saccharide ligand of N- acetyllactosamine. Individual galectins can recognize additional modifications, allowing for tissue and developmental specificity. Fifteen different mammalian galectins have been identified. Galectin-1 is a prototypical member of the galectin family with one carbohydrate recognition domain and the ability to form homodimers.
Galectin-1 has been implicated in a wide variety of biological pathways including cell proliferation, adhesion and migration, and apoptosis (Camby, Le Mercier et al.
2006). Recently, galectin-1 has been implicated in several nervous system functions including neurite outgrowth, resistance to glutamate toxicity, proliferation of neural stem cells, and production of inflammatory and neurotrophic factors after injury (Horie and
73
Kadoya 2004). Several studies have demonstrated that the functions of galectin-1 are dependent on cell type. Cell type specificity is not surprising, as different cell types express different subsets of galectin binding partners, and thus galectin binding could affect different downstream signaling pathways.
Both neurons and glia express galectin-1. Galectin-1 is expressed intracellularly and extracellularly, both on the cell surface as well as in the extracellular matrix.
Galectin-1 can exist as both a monomer and a homodimer. When expressed in oxidizing conditions, three disulfide bonds form in galectin-1. These three disulfide bonds keep galectin-1 in a monomeric state and restrict its ability to bind sugars (Cho and Cummings
1996). Alternatively, in reducing conditions, galectin-1 can exist as a monomer or dimer, and maintains the ability to bind sugars (Cho and Cummings 1995; Cho and Cummings
1995). Previous studies suggest that both oxidized and reduced galectin-1 can protect mature neurons from degeneration (Horie and Kadoya 2004; Ishibashi, Kuroiwa et al.
2007; Han, Xia et al. 2010).
This study investigates the involvement of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse. It is the first to specifically investigate galectin-1 binding partners within the central nervous system. The study reveals that galectin-1 is upregulated in the P347S mutant retina compared to control at 3 and 4 months at the RNA level, and at 2, 3 and 4 months at the protein level. Our study identifies tenascin-R, an extracellular matrix glycoprotein, as a galectin-1 binding partner in the nervous system. Our work with P347S mutant rhodopsin and galectin-1 knockout double transgenic mice suggests that galectin-1 may be able to modulate the rate of retinal degeneration in retinitis pigmentosa.
74
METHODS
Animals. All procedures were carried out in accordance with the Association for
Research in Vision and Ophthalmology statement concerning the use of animals in ophthalmic and vision research and within the guidelines of the Tufts Medical Center
Institutional Animal Care and Use Committee. P347S mutant rhodopsin mice on a wild- type and galectin-1 knockout background were maintained as independent lines under normal cyclic light (5 to 100 lux, in-cage readings). The P347S mutant rhodopsin transgene was maintained and studied in the heterozygous state. P347S, line C1 (Li,
Snyder et al. 1996) rhodopsin transgenic mice were used in these studies. Galectin-1 knockout mice generated previously (Poirier and Robertson 1993) were used in these studies.
Genotyping. Rhodopsin transgene positive animals were identified by PCR analysis.
The P347S rhodopsin mutant expressed a subcloned human rhodopsin gene. The forward control primer that recognized both human and mouse rhodopsin genes was CRH: 5’-
ATG CAG TTC CGG AAC TGG AA -3’. The reverse control primer that specifically recognized the mouse rhodopsin gene was MRH: 5’-TCA GGG ATT ACA CCA CTG
TCC-3’. The reverse primer that specifically recognized the P347S human rhodopsin gene was HRH: 5’-GAA CCT CAC TAA CGT GCC AGT-3’. Denaturing for one minute at 94C was followed by annealing for one minute at 64C then extension for 1 minute at 72C. This cycle was repeated 30 times, then followed by a final extension at
72C for 10 minutes. The amplification yielded a 400 and 100 bp gene product for the wild-type and P347S mutant rhodopsin alleles, respectively.
75
For PCR analysis, DNA primers used to amplify the Gal-1 gene were 6918: 5’-
AAA CTT CAG CCG GGA GAA AGG-3’ (forward primer); 6919: 5’-GAC CCC ATC
CCT ACA CCC CAG-3’ (reverse primer); and 6920: 5’-CAT TCA GGA CAT AGC
GGT GG-3’ (reverse primer). The amplification yielded a 390 and 150 bp gene product for the wild-type and null mutant alleles, respectively.
Quantitative RT-PCR. Total RNA from retinas was extracted using RNeasy mini kit according to the manufacturer's recommendations (Qiagen, Chatsworth, CA). Briefly, the frozen retinas were crushed, suspended in the guanidine thiocyanate buffer (Buffer RLT), and further homogenized using a QIAshredder column. The eluent was loaded onto a silica gel base column and the bound RNA was eluted with RNase-free water.
Total RNA (300 ηg) was reversed transcribed using the High Capacity kit
(Applied Biosystems (ABI), Foster City, CA) according to manufacturer's instructions.
Real-time PCR was performed (Mx4000 real-time PCR machine, Stratagene, La Jolla,
CA) in triplicates using 5 μL of cDNA (derived from 15 ηg total RNA), TaqMan MGB probes, primers specific for the selected genes, and TaqMan Universal PCR master mix
(ABI). Reactions performed in the absence of template served as negative controls. The
ABI primer sets used included: gapDH (Mm99999915_g1) and gal-1 (Mm00839408_g1).
For amplification, Amplitaq Gold DNA polymerase was activated (95°C for 10 min) and the reactions were subjected to 50 cycles involving denaturation (95°C for 15 s) and annealing plus extension (60°C for 1 min). The fluorescent signals were recorded using a
FAM detector and data analysis was performed using Mx4000 software version 2
(Stratagene). The FAM fluorescent signals were measured against the ROX (internal
76 reference dye) signal to normalize the non-PCR-related fluctuations; amplification plots showing the increase in FAM fluorescence with each cycle of PCR (ΔRn) were generated for all samples, and the threshold cycle values (Ct) were calculated from the amplification plots. The Ct value represents the cycle number at which the fluorescence was detectable above an arbitrary threshold, based on the variability of the baseline data during the first 15 cycles. All Ct values were obtained in the exponential phase.
Quantification data of each gene were normalized to the expression of a housekeeping gene, GAPDH. A value of 1.0 was assigned to the expression level of each gene in the normal, wild-type retinas. The values for P347S mutant rhodopsin retinas were expressed as a change in expression levels with respect to normal retinas. A minimum of three animals at each age and genotype were analyzed in triplicate.
SDS-PAGE and Immunoblots. Proteins were resolved by 4-20% SDS-PAGE and transferred onto PVDF membranes. The blots were washed two times for 10 min with
TBST (20 mM Tris-HCl (pH 7.4), 100 mM NaCl, and 0.1% Tween 20) and blocked with
5% nonfat dry milk powder (Bio-Rad) in TBST for 1 h at room temperature. Blots were then incubated with anti-galectin-1 (1:1000), anti-tenascin-R (1:1000) or anti--actin
(1:5000) overnight at 4 °C in blocking buffer. Following primary antibody incubations, immunoblots were incubated with HRP-linked secondary antibodies (either anti-rabbit or anti-mouse) and developed by ECL according to the manufacturer's instructions.
Densitometric analysis of immunoblots, where indicated, was performed using Alpha
Imager (Eastman Kodak Co.) in the linear range of detection. Absolute values were then normalized to total protein as indicated in the figure legends.
77
Sugar-binding column. Galectin-1 retinal binding proteins were isolated by chromatography of retinal extracts on a galectin-1 affinity column as described earlier
(Saravanan, Liu et al. 2009). Briefly, a galectin-1 affinity column was prepared by coupling 5 mg recombinant human Gal-1 to cyanogen-bromide-activated Sepharose 4B
(Sigma) according to the manufacturer's instructions. Whole retina lysates were incubated with the galectin-1 conjugated Sepharose beads, and a column was poured.
The unbound components were removed by washing the column with PBS containing
0.1% Triton X-100 (PBST). Galectin-1 specific binding partners were eluted using PBST containing 0.1 M β-lactose. To confirm sugar-binding specificity of the galectin-1– binding proteins, before elution with β-lactose, the column was eluted with PBST containing 0.1 M sucrose. Fractions eluted from the column with saccharides were dialyzed against water, lyophilized, and resolved by electrophoresis in reducing 4-20%
SDS-PAGE gel. Silver stained components detected on the gel were subjected to peptide analysis using a Thermo LTQ ion trap mass spectrometer. The identity of the galectin-1– binding protein of interest, tenascin-R, was also confirmed by immunoblot analysis using monoclonal anti-tenascin-R (1:1000) as the primary antibody and HRP-conjugated anti– mouse IgG (1:5,000) as a secondary antibody.
Histology. Eyes were oriented at the superior-most point of the eye with a cauterizing pen at the ora serrata. Eyes were then excised and rotated in 4% paraformaldehyde fixative for 4 hours or overnight at 4°C. After fixation, eyes were rinsed several times in
PBS and passed through a dehydrating series of ethanol, with alcohol concentrations
78 increasing from 30% to 90% by increments of 10% for 5-minute intervals. Eyes were immersed in 100% EtOH for 10 minutes and then embedded in paraffin. Eyes were bisected along the vertical meridian through the optic nerve head and stained with hematoxylin and eosin for light microscopy.
Immunohistochemistry. Enucleated eyes were rotated in 4% paraformaldehyde fixative for 4 hours at 4°C. After fixation, eyes were rinsed in PBS and then cryoprotected in
30% sucrose. Eyes were then embedded in OCT and frozen. 10 micron thick sections were cut using a cryostat. Frozen sections were rinsed in PBS for 5 minutes to remove the
OCT. Then sections were steamed with a puddle of 0.1 M sodium citrate buffer for 10 minutes in a food steamer. After steaming, sections were allowed to cool and rinsed in
PBS before incubation with blocking buffer (10% serum, 4% BSA, 0.01 % Triton X-100) for 20 minutes at room temperature. Sections were incubated with primary antibodies, anti-tenascin-R (1:1000) or anti-galectin-1 (1:1000), overnight in blocking buffer at 4°C.
Primary antibodies were detected with corresponding fluorescently labeled secondary antibodies. A negative control, in which the primary antibody was omitted, was included in each immunohistochemistry experiment.
Quantification of retinal degeneration. Retinal sections cut along the vertical meridian of the eye at the optic nerve head were analyzed. Morphological analysis was completed using two methods. First, the ONL thickness was determined by counting the rows of photoreceptor nuclei in the ONL from three different areas in the central and mid- peripheral retina across both the superior and inferior hemispheres. Counts were averaged
79 for each retina. Average counts from three or more animals were taken. A two-tailed
Student’s T-test was used to compare paired data from age-matched wild-type and mutant mice. Second, the ONL to INL ratio was calculated by measuring the ONL length and
INL length on images taken with a Zeiss Axioscope microscope and a Q-Imaging Retiga
200R Faste 1394 black and white CCD camera. Measurements were made on images using the Nikon Instruments NIS Elements software. Measurements were averaged for each retina. Average measurements from three or more animals were taken. Statistical analysis was performed using the one-way ANOVA.
Electroretinogram recording. Dark- and light-adapted electroretinograms (ERGs) were recorded as previously described (Sun et al 2010). Briefly, for ERG recording, mice were dark-adapted overnight and anesthetized with sodium pentobarbital (i.p.) before testing; both pupils of each animal were topically dilated with phenylephrine hydrochloride and cyclopentolate hydrochloride, and mice were then placed on a heated platform. Rod dominated responses were elicited in the dark with 10-μs flashes of white light (1.37 × 105 cd m−2) presented at intervals of 1 min in a Ganzfeld dome. Light- adapted, cone responses were elicited in the presence of a 41-cd m−2 rod-desensitizing white background with the same flashes (1.37 × 105 cd m−2) presented at intervals of
1 Hz. ERGs were monitored simultaneously from both eyes with a silver wire loop electrode in contact with each cornea topically anesthetized with proparacaine hydrochloride and wetted with Goniosol. A saline saturated cotton wick was placed in the mouth as the reference. An electrically shielded chamber served as ground.
80
All responses were differentially amplified at a gain of 1000 (−3 db at 2 and
300 Hz; AM502, Tektronix Instruments, Beaverton, OR, USA), digitized at 16-bit resolution with an adjustable peak-to-peak input amplitude (PCI-6251, National
Instruments, Austin, TX, USA) and displayed on a personal computer using custom software (Labview, version 8.2, National Instruments). Independently for each eye, cone
ERGs were conditioned by a 60-Hz notch filter and an adjustable artifact-reject window, summed (n=4–20).
RESULTS
Galectin-1 expression increases in the P347S mutant rhodopsin mouse retina.
Previous studies indicate that galectin-1 expression increases following neuronal injury induced by both physical trauma and the onset of nervous system disease (Kato,
Kurita et al. 2001; McGraw, Gaudet et al. 2005; Ishibashi, Kuroiwa et al. 2007; Plachta,
Annaheim et al. 2007; Gaudet, Leung et al. 2009; Kajitani, Nomaru et al. 2009; Craig,
Thummel et al. 2010; Han, Xia et al. 2010; Kurihara, Ueno et al. 2010; Qu, Wang et al.
2011). To determine if galectin-1 expression increases during the course of retinitis pigmentosa, we analyzed galectin-1 expression in the P347S mutant rhodopsin mouse model of RP. The time course of degeneration in the P347S mutant rhodopsin mouse model is well studied. At one month, degeneration is barely detectable, yet biochemical evidence suggests that rod photoreceptor cells have already begun to die. At 4 months of age, approximately half of the rod photoreceptor cells have degenerated. We chose to analyze galectin-1 transcript and protein expression levels at 1, 2, 3 and 4 months, as these time points represent early, mid- and late stages of retinal degeneration.
81
Quantitative RT-PCR was completed to compare the levels of galectin-1 transcript in wild-type and P347S mutant rhodopsin retinas (Figure 13). At 1 and 2 months there is no significant change in the levels of galectin-1 transcript between the two groups. Yet, at 3 and 4 months quantitative RT-PCR indicates that the average fold change in galectin-1 transcript levels are significantly increased in the P347S mutant rhodopsin retina compared to wild-type control.
Immunblots were completed to compare levels of galectin-1 protein in wild-type and P347S mutant rhodopsin retinas (Figure 14). At 1 month, the early stage of disease, there is no change in the levels of galectin-1 protein between the two groups. At 2, 3 and
4 months, during the mid and late stages of disease, galectin-1 levels are significantly increased in the P347S mutant rhodopsin retina compared to control.
Galectin-1 binds tenascin-R (TN-R) in a sugar dependent fashion.
By binding cell surface glycoproteins, galectin-1 is involved in many cellular processes. (Reviewed by (Camby, Le Mercier et al. 2006)). To determine what cellular processes galectin-1 might be involved in within the degenerating retina, we identified galectin-1 sugar-binding partners. To identify galectin-1 sugar-binding partners, a galectin-1 sugar-binding column was completed using whole retina lysates.
A galectin-1 affinity column was prepared, and whole retina lysates were passed over it. First, nonspecific binding partners were eluted using a sucrose rinse. Then, galectin-1 specific binding partners were eluted using a lactose rinse, as lactose is a competitor for galectin-1’s sugar-binding moiety. Sucrose and lactose lysates were run on a polyacrylamide gel which was then stained with Coomassie Blue. Two bands, at
82 approximately 175 kDa and 150 kDa, were detected within the lactose lysate but absent from the sucrose lysate (Figure 15A). These bands were cut and subjected to mass- spectrophotometry. Mass-spectrophotometry identified 10 unique peptides within the sequence for tenascin-R (Figure 16). Galectin-1 binding to TN-R was confirmed using a modified sugar-binding column, in which the sucrose and lactose lysates were used for an immunoblot that was probed with an antibody specific for TN-R. Tenascin-R is detected in the total and lactose lysate in both the wild-type and P347S mutant rhodopsin retina, but not within the nonspecific sucrose rinse (Figure 15B), confirming that TN-R is a galectin-1 binding partner in both the wild-type and mutant retina.
Tenascin-R localizes to the synaptic layers in the mouse retina.
Tenascin-R is an extracellular matrix glycoprotein expressed in the central nervous system by oligodendrocytes and a subpopulation of neurons. TN-R is known to localize to perineuronal nets, a specialized structure surrounding synapses. To determine where TN-R is localized in the retina we completed immunohistochemistry on retina sections from wild-type mice. In wild-type retinas, tenascin-R is localized to the outer plexiform layer (OPL) and inner plexiform layer (IPL). The OPL and IPL represent the synaptic layers of the retina.
To determine whether galectin-1 was required for proper localization of TN-R, we completed immunohistochemistry on retina sections from galectin-1 knockout mice. As in wild-type retinas, TN-R localized to the OPL and IPL in galectin-1 knockout retinas
(Figure 17), indicating that galectin-1 is not required for proper TN-R localization.
83
We did try to determine if TN-R colocalizes with galectin-1. However, we were unable to find an antibody for immunohistochemisty that was specific for galectin-1 (see appendix and Figure 26).
Tenascin-R expression levels in the P347S rhodopsin mutant mouse.
Tenascin-R has been implicated in many neuronal processes, including cell migration, neurite outgrowth, myelination and stabilization of synaptic contacts.
Changes within inner retina wiring have been well documented during retinal degeneration. Because TN-R plays a key role in stabilizing neuronal contacts and these contacts might be changing within the P347S mutant rhodopsin retina, TN-R expression levels were examined during the course of retinal degeneration (Figure 18). While it is not statistically significant, TN-R levels are increasing in the P347S mutant rhodopsin retina compared to control at 2 and 3 months. At 4 months the increase in TN-R expression in P347S mutant rhodopsin retina compared to wild-type is statistically significant (Figure 18). These results may suggest that at 4 months, the P347S rhodopsin mutant retina is drastically changing its synaptic contacts, either stabilizing the remaining contacts or making new contacts to compensate for those that have been lost.
Galectin-1 role in retinal degeneration observed in the P347S mutant rhodopsin mouse.
Galectin-1 has previously been shown to be involved in nervous system degeneration. Galectin-1 can protect from as well as exacerbate neural degeneration. To determine the role of galectin-1 in retinal degeneration, the rate of degeneration was examined in the P347S mutant rhodopsin mouse both in the presence and absence of
84 galectin-1. If galectin-1 causes retinal degeneration, then the rate of degeneration in the absence of galectin-1 will be slower than in its presence. If galectin-1 protects against retinal degeneration, then the rate of degeneration in the absence of galectin-1 will be faster than in its presence. The rate of retinal degeneration was assessed both morphologically, using retina sections at 2, 4 and 9 months and functionally, by measuring electroretinograms (ERG) at 4 and 9 months.
To remove galectin-1, the previously characterized galectin-1 knockout mouse
(Poirier and Robertson 1993) was crossed with the P347S mutant rhodopsin mouse.
First, we examined whether removal of galectin-1 alone causes degeneration. Galectin-1 knockout mice have the same outer nuclear layer thickness, outer nuclear layer to inner nuclear layer ratio, and electroretinogram as wild-type control mice at all ages studied
(Figure 19-21). Therefore, removal of galectin-1 alone does not cause degeneration.
The rate of degeneration in the P347S mutant rhodopsin retina in the presence and absence of galectin-1 was assessed morphologically at 2, 4, and 9 months. Two separate measurements were taken to assess the rate of degeneration. First, the outer nuclear layer was measured by counting the number of nuclei within a single column of the ONL.
Second, the ONL and INL length were measured, and the ratio of ONL to INL length was calculated.
At 2 months of age, the P347S rhodopsin mutant retina has only begun to lose photoreceptors, indicated by an ONL thickness of 8 nuclei/column compared to the wild- type ONL thickness of 10 nuclei/column (Figure 19B). At 2 months, there is no difference in either the ONL thickness (Figure 19B), or ONL to INL ratio (Figure 19C) between P347S retinas in the presence or absence of galectin-1. At 4 months, the P347S
85 mutant rhodopsin retina has undergone significant retinal degeneration, indicated by an
ONL thickness of 6 nuclei/column compared to wild-type of 10 nuclei/column (Figure
20B). Yet, at 4 months, there is still no difference in the ONL thickness (Figure 20B), or
ONL to INL ratio (Figure 20C) between P347S retinas in the presence or absence of galectin-1. At 9 months of age, the P347S mutant rhodopsin retina is at a late stage of degeneration with an ONL thickness of 5 nuclei/column compared to wild-type of 10 nuclei/column (Figure 21B). Again, at this time point, there is no difference in the ONL thickness (Figure 21B) or ONL to INL ratio (Figure 21C) between P347S retinas in the presence or absence of galectin-1. At all three time points studied (2, 4 and 9 months), the extent of degeneration was the same in the presence or absence of galectin-1, suggesting that galectin-1 does not affect the extent of degeneration measured morphologically.
In addition to assessing degeneration morphologically, we also assessed degeneration functionally by recording electroretinograms (ERGs). Electroretinograms measure the electrical response of the retina in response to light, and by analyzing different aspects of the electroretinogram, one can determine functionality of different components of the retina. For example, the scotopic ERG is recorded using lighting conditions that primarily activate rod photoreceptors. The scotopic ERG has an initial hyperpolarization, known as the A wave, which corresponds to the hyperpolarization of rod photoreceptor cells in response to light. The A wave is followed by a larger depolarization, known as the B wave. The B wave results because the inner retina depolarizes in response to photoreceptor cell signals. Because the A wave results from photoreceptor cells, it is a direct measure of their function. Likewise, because the B wave
86 results from inner retina signaling, it is a direct measure of inner retina function. It is also possible to record a light adapted, or photopic ERG, to measure the function of cone photoreceptor cells. Because cone photoreceptor cells are a small percentage of photoreceptors in the mouse retina, it is nearly impossible to detect the A wave in a photopic ERG, but you can record the B wave which serves as an indirect measure of the cone photoreceptor cell function. By comparing wave amplitudes it is possible to assess changes in function.
We recorded an initial set of ERGs using mice at 4 months of age. At 4 months of age, the galectin-1 knockout mice maintained normal retinal function compared to wild-type controls, recorded previously using the same experimental methods and setup.
Probably due to the small sample number (n =2), the galectin-1 knockout scotopic A wave amplitudes are on the low side of normal (Figure 22B). At 4 months, even though morphological data indicates that approximately half of the photoreceptor cells have been lost (Figure 20), P347S mutant rhodopsin mice have a normal ERG with an average scotopic A wave amplitude of 160 +/- 17 V (Figure 22B, and Table 1), and an average
B wave of 883 +/- 58 V (Figure 22C). Cone function was also normal with an average photopic B wave amplitude of 95 V.
At 4 months, the P347S mutant on the galectin-1 knockout background had a slightly, but not statistically significant, decreased average scotopic A wave amplitude of
123 +/- 25 V compared to the P347S rhodopsin mutant on a wild-type background
(Figure 22B). Compared to the P347S mutant on the wild-type background, the P347S rhodopsin mutant mice on the galectin-1 knockout background had similar scotopic B wave amplitudes (Figure 22C) as well as photopic B wave amplitudes (Figure 22D).
87
Additionally, all three genotypes studied, the galectin-1 knockout, the P347S mutant on the wild-type background, and the P347S rhodopsin mutant on the galectin-1 knockout background, had normal time-to-peak durations for both scotopic A and B waves (Figure
23).
While the ERG is the gold standard for measuring retinal function, classic ERG recordings are often unable to detect functional changes until nearly all of the photoreceptor cells are lost and the ONL thickness is less than 4 nuclei. Therefore, even though morphological data indicates that the P347S mutant rhodopsin retina has lost approximately half of its photoreceptor cells by 4 months, it is not surprising that the
ERG is unable to detect any functional change between the P347S mutant rhodopsin and wild-type. Because the P347S rhodopsin mutant did not display a degenerative ERG, characterized by decreased scotopic A and B wave amplitudes, we recorded ERGs at a later time point when degeneration would be more evident and the involvement of galectin-1 in retinal function could be measured with the ERG.
To assess galectin-1 involvement in the degeneration at a later stage of disease, we recorded ERGs at 9 months. First, it was determined that removal of galectin-1 did not affect retina function at this age. The galectin-1 knockout ERG is comparable to the wild-type ERG at 9 months. The scotopic A and B wave, and photopic B wave (Figure
24-25 and Table 1) are not different from wild-type at 9 months, confirming that removal of galectin-1 does not cause a change in function.
At 9 months of age, the P347S mutant rhodopsin has a decreased ERG compared to wild-type controls, indicating that it has undergone significant degeneration. The
P347S rhodopsin mutant has a decreased average A wave amplitude of 63 +/- 10 V
88 compared to wild-type of 200 V (Figure 24B). The P347S rhodopsin mutant also has a decreased average B wave amplitude of 403 +/- 37 V compared to wild-type of 700 V
(Figure 24C). The photopic ERG is normal in the P347S rhodopsin mutant mice compared to wild-type indicating that cone function is not changed at this stage of degeneration (Figure 24D).
After confirming that removal of galectin-1 alone does not change retinal function, and that the ERG could detect the degeneration in P347S rhodopsin mutant at 9 months, the degeneration of the P347S mutant rhodopsin mouse on the galectin-1 knockout background was assessed. While the small sample set does not give statistical significance, there is a trend that the P347S mutant rhodopsin in the absence of galectin-1 has a faster rate of degeneration. The P347S rhodopsin mutant on the galectin-1 knockout background has a slightly decreased average scotopic A wave 47 +/- 5 V compared to the P347S rhodopsin mutant on the wild-type background 63 +/- 10 V
(Figure 24B). The P347S mutant on the galectin-1 knockout background also has a slightly decreased average scotopic B wave 341 +/- 32 V compared to the P347S mutant rhodopsin alone 403 +/- 37 V (Figure 24C). Additionally, all three genotypes studied, galectin-1 knockout, P347S rhodopsin mutant on the wild-type background, and P347S rhodopsin mutant on the galectin-1 knockout background, had normal time-to-peak durations for both scotopic A and B waves (Figure 25). The photopic B wave of P347S mutant rhodopsin was not changed in the absence or presence of galectin-1 (Figure 24D), again indicating that at this stage cone photoreceptor cell function remains intact and only rod photoreceptor cells are dying.
