Rhodopsin Retinitis Pigmentosa: an in Vitro Study of the Cellular Fate of Wild

Rhodopsin Retinitis Pigmentosa: an in vitro study of the ceiiuiar fate of wiid type and mutant rhodopsin

Richard Sebastian Saiiba M.Sc.

A thesis submitted to the University of London for the degree of Doctor of Phiiosophy

September 2002

institute of Ophthaimoiogy, University Coiiege London ProQuest Number: U643855

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest.

ProQuest U643855

Published by ProQuest LLC(2016). Copyright of the Dissertation is held by the Author.

ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Abstract

Mutations in the photopigment, rhodopsin, are the major cause of autosomal dominant retinitis pigmentosa. The majority of mutations in rhodopsin lead to misfolding of the protein. Through the detailed examination of P23H and K296E mutant opsin processing in COS-7 cells, I have shown that the mutant protein does not accumulate in the Golgi, as previously thought, but that it forms aggregates that have many of the characteristic features of an aggresome. The aggregates form close to the centrosome and lead to the dispersal of the Golgi apparatus. Furthermore, these aggregates are ubiquitinated, recruit cellular chaperones and disrupt the intermediate filament network. Mutant opsin expression can disrupt the processing of normal opsin, as co-transfection revealed that the wild-type protein is recruited to mutant opsin aggregates. In SH-SY5Y cells mutant opsin forms multiple aggregates, which are ubiquitinated, but aggresomes are rarely seen. The degradation of mutant opsin is dependent on the proteasome machinery. Unlike the situation with AF508-CFTR, proteasome inhibition does not lead to a marked increase in aggresome formation in COS-7 cells but increases the levels of mutant opsin within the ER, suggesting that the proteasome is required for the efficient retro-translocation of the mutant protein. Inhibition of N-linked glycosylation with tunicamycin, leads to the selective retention of the mutant protein within the ER and increases the steady state level of mutant opsin. Glycosylation, however, has no influence on the biogenesis and targeting of wild- type opsin in cultured cells. This demonstrates that N-linked glycosylation is required for ER associated degradation of the mutant protein but is not essential for the quality control of opsin folding. The addition of 9-c/s-retinal to the media increased the amount of P23H, but not K296E, that was soluble and reached the plasma membrane. The expression of mutant opsin in COS-7 cells appears to induce the upregulation of the ER resident chaperone BiP, which is known to be induced upon ER stress. Thus, it appears that the expression of mutant opsin induces an unfolded protein response. These data show that rhodopsin autosomal dominant retinitis pigmentosa is similar to many other neurodegenerative diseases in which the formation of intracellular protein aggregates is linked to disease pathogenesis and suggest a mechanism for disease dominance. Acknowledgements:

I would like to express my gratitude to my mother and Marsha Nodelman for their unwavering support and understanding. I would like to thank Dr. T. Bailey and Dr. J. Van De Spuy for informative discussions on RT-PCR and protein biochemistry. I am grateful to Dr. R. Kopito, Dr. R. Molday, Dr. D. Oprian and Dr. R. Crouch for supplying vital reagents used in my work and Dr. P. Munro for help with my work involving electron microscopy. I would also like to thank my supervisors Dr. M. Cheetham and Professor P. Luthert for their supervision and support during my work in the department of Pathology. Contents Title 1 Abstract 2 Acknowledgements 3 Contents 4 Abbreviations 10

Chapter 1 Introduction 12 1.1 Structure of the Retina 14 1.2 Retinitis Pigmentosa 14 1.2.1 Rhodopsin RP 18 1.3 Rhodopsin Structure and Function: 22 1.3.1 Structure of rhodopsin 22 1.3.2 Function of rhodopsin 26 1.4 Phototransduction in rod ceils 27 1.5 Rhodopsin processing in vivo 29 1.5.1 Role of N-linked glycosylation 29 1.5.2 The role of ligand in opsin biogenesis 30 1.5.3 Molecular chaperones and rhodopsin biogenesis 30 1.6 Phenotypic behavior of rhodopsin RP mutants 31 1.6.1 In vitro studies 31 1.6.2 Transgenic models of Rhodopsin RP 32 1.7 N-iinked glycosylation and ER quality control 33 1.7.1 N-linked Glycosylation 33 1.7.2 ER Quality Control 37 1.8 The Ubiquitin Proteasome System (UPS) 41 1.8.1 The proteasome 41 1.8.2 Ubiquitin and proteasomal degradation 41 1.9 Neurodegenerative Diseases and Proteinaceous Intraceliuiar Inclusions 44 1.9.1 Neurodegenerative Diseases 44 1.9.2 Intracellular Proteinaceous Inclusions 45 1.9.3 Aggresomes 47 1.10 ER Stress and the Unfolded Protein Response 50 1.11 Aims of the Thesis 53 Chapter 2 Materials and Methods 54 2.1 Reagents 54 2.2 Molecular Biology 54 2.2.1 Site directed mutagenesis by overlap extension using PGR 54 2.2.2 Cloning of WT, K296E and P23H rhodopsin into pEGFP-N1 57 2.2.3 Restriction Enzyme Digests 58 2.2.4 Ligations and Transformations 59 2.2.5 RT-PCR of GADDI 53 59 2.2.6 Mini preparations of plasmid DNA 60 2.3 Cell Culture 61 2.3.1 Transfection of COS-7 and SH-SY5Y cells 61 2.3.2 Drug treatments of COS-7 and SH-SY5Y cells 61 2.4 Immunofluorescence and Electron Microscopy 61 2.4.1 Immunofluorescence microscopy 61 2.4.2 Electron Microscopy 63 2.5 Biochemistry 64 2.5.1 Preparation of COS-7 Cell Extracts 64 2.5.2 SDS-Polyacrylamide Gel Electrophoresis 64 2.5.3 Western Blotting 65 2.5.4 Deglycosylation of rod opsin 65