89
Again, the small sample set does not allow for statistical comparison, but a slight trend of decreased scotopic A and B wave amplitude is developing, suggesting that in the absence of galectin-1 the P347S rhodopsin mutant degenerates faster than in the presence of galectin-1. This trend may suggest that galectin-1 might provide protection from retinal degeneration.
DISCUSSION
Using the P347S mutant rhodopsin model of retinitis pigmentosa this study examined the involvement of galectin-1 in retinal degeneration. The P347S mutant rhodopsin model allowed for investigation of galectin-1 involvement at various stages of a long term degeneration of the central nervous system.
Galectin-1 is upregulated following nervous system injury or disease onset.
Our study indicates that galectin-1 levels increase both at the RNA and protein level following the onset of retinal degeneration in the P347S rhodopsin mutant mouse.
Our results are in agreement with several previous studies demonstrating upregulation of galectin-1 after the onset of disease or an injury to the nervous system. Galectin-1 levels increase in several models of neurodegeneration, including a zebra fish model of retinal degeneration (Craig, Thummel et al. 2010), a kainate model of seizure (Kajitani, Nomaru et al. 2009), and rodent models of both ischemic (Ishibashi, Kuroiwa et al. 2007; Qu,
Wang et al. 2011) and spinal cord injury (Han, Xia et al. 2010; Kurihara, Ueno et al.
2010). Galectin-1 has also been shown to accumulate within neurofilamentous lesions from ALS patient spinal cord samples (Kato, Kurita et al. 2001). Unlike previous studies,
90 ours was able to map the time course of galectin-1 expression over an extended period of degeneration. Our study indicates that galectin-1 levels are increased early in the disease, at two months, but also that levels continue to increase throughout the course of degeneration. These results suggest that galectin-1 may have an important function following insult to the nervous system.
Galectin-1 binding partners.
Many galectin-1 binding partners have been identified (Camby, Le Mercier et al.
2006), but this study is the first to identify galectin-1 binding partners specifically within the central nervous system. Previous reports identify several extracellular matrix molecules and cell surface proteins as galectin-1 sugar-binding partners, including laminin, fibronectin, and members of the integrin family. Additionally, gemin-4 (Park,
Voss et al. 2001), Ras (Paz, Haklai et al. 2001; Rotblat, Belanis et al. 2010) and Fos B
(Nishioka, Sakumi et al. 2002; Scott and Zhang 2002) have been identified as galectin-1 sugar independent binding partners. While previous galectin-1 binding partners had been identified, this study identified binding partners within the central nervous system to give insight into how galectin-1 could affect neuronal system function in degenerative conditions.
Using retina lysates and a galectin-1 affinity column, tenascin-R was identified as a galectin-1 sugar-binding partner. Tenascin-R is an extracellular matrix glycoprotein expressed in the nervous system in perineuronal nets. While our study did not investigate
TN-R function in retinal degeneration, previous research suggests that TN-R could be a modulator of neuronal degeneration.
91
In vitro studies suggest that TN-R has both adhesive and anti-adhesive properties for neurons and non-neuronal cells (Pesheva, Spiess et al. 1989; Taylor, Pesheva et al.
1993; Pesheva, Probstmeier et al. 1994). TN-R has also been found to both inhibit (Perez and Halfter 1993; Becker, Becker et al. 1999; Becker, Anliker et al. 2000; Becker,
Schweitzer et al. 2003) and enhance (Norenberg, Hubert et al. 1995; Xiao, Revest et al.
1998; Zacharias, Norenberg et al. 1999; Lang, Monzon-Mayor et al. 2008) neurite outgrowth. Given the contradictory nature of these results, it was further determined that
TN-R function was dependent on the cell type studied, and the expression of TN-R receptor splice variants (Bizzoca, Corsi et al. 2009). In vivo studies suggest that TN-R can restrict microglia migration by creating an inhibitory boundary (Sanchez-Lopez,
Cuadros et al. 2004). Additionally, following spinal cord injury, TN-R knockout mice had improved recovery compared to controls (Apostolova, Irintchev et al. 2006), possibly because TN-R restricted remodeling. Finally, TN-R has been implicated in fine motor control, as the TN-R knockout mice have decreased coordination on complicated behavioral tasks, like the rotorod (Freitag, Schachner et al. 2003; Montag-Sallaz and
Montag 2003).
Our study confirmed previous reports that in the retina, TN-R localizes to the outer and inner plexiform layers (Bartsch, Pesheva et al. 1993). The outer and inner plexiform layers represent the synaptic layers of the retina, containing synapses between the photoreceptor cells and inner retina cells, as well as synapses between inner retina cells, and between inner retina and ganglion cells. This study also demonstrates that galectin-1 is not required for TN-R localization as TN-R also localizes to the OPL and
IPL in the galectin-1 knockout retina.
92
Tenascin-R expression levels were trending to increase during the course of retinal degeneration. This trend has been seen previously in a mouse and a lizard model of optic nerve injury (Becker, Anliker et al. 2000; Lang, Monzon-Mayor et al. 2008).
While currently there is no data on why TN-R would increase following neuronal injury, it is possible that it increases to stabilize the remaining synaptic contacts. Interestingly, in salamanders TN-R expression levels decreased following optic nerve injury, and only once the optic nerve had regenerated did TN-R expression return to normal (Becker,
Becker et al. 1999). These results led to the hypothesis that the salamander, unlike the mouse, was able to regenerate because the expression of growth inhibitory molecules, including TN-R, decrease expression following injury.
Galectin-1 and neuronal degeneration.
This study was not able to definitively indicate galectin-1 involvement in the rate of retinal degeneration in the P347S mutant rhodopsin model of retinitis pigmentosa. The morphological results at 2, 4 and 9 months suggest that galectin-1 does not play a role in the rate of retinal degeneration. However, the functional data indicates the removal of galectin-1 may increase the rate of degeneration, seen by a slightly decreased amplitude of the scotopic A wave at 4 and 9 months. Unfortunately, we were unable to include enough animals to determine if this trend was statistically significant. This trend, together with previous studies on galectin-1 involvement in neuronal degeneration, demonstrates that galectin-1 is a modulator of neuronal degeneration. However, more work is needed to determine the function of galectin-1 in retinal degeneration.
93
Galectin-1 has been shown to both protect from and exacerbate neurodegeneration. To date, only one study supports the idea that galectin-1 can cause neurodegeneration in vitro and enhance degeneration in vivo (Plachta, Annaheim et al.
2007). However, there have been many in vitro and in vivo studies supporting the idea that galectin-1 provides a protective effect from neurodegeneration. Treatment of galectin-1 to dorsal root ganglion explants increased the number of axons in both the peripheral and central stumps (Horie, Inagaki et al. 1999). Application of galectin-1 to neuronal cultures also decreased glutamate toxicity (Lekishvili, Hesketh et al. 2006).
Additionally, galectin-1 has been shown to be neuroprotective in vivo. Galectin-1 treatment improved recovery following sciatic nerve injury (Kadoya, Oyanagi et al.
2005). Following induction of seizures by kainate administration, galectin-1 knockout mice have decreased numbers of proliferative cells in the dentate gyrus compared to controls, suggesting that removal of galectin-1 impairs recovery (Kajitani, Nomaru et al.
2009). Following ischemic injury, injection of galectin-1 decreases the area of brain damage, and improves functional recovery measured by behavioral tests (Ishibashi,
Kuroiwa et al. 2007; Qu, Wang et al. 2010; Qu, Wang et al. 2011). In both rodent and monkey models of spinal cord injury, galectin-1 treatment improved behavioral recovery
(Han, Xia et al. 2010; Yamane, Nakamura et al. 2010). Galectin-1 treatment also delayed disease progression and severity in the SOD1 transgenic mouse model of amyotrophic lateral sclerosis (Chang-Hong, Wada et al. 2005). Many studies support the idea that galectin-1 has a protective effect, demonstrated by improved recovery from injury, and delayed onset or decreased severity of disease.
94
Galectin-1 and neuronal function, the debate between oxidized and reduced galectin-1.
Galectin-1 has been shown to both protect from and exacerbate neurodegeneration. To explain these seemingly contradictory results, it was suggested that the oxidation state of galectin-1 could change its function (Horie and Kadoya 2004;
Plachta, Annaheim et al. 2007). It was noted that most studies supporting the idea that galectin-1 provides protection utilized the oxidized form that cannot bind sugars.
However, the study that suggests galectin-1 exacerbates degeneration utilized the reduced, sugar-binding form. These observations lead to the simplified explanation that oxidized galectin-1 provided protection, whereas reduced galectin-1 exacerbated degeneration. This explanation, while attractive, does not take into account biochemical studies on the dynamics between the oxidized and reduced states, as well as studies that demonstrate that the neuroprotective effect of galectin-1 can be independent of oxidation and dependent on sugar-binding.
Biochemical studies suggest that galectin-1 is able to switch between the oxidized and reduced states relatively quickly (Cho and Cummings 1995). In oxidizing environments, three disulfide bonds form within galectin-1, altering its structure and reducing access to the carbohydrate recognition domain (Kadoya and Horie 2005). If not kept in reducing conditions, galectin-1 will oxidize and lose its sugar-binding ability
(Cho and Cummings 1995). However, even in oxidizing conditions binding sugars can stabilize the reduced state of galectin-1 (Stowell, Cho et al. 2009). Given the complicated dynamics between the oxidized and reduced forms of galectin-1, it is impossible to determine which form of galectin-1 is being examined unless specifically tested by examining oxidation state or sugar dependency.
95
Examining whether galectin-1 is reduced or oxidized in vivo is difficult because the experiments needed may alone manipulate its oxidation state. One approach used to study the oxidation state of galectin-1 is to eliminate the possibility of oxidation.
Researchers have made a galectin-1 mutant that lacks the cysteine residues required to make the disulfide bonds that characterize in the oxidized state. This mutant, called CS-
Gal-1, cannot oxidize, but does bind sugars with the same affinity as wild-type galectin-1
(Cho and Cummings 1996). Injection of this CS-Gal-1 after an ischemic injury resulted in increased neurogenesis and behavioral recovery in mice (Ishibashi, Kuroiwa et al.
2007). Additionally, treatment with CS-Gal-1 following spinal cord injury resulted in improved behavioral outcome compared to controls (Han, Xia et al. 2010). Using a mutant form of galectin-1 that cannot oxidize, several studies clearly demonstrate that galectin-1 does not have to be oxidized to be neuroprotective.
While initial studies suggested that the neuroprotective effect of galectin-1 was sugar independent, recent studies have also suggested that it can be sugar dependent.
Galectin-1 provided protection from glutamate toxicity in vitro. This protection was shown to be sugar dependent, as it was blocked by the addition of lactose, a competitive inhibitor of galectin-1 sugar-binding (Lekishvili, Hesketh et al. 2006). Galectin-1 also provided protection in rat model of ischemic injury, indicated by decreased area of brain damage and increased neurological scores. Galectin-1 protection in this ischemia model was also dependent on sugar-binding (Qu, Wang et al. 2011). Therefore, despite initial reports that galectin-1 was only neuroprotective in its oxidized state that lacks the ability to bind sugars, it has been shown that the reduced, sugar-binding form of galectin-1 can
96 also be neuroprotective. As a result, it is important to identify galectin-1 sugar-binding partners in the nervous system.
Studies on the function of galectin-1 in neurodegeneration have generated many, but often conflicting results. All the studies do suggest that galectin-1 may be a modulator of neurodegeneration. While initial reports attempted to explain the differing results by the oxidation state of galectin-1, we would like to offer the alternative explanation that the difference in galectin-1 effect on neurodegeneration could be linked to different galectin-1 binding partners, whether they be sugar dependent or independent, present within the specific cases studied.
97
APPENDIX
Galectin-1 localization in wild-type and galectin-1 knockout mouse.
We did try to characterize the localization of galectin-1 in the P347S mutant rhodopsin retinas. Previously published data indicates that galectin-1 localizes to the inner segments, plexiform layers and ganglion cell layers in the wild-type and degenerating retina (Uehara, Ohba et al. 2001). However, we were unable to get either commercially available or homemade antibodies to work properly. Unlike previously published data, we tested the galectin-1 antibodies on retina sections from galectin-1 knockout mice (Figure 26A), and received the same signal as when using wild-type sections (Figure 26B). Additionally, this signal was consistent between the wild-type and galectin-1 knockout through a dilution series of the antibody. This data indicates that the galectin-1 antibodies are detecting something other than galectin-1, likely another galectin due to the homology between the galectin family members.
Galectin-1 sugar independent binding partners.
To determine galectin-1 sugar independent binding partners in the retina, an immunoprecipitation for galectin-1 using retina lysate was completed. First, it was confirmed that a commercially available antibody was able to pull down galectin-1 from retina lysate (Figure 27). Next, that antibody was used to pull down galectin-1 from a larger preparation of retina lysate. The IP eluate and IgG control were run on a polyacrylamide gel and silver stained (Figure 28A). A band at approximately 30 kDa was cut and sent to the proteomics core for protein identification. Mass-
98 spectrophotometry identified thymus cell antigen 1 (Thy-1) and -actin as galectin-1 binding partners (Figure 28B). To validate the mass spectrometry results, an IP was completed for galectin-1 followed by an immunoblot for thy-1 (Figure 29A). While thy-
1 was detected in the total retina lysate input, it is not detected in the galectin-1 IP, therefore classic IP was unable to validate the mass spectrometry results.
Since galectin interactions with binding partners can be very weak, a sugar- binding assay was also completed to try to validate the mass spectrometry results. Retina lysate was incubated with galectin-1 conjugated beads. Nonspecific binding partners were eluted using a sucrose wash, followed by elution of specific binding partners with a wash of lactose. The total retina lysate input, sucrose wash and lactose wash were used for an immunoblot which was probed using a thy-1 specific antibody (Figure 29B). Thy-
1 was detected in total retina lysate input, but not within the sucrose or lactose rinses, indicating that it does not bind galectin-1 in a sugar dependent fashion. Since both classic
IP and a sugar-binding assay were unable to validate the mass spectrometry results of galectin-1 and thy-1 binding, thy-1 was not pursued any further.
Galectin-1 sugar-binding with neuropilin-1 (NP-1) and tenascin-C (TN-C).
Galectin-1 has many previously characterized sugar dependent binding partners.
While confirming that galectin-1 binds tenascin-R in a sugar dependent fashion, two additional previously characterized galectin-1 binding partners were also assayed for their binding affinity to galectin-1 in the retina. Previously, neuropilin-1 (NP-1) and tenascin-
C (TN-C) have been shown to bind galectins in a sugar dependent fashion. Since both
NP-1 and TN-C have also been linked to various forms of nervous system disease, they
99 were added to the sugar-binding assays. Unfortunately, neither NP-1 nor TN-C are highly expressed within the adult or degenerating retina. NP-1 is detectable at its predicted molecular weight, 130 kDa, within total retina lysate. However, only lower molecular weight band is also detectable within the lactose elute, suggesting that perhaps a truncated form of NP-1 binds galectin-1 in a sugar dependent fashion (Figure 30A).
We were never able to rule out whether the lower molecular weight band was specific for
NP-1 or nonspecific. TN-C is barely detectable within the total retina lysate input, but it is not eluted with the lactose rinse from the galectin-1 sugar-binding assay (Figure 30B), indicating that it does not bind galectin-1 in a sugar dependent fashion.
100
CHAPTER 4: ANALYSIS OF GLYCOGENES AND THEIR
MOLECULAR PATHWAYS IN RHODOPSIN-MEDIATED RETINITIS
PIGMENTOSA BY MICROARRAY
101
ABSTRACT
Recently, several studies revealed a role for the sugar-binding galectin proteins in the degeneration of neural processes. These results suggest the possibility that alterations in glycogenes may play a role in retinal degeneration, an area that is relatively unexplored. In this study, we use a custom microarray to investigate differential expression of glycogenes in the P347S mutant rhodopsin mouse model of retinitis pigmentosa. We analyze four different time points to determine involvement of glycogenes at early, mid and late stages of retinal degeneration. Our initial analysis suggests that P347S mutant rhodopsin retinas have increased expression of genes that are critical for the regulation of microglia, including colony stimulating factor 1 (CSF1) and colony stimulating factor 1 receptor (CSF1R).
INTRODUCTION
Previous studies indicate that glycoproteins are involved in neuronal degeneration. Both galectin-1 and galectin-3, members of the galectin family of carbohydrate binding proteins, have been shown to be involved in neuronal death following injury and disease. Interestingly, depending on the cell type studied, both galectin-1 and galectin-3 can be neuroprotective or neurotoxic (Horie and Kadoya 2004;
Plachta, Annaheim et al. 2007; Narciso, Mietto Bde et al. 2009; Yan, Lang et al. 2009;
Kim, Kim et al. 2010). These differences in functional effect might be caused by cell type specific expression of different galectin binding partners (Camby, Le Mercier et al.
102
2006). By interacting with a different set of binding partners, galectins can have a variety of functional effects (Camby, Le Mercier et al. 2006; Laderach, Compagno et al. 2010).
Galectins have a high affinity for -galactoside, a sugar moiety added to many proteins and lipids during glycosylation. Glycosylation is the most common form of posttranscriptional modification, with over half of all proteins estimated to contain one or more glycan chains (Apweiler, Hermjakob et al. 1999; Lee, Lauc et al. 2005). Many genes, collectively called glycogenes, are responsible for glycosylation and recognition of glycosylated proteins. Glycogenes include glycosyltransferases, sugar nucleotide synthetases, sugar nucleotide transporters, and sugar-chain binding partners. To examine glycogene involvement in neuronal degeneration, this study investigates glycogene differential expression in a mouse model of retinitis pigmentosa.
Retinitis pigmentosa (RP) is a genetically heterogenous group of diseases that causes blindness. In the United States, approximately one in 4000 individuals is affected.
RP begins with the loss of night vision due to the loss of rod photoreceptor cells. The disease progresses slowly with the loss of peripheral vision, and eventually leads to complete debilitating blindness.
The first mutation associated with human RP was identified in the gene encoding rhodopsin, the G-protein coupled receptor of rod photoreceptor cells (Dryja, McGee et al.
1990; Dryja, McGee et al. 1990). Mutations within the rhodopsin gene account for 25% of autosomal dominant RP. Seven different point mutations of the proline at residue 347 in rhodopsin have been linked to human RP. We have selected to analyze the P347S rhodopsin mutation in this study, as it is one of a handful of rhodopsin mutations that have been produced in a transgenic mouse model (Li, Snyder et al. 1996).
103
The P347S rhodopsin mutation is one of the best characterized mutations, yet the mechanism by which the mutation causes degeneration is still unknown. The time course of degeneration in this mouse model has been carefully studied. In the P347S mouse, degeneration is barely detectable at 1 month of age, yet biochemical evidence suggests that the rod photoreceptor cells have already begun to die. At 4 months of age, approximately half of the rod photoreceptor cells have degenerated (Brill, Malanson et al.
2007). We have selected four time points (1, 2, 3, and 4 months of age) to analyze that target early, mid- and late stages of the retinal process in order to distinguish involvement of glycogenes at the various stages of retinal degeneration.
METHODS
Animals. All procedures were carried out in accordance with the Association for
Research in Vision and Ophthalmology statement concerning the use of animals in ophthalmic and vision research and within the guidelines of the Tufts Medical Center
Institutional Animal Care and Use Committee. The P347S mutant rhodopsin transgene was maintained and studied in the heterozygous state. P347S, line C1 (Li, Snyder et al.
1996) rhodopsin transgenic mice were used in these studies.
Genotyping. Rhodopsin transgene positive animals were identified by PCR analysis.
The P347S rhodopsin mutant expressed a subcloned human rhodopsin gene. The forward control primer that recognized both human and mouse rhodopsin genes was CRH: 5’-
ATG CAG TTC CGG AAC TGG AA -3’. The reverse control primer that specifically recognized the mouse rhodopsin gene was MRH: 5’-TCA GGG ATT ACA CCA CTG
104
TCC-3’. The reserve primer that specifically recognized the P347S human rhodopsin gene was HRH: 5’-GAA CCT CAC TAA CGT GCC AGT-3’. Denaturing for one minute at 94C was followed by annealing for one minute at 64C then extended for 1 minute at 72C. This cycle was repeated 30 times, then followed by a final extension at
72C for 10 minutes. The amplification yielded a 400 and 100 bp gene product for the wild-type and P347S mutant rhodopsin alleles, respectively.
Sample collection and RNA preparation. Total RNA from retinas was extracted using
RNeasy mini kit according to the manufacturer's recommendations (Qiagen, Chatsworth,
CA). Briefly, the frozen retinas were crushed, suspended in the guanidine thiocyanate buffer (Buffer RLT), and further homogenized using a QIAshredder column. The eluent was loaded onto a silica gel base column and the bound RNA was eluted with RNase-free water.
GLYCOv4 microarray. The glycogene microarray, GLYCOv4, is an oligonucleotide microarray, custom designed by Affymetrix (Santa Clara, CA) for the Consortium for
Functional Glycomics at the Scripps Institute, La Jolla, CA. The array contains approximately 2000 mouse and human glycogenes including glycosyltransferases, glycosidases, enzymes involved in nucleotide-sugar synthesis and transport, proteoglycans, and glycan binding proteins. A complete list of the probe sets and annotation for the GLYCOv4 oligonucleotide array is available at the Functional
Glycomics Gateway
(http://www.functionalglycomics.org/static/consortium/resources/resourcecoree.shtml).
105
The glycoarray has oligonucleotide probe sets consisting of 11 probe pairs to interrogate each targeted mRNA sequence. Each probe pair consists of one perfect match
(PM) and one mismatch (MM) 25-base oligonucleotide. The PM oligonucleotide is complementary to a given portion of the targeted gene. The MM oligonucleotide is identical in sequence to the PM probe, except for a single mismatched base at the 13th position (center) of the probe. The difference between the PM and MM probe signals among all the probe pairs for a given gene is used to calculate the hybridization signal.
This signal is a quantitative metric (a weighted average) calculated for each probe set that represents the relative abundance of the transcript. A one-sided Wilcoxon Signed Rank
Test is applied to this probe-pair intensity distribution to generate the p value. Thus, a given transcript can be categorized as present (p < 0.04), absent (p > 0.06), or marginal
(0.04 < p < 0.06) on the basis of the predefined p-value thresholds. More information of
Affymetrix algorithms is available at the following website: http://www.affymetrix.com/support/technical/technotes/statistical_reference_guide.pdf.
Hybridization and data analysis. To prepare hybridization probes, total RNA (100 ng) from each sample was amplified, and cDNA was synthesized according to a modified
Baugh/Harvard protocol. The resultant cDNAs were transcribed in vitro in the presence of biotin-labeled ribonucleotides and the labeled cRNA was hybridized to the GLYCOv4 microarray, and scanned using Affymetrix Scanner 3000. Quantitation of expression signal values, quantile normalization, and background subtraction were performed as described earlier (Diskin, Kumar et al. 2006). The gene expression patterns in the three replicates of each group (P347S mutant rhodopsin and wild-type) and time point (1, 2, 3,
106 and 4 months) were then analyzed by hierarchical clustering using BRB ArrayTools
3.2.2.
The differential expression of genes in P347S mutant rhodopsin compared to wild-type retinas was analyzed as described by Diskin et al. (Diskin, Kumar et al. 2006).
Briefly, the transformed expression values for the replicated probe sets were averaged to get a single expression value for each probe set on each array. Then, statistically significant changes in gene expression were identified using BRB ArrayTools 3.2.2 software. The class comparison test was conducted using a univariate alpha cutoff of
0.001 and a multivariate permutation-based false discovery rate calculation. The predicted proportion of false discoveries was preset at 10% and a false discovery rate calculation was set at a confidence level of 80%. The gene microarray data from this study has been deposited in the Consortium of Functional Glycomics database.
The Database for Annotation, Visualization and Integrated Discovery (DAVID) software (Diskin, Kumar et al. 2006), Genecards (http://genecards.org), Ingenuity
Pathway Analysis (IPA), and Entrez gene (http://www.ncbi.nlm.nih.gov/entrez) were used to gain insight into the biological functions of differentially expressed genes.
Quantitative RT-PCR. Total RNA (300 ηg) was reverse transcribed using the High
Capacity kit (Applied Biosystems (ABI), Foster City, CA) according to manufacturer's instructions. Real-time PCR was performed (Mx4000 real-time PCR machine,
Stratagene, La Jolla, CA) in triplicates using 5 μL of cDNA (derived from 15 ηg total
RNA), TaqMan MGB probes, primers specific for the selected genes, and TaqMan
Universal PCR master mix (ABI). Reactions performed in the absence of template served
107 as negative controls. The ABI primer sets used included: glyceraldehyde-3-phosphate dehydrogenase (gapDH) (Mm99999915_g1), colony stimulating factor 1 (CSF1)
(Mm00432686_m1), colony stimulating factor 1 receptor (CSF1R) (Mm01266652_m1), interleukin 6 signal transducer (IL6ST) (Mm00439668_m1), and CX3 chemokine receptor 1 (CX3CR1) (Mm00438354_m1). For amplification, Amplitaq Gold DNA polymerase was activated (95°C for 10 min) and the reactions were subjected to 50 cycles involving denaturation (95°C for 15 s) and annealing plus extension (60°C for 1 min).
The fluorescent signals were recorded using a FAM detector and data analysis was performed using Mx4000 software version 2 (Stratagene). The FAM fluorescent signals were measured against the ROX (internal reference dye) signal to normalize the non-
PCR-related fluctuations; amplification plots showing the increase in FAM fluorescence with each cycle of PCR (ΔRn) were generated for all samples, and the threshold cycle values (Ct) were calculated from the amplification plots. The Ct value represents the cycle number at which the fluorescence was detectable above an arbitrary threshold, based on the variability of the baseline data during the first 15 cycles. All Ct values were obtained in the exponential phase. Quantification data of each gene were normalized to the expression of a housekeeping gene, GAPDH. A value of 1.0 was assigned to the expression level of each gene in the wild-type retinas. The values for P347S mutant rhodopsin retinas were expressed as a change in expression levels with respect to wild- type retinas.
Immunohistochemistry. Enucleated eyes were rotated in 4% paraformaldehyde fixative for 4 hours at 4°C. After fixation, eyes were rinsed in PBS and passed through a
108 dehydrating series of ethanol, with alcohol concentrations increasing from 30% to 90% by increments of 10% for 5-minute intervals. Eyes were immersed in 100% EtOH for 10 minutes and then embedded in paraffin. Eyes were bisected along the vertical meridian through the optic nerve head. Sections were dewaxed with xylene and rehydrated with alcohol. Then sections were steamed with a puddle of 0.1 M sodium citrate buffer for 10 minutes in a food steamer. After steaming, sections were allowed to cool and rinsed in
PBS before incubation with blocking buffer (10% serum, 4% BSA, 0.01 % Triton X-100) for 20 minutes at room temperature. Sections were incubated with primary antibodies, including anti-CSF1 (1:1000) and anti-CSF1R (1:1000), overnight at 4°C. Primary antibodies were visualized using appropriate biotinylated secondary antibodies and peroxidase-based ABC kit (Vector Laboratories, Burlingame, CA) followed by incubation with diaminobenzidine (DAB). A negative control, in which only the primary antibody was omitted, was included in each immunohistochemistry experiment.
RESULTS
Gene expression profiles of P347S mutant rhodopsin retinas and wild-type retinas.
Previous studies revealed a role for the sugar-binding galectins proteins in the degeneration of neuronal processes (Horie and Kadoya 2004; Plachta, Annaheim et al.
2007; Narciso, Mietto Bde et al. 2009; Yan, Lang et al. 2009; Kim, Kim et al. 2010).
These results suggest the possibility that alterations in glycogenes may play a role in retinal degeneration. To identify glycogenes that are differentially expressed in P347S mutant rhodopsin retina compared to control a glycogene microarray was completed using samples taken at four time points during degeneration. The time points selected
109 were 1, 2, 3 and 4 months since they represent the beginning, middle, and late stages of disease for the P347S mutant rhodopsin mouse model of retinitis pigmentosa. An unsupervised hierarchical clustering by sample, which provides a step-wise analysis of the similarity in gene expression profiles between the individual samples, was used to examine the relationship between the data sets. The relationship is presented graphically in a dendrogram (Figure 31), where the closer samples cluster together, the more similar the gene expression profiles are. As shown in Figure 31, the P347S mutant rhodopsin retinas generally cluster together on one side of the dendrogram, whereas the wild-type control retinas cluster together on the other side of the dendrogram, suggesting that there are differences in expression profile between P347S mutant rhodopsin and control retinas. Additionally, the individual samples taken at the same time point from each genotype generally cluster together, suggesting that the glycogene expression profile changes during the course of degeneration.
Differentially expressed glycogenes in P347S mutant rhodopsin retinas compared to control retinas.
For this study, transcripts were considered differentially expressed when they had greater than 1.3-fold difference between the geometric mean signal of the wild-type control group and the P347S mutant rhodopsin group, as well as an adjusted p value of less than 0.05. The number of differentially expressed transcripts increased throughout the course of degeneration (Table 2). Of the 2000 genes on the microarray, only 4 transcripts were differentially expressed at 1 month. By 4 months, the latest time point studied, 117 transcripts were differentially expressed.
110
The transcripts were categorized based on their involvement in cellular processes or functions as determined by GeneCard, Entrez gene and DAVID software. To investigate potentially altered cellular processes or functions, the number of transcripts altered within each specific category was analyzed throughout the time course of degeneration (Figure 32, Table 3). Over all time points studied, the three most prominent categories of differentially expressed transcripts include glycan degradation, growth factors and receptors, and glycosyltransferases. At 1 month, of the 4 transcripts differentially expressed in the P347S mutant rhodopsin retinas, two are involved in glycan degradation, one is a growth factor, and one is a glycoprotein. At 2 months, the largest group of differentially expressed transcripts are glycosyltransferases (28%, or 9 transcripts), followed by growth factors and receptors (24%, or 8 transcripts). At 3 months, the largest single group of differentially expressed transcripts are growth factors and receptors (26%, or 16 transcripts), followed by glycosyltransferases (21%, or 13 transcripts). At 4 months, the largest single group of differentially expressed transcripts are glycosyltransferases (29%, or 34 transcripts) followed by growth factors and receptors (24% or 28 transcripts).
All of the differentially expressed transcripts are organized and listed in tables by category of cellular process or function (Tables 4-10). The tables also list the fold change for each transcript in P347S mutant rhodopsin retinas compared to control at each time point if it was statistically significant (adjusted p < 0.05). Note that many transcripts are differentially expressed at more than one time point, suggesting that after cellular functions change during degeneration, they remain changed throughout the course of degeneration. For transcripts that are differentially expressed at more than one time
111 point, note that they are either downregulated or upregulated at each time point, suggesting that the pathways that are altered during degeneration remain consistently downregulated or upregulated throughout degeneration.
Validation of differentially expressed genes by qRT-PCR.
We used three criteria to select a small subset of transcripts to further investigate.
First, because we were interested in processes that occur early in degeneration, we focused on transcripts that were differentially expressed at 1 and 2 months. This first criterion limited our list of 139 transcripts, to the 35 transcripts listed in Table 11. We also wanted to investigate transcripts that showed a trend of differential expression over several time points. Finally, we were interested in transcripts that had previously been implicated in neurodegeneration. These three criteria narrowed the list to four transcripts: chemokine (C-X3-C) receptor 1 (CX3CR1), colony stimulating factor 1 (CSF1), colony stimulating factor 1 receptor (CSF1R), and interleukin 6 signal transducer (IL6ST).
While none of the four selected transcripts encode genes responsible for glycosylation or the remodeling of glycan chains, and thus are not glycogenes, they all have been implicated in immune function by regulating microglia. Glycogenes play a critical role in immune function by creating glycans that allow for recognition of self and pathogens (Rabinovich, Ariel et al. 1999; Rabinovich and Ilarregui 2009; Rabinovich and
Toscano 2009; Ilarregui and Rabinovich 2010; Laderach, Compagno et al. 2010).
Therefore, even though the four selected transcripts are not glycogenes, we decided to further investigate their involvement in P347S mutant rhodopsin mediated degeneration,
112 because they met our previously determined criteria, and link to glycogenes through their involvement in modulating the inflammatory response.
All four selected transcripts have been implicated in microglia function.
Microglia are the resident immune cells of the central nervous system. Activation of microglia can be both neuroprotective and neurotoxic. Activated microglia provide neuronal protection by releasing neurotrophic and anti-inflammatory molecules, clearing toxic products or invading pathogens, and guiding stem cells to inflammatory sites to allow for neurogenesis. Activated microglia can also be neurotoxic through the release of cytotoxic substances such as nitric oxide, superoxide, and cytokines. While most studies on microglia function have been completed in brain pathologies including Alzheimer’s disease, Parkinson’s disease and Multiple Sclerosis (For review see (Kreutzberg 1996;
Streit 2002; van Rossum and Hanisch 2004; Kim and Joh 2006), recent studies have also suggested a role for microglia in retinal degenerations (for review see (Langmann 2007;
Karlstetter, Ebert et al. 2010). Therefore, in addition to our three criteria above, we were also interested in investigating these four transcripts to elucidate a role for microglia in the P347S mutant rhodopsin model of retinitis pigmentosa.
We used gene-specific qRT-PCR to confirm the differential expression of these four selected genes. GAPDH, a housekeeping gene, was used as a reference gene in the present study, as its expression was similar between P347S mutant rhodopsin and control retinas.
Chemokine (C-X3-C) receptor 1 (CX3CR1).
113
Primarily expressed by microglia CX3CR1 is a G-protein coupled receptor for fractalkine, a chemokine constitutively expressed by neurons in the CNS (Harrison, Jiang et al. 1998). CX3CR1 is important for sustaining normal microglia activity. While some studies suggest that CX3CR1 can provide protection from neurodegeneration (Cardona,
Pioro et al. 2006; Combadiere, Feumi et al. 2007), other studies suggest that CX3CR1 exacerbates neuronal death following injury (Zhuang, Kawasaki et al. 2007; Denes,
Ferenczi et al. 2008). The function of CX3CR1 has not been investigated in retinitis pigmentosa.
To further investigate the role of CX3CR in retinitis pigmentosa, we wanted to validate the microarray data which indicates that CX3CR1 is upregulated at 2, 3 and 4 months in the P347S mutant rhodopsin retina compared to control. To validate the microarray results, we completed quantitative RT-PCR. Unfortunately, while the microarray indicates that there is a statically significant fold change at 2, 3 and 4 months, the quantitative RT-PCR data was only able to validate a statistically significant increase at 3 months (Figure 33). Therefore, we decided not to further pursue CX3CR1 function in the P347S mutant rhodopsin mouse model of retinitis pigmentosa.
Colony stimulating factor 1(CSF1) and colony stimulating factor 1 receptor (CSF1R).
In the central nervous system, colony stimulating factor 1 (CSF1) is a cytokine that controls the production, differentiation and function of microglia by binding to its receptor, CSF1R. Several lines of evidence suggest that CSF1 and CSF1R play a role in neuronal-glial interaction in disease states and following injury (Murphy, Zhao et al.
2000; Takeuchi, Miyaishi et al. 2001; Kutza, Fields et al. 2002). Previous reports
114 indicate that CSF1 signaling through CSF1R is neuroprotective (Berezovskaya,
Maysinger et al. 1995; Berezovskaya, Maysinger et al. 1996; Wang, Berezovska et al.
1999; Murphy, Zhao et al. 2000; Vincent, Robinson et al. 2002; Mitrasinovic, Grattan et al. 2005). Therefore, we were interested in validating the microarray data, which indicates both CSF1 and CSF1R are upregulated in P347S mutant rhodopsin retinas compared to control at 2, 3 and 4 months.
To validate the microarray results for CSF1 and CSF1R we completed quantitative RT-PCR. The quantitative RT-PCR data indicates that both CSF1 and
CSF1R are upregulated at 1, 3 and 4 months (Figure 34, 36). Since the qRT-PCR validated the microarray results, we investigated whether CSF1 and CSF1R changed localization in response to retinal disease. To determine CSF1 and CSF1R localization, we completed immunohistochemistry on both wild-type and P347S mutant rhodopsin retinas. The immunohistochemistry data indicates that CSF1 (Figure 35) and CSF1R
(Figure 37) do not change localization in response to retinal disease as they are diffusely expressed throughout both the wild-type and P347S mutant rhodopsin retinas, particularly within the inner segments, and both the inner and outer plexiform layers. We did attempt to investigate protein level expression by immunoblot, but were unable to find a good antibody for this purpose.
Interleukin 6 signal transducer (IL6ST).
Interleukin 6 signal transducer, also known as glycoprotein 130 (gp130), is the common transmembrane receptor subunit for the IL-6 cytokine family. The IL-6 cytokine family includes IL-6, CNTF, LIF, oncostatin M, and cardiotrophin-1. Evidence
115 suggests that activated IL6ST is neuroprotective in both the CNS and PNS (Sendtner,
Schmalbruch et al. 1992; Hagg and Varon 1993; Cheema, Richards et al. 1994; Cheema,
Richards et al. 1994; Anderson, Panayotatos et al. 1996; Gadient, Lein et al. 1998). After neuronal injury IL-6 cytokines have been shown to modulate activation of astrocytes and microglia (Penkowa, Molinero et al. 2001; Lee, Seo et al. 2009). Additionally, the IL-6 cytokines, including CNTF, IL-6, LIF and cardiotrophin1, have been shown to protect against retinal degenerations (LaVail, Unoki et al. 1992; Cayouette, Behn et al. 1998;
LaVail, Yasumura et al. 1998; Chong, Alexander et al. 1999; Tao, Wen et al. 2002; Song,
Zhao et al. 2003; Ueki, Wang et al. 2008). Together these data demonstrate that the IL-6 cytokines provide protection from retinal degenerations, and suggest that IL6ST is critical for mediating their neuroprotective effect.
These previous reports and the microarray data, which indicates that IL6ST is upregulated in the P347S mutant rhodopsin retina at 1, 3 and 4 months, suggests that
IL6ST may play an important role in retinitis pigmentosa. Therefore we were interested in validating the microarray results for IL6ST. To validate the microarray results for
IL6ST we completed quantitative RT-PCR. Unfortunately, the qRT-PCR data was unable to validate IL6ST upregulation at any time point studied (Figure 38). Therefore, we did not further investigate IL6ST function in the P347S mutant rhodopsin model of retinitis pigmentosa.
DISCUSSION
The goal of the present study was to identify glycogenes that were differentially expressed in the P347S mutant rhodopsin retina compared to control to elucidate the
116 involvement of glycogenes and their cellular processes in retinitis pigmentosa. The microarray identified over 100 differentially expressed transcripts. During our initial investigation, we selected three criteria to limit the number of transcripts to further investigate. The three criteria were 1) differential expression at 1 or 2 months, 2) differential expression at 3 or more time points, and 3) previously characterized link to neurodegeneration. These three criteria limited our list to four transcripts: chemokine (C-
X3-C) receptor 1 (CX3CR1), colony stimulating factor 1 (CSF1), colony stimulating factor 1 receptor (CSF1R), and interleukin 6 signal transducer (IL6ST). We completed qRT-PCR for all four transcripts. While we were unable to validate differential expression for CX3CR1 and IL6ST, we were able to validate differential expression for
CSF1 and CSF1R.
Colony Stimulation Factor 1 and Colony Stimulating 1 Receptor
The qRT-PCR data indicates that CSF1 and CSF1R are increased by more than 2- fold in the P347S mutant rhodopsin retina compared to control at 1, 3 and 4 months. This upregulation implies that CSF1/CSF1R signaling may play an important role in retinitis pigmentosa. Previous studies demonstrate that CSF1 and CSF1R modulate neuronal- glial interaction in disease states and after injury. Following brain insult by either injury or disease, CSF1 is upregulated, and this upregulation is accompanied by a strong and selective induction of CSF1R on activated microglia (Murphy, Zhao et al. 2000;
Takeuchi, Miyaishi et al. 2001; Kutza, Fields et al. 2002). In vitro and in vivo studies suggest that microglia activation is mediated by CSF1/CSF1R signaling, resulting in increased microglia proliferation, cytokine expression and phagocytosis (Berezovskaya,
117
Maysinger et al. 1996; Mitrasinovic, Perez et al. 2001; Hao, Dheen et al. 2002; Li,
Pisalyaput et al. 2004; Mitrasinovic, Robinson et al. 2004). Activation of microglia via the CSF1/CSF1R pathway has been shown to be neuroprotective in both in vitro and in vivo models of disease and injury, including Alzheimer’s disease, excitotoxicity, traumatic injury and ischemic injury (Berezovskaya, Maysinger et al. 1995;
Berezovskaya, Maysinger et al. 1996; Wang, Berezovska et al. 1999; Murphy, Zhao et al.
2000; Vincent, Robinson et al. 2002; Mitrasinovic, Grattan et al. 2005).
While, the role of CSF1/CSF1R signaling has yet to be studied in the retina, several studies suggest that the pathway is activated following retinal injury. In a mouse model of acute blue light-induced retinal lesions, injured retinas had an increased expression of several cytokines, including CSF1 (Joly, Francke et al. 2009). In a model of diabetic retinopathy, CSF1 and CSF1R mRNA transcripts and protein levels were upregulated compared to controls. Additionally, CSF1 was in the nerve fiber co- localizing with GFAP, and CSF1R was in OX-42 labeled microglia and ganglion cells in diabetic retinas (Liu, Xu et al. 2009).
Given previous studies demonstrating a neuroprotective effect of CSF1/CSF1R signaling, and our own results indicating upregulation of CSF1 and CSF1R in a model of retinitis pigmentosa, more work needs be done to elucidate the role of these proteins in retinal degenerations.
Microglia and Retinal Degeneration
The microarray data suggests that microglia play an important role in retinal degeneration. All four selected transcripts, CX3CR1, CSF1, CSF1R and IL6ST, have
118 been implicated in modulating microglia function. While we were unable to validate the microarray data for CX3CR1 and IL6ST, it is interesting that all four transcripts share a common function of relating microglia activity. Additionally, the microarray data suggests the involvement of other inflammatory pathways in retinitis pigmentosa with the differential expression of several cytokines, interleukins and chemokines (Table 8-10) in the P347S mutant rhodopsin retina compared to control.
Several pieces of evidence suggest that microglia, and more generally inflammatory pathways, play an important role in rhodopsin-mediated retinitis pigmentosa. Often microglia activation is associated, and in some cases precedes, retinal degeneration (Zeiss and Johnson 2004; Zhang, Shen et al. 2005; Gehrig, Langmann et al.
2007). Following laser insult or white light damage, a large number of microglia contain drusen, a characteristic feature of age related macular degeneration (Combadiere, Feumi et al. 2007; Raoul, Feumi et al. 2008). A microarray completed using retinoschisin- deficient mice, a model of severe photoreceptor dystrophy, indicated high expression levels of microglia specific transcripts before the onset of apoptosis or degeneration
(Gehrig, Langmann et al. 2007). Finally, manipulation of microglia function in the rd model of retinitis pigmentosa, resulted in a changed rate of retinal degeneration. When microglia were stimulated, the rate of degeneration decreased. However, when microglia were depleted, the rate of degeneration increased (Sasahara, Otani et al. 2008). Many studies indicate that microglia are involved in retinal degenerations.
Activated microglia can provide neuroprotection through the release of neurotrophic factors including brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF), glia-derived neurotrophic factor (GDNF), 3'-nucleotidase
119
(NT3), basic fibroblast growth factor (bFGF), and galectin-1 (Carwile, Culbert et al.
1998; Yin, Cui et al. 2003; Yin, Henzl et al. 2006). However, activated microglia can also be neurotoxic through the release of factors that cause degeneration, including TNF- alpha, reactive oxygen species, nitric oxide, proteases, and excitatory amino acids (Boje and Arora 1992; Kreutzberg 1996). While it is still unknown what determines whether microglia are neuroprotective or neurotoxic, a recent study identified two separate populations of microglia from injured spinal cords. Investigators showed that the presence of lipopolysaccharide (LPS) and the proinflammatory cytokine interferon-γ
(IFNγ) resulted in “classically activated” microglia that were proinflammatory and neurotoxic. However, the presence of interleukin-4 (IL-4) or IL-13, resulted in
“alternatively activated” microglia that were anti-inflammatory and neuroprotective
(Kigerl, Gensel et al. 2009). This study suggests that it might be possible to isolate neuroprotective microglia and utilize them for treatment of neurodegenerations, including cases of retinitis pigmentosa.
Microglia involvement in the P347S mutant rhodopsin model of RP has yet to be directly demonstrated. Our study suggests that microglia, and more generally inflammatory pathways, may play a role in the degeneration observed in the P347S mutant rhodopsin model of retinitis pigmentosa. Therefore, it is important to further characterize microglia activation, migration and involvement in this model of retinal disease. First, microglia activation and migration could be studied by completing immunohistochemistry for microglia markers in P347S mutant rhodopsin retinas compared to control. Additionally, it would be interesting to determine microglia involvement in degeneration by modulating microglia function through the use of
120 cytokines and transgenics. Further work is needed to determine the role of microglia and whether they are neuroprotective or neurotoxic in the P347S mutant rhodopsin model of
RP.
Analysis of the Glycome.
Previous studies revealed a role for the sugar-binding galectin proteins in the degeneration of neuronal processes. These results suggested that glycogenes play an important role in neurodegeneration. Our microarray data is unique because it is the first to examine differential expression of glycogenes in a model of retinal degeneration.
During our initial study we selected three criteria to select a subset of transcripts to further investigate. While we selected our initial criteria to elucidate glycogene involvement in neurodegeneration, we did limit ourselves by focusing only on those transcripts that had previously been linked to degeneration. By eliminating that single criterion, one would be able to identify novel glycogenes involved in neurodegeneration.
Glycosylation is regulated by the actions of numerous genes that code for glycosyltransferases, glycosidases and other enzymes that synthesize and remodel glycan chains as well as accessory enzymes involved in the synthesis and transport of nucleotide sugars. Glycosylation patterns are known to be cell-type-specific. Additionally, aberrant glycosylation patterns impact cell behavior, morphology and function (Ohtsubo and
Marth 2006). Studies have shown that glycosylation patterns are altered in tumors and healing wounds (Saravanan, Cao et al. 2009; Kroes, He et al. 2010; Saravanan, Cao et al.
2010). To date, changes in glycosylation have not been examined in neurodegeneration,
121 but analysis of the microarray suggests that glycosylation may be altered in the P347S mutant rhodopsin retinas.
The microarray indicates that there are several differentially expressed transcripts involved in glycan degradation in the P347S mutant rhodopsin retina. At one month, before degeneration is even detected, the microarray indicates there are two differentially expressed transcripts involved in glycan degradation, beta-galactosidase (GLB1) and sulfatase 1 (SULF1) (Table 3). GLB1, which hydrolyzes the terminal beta-galactose from ganglioside substrates and other glycoconjugates, is down regulated in the P347S mutant rhodopsin retina at 1, 2, and 4 months by nearly 3-fold. Alternatively, SULF1 is upregulated at all time points in the P347S mutant rhodopsin retina. SULF1 selectively removes the 6-O-sulfate groups from heparan sulfate proteoglycans. Heparan sulfate proteoglycans act as coreceptors for numerous heparin-binding growth factors and cytokines. Additionally, heparan sulfate proteoglycans may play a role in the pathogenesis of Alzheimer’s disease by associating with the amyloid precursor protein
(van Horssen, Wesseling et al. 2003; Ariga, Miyatake et al. 2010). In addition to GBL1 and SULF1, the microarray detects eight other transcripts involved in glycan degradation that are differentially expressed in the diseased retina compared to control.
The microarray indicates there are 38 glycosyltransferases that are differentially expressed in the P347S mutant rhodopsin retina compared to control (Table 6). Three are differentially expressed at 2, 3 and 4 months: G3GALT1, B3GNT2, and ST8SIA3. The remaining 35 are differentially expressed at either 2 or only 1 time point. G3GALT1 is a member of the beta-1,3-galactosyltransferase (beta3GalT) gene family which synthesize type 1 carbohydrate chains. B3GNT2 is a member of the beta-1,3-N-
122 acetylglucosaminyltransferase family. B3GNT2 is involved in the synthesis of poly-N- acetyllactosamine chains. ST8SIA3 belongs to a family of sialyltransferases that form sialyl-alpha-2,8-sialyl-R linkages at the nonreducing termini of glycoconjugates. The actions of the 38 differentially expressed glycosyltransferases could change the glycosylation pattern in the P347S mutant retina compared to control.
The glycogene microarray indicates that there are many differentially expressed transcripts that could alter the glycosylation pattern in the degenerating retina compared to control. Further analysis would need to be done in order to validate differential expression. Given the number of differentially expressed transcripts involved in glycosylation, it is reasonable to predict that the glycosylation patterns between the diseased and healthy retina would be different. Since altered glycosylation patterns can change cell morphology, behavior and function, it would be of interest to further examine these transcripts and glycosylation pattern in the P347S mutant rhodopsin retina compared to control to determine if alterations in glycosylation have an effect on disease progression.
123
CHAPTER 5: A DISCUSSION OF THE MOLECULAR MECHANISMS
UNDERLYING P347S MUTANT RHODOPSIN-MEDIATED
RETINITIS PIGMENTOSA
124
ABSTRACT
Retinitis pigmentosa (RP) is a genetically heterogeneous group of diseases that causes blindness. RP can be inherited as an X-linked, autosomal recessive or autosomal dominant disease. Mutations within the rhodopsin gene account for approximately 25% of autosomal dominantly inherited RP cases. Therefore understanding the mechanisms causing rhodopsin-mediated RP has a significant health impact. To date, results from multiple labs indicate that rhodopsin-mediated RP pathogenesis does not share a common mechanism of degeneration. There is strong evidence that multiple mechanisms are involved, including protein misfolding, mislocalization, release of toxic products and aberrant signaling. This thesis investigated the molecular mechanisms involved in the retinal degeneration of P347S mutant rhodopsin mouse model of retinitis pigmentosa.
This chapter will discuss the major findings of the work and highlight new areas for future investigation.
INTRODUCTION
Autosomal dominant retinitis pigmentosa (ADRP) is a genetically heterogeneous group of diseases that causes blindness. Mutations in several proteins within the phototransduction cascade have been linked to ADRP (Daiger 1996-2007). Mutations within rhodopsin are the most common cause of ADRP. More than 100 different mutations within rhodopsin have been identified to cause retinitis pigmentosa (Daiger
1996-2007). Seven different point mutations of the proline at residue 347 in rhodopsin
125 have been linked to human RP. This study utilized one such mutation, the P347S mutant rhodopsin as a model for retinitis pigmentosa.
Located in the c-terminal cytoplasmic tail, the P347S mutation in rhodopsin has been well studied, yet the mechanism causing retinal degeneration is still under debate.
In vitro studies suggest the P347S mutant rhodopsin protein functions normally. In vitro
P347S mutant rhodopsin has been found to regenerate with 11-cis-retinal, have a normal spectral response to light, activate -transducin, be phosphorylated by rhodopsin kinase, and bind arrestin (Sung, Schneider et al. 1991). In vivo evidence suggests that the majority of P347S mutant rhodopsin localizes properly to the rod outer segments, however some of the mutant rhodopsin also accumulates in extracellular vesicles near the junction of inner and outer segments (Li, Snyder et al. 1996). Additionally, the creation of transgenic chimeric mice expressing P347S mutant rhodopsin in only a subset of photoreceptor cells demonstrated that retinal degeneration was through a non- autonomous pathway (Huang, Gaitan et al. 1993). The P347S rhodopsin mutant mouse model is one of the best characterized models of ADRP, yet the mechanism causing its degeneration is still under debate.
This thesis utilized the P347S rhodopsin mutant mouse model to investigate the involvement of three molecular mechanisms in the pathogenesis of rhodopsin-mediated retinitis pigmentosa. First, the involvement of persistent photosignaling was tested using the transducin knockout mouse which blocks the phototransduction cascade. Second, the involvement of galectin-1 was tested using the galectin-1 knockout mouse. Third, the involvement of glycogenes was investigating using a custom gene microarray.
126
DISCUSSION
Persistent photosignaling
Persistent photosignaling has previously been shown to cause retinal degeneration
(Hao, Wenzel et al. 2002; Woodruff, Wang et al. 2003). Furthermore, degeneration caused by persistent photosignaling can be rescued by introducing a null mutation for the
-subunit of transducin which prevents G-protein mediated signaling (Hao, Wenzel et al.
2002; Woodruff, Wang et al. 2003). In chapter 2, we investigate the hypothesis that the retinal degeneration observed in the P347S mutant rhodopsin mouse is due to persistent photosignaling. To test this hypothesis, we placed the P347S mutant rhodopsin onto an
-transducin null background. The results of this experiment demonstrate that removal of -transducin increases the rate of retinal degeneration in both the P347S mutant rhodopsin and the K296E mutant rhodopsin mouse model (Figure 8). Therefore, persistent photosignaling is not a mechanism causing retinal degeneration for these two mutant rhodopsins. Additionally, the results suggest that transducin activity provides some level of protection from degeneration for the P347S and K296E mutant rhodopsin mediated RP.
The presence of transducin could provide protection from degeneration by restricting the ability of rhodopsin to form aberrant protein complexes. When
transducin is bound to rhodopsin, arrestin cannot bind (Krupnick, Gurevich et al.
1997). The formation of stable rhodopsin-arrestin complexes is reported to cause retinal degeneration in drosophila and mice (Alloway, Howard et al. 2000; Kiselev, Socolich et al. 2000; Chen, Shi et al. 2006). Therefore, we hypothesized in the absence of
transducin, P347S mutant rhodopsin preferentially formed a complex with arrestin
127 which lead to the increased rate of degeneration. To test this hypothesis, we placed the
P347S mutant rhodopsin, transducin null mice onto an arrestin knockout background.
The results of this experiment demonstrate that removal of arrestin does not provide protection from degeneration (Figure 9). Therefore, rhodopsin-arrestin complexes are not causing the increased rate of retinal degeneration observed in the P347S mutant rhodopsin on the -transducin null background.
The presence of -transducin could also provide protection from degeneration by stabilizing light-activated P347S metarhodopsin. To assess this possibility, animals were dark-reared to block the light-initiated conversion of rhodopsin to metarhodopsin. Dark- rearing does provide some, but not complete, protection for the P347S mutant rhodopsin mice on the -transducin null background (Figure 10), suggesting that indeed transducin provides protection by stabilizing metarhodopsin. Interestingly, dark-rearing does not provide any protection for the P347S mutant rhodopsin mice in the presence of - transducin (Figure 10), suggesting that a portion of the P347S mutant rhodopsin mediated degeneration is caused by a light independent mechanism.
The presence of -transducin could provide protection from degeneration by restricting the ability of rhodopsin to signal through non-canonical pathways which might lead to degeneration. The involvement of an abnormal rhodopsin signaling pathway has been investigated with a separate line of P347S mutant rhodopsin mice (Chang, Hao et al.
1993). The P347S mutant rhodopsin retina had increased levels of cAMP compared to wild-type retinas (Weiss, Hao et al. 1995). Since cAMP levels are not normally regulated through the canonical phototransduction pathway, increased levels of cAMP in rhodopsin mutant mice suggested the involvement of a non-canonical pathway in the P347S mutant
128 rhodopsin retina. These data suggest the mutant rhodopsin might be activating a signaling pathway independent of transducin, the photoreceptor-specific G-protein.
While studies are just beginning to identify transducin-independent rhodopsin signaling pathways, many studies have identified other G-protein coupled receptors that signal independently of their G-proteins. These non-canonical signaling G-proteins include metabotropic glutamate (Heuss, Scanziani et al. 1999), muscarinic (Rolland,
Henquin et al. 2002), mu opiod (Pak, O'Dowd et al. 1999), 2-adrenergic (Wei, Ahn et al.
2003), platelet activating factor (Lukashova, Asselin et al. 2001; Chen, Dupre et al. 2002) and angiotensin II type 1a receptors (Wei, Ahn et al. 2003).
New studies are underway to identify other non-canonical rhodopsin-dependent
-transducin-independent signaling pathways. Currently, only one published study has been able to identify a signaling pathway initiated by light exposure to rhodopsin and independent of transducin activity. This study found that the phosphorylation of retinal insulin receptor was independent of -transducin but dependent on rhodopsin activation
(Rajala, Anderson et al. 2007). Activation of insulin receptor has been implicated in neuroprotection (Yu, Rajala et al. 2004) as well as diabetic retinopathy (Reiter, Wu et al.
2006). Unpublished work from our lab suggests that protein kinase D (PKD) is phosphorylated in response to light. PKD phosphorylation is rhodopsin dependent, yet
-transducin independent. PKD activity has been implicated in many cellular processes including apoptosis, proliferation, and cellular adhesion and migration (Jaggi, Du et al.
2007). It is currently not known whether either of these non-canonical signaling pathways is involved in the increase of degeneration in the P347S mutant rhodopsin mice in the absence of -transducin. However, the absence of -transducin may allow non-
129 canonical pathways to increase signaling which may result in an increased rate of retinal degeneration. The involvement of non-canonical pathway is retinal degeneration is an area that requires further examination.
Chromophore Toxicity
The accelerated degeneration observed in P347S transgenic mice in the absence of
-transducin may result from destabilization of light-activated P347S metarhodopsin, and a faster release of toxic free all-trans-retinal. Unfortunately, it is technically impossible to directly measure the levels of free all-trans-retinal because it is highly reactive and is immediately converted to other species. Free all-trans-retinal reacts with phosphatidylethanolamine to form a family of toxic bis-retinoid fluorophores. Therefore, as an indirect measure of the level of free all-trans-retinal, the accumulation of bis- retinoid fluorophores, including A2E and its precursors, was assessed in the P347S retina compared to wild-type. A2E and three of its precursors were elevated in the P347S retina compared to control (Figure 11). Additionally, lipofuscin granule density, an indicator of cellular waste, is increased in the P347S retina compared to control (Figure 11). Together these results suggest that P347S mutant rhodopsin is destabilized and has a faster release of toxic free all-trans-retinal.
As a more direct test of chromophore toxicity, we placed the P347S expressing mice onto an Rpe65 null genetic background. The Rpe65 enzyme is critical for the recycling of chromophore and, in its absence cells cannot produce 11-cis-retinal
(Redmond, Yu et al. 1998). Without 11-cis-retinal, all-trans-retinal cannot be formed.
Therefore, if the degeneration present in the P347S mutant is caused by increased release
130 of all-trans-retinal, then placing the mice on the Rpe65 null background should provide protection. However, placing the mice on the Rpe65 null background did not protect from degeneration (Figure 12). These results suggest that chromophore toxicity alone does not account for the degeneration observed in the P347S mutant rhodopsin model of retinitis pigmentosa.
While our initial studies measuring the levels of bis-retinoid fluorophores suggest that P347S mutant rhodopsin is destabilized and has a faster release of toxic free all- trans-retinal, our follow up studies with Rpe65 null mice suggest that chromophore toxicity is not involved in degeneration. These conflicting results may be explained by examining the role 11-cis-retinal plays in opsin stabilization. It has long been known that
11-cis-retinal binding increases opsin thermostability (Hubbard 1958). The chromophore is a strong inverse agonist for rhodopsin, stabilizing the inactive form (Black and
Shankley 1995; Bond, Leff et al. 1995). Without both 11-cis-retinal and -transducin, it is possible that P347S mutant rhodopsin is constitutively signaling through a non- canonical pathway which is causing degeneration. Finally, multiple studies have shown an association between poor chromophore binding and rhodopsin mislocalization
(Kaushal, Ridge et al. 1994; Li, Sandberg et al. 1998). Therefore it is possible that the
Rpe65 null background did not rescue degeneration, because it not only blocked the formation of toxic free all-trans-retinal, but also all the beneficial effects of 11-cis-retinal, including stabilization of inactive rhodopsin, and aiding in folding and localization.
Further examination is required to determine the necessity of chromophore in stabilizing the inactive form of P347S mutant rhodopsin, as well as its folding and localization.
131
Involvement of Galectin-1
Galectin-1 is a prototypical member of the galectin family, a group of lectins with affinity for -galactosides. Galectin-1 has been shown to be involved in many signaling pathways including regulation of cell growth and cell migration by altering adhesion, motility and invasion (Camby, Le Mercier et al. 2006). Additionally, previous studies revealed a role for galectin-1 in the degeneration of neuronal processes (Chang-Hong,
Wada et al. 2005; Sakaguchi, Shingo et al. 2006; Ishibashi, Kuroiwa et al. 2007; Jin, Na et al. 2007; Plachta, Annaheim et al. 2007; Kajitani, Nomaru et al. 2009). Chapter 3 examined the involvement of galectin-1 in the retinal degeneration in the P347S mutant rhodopsin mouse model of retinitis pigmentosa.
Previous studies demonstrate that galectin-1 expression increases following neuronal damage (Ishibashi, Kuroiwa et al. 2007; Kajitani, Nomaru et al. 2009; Craig,
Thummel et al. 2010; Han, Xia et al. 2010; Kurihara, Ueno et al. 2010; Qu, Wang et al.
2011). Therefore, galectin-1 RNA and protein expression levels were examined in the
P347S retina compared to control. Galectin-1 RNA was upregulated in the mutant retina at the 3 and 4 month time points (Figure 13). Additionally, galectin-1 protein levels were upregulated in the mutant retina compared to control at 2, 3 and 4 months (Figure 14).
Since we were most interested in comparing galectin-1 expression in mutant to control, we did not examine galectin-1 developmental expression in wild-type. Although, preliminary data suggests that galectin-1 protein levels are developmentally regulated and increase with time in both the wild-type and mutant retinas. It would be interesting to examine galectin-1 developmental expression in wild-type.
132
Galectin-1 Neuronal Binding Partners
Galectins exert control over cellular functions by binding to glycosylated proteins.
Several previous studies have identified galectin-1 binding partners (Yang, Rabinovich et al. 2008; Rabinovich and Toscano 2009). However, our study was the first to identify a neuronal specific galectin-1 binding partner. By passing retina lysate through a galectin-
1 affinity column, we identified tenascin-R (TN-R) as a neuronal specific galectin-1 sugar-binding partner (Figure 15).
TN-R is an extracellular matrix molecule expressed in the central nervous system in perineuronal nets. Perineuronal nets, once thought to be an artifact of staining techniques, are now recognized as specialized structures which surround cell bodies and synapses in the central nervous system (Celio, Spreafico et al. 1998). TN-R is an essential component of perineuronal nets. When TN-R is removed, perineuronal nets do not form properly (Weber, Bartsch et al. 1999). Perineuronal nets and their molecular components, including TN-R, have been found to play a critical role in synaptic plasticity and in pathological conditions such a trauma (Bukalo, Schachner et al. 2001; Apostolova,
Irintchev et al. 2006; Galtrey and Fawcett 2007; Gogolla, Caroni et al. 2009; Wakao,
Imagama et al. 2011). During development, an increase in chondroitin sulfate proteoglycan (CSPG) expression within perineuronal nets coincides with the end of the critical period and a decrease in synaptic plasticity. This observation led to the hypothesis that the main functions of perineuronal nets are to maintain tissue architecture
(Margolis and Margolis 1993) and to regulate stabilization of synapses (Kalb and
Hockfield 1988). Therefore, it is reasonable to predict that TN-R, within both the wild- type and diseased retina, is stabilizing synaptic contacts. Furthermore, it is possible that
133 disrupting TN-R function in the diseased retina could lead to an exacerbated phenotype, as removal of TN-R could destabilize already stressed synaptic contacts. Experiments to test this possibility could be completed using the TN-R knockout mice previous created
(Weber, Bartsch et al. 1999), and would give further insight into the role of TN-R in the retina.
TN-R and galectin-1 both effect neurite outgrowth. To date, their effect on each other in models of neuronal regeneration has not been studied. Since we demonstrate an association between galectin-1 and TN-R in the retina, it would be interesting to examine their functional effects on each other’s ability to affect neurite outgrowth and modulate neuronal regeneration. Their combined functional effect could be studied in vitro using dorsal root ganglion explants or hippocampal cultures. To determine if the protective effect of galectin-1 is dependent on TN-R in vivo, one could compare the effect of galectin-1 treatment following a neuronal insult to TN-R knockout versus control. These in vivo studies would need to be tightly controlled as the TN-R knockout alone has a unique phenotype following neuronal injury (Apostolova, Irintchev et al. 2006).
Interestingly, both TN-R and galectin-1 have given conflicting results on their functional effect on neurite outgrowth. Some studies suggest they can enhance growth while other studies suggest they inhibit growth. Both sets of data have been shown to be dependent on the cell type and system studied. The conflicting results of TN-R function were shown to be dependent on the cell-type specific expression of different TN-R receptor splice variants (Bizzoca, Corsi et al. 2009). The varying results on galectin-1 function were initially explained as dependent on galectin-1 oxidation state and thus whether or not it could bind sugars. As discussed in chapter 3, this argument does not
134 take into account the biochemical studies on the dynamics between the oxidized and reduced states, as well as studies that demonstrate that the neuroprotective effect of galectin-1 can be independent of oxidation and dependent on sugar-binding. While initial reports attempted to explain the differing results by the oxidation state of galectin-1, we offer the alternative explanation that the difference in galectin-1 effect on neurodegeneration could be linked to different galectin-1 binding partners present within the specific cases studied. Therefore it would also be interesting to study the TN-R receptor splice variants relative to galectin-1 function. As different splice variants of TN-
R receptors are expressed, TN-R binding is altered, which might affect TN-R glycosylation and thus the ability of galectin-1 to bind.
While our study was the first to identify a neuronal specific sugar-binding partner for galectin-1, there is still much work to be done to determine the functional effect of
TN-R and galectin-1 binding. Additionally, galectin-1 likely has many binding partners within the nervous system that still need to be identified. It was quite possible that galectin-1 involvement in the degeneration of neuronal processes could be mediated through several different binding partners. Therefore, it would be of interest to continue to identify neuronal specific galectin-1 binding partners and investigate their combined functional effect on neuronal degeneration.
Galectin-1 functional effect on retinal degeneration
Previous studies indicate that galectin-1 is involved in neuronal degeneration. We hypothesized that galectin-1 was involved in the rate of retinal degeneration observed in the P347S mutant rhodopsin mediated retinitis pigmentosa. This hypothesis was tested
135 by examining the rate of degeneration in the presence and absence of galectin-1 with both morphological and functional assessments. Retinal morphology was examined at 2, 4 and 9 months using both outer nuclear layer thickness and outer nuclear layer to inner nuclear layer ratio. At all time points studied, both assessments of morphology indicate that the rate of degeneration in the absence of galectin-1 is not statistically different from the rate of degeneration in the presence of galectin-1 (Figures 19-21). These morphological results suggest that galectin-1 is not involved in the rate of P347S mutant rhodopsin mediated retinitis pigmentosa.
While morphological assessments can be useful in determining gross abnormalities in retinal morphology, they do not measure changes in retina functional ability. Since we were also interested in determining if galectin-1 had an effect on retina function during degeneration, we recorded electroretinograms in the presence and absence of galectin-1. Our preliminary studies indicate that the scotopic A wave amplitude is slightly decreased in the absence of galectin-1 at both 4 and 9 months
(Figures 22B and 24B), suggesting that galectin-1 may modulate the rate of retinal degeneration. Since the decrease in A wave amplitude is small, an additional twenty mice would need to be examined in order to determine statistical significance.
Unfortunately, the galectin-1 knockout mice are poor breeders and we did not have the number of mice necessary to complete the functional experiments to be able to determine statistical significance. Although, with the slight trend our preliminary data suggest, and previous reports on galectin-1 involvement in degeneration of neuronal processes, it is important to further examine galectin-1 involvement in other models of retinal degeneration.
136
Galectin-1 has been found to be involved in the pathology of numerous models of neurodegeneration, including amyotrophic lateral sclerosis, multiple sclerosis, spinal cord injury and ischemic injury. Therefore, it is likely that galectin-1 is involved in other models of neurodegeneration including retinal degeneration. The previous studies investigating galectin-1 involvement in neurodegeneration all used experimental models with faster degeneration than the P347S mutant rhodopsin mouse model. On average, experimental end points are within one month of galectin-1 treatment, with only one study having an end point of 3 months (Yamane, Nakamura et al. 2010). We thought the slower degeneration of the P347S mutant mouse model would benefit our study by allowing us to focus on early, mid and late involvement of galectin-1 in retinal degeneration. However, it is possible that with the slower time course of degeneration, other mechanisms were able to compensate for lack of galectin-1 in the galectin-1 knockout mouse. Therefore it would be interesting to examine galectin-1 involvement in faster models of retinal degeneration like the retinal degeneration (rd) mouse model of retinitis pigmentosa or a model of light-induced degeneration.
Using multiple experimental models of retinal degeneration, it would also be interesting to examine the effect of galectin-1 treatment on the rate of degeneration. The most convincing results demonstrating galectin-1 involvement in neurodegeneration are from studies in which galectin-1 treatment delayed the onset, decreased the severity, and improved functional recovery following neuronal insult (Chang-Hong, Wada et al. 2005;
Kadoya, Oyanagi et al. 2005; Ishibashi, Kuroiwa et al. 2007; Han, Xia et al. 2010; Qu,
Wang et al. 2010; Yamane, Nakamura et al. 2010; Qu, Wang et al. 2011). Therefore, it would be interesting to test the effect on galectin-1 treatment in models of retinal
137 degeneration. These experiments could be completed by directly injecting recombinant galectin-1 into the subretinal space. Initial experiments would need to be completed to determine rate of diffusion and stability of galectin-1 in the retina. The dosage of galectin-1 for treatment could be modeled on previous studies, but given that galectin-1 has not been used in the retina, a dose response curve would also need to be completed.
Once these preliminary experiments are completed, it would be possible to treat multiple mouse models of retinal degeneration with recombinant galectin-1 to determine its effect on morphology and function. In addition to treatment via injection studies, the effect of galectin-1 on neurodegeneration could also be tested with transgenic approaches. It would be possible to examine galectin-1 effect on neurodegeneration through injection of an adenovirus expressing galectin-1 transcript or by creating a transgenic mouse overexpressing galectin-1 in either retina or neuronal specific fashion. A combination of transgenic and classical drug injection approaches could be used to examine the effect of galectin-1 treatment on retinal degeneration and further our understanding of its involvement in retinal disease.
Galectin-1 influence on microglia function
Another approach to investigate galectin-1 involvement in retinal degeneration is to examine the ability of galectin-1 to modulate microglia function in the retina. Previous studies investigating the involvement of galectin-1 in neuronal degeneration found that galectin-1 directly affects microglia, the resident macrophage and immune cell of the
CNS. Changes in microglia are known to alter neuronal function. In addition, microglia
138 are involved in several models of retinal degenerations. Therefore, it is important to understand the role of galectin-1 in mediating microglia response in retinal degenerations.
Several previous studies demonstrate galectin-1 directly affects the function of microglia. Following injury to the central nervous system, galectin-1 is first seen upregulated in microglia cells (Kurihara, Ueno et al. 2010). Galectin-1 is both necessary and sufficient to increase macrophage accumulation following peripheral nerve injury
(Gaudet, Leung et al. 2009). Galectin-1 is upregulated in activated macrophages and can inhibit nitric oxide production by decreasing expression of iNOS (Correa, Sotomayor et al. 2003). Galectin-1 has also been shown to associate with macrophage complement receptor 3 (Avni, Pur et al. 1998). Using a fluorescently tagged galectin-1, researchers were able to demonstrate that galectin-1 bound to macrophages and not to DRG neurons or Schwann cells in vitro (Horie, Kadoya et al. 2004). Galectin-1 binding to macrophages also stimulated changes in tyrosine phosphorylation. Finally, after galectin-
1 treatment, macrophage conditioned media increased axonal regeneration in DRG explants (Horie, Kadoya et al. 2004). Together these results suggest that galectin-1 may affect neuronal survival by directly affecting microglia function.
Many studies, including the glycogene microarray discussed in chapter 4, indicate that microglia are involved in retinal degenerations. Often microglia activation is associated with, and in some cases precedes, retinal degeneration (Zeiss and Johnson
2004; Zhang, Shen et al. 2005; Gehrig, Langmann et al. 2007). After light damage, microglia contain drusen, a characteristic feature of age related macular degeneration,
(Combadiere, Feumi et al. 2007; Raoul, Feumi et al. 2008). Using a mouse model of severe photoreceptor dystrophy, a microarray detected high expression levels of
139 microglia specific transcripts before the onset of apoptosis or degeneration (Gehrig,
Langmann et al. 2007). Finally, manipulation of microglia function can change the rate of retinal degeneration. When microglia were stimulated, the rate of degeneration decreases. However, when microglia were depleted, the rate of degeneration increases
(Sasahara, Otani et al. 2008). The results of several studies support the idea that microglia are involved in retinal degenerations.
Activated microglia can be neuroprotective or neurotoxic. Activated microglia provide neuronal protection by releasing neurotrophic and anti-inflammatory molecules, clearing toxic products or invading pathogens, and guiding stem cells to inflammatory sites to allow for neurogenesis. Activated microglia can also be neurotoxic through the release of cytotoxic substances such as nitric oxide, superoxide, and cytokines. In the periphery, galectin-1 not only suppresses the secretion of pro-inflammatory cytokines, including IL-2 (Rabinovich, Ariel et al. 1999), but galectin-1 also induces the release of interleukin-10 (IL-10) an anti-inflammatory molecule (van der Leij, van den Berg et al.
2004). Therefore, it is possible that galectin-1 could modulate retina microglia to suppress the release of pro-inflammatory cytokines and induce the release of anti- inflammatory cytokines to adopt a neuroprotective role. It is not only important to determine the effect of galectin-1 on retina microglia, but also to determine whether this effect is neuroprotective or neurotoxic.
Because galectin-1 has been shown to mediate microglia function, and microglia have been previously linked to other models of retinal degeneration, it would be interesting to investigate galectin-1 effect on microglia function in other models of retinal disease. Good experimental models for these studies will have previously characterized
140 microglia involvement, including age related macular degeneration, rd mouse model of retinal degeneration, and light induced degeneration (Langmann 2007).
Glycogene involvement in retinal degeneration
Several previous studies indicate that both galectin-1 and galectin-3 are involved in the degeneration of neuronal processes (Horie and Kadoya 2004; Plachta, Annaheim et al. 2007; Narciso, Mietto Bde et al. 2009; Yan, Lang et al. 2009; Kim, Kim et al. 2010).
Since galectins exert their primary functional affects through binding glycans, these results suggest the possibility that alterations in glycans and more specifically, the genes responsible for synthesizing these sugar chains, might be involved in neuronal degeneration. To investigate this possibility a glycogene microarray was completed using retina lysate from P347S mutant rhodopsin mice compared to control at four different time points to target early, mid and late degeneration.
Analysis of the glycogene microarray revealed several novel findings on glycogene expression during retinal degeneration. First, the number of differentially expressed transcripts, defined as having an adjusted p value of greater than 0.01 and a fold change of greater than 1.3-fold, increased as degeneration progressed. At one month only 4 transcripts were differentially expressed, but by 4 months over 100 transcripts were differentially expressed (Table 2). Additionally, at each time point studied over half of all differentially expressed transcripts were categorized as glycosyltransferases, growth factor and receptors, and involved in glycan degradation (Figure 32).
To select a handful of transcripts to further investigate we decided on three criteria. The three criteria were 1) differential expression at 1 or 2 months, 2) differential
141 expression at 3 or more time points, and 3) previously characterized involvement in neurodegeneration. These three criteria limited our list to four transcripts: chemokine (C-
X3-C) receptor 1 (CX3CR1), colony stimulating factor 1 (CSF1), colony stimulating factor 1 receptor (CSF1R), and interleukin 6 signal transducer (IL6ST). We completed qRT-PCR for all four transcripts. While we were unable to validate differential expression for CX3CR1 and IL6ST (Figures 33, 38), we were able to validate differentially expression for CSF1 and CSF1R (Figures 34, 36).
Examination of the glycome
Previous studies indicate that glycans and the glycogenes responsible for glycan synthesis have altered expression patterns in pathological settings (Diskin, Kumar et al.
2006; Kroes, He et al. 2010; Saravanan, Cao et al. 2010). Additionally, alterations in glycosylation patterns can widely alter protein function (Lowe and Marth 2003). The data from the microarray can be used to characterize differential expression of glycogenes to further our understanding of the role of altered glycosylation patterns in neurodegeneration.
Since we focused only on those transcripts that had previously been linked to neuronal degeneration, our initial criteria to select a limited number of transcripts did restrict the investigation from finding novel glycogenes involved in degeneration. By eliminating that single criterion, or by creating a new set of criteria, one would be able to further investigate the glycogene microarray data and could find novel glycogenes involved in the degeneration of neuronal processes.
142
One possible approach to further examine the microarray data would be to examine differentially expressed transcripts specifically involved in glycan synthesis and degradation. Changes in expression in these enzymes have previously been shown in pathological conditions (Diskin, Kumar et al. 2006; Kroes, He et al. 2010; Saravanan,
Cao et al. 2010). Therefore, it would be interesting to examine their role in neuronal degeneration. The microarray indicates that several transcripts involved in glycan transfer and degradation are altered in the P347S retina compared to control (Tables 4 and 6). Altered expression of these enzymes could result in a different set of glycan structures being created and attached to proteins and lipids. Changes in glycan structures could affect pathology in two ways. First, altered glycosylation patterns have been shown to directly affect protein function (Lowe and Marth 2003). Second, a unique set of glycan structures could recruit a different set of glycan-binding partners, which could have different roles and functional effects upon binding sugar substrates. Altering the glycan structures attached to proteins, could not only alter the function of that glycosylated protein, but also alter the binding of additional proteins and therefore downstream signal events. Therefore, alterations in the enzymes involved in the transfer and degradation of glycan moieties could have vast functional effects.
The microarray presented in chapter 4 has a wealth of data still to be examined.
The microarray is unique because it is the first to investigate differential expression of glycogenes in a model of neuronal degeneration. It would be interesting to further examine trends of altered glycogenes within the microarray and compare them to trends within other models of retinal degeneration to determine common pathways of glycogene involvement in degeneration.
143
CONCLUSION
Retinitis pigmentosa (RP) is a heterogeneous group of disease that causes blindness. Approximately 25% of autosomal dominantly inherited RP cases are caused by mutations within rhodopsin. This thesis examined the molecular mechanisms underlying the degeneration of the P347S mutant rhodopsin model of retinitis pigmentosa. Through the use of transgenic animal models the involvement of persistent photosignaling, aberrant rhodopsin-arrestin complexes, chromophore toxicity and galectin-1 in retinal degeneration were investigated. Additionally, the involvement of glycogenes was investigated with a custom gene microarray.
144
FIGURES
Figure 1: Schematic of the mammalian eye
The mammalian eye is responsible for converting the light energy of the environment into a neural response that can be decoded by the brain. Light enters the eye via the transparent cornea, passes through the anterior chamber and lens until it falls on the retina, the specialized tissue at the back of the eye that contains the light sensitive photoreceptor cells.
145
Figure 2: Retina structure
The retina is a highly organized tissue with cells organized into specific layers. Rod and cone photoreceptor outer segments contain membranous discs packed with rhodopsin.
The nuclei for the photoreceptor cells are located within the outer nuclear layer (ONL).
Because photoreceptor cells are lost during retinal degeneration, counting the number of nuclei in a column in the ONL is a measure of the extent of degeneration. Photoreceptor cells synapse onto second order neurons in the outer plexiform layer (OPL). The inner nuclear layer (INL) contains the bipolar, amacrine and horizontal cells. The ganglion cells, whose axons form the optic nerve, are located within the ganglion cell layer (GCL).
Scale bar = 10 m.
146
Figure 3: Phototransduction cascade
Activation by a single photon of light produces a conformational change in rhodopsin by converting 11-cis-retinal to all-trans-retinal. This isomerization creates a conformational change in rhodopsin, which then binds transducin, the rod specific heterotrimeric G- protein. The substitution of GDP by GTP on the transducin -subunit activates transducin. Activation of transducin causes the dissociation of transducin complex from the active GTP-bound -subunit. The activated -subunit of transducin then activates cyclic guanosine monophosphate phosphodiesterase (cGMP PDE) by binding the phosphodiesterase -subunit, removing its inhibitory effect on the catalytic cGMP phosphodiesterase - and -subunits. Activated cGMP PDE hydrolyzes cGMP to
5’GMP, lowering the intracellular levels of cGMP. The decrease in cGMP
147 concentrations results in the closure of the cyclic nucleotide gated (CNG) channels, blocking the inward flow of sodium and calcium. The decreased flow of sodium and calcium hyperpolarizes the membrane. This change in membrane potential is transmitted as a neural signal through the secondary neurons of the retina to the ganglion cells and, ultimately to rest of the brain. Rhodopsin is inactivated by phosphorylation by rhodopsin kinase, which uncouples rhodopsin and transducin, followed by binding of visual arrestin to completely quench rhodopsin signaling activity.
148
Figure 4: Rhodopsin secondary structure and point mutations associated with ADRP
Secondary structure of rhodopsin with point mutations that have been associated with retinitis pigmentosa are colored in black. Note that the mutations are found within every domain of the protein, including the cytoplasmic, transmembrane, and intradiscal domains. The three mutations discussed in further detail are labeled, P23, K296 and
P347.
149
Figure 5: P23H difference spectra
Photobleaching difference spectra of wild-type rhodopsin and P23H mutant rhodopsin expressed in a heterologous cell system. The 495 nm absorbance peak observed in the wild-type rhodopsin is not seen in P23H mutant rhodopsin, indicating the mutant protein is misfolded and not able to respond to light. Adapted from (Sung, Schneider et al.
1991).
150
Figure 6: Cone cell death mechanisms: involvement of rod-derived trophic factor
Estimated total number of cones (A) and rods (B) in retinas from 5 week old RP mouse model cultured in DMEM alone (grey bars) or co-cultured with retinas from 8-day-old wild-type mice (black bars). Estimated number of cone cells increased when cultured with cells from rod containing retinas. Error bars represent standard deviation. *P <
0.0001. Adapted from (Mohand-Said, Deudon-Combe et al. 1998).
151
Figure 7: Gene therapy: The use of ribozymes to treat P23H rats
Measurements of the outer nuclear layer (ONL) thickness in P130 rats after late-stage ribozyme injection at P30 and P45. ONL thickness from WT nontransgenic rats (open squares) and the uninjected P23H transgenic (Tg) rats (filled squares) is plotted against age. P23H rats were injected subretinally with rAAV vectors carrying either hairpin (A) or hammerhead (B) ribozymes at either P30 (closed circles) or P45 (open triangles). For comparison, rats injected at P15 and analyzed at P130 are shown (filled triangle).
Adapted from (LaVail, Yasumura et al. 2000).
152
Figure 8: Comparison of rhodopsin mutant mice on wild-type (Tr ) and -
transducin null (Tr ) genetic backgrounds.
153
+/+ A. Time course of degeneration in Tr ( ) and Tr ( ) mice (left panel); retinal
morphology at 3-months (right panels). B. VPP, Tr ( ) and VPP, Tr ( )
degeneration kinetics (left); retinal morphology at 3-months (right). C. K296E, Tr ( )
and K296E, Tr ( ) degeneration kinetics (left); retinal morphology at 3-months
(right); D. P347S, Tr ( ) and P347S, Tr ( ) degeneration kinetics (left); retinal morphology at 6 months (right). Outer segment (OS), inner segment (IS), outer nuclear layer (ONL), inner nuclear layer (INL). n= number of animals of each genotype. Error bars represent standard deviation. Scale bar = 10 m.
154
Figure 9: Role of rhodopsin-arrestin complexes in P347S mutant rhodopsin degeneration.
The outer nuclear layer thickness of 4-month P347S mutant rhodopsin mice in the presence or absence of arrestin and transducin is plotted. The absence of -transducin, or arrestin and -transducin did not contribute to degeneration (compare 3 left columns).
Loss of -transducin (Tr -/-) in combination with the P347S rhodopsin mutation (P347S
+) produced a statistically significant (one-way ANOVA) loss of outer nuclear layer thickness (compare columns 4 and 5). The combined loss of -transducin and arrestin did not protect from degeneration ( compare columns 5 and 6). Error bars represent s.e.m.
155
Figure 10: -transducin stabilizes P347S metarhodopsin.
A. Retinal morphology of P347S (Pro347Ser) mutant rhodopsin mice on a Tr or
+/+ Tr genetic background reared in cyclic light or dark-reared at 4 months. P347S, Tra mice showed similar degrees of degeneration whether reared in cyclic light or darkness.
Dark rearing provided protection from degeneration, but only in P347S, Tr retinas. B.
Histogram comparing outer nuclear layer (ONL) thickness in cyclic-light (gray) and
dark-reared animals (black) on the Tr or Tr genetic background. Error bars show s.e.m. Scale bar = 10 m.
156
Figure 11: Comparison of A2E and A2E-precursor levels in wild-type and P347S mutant rhodopsin retinas
Comparison of A2E and A2E-precursor levels in A. 61-day old wild-type mice (gray) and
littermate P347S, Tr transgenic mice (black) show statistically significant differences
+/+ for four bis-retinoids. A2E was present at 4.8 versus 1.0 pmol/eye in P347S, Tr and
WT mice, respectively. B. Lipofuscin granule density. Four mice of each genotype were surveyed by measuring square microns of lipofuscin granule per square micron of
RPE cytoplasm. Error bars represent standard deviation. P347S mutant mice (black) had a higher density of granules than wild-type control mice (gray).
157
Figure 12: P347S mutant rhodopsin protein has decreased stability
The role of 11-cis-retinal in stabilizing P347S mutant rhodopsin. The outer nuclear layer
(ONL) thickness in retinas from 4-month old P347S mutant rhodopsin mice in the presence or absence of transducin and Rpe65 are plotted. The absence of -transducin alone or -transducin and Rpe65 did not cause degeneration (compare 3 left columns).
The presence of P347S mutant rhodopsin caused degeneration (compare first and fourth columns). Loss of -transducin in P347S mutant rhodopsin mice produced a statistically significant decrease in ONL thickness (compare fourth and fifth columns). The combined loss of -transducin and Rpe65 did not rescue degeneration (compare fifth and sixth columns). Error bars show s.e.m. *P < 0.005
158
Figure 13: Comparison of galectin-1 RNA levels in P347S mutant rhodopsin retinas compared to control
Characterization of galectin-1 RNA transcript in P347S mutant rhodopsin mice.
Quantitative RT-PCR indicates a significant increase in the average fold change in galectin-1 transcript in the P347S mutant retina compared to control at 3 and 4 months, calculated using the DDCT method. n = 3 for all groups. *p < 0.05. Error bars represent s.e.m.
159
Figure 14: Galectin-1 protein expression is increased in the P347S mutant rhodopsin retina compared to control
Characterization of galectin-1 protein levels in wild-type and P347S mutant rhodopsin retina. A. Representative immunoblots of wild-type and P347S mutant rhodopsin whole retina lysates. Each lane contains retina lysate from 3 animals. B. Average fold change of galectin-1 protein levels in P347S mutant rhodopsin retinas compared to control corrected for loading with -actin. Galectin-1 levels of expression are significantly increased in the P347S mutant retina compared to control at 2, 3 and 4 months. * p <0.05.
Error bars represent s.e.m.
160
Figure 15: Galectin-1 sugar-binding column identifies tenascin-R (TN-R) as a galectin-1 binding partner
Identification of galectin-1 binding partners in the retina. A. A sugar-binding column was completed to identify galectin-1 sugar-binding partners using whole retina lysates.
Nonspecific binding partners were eluted using sucrose. Specific binding partners were eluted using lactose. Sucrose and lactose lysates were run on a polyacrylamide gel which was then silver-stained. Bands at approximately 175 kDa and 150 kDa, were cut from the lactose lysate and subjected to mass-spectrophotometry. Mass-spectrophotometry results identified 10 unique peptides within tenascin-R. B. Confirmation of galectin-1 binding with tenascin-R. Total retina lysate was incubated with galectin-1 conjugated beads.
Nonspecific binding partners were eluted using sucrose. Specific binding partners were eluted using lactose. Elutions were then subjected to SDS-PAGE. The immunoblot was
161 probed using an antibody for tenascin-R. TN-R is present in the total lysate as well as the lactose lysate, but absent from the sucrose lysate, indicating that TN-R is a galectin-1 sugar-binding partner. The two bands in the TN-R immunoblot have been previously reported by the manufacturer, BD Biosciences. The topmost band is representative of
TN-R.
162
Figure 16: Mass-spectrophotometry identified peptides within TN-R Sequence
Mass-spectrophotometry results identify TN-R as galectin-1 sugar-binding partner. A sugar-binding column was completed to identify galectin-1 binding partners. Silver- stained bands appearing only in the galectin-1 specific rinse were cut and subjected to mass-spectrophotometry. Mass-spec identified 10 unique peptides, underlined in green, within the sequence for tenascin-R. Each green line indicates a unique peptide identified, duplicate lines indicated duplicate identifications of the same peptide.
163
Figure 17: Tenascin-R localizes to synaptic layers in the wild-type and galectin-1 knockout retina
Tenascin-R localization within the retina. Frozen retina sections were stained with antibodies for tenascin-R, then imaged using a fluorescent microscope. TN-R localizes within the outer plexiform (OPL) and inner plexiform (IPL) layers in both the wild-type and galectin-1 knockout mouse retinas.
164
Figure 18: Tenascin-R expression in wild-type and P347S mutant rhodopsin retinas
Characterization of tenascin-R protein levels in wild-type and P347S mutant rhodopsin retina. A. Representative immunoblots of wild-type and P347S mutant rhodopsin whole retina lysates. Each lane contains retina lysate from 3 animals. B. Average fold change of tenascin-R protein levels in P347S mutant rhodopsin retinas compared to control corrected for loading with -actin. A trend is appearing that tenascin-R levels are increasing at 2 and 3 months in the mutant retina compared to control. However, the change at 2 and 3 months is not statistically significant. Tenascin-R levels of expression are significantly increased in the P347S mutant rhodopsin retina compared to control at 4 months. * p <0.05. Error bars represent s.e.m.
165
Figure 19: Morphology of P347S mutant rhodopsin retinas in the presence and absence of galectin-1 at 2M
166
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse. A. Retinal morphology of wild-type, galectin-1 knockout, and P347S mutant rhodopsin mice on a wild-type or galectin-1 knockout background at 2 months. B. Outer nuclear layer (ONL) thickness, measured in number of nuclei per single ONL column, in retinas from 2-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. C. The ratio of ONL to INL thickness in retinas from 2-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. The absence of galectin-
1 alone (compare first and second columns) does not cause degeneration. The presence of P347S mutant rhodopsin is starting to cause degeneration (compare first and third columns), noted by the slightly decreased ONL thickness and ONL: INL of P347S mutant rhodopsin retinas compared to control (compare first and third columns). Loss galectin-1 in P347S mutant rhodopsin mice produced no difference in ONL thickness
(compare third and four columns). Error bars represent s.e.m. n = number of animals of each genotype.
167
Figure 20: Morphology of P347S mutant rhodopsin retinas in the presence and absence of galectin-1 at 4M
168
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse. A. Retinal morphology of wild-type, galectin-1 knockout, and P347S mutant rhodopsin mice on a wild-type or galectin-1 knockout background at 4 months. B. Outer nuclear layer (ONL) thickness, measured in number of nuclei per single ONL column, in retinas from 4-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. C. The ratio of ONL to INL thickness in retinas from 4-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. The absence of galectin-
1 alone (compare first and second columns) does not cause degeneration. The presence of P347S rhodopsin caused degeneration (compare first and third columns). Loss galectin-1 in P347S rhodopsin mice produced no difference in ONL thickness (compare third and four columns). Error bars represent s.e.m. n = number of animals of each genotype. *p < 0.05.
169
Figure 21: Morphology of P347S mutant rhodopsin retinas in the presence and absence of galectin-1 at 9M
170
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse. A. Retinal morphology of wild-type, galectin-1 knockout, and P347S mutant rhodopsin mice on a wild-type or galectin-1 knockout background at 9 months. B. Outer nuclear layer (ONL) thickness, measured in number of nuclei per single ONL column, in retinas from 9-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. C. The ratio of ONL to INL thickness in retinas from 9-month old P347S mutant rhodopsin mice in the presence or absence of galectin-1. The absence of galectin-
1 alone (compare first and second columns) does not cause degeneration. The presence of P347S rhodopsin caused degeneration (compare first and third columns). Loss galectin-1 in P347S rhodopsin mice produced no difference in ONL thickness (compare third and four columns). Error bars represent s.e.m. n = number of animals of each genotype. *p < 0.05.
171
Figure 22: Involvement of galectin-1 in retinal degeneration measured with electroretinogram at 4M
172
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse assessed with electroretinogram at 4 months.
A. Scotopic and photopic electroretinograms of wild-type (recorded previously), galectin-1 knockout; P347S rhodopsin mutant; and P347S rhodopsin mutant, galectin-1 knockout double mutant at 4 months.
B–D. Box plots of wave amplitudes. The bottom of the box indicates the 25th percentile, the top of the box indicates the 75th percentile, and the line in the middle of the box represents the average amplitude. The vertical lines indicate the minimum and maximum amplitudes.
B. Box plot of scotopic A wave amplitude. Galectin-1 knockout mice have normal A wave amplitudes of 180 +/- 20 V. At 4 months P347S rhodopsin mice also have relatively normal A wave amplitudes, average 160 +/- 9 V. Finally, at 4 months the loss of galectin-1 in P347S mutant rhodopsin mice produced a slight, but not statistically significant (one-way ANOVA), decrease in A wave amplitude, average 123 +/- 11 V.
C. Box plot of scotopic B wave amplitudes. Galectin-1 knockout mice have normal B wave amplitudes of 600 +/- 10 V. At 4 months P347S rhodopsin mice also have normal
B wave amplitudes, average 883 +/- 29 V. The loss of galectin-1 in P347S mutant rhodopsin mice produced no difference in B wave amplitude, average 800 +/- 85 V.
D. Box plot of photopic B wave amplitudes. Galectin-1 knockout mice have normal B wave amplitudes of 70 +/- 20 V. At 4 months P347S rhodopsin mice also have normal photopic B wave amplitudes, average 95 V. The loss of galectin-1 in P347S mutant
173 rhodopsin mice produced no difference in photopic B wave amplitude, average 95 +/- 3
V.
174
Figure 23: Time-to-Peak for 4M ERG
Box plots of duration to time-to-peak for ERG recorded at 4 months. The bottom of the box indicates the 25th percentile, the top of the box indicates the 75th percentile, and the line in the middle of the box indicates the median time-to-peak. The vertical lines indicate the minimum and maximum values.
A. Time-to-peak for scotopic A wave. All mice studied have normal time-to-peak for scotopic A wave (one-way ANOVA). Wild-type mice were recorded as part of a previous study. Galectin-1 knockout mice average time-to-peak is 18.7 msec. P347S rhodopsin mice average time-to-peak is 16.4 +/- 0.5 msec. Finally, the loss of galectin-1 in P347S mutant rhodopsin mice produced no difference in time-to-peak for scotopic A wave with an average of 17.9 +/- 0.4 msec.
B. Time to peak for the scotopic B wave. All mice studied have normal time-to-peak for scotopic B wave (one-way ANOVA). Wild-type mice were recorded as part of a previous study. Galectin-1 knockout mice average time-to-peak is 51.6 +/- 2.4 msec. P347S rhodopsin mice average time-to-peak is 47.9 +/- 1.0 msec. Finally, the loss of galectin-1
175 in P347S mutant rhodopsin mice produced no difference in time-to-peak for scotopic B wave with an average of 50.0 +/- 1.4 msec.
176
Figure 24: Involvement of galectin-1 in retinal degeneration measured with electroretinogram at 9M
177
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin retina assessed with electroretinogram at 9 months.
A. Scotopic and photopic electroretinograms of wild-type, galectin-1 knockout; P347S rhodopsin mutant; and P347S rhodopsin mutant, galectin-1 knockout double mutant mice at 9 months.
B–D. Box plots of wave amplitudes. The bottom of the box indicates the 25th percentile, the top of the box indicates the 75th percentile, and the line in the middle of the box represents the average amplitude. The vertical lines indicate the minimum and maximum amplitudes.
B. Box plot of scotopic A wave amplitudes. At 9 months, wild-type mice have A wave amplitudes of 200 V. Compared to wild-type, galectin-1 knockout mice have normal A wave amplitudes of 205 +/- 5 V. Compared to wild-type, P347S rhodopsin mice have decreased A wave amplitudes, average 63 +/- 10 V. The loss of galectin-1 in P347S mutant rhodopsin mice produced a slight, but not statistically significant (one-way
ANOVA), decrease in A wave amplitude, average 47 +/- 5 V, compared to P347S mutant rhodopsin alone.
C. Box plot of scotopic B wave amplitudes. At 9 months, wild-type mice have B wave amplitudes of 700 V. Compared to wild-type, galectin-1 knockout mice have normal A wave amplitudes of 743 +/- 8 V. Compared to wild-type, P347S rhodopsin mice have decreased B wave amplitudes, average 403 +/- 37 V. The loss of galectin-1 in P347S
178 mutant rhodopsin mice produced a slight but not statistically significant decrease in B wave amplitude, average 341 +/- 32 V, compared to P347S mutant rhodopsin alone.
D. Box plot of photopic B wave amplitudes. All mice studied have normal photopic B wave amplitudes, indicating that at 9 months the degeneration observed is specifically rod photoreceptor cell specific. Wild-type mice have a photopic B wave amplitude of
55V. Galectin-1 knockout mice have a photopic B wave amplitude of 65 V. P347S rhodopsin mice have a photopic B wave amplitude of 61 V +/- 3 V. Finally, compared to P347S mutant rhodopsin alone, the loss of galectin-1 in P347S mutant rhodopsin mice produced no difference in photopic B wave amplitude, average 69 V +/- 10 V.
179
Figure 25: Time-to-peak for 9M ERG
Box plots of duration to time-to-peak for ERG recorded at 9 months. The bottom of the box indicates the 25th percentile, the top of the box indicates the 75th percentile, and the line in the middle of the box indicates the median time-to-peak. The vertical lines indicate the minimum and maximum time-to-peak.
A. Time-to-peak for scotopic A wave. All mice studied have normal time-to-peak for scotopic A wave (one-way ANOVA). Wild-type mice scotopic A wave time-to-peak is
16.4 msec. Galectin-1 knockout mice average time-to-peak is 19.1 +/- 0.4 msec. P347S rhodopsin mice average time-to-peak is 15.4 +/- 0.8 msec. Finally, the loss of galectin-1 in P347S mutant rhodopsin mice produced no difference in time-to-peak for scotopic A wave with an average of 17.6 +/- 2.1 msec.
B. Time to peak for the scotopic B wave. All mice studied have normal time-to-peak for scotopic B wave (one-way ANOVA). Wild-type mice scotopic B wave time-to-peak is
43.7 msec. Galectin-1 knockout mice average time-to-peak is 50.0 +/- 0.8 msec. P347S rhodopsin mice average time-to-peak is 51.2 +/- 2.9 msec. Finally, the loss of galectin-1
180 in P347S mutant rhodopsin mice produced no difference in time-to-peak for scotopic B wave with an average of 51.1 +/- 2.2 msec.
181
Table 1: Compiled ERG Data
The role of galectin-1 in the retinal degeneration observed in the P347S mutant rhodopsin mouse assessed with electroretinogram at 4 and 9 months. The tables list the average amplitudes of scotopic A wave, B wave, and photopic B wave for each genotype plus or minus the standard error. Included in the tables is the number of animals assessed of each genotype at each age. (The wild-type mice included at the 4M age point were recorded as part of a previous study.)
182
Figure 26: Galectin-1 localization in wild-type control and galectin-1 knockout mice.
A. Immunoblot of wild-type, galectin-1 heterozygous (+/-) and galectin-1 homozygous knockout (-/-) retina lysates, demonstrating that the galectin-1 knockout mice are not expressing galectin-1 protein. B. Immunofluorescence using anti-galectin-1 antibodies
(green) on wild-type and galectin-1 knockout retina sections. The same signal is detected in wild-type and galectin-1 knockout indicating that for immunohistochemistry, the galectin-1 antibodies are not specific for galectin-1.
183
Figure 27: Verification of galectin-1 antibody for immunoprecipitation.
The commercially available goat polyclonal galectin-1 antibody, used in previous immunoblots, was tested for its ability to be used to immunoprecipitate (IP) galectin-1 using retina lysates from wild-type mice. Total lysate and IP lysate was then used for an immunoblot, which was then probed for galectin-1 using the homemade rabbit polyclonal galectin-1 antibody made by a collaborator. The goat polyclonal antibody does pull down galectin-1, so it will be used for IP followed by mass-spectrophotometry to identify galectin-1 binding partners in the retina.
184
Figure 28: Identification of sugar independent galectin-1 binding partners
Identification of sugar independent galectin-1 binding partners in the retina. A. An immuneprecipitation for galectin-1 using the goat polyclonal galectin-1 antibody characterized previously, was completed to identify galectin-1 binding partners in whole retina lysates. A band at approximately 30 kDa was cut and subjected to mass- spectrophotometry. B. Mass-spec identified both Thy-1 and actin as galectin-1 binding partners. Thy-1 and -actin sequences are shown, with peptides detected by mass-spec underlined in dark green. Four peptides were observed for Thy-1 and two peptides were observed for -actin.
185
Figure 29: Validation of galectin-1 binding with thy-1.
A. An immunoprecipitation for galectin-1 using the goat polyclonal galectin-1 antibody previously characterized was completed using whole retina lysates from wild-type adult mice. An immunoblot was then run and probed using a rabbit polyclonal antibody for thy-1. Thy-1 is detected in the input lysate but not within the galectin-1 IP, indicating that thy-1 is not a galectin-1 binding partner. The band at 25 kDa is believed to be light chain
IgG.
B. Total retina lysate was incubated with galectin-1 conjugated beads. Nonspecific binding partners were eluted using sucrose. Specific binding partners were eluted using lactose. An immunoblot was probed using an antibody for thy-1. Thy-1 is detected in the input lysate but not within the lactose elution, indicating that thy-1 is not a galectin-1 sugar-binding partner.
186
Figure 30: Confirmation of galectin-1 sugar-binding with neuropilin-1 (NP-1) and tenascin-C (TN-C)
Confirmation of galectin-1 sugar-binding with neuropilin-1 (NP-1) and tenascin-C (TN-
C). Total retina lysate was incubated with galectin-1 conjugated beads. Nonspecific binding partners were eluted using sucrose. Specific binding partners were eluted using lactose. Sucrose and lactose elutions were used for immunoblots. Immunoblots were probed using antibodies for NP-1 (A) and TN-C (B).
A. NP-1 was detected at the predicted molecular weight of 130 kDa within input total retina lysate, and a truncated form, approximately 75 kDa, was detected within the lactose elution.
B. TN-C was very weakly detected within input total retina lysate and not detected within either the sucrose or lactose elution of wild-type or P347S mutant rhodopsin retinas.
187
Figure 31: Dendrogram showing hierarchical clustering analysis of mutant and control retinas.
Dendrogram showing unsupervised hierarchical clustering analysis of P347S mutant (M) rhodopsin and wild-type control (C) retinal samples. The individual samples are clustered in branches of the dendrogram based on overall similarity in patterns of gene expression.
Generally, P347S mutant rhodopsin retinas (M) cluster together and control retinas (C) cluster together. The three replicates for each genotype at each timepoint are shown.
188
Table 2: Number of differentially expressed transcripts in P347S mutant rhodopsin retina compared to control
Number of differentially expressed transcripts at all time points studied. Genes were considered differentially expressed when there is a difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
189
Figure 32: Categories of differentially expressed transcripts in P347S mutant rhodopsin retina compared to control
Categories of differentially expressed transcripts at all time points studied. Genes were considered differentially expressed when there is a difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3). The miscellaneous group includes transcripts from adhesion molecule, CBP:C-Type lectin, CBP:I-Type lectin, cytokine, galectin, glycoprotein, notch pathway, and nuclear sugar categories.
190
Table 3: Categories of differentially expressed transcripts in P347S mutant rhodopsin retina over time
The table indicates the number of differentially expressed transcripts in each category at each time point studied. Genes were considered differentially expressed when there is a difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3). A value of zero indicates that there were not any differentially expressed transcripts for the category at the time point.
191
Table 4: Differentially expressed transcripts involved in glycan degradation
Fold Change Gene Title Common Name 1M 2M 3M 4M AGA Aspartylglucosaminidase 1.4 1.4 ARSA Arylsulfatase A 1.4 ARSB Arylsulfatase B 1.7 GLB1 Beta-Galactosidase (lactase) -2.7 -2.7 -3.0 GUSB Glucuronidase beta 1.4 HYAL1 Hyaluronoglucosaminidase 1 1.4 LAMP2 Lysosomal membrane glycoprotein 2 -1.7 -1.4 MAN2A1 Mannosidase 2 alpha 1 1.4 1.5 MAN2A2 Mannosidase 2 alpha 2 -2.5 -2.2 SULF1 Sulfatase 1 1.6 1.4 1.7 1.9
Transcripts involved in glycan degradation that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
192
Table 5: Growth factor and receptor differentially expressed transcripts
Fold Change Gene Common Name 1M 2M 3M 4M Title ACVR1 Activin A receptor, type 1 1.4 1.6 1.5 AREG Amphiregulin 1.4 BMP3 Bone morphogenetic protein 3 1.6 BMPR1A Bone morphogenetic protein receptor type 1A -1.4 BMPR1B Bone morphogenetic protein receptor type 1B 1.4 1.6 BTC Betacellulin epidermal growth factor family 4.2 2.8 2.6 2.6 CSF1 Colony stimulating factor 1 1.3 1.3 1.3 CSF1R Colony stimulating factor 1 receptor 1.7 2.0 2.2 EGFR Epidermal growth factor receptor isoform 1 1.6 1.9 FGF1 Fibroblast growth factor 1 1.7 FGF13 Fibroblast growth factor 13 1.4 FGF2 Fibroblast growth factor 2 1.3 FGF5 Fibroblast growth factor 5 1.4 FIBP FGF intracellular binding protein -1.4 FIGF C-fos induced growth factor -1.3 FST Follistatin 1.5 FZD1 Frizzled 1 1.6 1.5 FZD7 Frizzled 7 1.9 2.1 HBEGF Heparin-binding EGF-like growth factor 1.3 IGF1 Insulin-like growth factor 1 1.3 1.3 IGF2 Insulin-like growth factor 2 1.3 IGFBP5 Insulin-like growth factor binding protein 5 1.8 1.5 MAGI2 Membrane associated guanylate kinase WW and -1.4 -1.4 PDZ MET Met proto-oncogene 2.0 2.0 ODZ4 Odd Oz/ten-m homolog 4 1.4 PDGFC PDGFC 1.3 SHH Sonic hedgehog 1.4 SMO Smoothened 1.4 1.4 1.4 TGFB2 Transforming growth factor beta 2 1.3 1.4 TGFBR2 Transforming growth factor beta receptor II 1.9 2.0 TGFBR3 Transforming growth factor beta receptor III 1.7 1.5
Growth factor and receptor transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time
193 point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
194
Table 6: Glycosyltransferases differentially expressed transcripts
Fold Change Gene Title Common Name 1M 2M 3M 4M B3GALT1 UDP-Gal:betaGlcNAc beta 1,3- 1.3 1.5 1.6 galactosyltransferase, polypeptide 1 B3GAT2 Beta-1,3-glucuronyltransferase 2 1.8 1.9 (glucuronosyltransferase S) B3GNT2 UDP-GlcNAc:betaGal beta-1,3-N- 1.5 1.4 1.9 acetylglucosaminyltransferase 2 B4GALT1 UDP-Gal:betaGlcNAc beta 1,4- -1.3 -1.5 galactosyltransferase, polypeptide 1 B4GALT4 UDP-Gal:betaGlcNAc beta 1,4- 1.4 galactosyltransferase, polypeptide 4 B4GALT5 UDP-Gal:betaGlcNAc beta 1,4- 1.4 1.8 galactosyltransferase, polypeptide 5 B4GALT6 UDP-Gal:betaGlcNAc beta 1,4- 1.7 galactosyltransferase, polypeptide 6 CHST10 Carbohydrate sulfotransferase 10 1.3 CHST11 Carbohydrate sulfotransferase 11 1.3 CHST2 Carbohydrate sulfotransferase 2 1.6 1.5 CHST3 Carbohydrate (chondroitin 6/keratan) -1.5 -1.6 sulfotransferase 3 CHSY1 Carbohydrate (chondroitin) synthase 1 -1.4 CSGALNACT1 CSGalNAcT1/ChGalNAcT1 1.4 DSEL Dermatan sulfate epimerase-like; NCAG1 similar 1.4 to sulfotransferase EXT1 Exostosin 1 -1.8 FUT8 Fucosyltransferase 8 1.4 FUT9 Fucosyltransferase 9 1.3 1.4 GALNT13 UDP-N-acetyl-alpha-D-galactosamine:polypeptide 1.6 1.6 N-acetylgalactosaminyltransferase 13 GALNT4 UDP-N-acetyl-alpha-D-galactosamine:polypeptide -1.4 -1.6 N-acetylgalactosaminyltransferase 4 GALNT7 UDP-N-acetyl-alpha-D-galactosamine: -1.5 polypeptide N-acetylgalactosaminyltransferase 7 GCNT2 Glucosaminyl (N-acetyl) transferase 2, I-branching 1.5 enzyme HS3ST1 Heparan sulfate (glucosamine) 3-O- 1.3 1.3 sulfotransferase 1 HS3ST5 Heparan sulfate (glucosamine) 3-O- 1.4 sulfotransferase 5 HS6ST2 Heparan sulfate 6-O-sulfotransferase 2 1.4 1.7 MGAT4C Mannoside acetylglucosaminyltransferase 4c 1.3 1.3 MGAT5 Mannoside acetylglucosaminyltransferase 5 -1.4
195
POFUT1 Protein O-fucosyltransferase 1 -1.5 POMT2 Protein-O-mannosyltransferase 2 -1.4 ST3GAL1 ST3 beta-galactoside alpha-2,3-sialyltransferase 1 -1.5 ST3GAL6 ST3 beta-galactoside alpha-2,3-sialyltransferase 6 1.5 ST6GAL1 Beta galactoside alpha 2,6 sialyltransferase 1 1.4 ST6GALNAC2 ST6 (alpha-N-acetyl-neuraminyl-2,3-beta- -1.4 galactosyl-1,3)-N-acetylgalactosaminide alpha- 2,6-sialyltransferase 2 ST6GALNAC4 ST6 (alpha-N-acetyl-neuraminyl-2,3-beta- 1.6 galactosyl-1,3)-N-acetylgalactosaminide alpha- 2,6-sialyltransferase 4 ST8SIA1 ST8 alpha-N-acetyl-neuraminide alpha-2,8- -1.6 sialyltransferase 1 ST8SIA3 ST8 alpha-N-acetyl-neuraminide alpha-2,8- 1.3 1.4 1.6 sialyltransferase 3 UGCG UDP-glucose ceramide glucosyltransferase 1.4 UST Uronyl-2-sulfotransferase 1.4 XYLT1 Xylosyltransferase I -1.5 -1.5
Glycosyltransferases transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
196
Table 7: Mouse Housekeeping differentially expressed transcripts
Fold Change Gene Title Common Name 1M 2M 3M 4M ANKRD17 Ankyrin repeat domain 17 -1.4 CKS2 CDC28 protein kinase regulatory subunit -1.9 2 COPG Coatomer protein complex, subunit 1.4 gamma COX18 COX18 cytochrome c oxidase assembly -1.5 homolog DHRS1 Dehydrogenase/reductase (SDR family) 1.4 member 1 DLG1 Discs, large homolog 1 (Drosophila) -1.3 FRYL Furry homolog-like isoform 1 -1.4 GOLM1 Golgi membrane protein 1 1.6 1.5 HBXIP Hepatitis B virus x interacting protein 1.4 PRDM1 PR domain containing 1 with ZNF domain -1.6 -1.4 S100A10 S100 calcium binding protein A10 1.7 XPO7 Exportin 7 -1.4
Mouse housekeeping transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
197
Table 8: Chemokine differentially expressed transcripts
Fold Change Gene Title Common Name 1M 2M 3M 4M CCL2 Chemokine (C-C motif) ligand 12 1.8 CCR5 Chemokine (C-C motif) receptor 5 1.3 1.5 CX3CL1 Chemokine (C-X3-C motif) ligand 1 1.4 CX3CR1 Chemokine (C-X3-C) receptor 1 2.1 1.9 2.2 CXCL12 Chemokine (C-X-C motif) ligand 12; stromal cell 2.0 1.8 derived factor 1 isoform alpha KIT C-kit 1.4 PTPRT Protein tyrosine phosphatase receptor type T 1.6
Chemokine transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
198
Table 9: Interleukin and receptors differentially expressed transcripts
Fold Change Gene Title Common Name 1M 2M 3M 4M IL10RB Interleukin 10 receptor beta 1.6 1.5 IL13RA2 Interleukin 13 receptor alpha 2 1.8 IL17RD Interleukin 17 receptor D 1.3 IL1R1 Interleukin 1 receptor type I 1.6 1.4 IL33 Interleukin 33 1.7 IL6ST Interleukin-6 (gp130); interleukin 6 signal 1.5 1.5 transducer
Interleukin and receptor transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
199
Table 10: Additional differentially expressed transcripts
Fold Change Category Gene Common Name 1M 2M 3M 4M Title Adhesion Molecule CD43 CD34 antigen 1.6 CD48 CD48 antigen; BCM1 1.8 1.7 GLYCAM1 Glycosylation dependent cell 3.9 5.7 4.0 adhesion molecule SELPLG Selectin platelet (p-selectin) 1.6 1.3 1.6 ligand CBP:C-Type Lectin CHODL Chondrolectin; Layilin homolog 1.7 CLEC2D Clec2d 2.0 CLEC7A Dendritic cell-associated C-type 2.0 2.6 lectin 1; Dectin-1 FREM1 Fras1 related extracellular matrix 1.4 1.3 protein 1 NCAN Neurocan 1.3 PLA2R1 Phospholipase A2 receptor 1 -2.8 -1.7 CBP:I-Type Lectin CD33 Siglec-H 1.4 1.4 CD83 CD83 1.4 ICAM1 Intercellular adhesion molecule 2.6 3.1 VCAM1 Vascular cell adhesion molecule 1 1.8 2.5 Cytokine GFRA2 Glial cell line derived 1.4 neurotrophic factor family receptor alpha 2 IFNAR2 Interferon (alpha and beta) 1.4 1.6 receptor 2 Galectin Gal-9 Lectin galactose binding soluble 1.9 9; Galectin 9 Glycoproteins EMR1 EGF-like module containing, 1.9 1.8 mucin-like, hormone receptor- like sequence 1 MCAM Melanoma cell adhesion 1.6 molecule MUC1 Mucin 1 transmembrane 1.4 OVGP1 Oviductal glycoprotein 1 -1.5 -1.4 Notch pathway DLL1 Delta-like 1 1.4 HES1 Hairy and enhancer of split 1 1.3 1.3 JAG1 Jagged1 -1.4 -1.4 NOTCH1 Notch gene homolog 1 1.6 1.8 NOTCH2 Notch gene homolog 2 1.5 1.8 Nuc. Sugar PAPSS2 PAPS synthetase-2; 3'- -2.0 phosphoadenosine 5'- phosphosulfate synthase 2
200
PGM2 Phosphoglucomutase 2 1.4 PGM2L1 Phosphoglucomutase 2-like 1 1.4 SLC35F1 Solute carrier family 35 member 1.4 1.5 F1 Proteoglycans AGRN Agrin 1.4 1.4 GPC3 Glypican-3, OCI-5 1.8 PTPRZ1 Phosphacan, protein tyrosine 1.4 phosphatase receptor type Z SPOCK1 Testican-1, sparc/osteonectin 1.4 cwcv and kazal-like domains SRGN Serglycin 1.8
Additional transcripts that are differentially expressed in P347S mutant rhodopsin retinas compared to control. Transcripts are sorted by category. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
201
Table 11: Trends of transcripts differentially expressed at 1 and 2 Months
Fold Change Category Gene Title Common Name 1M 2M 3M 4M Adhesion GLYCAM1 Glycosylation dependent cell 3.9 5.7 4.0 Molecule adhesion molecule SELPLG Selectin platelet (p-selectin) 1.6 1.3 1.6 ligand CBP:C-Type Lectin CLEC7A Dendritic cell-associated C-type 2.0 2.6 lectin 1; Dectin-1 PLA2R1 Phospholipase A2 receptor 1 -2.8 -1.7 Chemokine CX3CR1 Chemokine (C-X3-C) receptor 1 2.1 1.9 2.2 Glycan GLB1 Beta-Galactosidase (lactase) -2.7 -2.7 -3.0 Degradation LAMP2 Lysosomal membrane -1.7 -1.4 glycoprotein 2 MAN2A2 Mannosidase 2 alpha 2 -2.5 -2.2 SULF1 Sulfatase 1 1.6 1.4 1.7 1.9 Glycosyltrans- B3GALT1 UDP-Gal:betaGlcNAc beta 1,3- 1.3 1.5 1.6 ferases galactosyltransferase, polypeptide 1 B3GNT2 UDP-GlcNAc:betaGal beta-1,3- 1.5 1.4 1.9 N-acetylglucosaminyltransferase 2 B4GALT1 UDP-Gal:betaGlcNAc beta 1,4- -1.3 -1.5 galactosyltransferase, polypeptide 1 CHST3 Carbohydrate (chondroitin -1.5 -1.6 6/keratan) sulfotransferase 3 GALNT4 UDP-N-acetyl-alpha-D- -1.4 -1.6 galactosamine:polypeptide N- acetylgalactosaminyltransferase 4 POMT2 Protein-O-mannosyltransferase -1.4 2 ST6GALNA ST6 (alpha-N-acetyl-neuraminyl- -1.4 C2 2,3-beta-galactosyl-1,3)-N- acetylgalactosaminide alpha- 2,6-sialyltransferase 2 ST8SIA3 ST8 alpha-N-acetyl-neuraminide 1.3 1.4 1.6 alpha-2,8-sialyltransferase 3 XYLT1 Xylosyltransferase I -1.5 -1.5 Glycoproteins MCAM Melanoma cell adhesion 1.6 molecule OVGP1 Oviductal glycoprotein 1 -1.5 -1.4
202
Growth Factors & ACVR1 Activin A receptor, type 1 1.4 1.6 1.5 Receptors BTC Betacellulin epidermal growth 4.2 2.8 2.6 2.6 factor family CSF1 Colony stimulating factor 1 1.3 1.3 1.3 CSF1R Colony stimulating factor 1 1.7 2.0 2.2 receptor FIBP FGF intracellular binding protein -1.4 FIGF C-fos induced growth factor -1.3 MAGI2 Membrane associated guanylate -1.4 -1.4 kinase WW and PDZ SMO Smoothened 1.4 1.4 1.4 Interleukin & IL6ST Interleukin 6 signal transducer 1.4 1.5 1.5 Receptors Mouse ANKRD17 Ankyrin repeat domain 17 -1.4 Housekeeping COX18 COX18 cytochrome c oxidase -1.5 assembly homolog HBXIP Hepatitis B virus x interacting 1.4 protein PRDM1 PR domain containing 1 with -1.6 -1.4 ZNF domain Notch pathway JAG1 Jagged1 -1.4 -1.4 Nuc. Sugar PAPSS2 PAPS synthetase-2; 3'- -2.0 phosphoadenosine 5'- phosphosulfate synthase 2
Transcripts differentially expressed in P347S mutant rhodopsin retinas compared to control retinas at 1 and 2 months. Transcripts are sorted by category. Magnitude fold change is listed for each time point assayed if the adjusted p value was < 0.05. Genes were considered differentially expressed when there is a net difference of > 1.3-fold between the geometric mean signal of the wild-type control group (n=3) and the P347S mutant rhodopsin group (n=3).
203
Figure 33: Comparison of microarray and quantitative RT-PCR data for CX3 chemokine receptor (CX3CR1)
Comparison of microarray and quantitative RT-PCR data for CX3CR1. The glycogene microarray indicates a significant increase in the average fold change in CX3CR1 transcript in the P347S mutant rhodopsin retina compared to control at 2, 3 and 4 months.
Quantitative RT-PCR indicates a significant increase in the average fold change in
CX3CR1 transcript in the P347S mutant rhodopsin retina compared to control at 3 months, calculated using the DDCT method. n = 3 for all groups. * p <0.05. Error bars represent standard error from the mean.
204
Figure 34: Comparison of microarray and quantitative RT-PCR data for colony stimulating factor 1 (CSF1)
Comparison of microarray and quantitative RT-PCR data for CSF1. The glycogene microarray indicates a significant increase in the average fold change in CSF1 transcript in the P347S mutant rhodopsin retina compared to control at 2, 3 and 4 months.
Quantitative RT-PCR indicates a significant increase in the average fold change in CSF1 transcript in the P347S mutant rhodopsin retina compared to control at 1, 3 and 4 months, calculated using the DDCT method. Note that both qRT-PCR and the microarray indicate a similar trend of CSF1 upregulation in the P347S mutant rhodopsin retina compared to control. n = 3 for all groups. * p <0.05. Error bars represent standard error from the mean.
205
Figure 35: Localization of colony stimulating factor 1 (CSF1) in wild-type and
P347S mutant rhodopsin retinas
CSF1 localization in wild-type control (WT) and P347S mutant rhodopsin retinas at 4 months of age. The DAB method for immunohistochemistry was used on paraffin embedded sections to localize CSF1 in the retina. CSF1 is localized to the inner segment
(IS), inner and outer plexiform layers as well as the ganglion cell layer (GCL) in both the wild-type control and P347S mutant rhodopsin retina. To control for nonspecific binding of the secondary antibody, a negative control was included on which only the primary antibody was omitted.
206
Figure 36: Comparison of microarray and quantitative RT-PCR data for colony stimulating factor 1 receptor (CSF1R)
Comparison of microarray and quantitative RT-PCR data for CSF1R. The glycogene microarray indicates a significant increase in the average fold change in CSF1R transcript in the P347S mutant rhodopsin retina compared to control at 2, 3 and 4 months.
Quantitative RT-PCR indicates a significant increase in the average fold change in
CSF1R transcript in the P347S mutant rhodopsin retina compared to control at 1, 3 and 4 months, calculated using the DDCT method. Note that both qRT-PCR and the microarray indicate a similar trend of CSF1R upregulation in the P347S mutant rhodopsin retina compared to control. n = 3 for all groups. * p <0.05. Error bars represent standard error from the mean.
207
Figure 37: Localization of colony stimulating factor 1 receptor (CSF1R) in wild-type and P347S mutant rhodopsin retina
CSF1R localization in wild-type control (WT) and P347S mutant rhodopsin retinas at 4 months of age. The DAB method for immunohistochemistry was used on paraffin embedded sections to localize CSF1R in the retina. CSF1R is localized to the inner segment (IS), inner and outer plexiform layers as well as the ganglion cell layer (GCL) in both the wild-type control and P347S mutant rhodopsin retina. To control for nonspecific binding of the secondary antibody, a negative control was included on which only the primary antibody was omitted.
208
Figure 38: Comparison of microarray and quantitative RT-PCR data for interleukin 6 signal transducer (IL6ST)
Comparison of microarray and quantitative RT-PCR data for IL6ST. The glycogene microarray indicates a significant increase in the average fold change in IL6ST transcript in the P347S mutant rhodopsin retina compared to control at 1, 3 and 4 months.
Quantitative RT-PCR does not detect a significant increase in IL6ST transcript at any time point studied, calculated using the DDCT method. n = 3 for all groups. * p <0.05.
Error bars represent standard error from the mean.
209
BIBLIOGRAPHY
Alfinito, P. D. and E. Townes-Anderson (2002). "Activation of mislocalized opsin kills rod cells: a novel mechanism for rod cell death in retinal disease." Proc Natl Acad Sci U S A 99(8): 5655-5660. Alloway, P. G. and P. J. Dolph (1999). "A role for the light-dependent phosphorylation of visual arrestin." Proc Natl Acad Sci U S A 96(11): 6072-6077. Alloway, P. G., L. Howard, et al. (2000). "The formation of stable rhodopsin-arrestin complexes induces apoptosis and photoreceptor cell degeneration." Neuron 28(1): 129-138. Ambati, J., A. Anand, et al. (2003). "An animal model of age-related macular degeneration in senescent Ccl-2- or Ccr-2-deficient mice." Nat Med 9(11): 1390- 1397. Anderson, K. D., N. Panayotatos, et al. (1996). "Ciliary neurotrophic factor protects striatal output neurons in an animal model of Huntington disease." Proc Natl Acad Sci U S A 93(14): 7346-7351. Apostolova, I., A. Irintchev, et al. (2006). "Tenascin-R restricts posttraumatic remodeling of motoneuron innervation and functional recovery after spinal cord injury in adult mice." J Neurosci 26(30): 7849-7859. Apweiler, R., H. Hermjakob, et al. (1999). "On the frequency of protein glycosylation, as deduced from analysis of the SWISS-PROT database." Biochim Biophys Acta 1473(1): 4-8. Ariga, T., T. Miyatake, et al. (2010). "Role of proteoglycans and glycosaminoglycans in the pathogenesis of Alzheimer's disease and related disorders: amyloidogenesis and therapeutic strategies--a review." J Neurosci Res 88(11): 2303-2315. Avni, O., Z. Pur, et al. (1998). "Complement receptor 3 of macrophages is associated with galectin-1-like protein." J Immunol 160(12): 6151-6158. Bartsch, U., P. Pesheva, et al. (1993). "Expression of janusin (J1-160/180) in the retina and optic nerve of the developing and adult mouse." Glia 9(1): 57-69. Becker, C. G., T. Becker, et al. (1999). "Tenascin-R inhibits the growth of optic fibers in vitro but is rapidly eliminated during nerve regeneration in the salamander Pleurodeles waltl." J Neurosci 19(2): 813-827. Becker, C. G., J. Schweitzer, et al. (2003). "Tenascin-R as a repellent guidance molecule for developing optic axons in zebrafish." J Neurosci 23(15): 6232-6237. Becker, T., B. Anliker, et al. (2000). "Tenascin-R inhibits regrowth of optic fibers in vitro and persists in the optic nerve of mice after injury." Glia 29(4): 330-346. Berezovskaya, O., D. Maysinger, et al. (1995). "The hematopoietic cytokine, colony- stimulating factor 1, is also a growth factor in the CNS: congenital absence of CSF-1 in mice results in abnormal microglial response and increased neuron vulnerability to injury." Int J Dev Neurosci 13(3-4): 285-299. Berezovskaya, O., D. Maysinger, et al. (1996). "Colony stimulating factor-1 potentiates neuronal survival in cerebral cortex ischemic lesion." Acta Neuropathol 92(5): 479-486. Berson, E. L., B. Rosner, et al. (1993). "A randomized trial of vitamin A and vitamin E supplementation for retinitis pigmentosa." Arch Ophthalmol 111(6): 761-772.
210
Berson, E. L., B. Rosner, et al. (1993). "Vitamin A supplementation for retinitis pigmentosa." Arch Ophthalmol 111(11): 1456-1459. Bizzoca, A., P. Corsi, et al. (2009). "The mouse F3/contactin glycoprotein: structural features, functional properties and developmental significance of its regulated expression." Cell Adh Migr 3(1): 53-63. Black, J. W. and N. P. Shankley (1995). "Drug receptors. Inverse agonists exposed." Nature 374(6519): 214-215. Boje, K. M. and P. K. Arora (1992). "Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death." Brain Res 587(2): 250-256. Bond, R. A., P. Leff, et al. (1995). "Physiological effects of inverse agonists in transgenic mice with myocardial overexpression of the beta 2-adrenoceptor." Nature 374(6519): 272-276. Brill, E., K. M. Malanson, et al. (2007). "A novel form of transducin-dependent retinal degeneration: accelerated retinal degeneration in the absence of rod transducin." Invest Ophthalmol Vis Sci 48(12): 5445-5453. Bui, T. V., Y. Han, et al. (2006). "Characterization of native retinal fluorophores involved in biosynthesis of A2E and lipofuscin-associated retinopathies." J Biol Chem 281(26): 18112-18119. Bukalo, O., M. Schachner, et al. (2001). "Modification of extracellular matrix by enzymatic removal of chondroitin sulfate and by lack of tenascin-R differentially affects several forms of synaptic plasticity in the hippocampus." Neuroscience 104(2): 359-369. Calvert, P. D., N. V. Krasnoperova, et al. (2000). "Phototransduction in transgenic mice after targeted deletion of the rod transducin subunit." Proc. Natl. Acad. Sci. USA 97(25): 13913-13918. Camby, I., M. Le Mercier, et al. (2006). "Galectin-1: a small protein with major functions." Glycobiology 16(11): 137R-157R. Canola, K., B. Angenieux, et al. (2007). "Retinal stem cells transplanted into models of late stages of retinitis pigmentosa preferentially adopt a glial or a retinal ganglion cell fate." Invest Ophthalmol Vis Sci 48(1): 446-454. Cardona, A. E., E. P. Pioro, et al. (2006). "Control of microglial neurotoxicity by the fractalkine receptor." Nat Neurosci 9(7): 917-924. Carwile, M. E., R. B. Culbert, et al. (1998). "Rod outer segment maintenance is enhanced in the presence of bFGF, CNTF and GDNF." Exp Eye Res 66(6): 791-805. Cayouette, M., D. Behn, et al. (1998). "Intraocular gene transfer of ciliary neurotrophic factor prevents death and increases responsiveness of rod photoreceptors in the retinal degeneration slow mouse." J Neurosci 18(22): 9282-9293. Celio, M. R., R. Spreafico, et al. (1998). "Perineuronal nets: past and present." Trends Neurosci 21(12): 510-515. Chacko, D. M., J. A. Rogers, et al. (2000). "Survival and differentiation of cultured retinal progenitors transplanted in the subretinal space of the rat." Biochem Biophys Res Commun 268(3): 842-846. Chadderton, N., S. Millington-Ward, et al. (2009). "Improved Retinal Function in a Mouse Model of Dominant Retinitis Pigmentosa Following AAV-delivered Gene Therapy." Mol Ther.
211
Chalmel, F., T. Leveillard, et al. (2007). "Rod-derived Cone Viability Factor-2 is a novel bifunctional-thioredoxin-like protein with therapeutic potential." BMC Mol Biol 8: 74. Chang-Hong, R., M. Wada, et al. (2005). "Neuroprotective effect of oxidized galectin-1 in a transgenic mouse model of amyotrophic lateral sclerosis." Exp Neurol 194(1): 203-211. Chang, G. Q., Y. Hao, et al. (1993). "Apoptosis: final common pathway of photoreceptor death in rd, rds, and rhodopsin mutant mice." Neuron 11(4): 595-605. Cheema, S. S., L. Richards, et al. (1994). "Leukemia inhibitory factor prevents the death of axotomised sensory neurons in the dorsal root ganglia of the neonatal rat." J Neurosci Res 37(2): 213-218. Cheema, S. S., L. J. Richards, et al. (1994). "Leukaemia inhibitory factor rescues motoneurones from axotomy-induced cell death." Neuroreport 5(8): 989-992. Chen, C. K., M. E. Burns, et al. (1999). "Abnormal photoresponses and light-induced apoptosis in rods lacking rhodopsin kinase." Proc Natl Acad Sci U S A 96(7): 3718-3722. Chen, J., G. Shi, et al. (2006). "Stable rhodopsin/arrestin complex leads to retinal degeneration in a transgenic mouse model of autosomal dominant retinitis pigmentosa." J Neurosci 26(46): 11929-11937. Chen, Z., D. J. Dupre, et al. (2002). "Agonist-induced internalization of the platelet- activating factor receptor is dependent on arrestins but independent of G-protein activation. Role of the C terminus and the (D/N)PXXY motif." J Biol Chem 277(9): 7356-7362. Cho, M. and R. D. Cummings (1995). "Galectin-1, a beta-galactoside-binding lectin in Chinese hamster ovary cells. I. Physical and chemical characterization." J Biol Chem 270(10): 5198-5206. Cho, M. and R. D. Cummings (1995). "Galectin-1, a beta-galactoside-binding lectin in Chinese hamster ovary cells. II. Localization and biosynthesis." J Biol Chem 270(10): 5207-5212. Cho, M. and R. D. Cummings (1996). "Characterization of monomeric forms of galectin- 1 generated by site-directed mutagenesis." Biochemistry 35(40): 13081-13088. Chong, N. H., R. A. Alexander, et al. (1999). "Repeated injections of a ciliary neurotrophic factor analogue leading to long-term photoreceptor survival in hereditary retinal degeneration." Invest Ophthalmol Vis Sci 40(6): 1298-1305. Cobbs, W. H., A. E. Barkdoll, 3rd, et al. (1985). "Cyclic GMP increases photocurrent and light sensitivity of retinal cones." Nature 317(6032): 64-66. Cohen, G. B., D. D. Oprian, et al. (1992). "Mechanism of activation and inactivation of opsin: role of Glu113 and Lys296." Biochemistry 31(50): 12592-12601. Coles, B. L., B. Angenieux, et al. (2004). "Facile isolation and the characterization of human retinal stem cells." Proc Natl Acad Sci U S A 101(44): 15772-15777. Collins, B. E., O. Blixt, et al. (2004). "Masking of CD22 by cis ligands does not prevent redistribution of CD22 to sites of cell contact." Proc Natl Acad Sci U S A 101(16): 6104-6109. Combadiere, C., C. Feumi, et al. (2007). "CX3CR1-dependent subretinal microglia cell accumulation is associated with cardinal features of age-related macular degeneration." J Clin Invest 117(10): 2920-2928.
212
Comelli, E. M., S. R. Head, et al. (2006). "A focused microarray approach to functional glycomics: transcriptional regulation of the glycome." Glycobiology 16(2): 117- 131. Correa, S. G., C. E. Sotomayor, et al. (2003). "Opposite effects of galectin-1 on alternative metabolic pathways of L-arginine in resident, inflammatory, and activated macrophages." Glycobiology 13(2): 119-128. Craig, S. E., R. Thummel, et al. (2010). "The zebrafish galectin Drgal1-l2 is expressed by proliferating Muller glia and photoreceptor progenitors and regulates the regeneration of rod photoreceptors." Invest Ophthalmol Vis Sci 51(6): 3244-3252. Daiger, S. P. (1996-2007). "RetNet: Retinal Information Network." from http://www.sph.uth.tmc.edu/Retnet/. Danciger, M., J. Lyon, et al. (2004). "New retinal light damage QTL in mice with the light-sensitive RPE65 LEU variant." Mamm Genome 15(4): 277-283. Denes, A., S. Ferenczi, et al. (2008). "Role of CX3CR1 (fractalkine receptor) in brain damage and inflammation induced by focal cerebral ischemia in mouse." J Cereb Blood Flow Metab 28(10): 1707-1721. Deretic, D. (2006). "A role for rhodopsin in a signal transduction cascade that regulates membrane trafficking and photoreceptor polarity." Vision Res 46: 4427-4433. Deretic, D., S. Schmerl, et al. (1998). "Regulation of sorting and post-Golgi trafficking of rhodopsin by its C-terminal sequence QVS(A)PA." Proc Natl Acad Sci U S A 95(18): 10620-10625. Diskin, S., J. Kumar, et al. (2006). "Detection of differentially expressed glycogenes in trabecular meshwork of eyes with primary open-angle glaucoma." Invest Ophthalmol Vis Sci 47(4): 1491-1499. Dryja, T. P., L. B. Hahn, et al. (1996). "Missense mutation in the gene encoding the alpha subunit of rod transducin in the Nougaret form of congenital stationary night blindness." Nat Genet 13(3): 358-360. Dryja, T. P., T. L. McGee, et al. (1990). "Mutations within the rhodopsin gene in patients with autosomal dominant retinitis pigmentosa." N Engl J Med 323(19): 1302- 1307. Dryja, T. P., T. L. McGee, et al. (1990). "A point mutation of the rhodopsin gene in one form of retinitis pigmentosa." Nature 343(6256): 364-366. Ebrey, T. and Y. Koutalos (2001). "Vertebrate photoreceptors." Prog Retin Eye Res 20(1): 49-94. Fesenko, E. E., S. S. Kolesnikov, et al. (1985). "Induction by cyclic GMP of cationic conductance in plasma membrane of retinal rod outer segment." Nature 313(6000): 310-313. Fishkin, N. E., J. R. Sparrow, et al. (2005). "Isolation and characterization of a retinal pigment epithelial cell fluorophore: an all-trans-retinal dimer conjugate." Proc Natl Acad Sci U S A 102(20): 7091-7096. Fliesler, S. J. and S. F. Basinger (1985). "Tunicamycin blocks the incorporation of opsin into retinal rod outer segment membranes." Proc Natl Acad Sci U S A 82(4): 1116-1120. Frederick, J. M., N. V. Krasnoperova, et al. (2001). "Mutant rhodopsin transgene expression on a null background." Invest Ophthalmol Vis Sci 42(3): 826-833.
213
Freeze, H. H. (2006). "Genetic defects in the human glycome." Nat Rev Genet 7(7): 537- 551. Freitag, S., M. Schachner, et al. (2003). "Behavioral alterations in mice deficient for the extracellular matrix glycoprotein tenascin-R." Behav Brain Res 145(1-2): 189- 207. Gadient, R. A., P. Lein, et al. (1998). "Effect of leukemia inhibitory factor (LIF) on the morphology and survival of cultured hippocampal neurons and glial cells." Brain Res 798(1-2): 140-146. Galtrey, C. M. and J. W. Fawcett (2007). "The role of chondroitin sulfate proteoglycans in regeneration and plasticity in the central nervous system." Brain Res Rev 54(1): 1-18. Galy, A., M. J. Roux, et al. (2005). "Rhodopsin maturation defects induce photoreceptor death by apoptosis: a fly model for RhodopsinPro23His human retinitis pigmentosa." Hum Mol Genet 14(17): 2547-2557. Gaudet, A. D., M. Leung, et al. (2009). "A role for galectin-1 in the immune response to peripheral nerve injury." Exp Neurol 220(2): 320-327. Gehrig, A., T. Langmann, et al. (2007). "Genome-wide expression profiling of the retinoschisin-deficient retina in early postnatal mouse development." Invest Ophthalmol Vis Sci 48(2): 891-900. Ghosh, F., K. Engelsberg, et al. (2007). "Long-term neuroretinal full-thickness transplants in a large animal model of severe retinitis pigmentosa." Graefes Arch Clin Exp Ophthalmol 245(6): 835-846. Gogolla, N., P. Caroni, et al. (2009). "Perineuronal nets protect fear memories from erasure." Science 325(5945): 1258-1261. Gross, A. K., V. R. Rao, et al. (2003). "Characterization of rhodopsin congenital night blindness mutant T94I." Biochemistry 42(7): 2009-2015. Gupta, N., K. E. Brown, et al. (2003). "Activated microglia in human retinitis pigmentosa, late-onset retinal degeneration, and age-related macular degeneration." Exp Eye Res 76(4): 463-471. Hagg, T. and S. Varon (1993). "Ciliary neurotrophic factor prevents degeneration of adult rat substantia nigra dopaminergic neurons in vivo." Proc Natl Acad Sci U S A 90(13): 6315-6319. Han, H., Y. Xia, et al. (2010). "Synergistic effects of galectin-1 and reactive astrocytes on functional recovery after contusive spinal cord injury." Arch Orthop Trauma Surg. Han, M., J. Lou, et al. (1997). "Partial agonist activity of 11-cis-retinal in rhodopsin mutants." J Biol Chem 272(37): 23081-23085. Han, M., S. O. Smith, et al. (1998). "Constitutive activation of opsin by mutation of methionine 257 on transmembrane helix 6." Biochemistry 37(22): 8253-8261. Han, S., B. E. Collins, et al. (2005). "Homomultimeric complexes of CD22 in B cells revealed by protein-glycan cross-linking." Nat Chem Biol 1(2): 93-97. Hao, A. J., S. T. Dheen, et al. (2002). "Expression of macrophage colony-stimulating factor and its receptor in microglia activation is linked to teratogen-induced neuronal damage." Neuroscience 112(4): 889-900. Hao, W., A. Wenzel, et al. (2002). "Evidence for two apoptotic pathways in light-induced retinal degeneration." Nat. Genet. 32: 254-260.
214
Hao, W., A. Wenzel, et al. (2002). "Evidence for two apoptotic pathways in light-induced retinal degeneration." Nat Genet 32(2): 254-260. Harrison, J. K., Y. Jiang, et al. (1998). "Role for neuronally derived fractalkine in mediating interactions between neurons and CX3CR1-expressing microglia." Proc Natl Acad Sci U S A 95(18): 10896-10901. Haynes, L. W., A. R. Kay, et al. (1986). "Single cyclic GMP-activated channel activity in excised patches of rod outer segment membrane." Nature 321(6065): 66-70. Heuss, C., M. Scanziani, et al. (1999). "G-protein-independent signaling mediated by metabotropic glutamate receptors." Nat Neurosci 2(12): 1070-1077. Horie, H., Y. Inagaki, et al. (1999). "Galectin-1 regulates initial axonal growth in peripheral nerves after axotomy." J Neurosci 19(22): 9964-9974. Horie, H. and T. Kadoya (2004). "Galectin-1 plays essential roles in adult mammalian nervous tissues. Roles of oxidized galectin-1." Glycoconj J 19(7-9): 479-489. Horie, H., T. Kadoya, et al. (2004). "Oxidized galectin-1 stimulates macrophages to promote axonal regeneration in peripheral nerves after axotomy." J Neurosci 24(8): 1873-1880. Huang, P. C., A. E. Gaitan, et al. (1993). "Cellular interactions implicated in the mechanism of photoreceptor degeneration in transgenic mice expressing a mutant rhodopsin gene." Proc Natl Acad Sci U S A 90(18): 8484-8488. Hubbard, R. (1958). "The thermal stability of rhodopsin and opsin." J Gen Physiol 42(2): 259-280. Iakhine, R., I. Chorna-Ornan, et al. (2004). "Novel dominant rhodopsin mutation triggers two mechanisms of retinal degeneration and photoreceptor desensitization." J Neurosci 24(10): 2516-2526. Ilarregui, J. M. and G. A. Rabinovich (2010). "Tolerogenic dendritic cells in the control of autoimmune neuroinflammation: an emerging role of protein-glycan interactions." Neuroimmunomodulation 17(3): 157-160. Illing, M. E., R. S. Rajan, et al. (2002). "A rhodopsin mutant linked to autosomal dominant retinitis pigmentosa is prone to aggregate and interacts with the ubiquitin proteasome system." J Biol Chem 277(37): 34150-34160. Ishibashi, S., T. Kuroiwa, et al. (2007). "Galectin-1 regulates neurogenesis in the subventricular zone and promotes functional recovery after stroke." Exp Neurol 207(2): 302-313. Jaggi, M., C. Du, et al. (2007). "Protein kinase D1: a protein of emerging translational interest." Front Biosci 12: 3757-3767. Jang, Y. P., H. Matsuda, et al. (2005). "Characterization of peroxy-A2E and furan-A2E photooxidation products and detection in human and mouse retinal pigment epithelial cell lipofuscin." J Biol Chem 280(48): 39732-39739. Jin, J. K., Y. J. Na, et al. (2007). "Galectin-3 expression is correlated with abnormal prion protein accumulation in murine scrapie." Neurosci Lett 420(2): 138-143. Jin, S., M. C. Cornwall, et al. (2003). "Opsin activation as a cause of congenital night blindness." Nat Neurosci 6(7): 731-735. Joly, S., M. Francke, et al. (2009). "Cooperative phagocytes: resident microglia and bone marrow immigrants remove dead photoreceptors in retinal lesions." Am J Pathol 174(6): 2310-2323.
215
Kadoya, T. and H. Horie (2005). "Structural and functional studies of galectin-1: a novel axonal regeneration-promoting activity for oxidized galectin-1." Curr Drug Targets 6(4): 375-383. Kadoya, T., K. Oyanagi, et al. (2005). "Oxidized galectin-1 advances the functional recovery after peripheral nerve injury." Neurosci Lett 380(3): 284-288. Kajitani, K., H. Nomaru, et al. (2009). "Galectin-1 promotes basal and kainate-induced proliferation of neural progenitors in the dentate gyrus of adult mouse hippocampus." Cell Death Differ 16(3): 417-427. Kalb, R. G. and S. Hockfield (1988). "Molecular evidence for early activity-dependent development of hamster motor neurons." J Neurosci 8(7): 2350-2360. Karan, G., C. Lillo, et al. (2005). "Lipofuscin accumulation, abnormal electrophysiology, and photoreceptor degeneration in mutant ELOVL4 transgenic mice: a model for macular degeneration." Proc Natl Acad Sci U S A 102(11): 4164-4169. Karlstetter, M., S. Ebert, et al. (2010). "Microglia in the healthy and degenerating retina: insights from novel mouse models." Immunobiology 215(9-10): 685-691. Kato, T., K. Kurita, et al. (2001). "Galectin-1 is a component of neurofilamentous lesions in sporadic and familial amyotrophic lateral sclerosis." Biochem Biophys Res Commun 282(1): 166-172. Kato, T., C. H. Ren, et al. (2005). "Galectin-1 as a potential therapeutic agent for amyotrophic lateral sclerosis." Curr Drug Targets 6(4): 407-418. Katz, M. L., G. E. Eldred, et al. (1987). "Lipofuscin autofluorescence: evidence for vitamin A involvement in the retina." Mech Ageing Dev 39(1): 81-90. Kaushal, S., K. D. Ridge, et al. (1994). "Structure and function in rhodopsin: the role of asparagine-linked glycosylation." Proc Natl Acad Sci U S A 91(9): 4024-4028. Kigerl, K. A., J. C. Gensel, et al. (2009). "Identification of two distinct macrophage subsets with divergent effects causing either neurotoxicity or regeneration in the injured mouse spinal cord." J Neurosci 29(43): 13435-13444. Kim, J. M., C. Altenbach, et al. (1997). "Structure and function in rhodopsin: rhodopsin mutants with a neutral amino acid at E134 have a partially activated conformation in the dark state." Proc Natl Acad Sci U S A 94(26): 14273-14278. Kim, J. Y., D. H. Kim, et al. (2010). "Galectin-3 secreted by human umbilical cord blood-derived mesenchymal stem cells reduces amyloid-beta42 neurotoxicity in vitro." FEBS Lett 584(16): 3601-3608. Kim, S. R., K. Nakanishi, et al. (2006). "Photooxidation of A2-PE, a photoreceptor outer segment fluorophore, and protection by lutein and zeaxanthin." Exp Eye Res 82(5): 828-839. Kim, Y. S. and T. H. Joh (2006). "Microglia, major player in the brain inflammation: their roles in the pathogenesis of Parkinson's disease." Exp Mol Med 38(4): 333- 347. Kiselev, A., M. Socolich, et al. (2000). "A molecular pathway for light-dependent photoreceptor apoptosis in Drosophila." Neuron 28(1): 139-152. Klassen, H. J., T. F. Ng, et al. (2004). "Multipotent retinal progenitors express developmental markers, differentiate into retinal neurons, and preserve light- mediated behavior." Invest Ophthalmol Vis Sci 45(11): 4167-4173. Komeima, K., B. S. Rogers, et al. (2007). "Antioxidants slow photoreceptor cell death in mouse models of retinitis pigmentosa." J Cell Physiol 213(3): 809-815.
216
Komeima, K., B. S. Rogers, et al. (2006). "Antioxidants reduce cone cell death in a model of retinitis pigmentosa." Proc Natl Acad Sci U S A 103(30): 11300-11305. Kono, M., P. W. Goletz, et al. (2008). "11-cis- and all-trans-retinols can activate rod opsin: rational design of the visual cycle." Biochemistry 47(28): 7567-7571. Kreutzberg, G. W. (1996). "Microglia: a sensor for pathological events in the CNS." Trends Neurosci 19(8): 312-318. Kroes, R. A., H. He, et al. (2010). "Overexpression of ST6GalNAcV, a ganglioside- specific alpha2,6-sialyltransferase, inhibits glioma growth in vivo." Proc Natl Acad Sci U S A 107(28): 12646-12651. Krupnick, J. G., V. V. Gurevich, et al. (1997). "Mechanism of quenching of phototransduction. Binding competition between arrestin and transducin for phosphorhodopsin." J Biol Chem 272(29): 18125-18131. Kurihara, D., M. Ueno, et al. (2010). "Expression of galectin-1 in immune cells and glial cells after spinal cord injury." Neurosci Res 66(3): 265-270. Kutza, J., K. Fields, et al. (2002). "Inhibition of HIV replication and macrophage colony- stimulating factor production in human macrophages by antiretroviral agents." AIDS Res Hum Retroviruses 18(9): 619-625. Laderach, D. J., D. Compagno, et al. (2010). "Dissecting the signal transduction pathways triggered by galectin-glycan interactions in physiological and pathological settings." IUBMB Life 62(1): 1-13. Lang, D. M., M. Monzon-Mayor, et al. (2008). "Tenascin-R and axon growth-promoting molecules are up-regulated in the regenerating visual pathway of the lizard (Gallotia galloti)." Dev Neurobiol 68(7): 899-916. Langmann, T. (2007). "Microglia activation in retinal degeneration." J Leukoc Biol 81(6): 1345-1351. LaVail, M. M., K. Unoki, et al. (1992). "Multiple growth factors, cytokines, and neurotrophins rescue photoreceptors from the damaging effects of constant light." Proc Natl Acad Sci U S A 89(23): 11249-11253. LaVail, M. M., D. Yasumura, et al. (2000). "Ribozyme rescue of photoreceptor cells in P23H transgenic rats: long-term survival and late-stage therapy." Proc Natl Acad Sci U S A 97(21): 11488-11493. LaVail, M. M., D. Yasumura, et al. (1998). "Protection of mouse photoreceptors by survival factors in retinal degenerations." Invest Ophthalmol Vis Sci 39(3): 592- 602. Lee, H. K., I. A. Seo, et al. (2009). "Interleukin-6 is required for the early induction of glial fibrillary acidic protein in Schwann cells during Wallerian degeneration." J Neurochem 108(3): 776-786. Lee, R. T., G. Lauc, et al. (2005). "Glycoproteomics: protein modifications for versatile functions. Meeting on glycoproteomics." EMBO Rep 6(11): 1018-1022. Lee, S. J., H. Xu, et al. (2003). "Light adaptation through phosphoinositide-regulated translocation of Drosophila visual arrestin." Neuron 39(1): 121-132. Lekishvili, T., S. Hesketh, et al. (2006). "Mouse galectin-1 inhibits the toxicity of glutamate by modifying NR1 NMDA receptor expression." Eur J Neurosci 24(11): 3017-3025. Lem, J. and G. L. Fain (2004). "Constitutive opsin signaling: night blindness or retinal degeneration?" Trends Mol Med 10(4): 150-157.
217
Leonard, K. C., D. Petrin, et al. (2007). "XIAP protection of photoreceptors in animal models of retinitis pigmentosa." PLoS ONE 2(3): e314. Leveillard, T., S. Mohand-Said, et al. (2004). "Identification and characterization of rod- derived cone viability factor." Nat Genet 36(7): 755-759. Li, M., K. Pisalyaput, et al. (2004). "Macrophage colony stimulatory factor and interferon-gamma trigger distinct mechanisms for augmentation of beta-amyloid- induced microglia-mediated neurotoxicity." J Neurochem 91(3): 623-633. Li, T., W. K. Franson, et al. (1995). "Constitutive activation of phototransduction by K296E opsin is not a cause of photoreceptor degeneration." Proc. Natl. Acad. Sci. USA 92: 3551-3555. Li, T., W. K. Franson, et al. (1995). "Constitutive activation of phototransduction by K296E opsin is not a cause of photoreceptor degeneration." Proc Natl Acad Sci U S A 92(8): 3551-3555. Li, T., M. A. Sandberg, et al. (1998). "Effect of vitamin A supplementation on rhodopsin mutants threonine-17 --> methionine and proline-347 --> serine in transgenic mice and in cell cultures." Proc Natl Acad Sci U S A 95(20): 11933-11938. Li, T., W. K. Snyder, et al. (1996). "Transgenic mice carrying the dominant rhodopsin mutation P347S: Evidence for defective vectorial transport of rhodopsin to the outer segments." Proc. Natl. Acad. Sci. USA 93: 14176-14181. Li, T., W. K. Snyder, et al. (1996). "Transgenic mice carrying the dominant rhodopsin mutation P347S: evidence for defective vectorial transport of rhodopsin to the outer segments." Proc Natl Acad Sci U S A 93(24): 14176-14181. Liu, J., Y. Itagaki, et al. (2000). "The biosynthesis of A2E, a fluorophore of aging retina, involves the formation of the precursor, A2-PE, in the photoreceptor outer segment membrane." J Biol Chem 275(38): 29354-29360. Liu, W., G. Z. Xu, et al. (2009). "Expression of macrophage colony-stimulating factor (M-CSF) and its receptor in streptozotocin-induced diabetic rats." Curr Eye Res 34(2): 123-133. Liu, X., P. Garriga, et al. (1996). "Structure and function in rhodopsin: correct folding and misfolding in two point mutants in the intradiscal domain of rhodopsin identified in retinitis pigmentosa." Proc Natl Acad Sci U S A 93(10): 4554-4559. Liu, X., T.-H. Wu, et al. (1997). "Defective phototransductive disk membrane morphogenesis in transgenic mice expressing opsin with a mutated N-terminal domain." J. Cell Sci. 110: 2589-2597. Lowe, J. B. and J. D. Marth (2003). "A genetic approach to Mammalian glycan function." Annu Rev Biochem 72: 643-691. Lukashova, V., C. Asselin, et al. (2001). "G-protein-independent activation of Tyk2 by the platelet-activating factor receptor." J Biol Chem 276(26): 24113-24121. MacLaren, R. E., R. A. Pearson, et al. (2006). "Retinal repair by transplantation of photoreceptor precursors." Nature 444(7116): 203-207. Maeda, A., T. Maeda, et al. (2005). "Role of photoreceptor-specific retinol dehydrogenase in the retinoid cycle in vivo." J Biol Chem 280(19): 18822-18832. Malanson, K. M. and J. Lem (2009). "Chapter 1 rhodopsin-mediated retinitis pigmentosa." Prog Mol Biol Transl Sci 88: 1-31. Margolis, R. K. and R. U. Margolis (1993). "Nervous tissue proteoglycans." Experientia 49(5): 429-446.
218
Mata, N. L., J. Weng, et al. (2000). "Biosynthesis of a major lipofuscin fluorophore in mice and humans with ABCR-mediated retinal and macular degeneration." Proc Natl Acad Sci U S A 97(13): 7154-7159. McGraw, J., A. D. Gaudet, et al. (2005). "Regulation of neuronal and glial galectin-1 expression by peripheral and central axotomy of rat primary afferent neurons." Exp Neurol 195(1): 103-114. Mendes, H. F., J. van der Spuy, et al. (2005). "Mechanisms of cell death in rhodopsin retinitis pigmentosa: implications for therapy." Trends Mol Med 11(4): 177-185. Mitrasinovic, O. M., A. Grattan, et al. (2005). "Microglia overexpressing the macrophage colony-stimulating factor receptor are neuroprotective in a microglial- hippocampal organotypic coculture system." J Neurosci 25(17): 4442-4451. Mitrasinovic, O. M., G. V. Perez, et al. (2001). "Overexpression of macrophage colony- stimulating factor receptor on microglial cells induces an inflammatory response." J Biol Chem 276(32): 30142-30149. Mitrasinovic, O. M., C. C. Robinson, et al. (2004). "Biolistic expression of the macrophage colony stimulating factor receptor in organotypic cultures induces an inflammatory response." J Neurosci Res 77(3): 420-429. Mohand-Said, S., A. Deudon-Combe, et al. (1998). "Normal retina releases a diffusible factor stimulating cone survival in the retinal degeneration mouse." Proc Natl Acad Sci U S A 95(14): 8357-8362. Montag-Sallaz, M. and D. Montag (2003). "Severe cognitive and motor coordination deficits in tenascin-R-deficient mice." Genes Brain Behav 2(1): 20-31. Muchowski, P. J. (2002). "Protein misfolding, amyloid formation, and neurodegeneration: a critical role for molecular chaperones?" Neuron 35(1): 9-12. Muradov, K. G. and N. O. Artemyev (2000). "Loss of the effector function in a transducin-alpha mutant associated with Nougaret night blindness." J Biol Chem 275(10): 6969-6974. Murphy, G. M., Jr., F. Zhao, et al. (2000). "Expression of macrophage colony-stimulating factor receptor is increased in the AbetaPP(V717F) transgenic mouse model of Alzheimer's disease." Am J Pathol 157(3): 895-904. Naash, M. I., J. G. Hollyfield, et al. (1993). "Simulation of human autosomal dominant retinitis pigmentosa in transgenic mice expressing a mutated murine opsin gene." Proc Natl Acad Sci U S A 90: 5499-5503. Naash, M. I., N. S. Peachey, et al. (1996). "Light-Induced Acceleration of Photoreceptor Degeneration in Transgenic Mice Expressing Mutant Rhodopsin." Invest. Ophthalmol. Vis. Sci. 37(5): 775-782. Naash, M. I., H. Ripps, et al. (1996). "Polygenic disease and retinitis pigmentosa: albinism exacerbates photoreceptor degeneration induced by the expression of a mutant opsin in transgenic mice." J Neurosci 16(24): 7853-7858. Narciso, M. S., S. Mietto Bde, et al. (2009). "Sciatic nerve regeneration is accelerated in galectin-3 knockout mice." Exp Neurol 217(1): 7-15. Nishioka, T., K. Sakumi, et al. (2002). "FosB gene products trigger cell proliferation and morphological alteration with an increased expression of a novel processed form of galectin-1 in the rat 3Y1 embryo cell line." J Biochem 131(5): 653-661.
219
Noorwez, S. M., V. Kuksa, et al. (2003). "Pharmacological chaperone-mediated in vivo folding and stabilization of the P23H-opsin mutant associated with autosomal dominant retinitis pigmentosa." J Biol Chem 278(16): 14442-14450. Noorwez, S. M., R. Malhotra, et al. (2004). "Retinoids assist the cellular folding of the autosomal dominant retinitis pigmentosa opsin mutant P23H." J Biol Chem 279(16): 16278-16284. Norenberg, U., M. Hubert, et al. (1995). "Characterization of functional domains of the tenascin-R (restrictin) polypeptide: cell attachment site, binding with F11, and enhancement of F11-mediated neurite outgrowth by tenascin-R." J Cell Biol 130(2): 473-484. O'Donnell, N., N. E. Zachara, et al. (2004). "Ogt-dependent X-chromosome-linked protein glycosylation is a requisite modification in somatic cell function and embryo viability." Mol Cell Biol 24(4): 1680-1690. O'Reilly, M., S. Millington-Ward, et al. (2008). "A transgenic mouse model for gene therapy of rhodopsin-linked Retinitis Pigmentosa." Vision Res 48(3): 386-391. O'Reilly, M., A. Palfi, et al. (2007). "RNA interference-mediated suppression and replacement of human rhodopsin in vivo." Am J Hum Genet 81(1): 127-135. Ohguro, H. (1997). "High levels of rhodopsin phosphorylation in missense mutations of C-terminal region of rhodopsin." FEBS Lett 413(3): 433-435. Ohtsubo, K. and J. D. Marth (2006). "Glycosylation in cellular mechanisms of health and disease." Cell 126(5): 855-867. Ohtsubo, K., S. Takamatsu, et al. (2005). "Dietary and genetic control of glucose transporter 2 glycosylation promotes insulin secretion in suppressing diabetes." Cell 123(7): 1307-1321. Olsson, J. E., J. W. Gordon, et al. (1992). "Transgenic Mice with a Rhodopsin Mutation (Pro23His): A Mouse Model of Autosomal Dominant Retinitis Pigmentosa." Neuron 9: 815-830. Olsson, J. E., J. W. Gordon, et al. (1992). "Transgenic mice with a rhodopsin mutation (Pro23His): a mouse model of autosomal dominant retinitis pigmentosa." Neuron 9(5): 815-830. Oprian, D. D. (1992). "The ligand-binding domain of rhodopsin and other G protein- linked receptors." J Bioenerg Biomembr 24(2): 211-217. Orem, N. R., L. Xia, et al. (2006). "An essential role for endocytosis of rhodopsin through interaction of visual arrestin with the AP-2 adaptor." J Cell Sci 119(Pt 15): 3141-3148. Owens, S. L., F. W. Fitzke, et al. (1994). "Ocular manifestations in autosomal dominant retinitis pigmentosa with a Lys-296-Glu rhodopsin mutation at the retinal binding site." Br J Ophthalmol 78(5): 353-358. Pak, Y., B. F. O'Dowd, et al. (1999). "Agonist-induced, G protein-dependent and - independent down-regulation of the mu opioid receptor. The receptor is a direct substrate for protein-tyrosine kinase." J Biol Chem 274(39): 27610-27616. Parish, C. A., M. Hashimoto, et al. (1998). "Isolation and one-step preparation of A2E and iso-A2E, fluorophores from human retinal pigment epithelium." Proc Natl Acad Sci U S A 95(25): 14609-14613.
220
Park, J. W., P. G. Voss, et al. (2001). "Association of galectin-1 and galectin-3 with Gemin4 in complexes containing the SMN protein." Nucleic Acids Res 29(17): 3595-3602. Paz, A., R. Haklai, et al. (2001). "Galectin-1 binds oncogenic H-Ras to mediate Ras membrane anchorage and cell transformation." Oncogene 20(51): 7486-7493. Penkowa, M., A. Molinero, et al. (2001). "Interleukin-6 deficiency reduces the brain inflammatory response and increases oxidative stress and neurodegeneration after kainic acid-induced seizures." Neuroscience 102(4): 805-818. Pepperberg, D. R. and R. K. Crouch (2001). "An illuminating new step in visual-pigment regeneration." Lancet 358(9299): 2098-2099. Perez, R. G. and W. Halfter (1993). "Tenascin in the developing chick visual system: distribution and potential role as a modulator of retinal axon growth." Dev Biol 156(1): 278-292. Pesheva, P., R. Probstmeier, et al. (1994). "Tenascin-R (J1 160/180 inhibits fibronectin- mediated cell adhesion--functional relatedness to tenascin-C." J Cell Sci 107 ( Pt 8): 2323-2333. Pesheva, P., E. Spiess, et al. (1989). "J1-160 and J1-180 are oligodendrocyte-secreted nonpermissive substrates for cell adhesion." J Cell Biol 109(4 Pt 1): 1765-1778. Plachta, N., C. Annaheim, et al. (2007). "Identification of a lectin causing the degeneration of neuronal processes using engineered embryonic stem cells." Nat Neurosci 10(6): 712-719. Poirier, F. and E. J. Robertson (1993). "Normal development of mice carrying a null mutation in the gene encoding the L14 S-type lectin." Development 119(4): 1229- 1236. Punzo, C., K. Kornacker, et al. (2009). "Stimulation of the insulin/mTOR pathway delays cone death in a mouse model of retinitis pigmentosa." Nat Neurosci 12(1): 44-52. Qu, W. S., Y. H. Wang, et al. (2011). "Galectin-1 attenuates astrogliosis-associated injuries and improves recovery of rats following focal cerebral ischemia." J Neurochem 116(2): 217-226. Qu, W. S., Y. H. Wang, et al. (2010). "Galectin-1 enhances astrocytic BDNF production and improves functional outcome in rats following ischemia." Neurochem Res 35(11): 1716-1724. Rabinovich, G. A., A. Ariel, et al. (1999). "Specific inhibition of T-cell adhesion to extracellular matrix and proinflammatory cytokine secretion by human recombinant galectin-1." Immunology 97(1): 100-106. Rabinovich, G. A. and J. M. Ilarregui (2009). "Conveying glycan information into T-cell homeostatic programs: a challenging role for galectin-1 in inflammatory and tumor microenvironments." Immunol Rev 230(1): 144-159. Rabinovich, G. A. and M. A. Toscano (2009). "Turning 'sweet' on immunity: galectin- glycan interactions in immune tolerance and inflammation." Nat Rev Immunol 9(5): 338-352. Rajala, A., R. E. Anderson, et al. (2007). "G-protein-coupled receptor rhodopsin regulates the phosphorylation of retinal insulin receptor." J Biol Chem 282(13): 9865-9873. Rajan, R. S. and R. R. Kopito (2005). "Suppression of wild-type rhodopsin maturation by mutants linked to autosomal dominant retinitis pigmentosa." J Biol Chem 280(2): 1284-1291.
221
Rakoczy, P. E., D. Zhang, et al. (2002). "Progressive age-related changes similar to age- related macular degeneration in a transgenic mouse model." Am J Pathol 161(4): 1515-1524. Ramon, E., L. J. del Valle, et al. (2003). "Unusual thermal and conformational properties of the rhodopsin congenital night blindness mutant Thr-94 --> Ile." J Biol Chem 278(8): 6427-6432. Rao, V. R., G. B. Cohen, et al. (1994). "Rhodopsin mutation G90D and a molecular mechanism for congenital night blindness." Nature 367(6464): 639-642. Rao, V. R. and D. D. Oprian (1996). "Activating mutations of rhodopsin and other G protein-coupled receptors." Annu Rev Biophys Biomol Struct 25: 287-314. Raoul, W., C. Feumi, et al. (2008). "Lipid-bloated subretinal microglial cells are at the origin of drusen appearance in CX3CR1-deficient mice." Ophthalmic Res 40(3- 4): 115-119. Redmond, T. M., S. Yu, et al. (1998). "Rpe65 is necessary for production of 11-cis- vitamin A in the retinal visual cycle." Nat Genet 20(4): 344-351. Reiter, C. E., X. Wu, et al. (2006). "Diabetes reduces basal retinal insulin receptor signaling: reversal with systemic and local insulin." Diabetes 55(4): 1148-1156. Rim, J. and D. D. Oprian (1995). "Constitutive activation of opsin: interaction of mutants with rhodopsin kinase and arrestin." Biochemistry 34(37): 11938-11945. Ripps, H. (2002). "Cell death in retinitis pigmentosa: gap junctions and the 'bystander' effect." Exp Eye Res 74(3): 327-336. Robinson, P. R., J. Buczylko, et al. (1994). "Opsins with mutations at the site of chromophore attachment constitutively activate transducin but are not phosphorylated by rhodopsin kinase." Proc. Natl. Acad. Sci. USA 91: 5411-5415. Robinson, P. R., G. B. Cohen, et al. (1992). "Constitutively active mutants of rhodopsin." Neuron 9(4): 719-725. Rolland, J. F., J. C. Henquin, et al. (2002). "G protein-independent activation of an inward Na(+) current by muscarinic receptors in mouse pancreatic beta-cells." J Biol Chem 277(41): 38373-38380. Roof, D. J., M. Adamian, et al. (1994). "Rhodopsin accumulation at abnormal sites in retinas of mice with a human P23H rhodopsin transgene." Invest Ophthalmol Vis Sci 35(12): 4049-4062. Rotblat, B., L. Belanis, et al. (2010). "H-Ras nanocluster stability regulates the magnitude of MAPK signal output." PLoS ONE 5(8): e11991. Saari, J. C., G. G. Garwin, et al. (1998). "Reduction of all-trans-retinal limits regeneration of visual pigment in mice." Vision Res 38(10): 1325-1333. Sakaguchi, M., T. Shingo, et al. (2006). "A carbohydrate-binding protein, Galectin-1, promotes proliferation of adult neural stem cells." Proc Natl Acad Sci U S A 103(18): 7112-7117. Samardzija, M., A. Wenzel, et al. (2006). "Rpe65 as a modifier gene for inherited retinal degeneration." Eur J Neurosci 23(4): 1028-1034. Sanchez-Lopez, A., M. A. Cuadros, et al. (2004). "Radial migration of developing microglial cells in quail retina: a confocal microscopy study." Glia 46(3): 261- 273.
222
Saravanan, C., Z. Cao, et al. (2009). "Detection of differentially expressed wound- healing-related glycogenes in galectin-3-deficient mice." Invest Ophthalmol Vis Sci 50(12): 5690-5696. Saravanan, C., Z. Cao, et al. (2010). "Analysis of differential expression of glycosyltransferases in healing corneas by glycogene microarrays." Glycobiology 20(1): 13-23. Saravanan, C., F. T. Liu, et al. (2009). "Galectin-3 promotes lamellipodia formation in epithelial cells by interacting with complex N-glycans on alpha3beta1 integrin." J Cell Sci 122(Pt 20): 3684-3693. Sasahara, M., A. Otani, et al. (2008). "Activation of bone marrow-derived microglia promotes photoreceptor survival in inherited retinal degeneration." Am J Pathol 172(6): 1693-1703. Scott, K. and J. Zhang (2002). "Partial identification by site-directed mutagenesis of a cell growth inhibitory site on the human galectin-1 molecule." BMC Cell Biol 3: 3. Sendtner, M., H. Schmalbruch, et al. (1992). "Ciliary neurotrophic factor prevents degeneration of motor neurons in mouse mutant progressive motor neuronopathy." Nature 358(6386): 502-504. Shen, J., X. Yang, et al. (2005). "Oxidative damage is a potential cause of cone cell death in retinitis pigmentosa." J Cell Physiol 203(3): 457-464. Sieving, P. A., M. L. Fowler, et al. (2001). "Constitutive "light" adaptation in rods from G90D rhodopsin: a mechanism for human congenital nightblindness without rod cell loss." J Neurosci 21(15): 5449-5460. Sieving, P. A., J. E. Richards, et al. (1995). "Dark-light: model for nightblindness from the human rhodopsin Gly-90-->Asp mutation." Proc Natl Acad Sci U S A 92(3): 880-884. Song, Y., L. Zhao, et al. (2003). "Photoreceptor protection by cardiotrophin-1 in transgenic rats with the rhodopsin mutation s334ter." Invest Ophthalmol Vis Sci 44(9): 4069-4075. Soto, C. and L. D. Estrada (2008). "Protein misfolding and neurodegeneration." Arch Neurol 65(2): 184-189. Sparrow, J. and M. Boulton (2005). "RPE lipofuscin and its role in retinal pathobiology." Exp Eye Res. 80((5)): 595-606. Steinberg, R. H. (1994). "Survival factors in retinal degenerations." Curr Opin Neurobiol 4(4): 515-524. Stowell, S. R., M. Cho, et al. (2009). "Ligand reduces galectin-1 sensitivity to oxidative inactivation by enhancing dimer formation." J Biol Chem 284(8): 4989-4999. Streichert, L. C., C. D. Birnbach, et al. (1999). "A diffusible factor from normal retinal cells promotes rod photoreceptor survival in an in vitro model of retinitis pigmentosa." J Neurobiol 39(4): 475-490. Streit, W. J. (2002). "Microglia as neuroprotective, immunocompetent cells of the CNS." Glia 40(2): 133-139. Sung, C. H., C. M. Davenport, et al. (1993). "Rhodopsin mutations responsible for autosomal dominant retinitis pigmentosa. Clustering of functional classes along the polypeptide chain." J Biol Chem 268(35): 26645-26649.
223
Sung, C. H., B. G. Schneider, et al. (1991). "Functional heterogeneity of mutant rhodopsins responsible for autosomal dominant retinitis pigmentosa." Proc Natl Acad Sci U S A 88(19): 8840-8844. Suzumura, A., H. Takeuchi, et al. (2006). "Roles of glia-derived cytokines on neuronal degeneration and regeneration." Ann N Y Acad Sci 1088: 219-229. Takeuchi, A., O. Miyaishi, et al. (2001). "Macrophage colony-stimulating factor is expressed in neuron and microglia after focal brain injury." J Neurosci Res 65(1): 38-44. Tam, B. M. and O. L. Moritz (2006). "Characterization of rhodopsin P23H-induced retinal degeneration in a Xenopus laevis model of retinitis pigmentosa." Invest Ophthalmol Vis Sci 47(8): 3234-3241. Tam, B. M. and O. L. Moritz (2007). "Dark rearing rescues P23H rhodopsin-induced retinal degeneration in a transgenic Xenopus laevis model of retinitis pigmentosa: a chromophore-dependent mechanism characterized by production of N- terminally truncated mutant rhodopsin." J Neurosci 27(34): 9043-9053. Tam, B. M., G. Xie, et al. (2006). "Mislocalized rhodopsin does not require activation to cause retinal degeneration and neurite outgrowth in Xenopus laevis." J Neurosci 26(1): 203-209. Tao, W., R. Wen, et al. (2002). "Encapsulated cell-based delivery of CNTF reduces photoreceptor degeneration in animal models of retinitis pigmentosa." Invest Ophthalmol Vis Sci 43(10): 3292-3298. Taylor, J., P. Pesheva, et al. (1993). "Influence of janusin and tenascin on growth cone behavior in vitro." J Neurosci Res 35(4): 347-362. Uehara, F., N. Ohba, et al. (2001). "Isolation and characterization of galectins in the mammalian retina." Invest Ophthalmol Vis Sci 42(10): 2164-2172. Ueki, Y., J. Wang, et al. (2008). "STAT3 activation in photoreceptors by leukemia inhibitory factor is associated with protection from light damage." J Neurochem 105(3): 784-796. van der Leij, J., A. van den Berg, et al. (2004). "Dimeric galectin-1 induces IL-10 production in T-lymphocytes: an important tool in the regulation of the immune response." J Pathol 204(5): 511-518. van Horssen, J., P. Wesseling, et al. (2003). "Heparan sulphate proteoglycans in Alzheimer's disease and amyloid-related disorders." Lancet Neurol 2(8): 482-492. van Rossum, D. and U. K. Hanisch (2004). "Microglia." Metab Brain Dis 19(3-4): 393- 411. Vincent, V. A., C. C. Robinson, et al. (2002). "Macrophage colony stimulating factor prevents NMDA-induced neuronal death in hippocampal organotypic cultures." J Neurochem 82(6): 1388-1397. Wakao, N., S. Imagama, et al. (2011). "Hyaluronan oligosaccharides promote functional recovery after spinal cord injury in rats." Neurosci Lett 488(3): 299-304. Wang, M., T. T. Lam, et al. (1997). "Expression of a mutant opsin gene increases the susceptibility of the retina to light damage." Vis. Neurosci. 14: 55-62. Wang, T. and C. Montell (2007). "Phototransduction and retinal degeneration in Drosophila." Pflugers Arch. Wang, T., H. Xu, et al. (2005). "Light activation, adaptation, and cell survival functions of the Na+/Ca2+ exchanger CalX." Neuron 45(3): 367-378.
224
Wang, Y., O. Berezovska, et al. (1999). "Expression of colony stimulating factor-1 receptor (CSF-1R) by CNS neurons in mice." J Neurosci Res 57(5): 616-632. Weber, P., U. Bartsch, et al. (1999). "Mice deficient for tenascin-R display alterations of the extracellular matrix and decreased axonal conduction velocities in the CNS." J Neurosci 19(11): 4245-4262. Wei, H., S. Ahn, et al. (2003). "Independent beta-arrestin 2 and G protein-mediated pathways for angiotensin II activation of extracellular signal-regulated kinases 1 and 2." Proc Natl Acad Sci U S A 100(19): 10782-10787. Weiss, E. R., Y. Hao, et al. (1995). "Altered cAMP Levels in Retinas from Transgenic Mice Expressing a Rhodopsin Mutant." Biochem. Biophys. Res. Comm. 216(3): 755-761. Weiss, E. R., Y. Hao, et al. (1995). "Altered cAMP levels in retinas from transgenic mice expressing a rhodopsin mutant." Biochem Biophys Res Commun 216(3): 755- 761. Weng, J., N. L. Mata, et al. (1999). "Insights into the function of Rim protein in photoreceptors and etiology of Stargardt's disease from the phenotype in abcr knockout mice." Cell 98(1): 13-23. White, D. A., J. J. Fritz, et al. (2007). "Increased sensitivity to light-induced damage in a mouse model of autosomal dominant retinal disease." Invest Ophthalmol Vis Sci 48(5): 1942-1951. Williams, A. J. and H. L. Paulson (2008). "Polyglutamine neurodegeneration: protein misfolding revisited." Trends Neurosci 31(10): 521-528. Woodruff, M. L., Z. Wang, et al. (2003). "Spontaneous activity of opsin apoprotein is a cause of Leber congenital amaurosis." Nat Genet 35(2): 158-164. Woodruff, M. L., Z. Wang, et al. (2003). "Spontaneous activity of opsin apoprotein is a cause of Leber congenital amaurosis." Nat. Genet. 35(2): 158-164. Wu, T. H., T. D. Ting, et al. (1998). "Opsin localization and rhodopsin photochemistry in a transgenic mouse model of retinitis pigmentosa." Neuroscience 87(3): 709-717. Xiao, Z. C., J. M. Revest, et al. (1998). "Defasciculation of neurites is mediated by tenascin-R and its neuronal receptor F3/11." J Neurosci Res 52(4): 390-404. Xu, J., R. L. Dodd, et al. (1997). "Prolonged photoresponses in transgenic mouse rods lacking arrestin." Nature 389(6650): 505-509. Yamane, J., M. Nakamura, et al. (2010). "Transplantation of galectin-1-expressing human neural stem cells into the injured spinal cord of adult common marmosets." J Neurosci Res 88(7): 1394-1405. Yan, Y. P., B. T. Lang, et al. (2009). "Galectin-3 mediates post-ischemic tissue remodeling." Brain Res 1288: 116-124. Yang, R. Y., G. A. Rabinovich, et al. (2008). "Galectins: structure, function and therapeutic potential." Expert Rev Mol Med 10: e17. Yin, Y., Q. Cui, et al. (2003). "Macrophage-derived factors stimulate optic nerve regeneration." J Neurosci 23(6): 2284-2293. Yin, Y., M. T. Henzl, et al. (2006). "Oncomodulin is a macrophage-derived signal for axon regeneration in retinal ganglion cells." Nat Neurosci 9(6): 843-852. Yu, D. Y., S. Cringle, et al. (2004). "Photoreceptor death, trophic factor expression, retinal oxygen status, and photoreceptor function in the P23H rat." Invest Ophthalmol Vis Sci 45(6): 2013-2019.
225
Yu, D. Y., S. J. Cringle, et al. (2000). "Intraretinal oxygen levels before and after photoreceptor loss in the RCS rat." Invest Ophthalmol Vis Sci 41(12): 3999-4006. Yu, X., R. V. Rajala, et al. (2004). "Involvement of insulin/phosphoinositide 3- kinase/Akt signal pathway in 17 beta-estradiol-mediated neuroprotection." J Biol Chem 279(13): 13086-13094. Zachara, N. E. and G. W. Hart (2004). "O-GlcNAc a sensor of cellular state: the role of nucleocytoplasmic glycosylation in modulating cellular function in response to nutrition and stress." Biochim Biophys Acta 1673(1-2): 13-28. Zacharias, U., U. Norenberg, et al. (1999). "Functional interactions of the immunoglobulin superfamily member F11 are differentially regulated by the extracellular matrix proteins tenascin-R and tenascin-C." J Biol Chem 274(34): 24357-24365. Zeiss, C. J. and E. A. Johnson (2004). "Proliferation of microglia, but not photoreceptors, in the outer nuclear layer of the rd-1 mouse." Invest Ophthalmol Vis Sci 45(3): 971-976. Zhang, C., J. K. Shen, et al. (2005). "Activation of microglia and chemokines in light- induced retinal degeneration." Mol Vis 11: 887-895. Zhang, D., M. Brankov, et al. (2005). "Correlation between inactive cathepsin D expression and retinal changes in mcd2/mcd2 transgenic mice." Invest Ophthalmol Vis Sci 46(9): 3031-3038. Zhang, F., K. Su, et al. (2003). "O-GlcNAc modification is an endogenous inhibitor of the proteasome." Cell 115(6): 715-725. Zhu, L., G. F. Jang, et al. (2004). "A naturally occurring mutation of the opsin gene (T4R) in dogs affects glycosylation and stability of the G protein-coupled receptor." J Biol Chem 279(51): 53828-53839. Zhuang, Z. Y., Y. Kawasaki, et al. (2007). "Role of the CX3CR1/p38 MAPK pathway in spinal microglia for the development of neuropathic pain following nerve injury- induced cleavage of fractalkine." Brain Behav Immun 21(5): 642-651. Zhukovsky, E. A., P. R. Robinson, et al. (1991). "Transducin activation by rhodopsin without a covalent bond to the 11-cis-retinal chromophore." Science 251(4993): 558-560. Zvyaga, T. A., K. Fahmy, et al. (1996). "Characterization of the mutant visual pigment responsible for congenital night blindness: a biochemical and Fourier-transform infrared spectroscopy study." Biochemistry 35(23): 7536-7545.