Chapter 3: Intracellular Inclusions of RP mutants of rhodopsin 67 3.1 Introduction 67 3.2 Results 68 3.2.1 Generation of P23H Rhodopsin and GFP tagged rhodopsin constructs 68 3.2.2 The localisation of WT and mutant rod opsin in COS-7 cells using immunofluorescence microscopy 69 3.2.3 Mutant rod opsin inclusions co-localise with c-myc-tagged ubiquitin in COS-7 cells 73 3.2.4 Mutant rod opsin inclusions alter the normal distribution of the intermediate filament vimentin 76 3.2.5 Cytosolic chaperones are recruited to mutant rod opsin aggresomes 76 3.2.6 The ultrastruture of P23H rod opsin-GFP aggresomes 80 3.2.7 Comparison of the numbers of WT and mutant rod opsin aggresomes in COS-7 cells over time 80 3.2.8 The influence of P23H rod opsin on the number of WT rod opsin-GFP aggresomes in COS-7 cells 83 3.2.9 The localisation of WT and mutant GFP-tagged rod opsin in SH-SY5Y cells using immunofluorescence microscopy 83 3.2.10 Mutant rod opsin-GFP inclusions sequester ubiquitin in SH-SY5Y cells 84 3.3 Discussion 88

Chapter 4: N-linked glycosylation of rod opsin and ERAD 95 4.1 Introduction 95 4.2 Results 97 4.2.1 The glycoforms of wild type rod opsin processed in COS-7 cells 97 4.2.2 The different glycoforms of WT, K296E and P23H rod opsin in COS-7 cells 100 4.2.3 N-linked glycosylation of mutant rod opsin is required for ERAD in COS-7 cells 103 4.2.4 Tunicamycin induces the accumulation of mutant rod opsin in the ER of SH-SY5Y cells 106 4.3 Discussion 108

Chapter 5: The degradation of mutant opsin is dependent on the proteasome 113 5.1 Introduction 113 5.2 Results 114 5.2.1 MG-132 inhibits the degradation of mutant rod opsin in COS-7 cells 114 5.2.2 The influence of MG-132 on mutant rod opsin aggresome numbers 118 5.2.3 Detection of WT and mutant rod opsins in COS-7 cells following MG-132 treatment using immunofluorescence microscopy 120 5.2.4 The influence of MG-132 on mutant opsin degradation in SH-SY5Y cells 125 5.3 Discussion 127

Chapter 6: The influence of ligand and temperature on the processing and trafficking of RP mutants of rhodopsin in vitro 130 6.1 Introduction 131 6.2 Results 131 6.2.1 The influence of 9-c/s-retinal on P23H rod opsin processing and trafficking in COS-7 cells 131 6 6.2.2 The influence of a low temperature incubation on the processing of P23H rod opsin in COS-7 cells 133 6.3 Discussion 139

Chapter 7: RP mutants of rhodopsin and ER stress in cuitured celis 144 7.1 Introduction 144 7.2 Results 144 7.2.1 Detection of the Gaddi 53 transcript in SH-SY5Y cells expressing the K296E rhodopsin gene 144 7.2.2 Detection of the Gaddi 53 protein in SH-SY5Y cells expressing mutant rhodopsin-GFP 145 7.2.3 The influence of GFP-tagged WT and GFP-tagged mutant rod opsin on BiP protein levels in COS-7 cells 149 7.3 Discussion 151

General discussion 154

List of Figures: Figure 1.1 Simple organisation of the retina 13 Figure 1.2 Fundus photographs of a normal human retina and RP retina 17 Figure 1.3 Secondary structure diagram of bovine rhodopsin 23 Figure 1.4 Structure of 11-c/s-retinal, all-frans-retinal and all-frans-retinol 24 Figure 1.5 The absorbance spectrum of bovine rhodopsin 25 Figure 1.6 Intermediates in the photolysis of rhodopsin 27 Figure 1.7 Diagramatic representation of the N-linked core oligosaccharide added to growing polypetides within the ER lumen 36 Figure 1.8 Diagrammatic representation of the biosynthesis of the dolichol pyrophosphoryl oligosaccharide precursor of N-linked oligosacharides 36 Figure 1.9 Calnexin/Calreticulin mediated degradation of misfolded glycoproteins 40 Figure 2.1 The pMT3 mammalian expression vector 55 Figure 3.1 Construction of P23H rhodopsin by PCR mutagenesis 70 Figure 3.2 Cloning of WT, K296E and P23H rhodopsin into pEGFP-N1 71 Figure 3.3 The localisation of WT and mutant rod opsins in COS-7 cells 72 Figure 3.4 Inclusions of mutant rod opsin disrupt the Golgi apparatus in COS-7 cells 74 Figure 3.5 Dispersal of the Golgi with BFA does not disrupt mutant rod opsin inclusions 75

7 Figure 3.6 Mutant rod opsin inclusions are ubiquitinated in COS-7 cells 77 Figure 3.7 Mutant rod opsin inclusions disrupt the normal vimentin network in COS-7 cells 78 Figure 3.8 Mutant rod opsin-GFP aggresomes recruit molecular chaperones 79 Figure 3.9 The ultrastructure of P23H rod opsin-GFP aggresomes 81 Figure 3.10 The numbers of mutant rod opsin-GFP aggresomes in COS-7 cells increase with time 82 Figure 3.11P23H rod opsin increases the numbers of WT rod opsin-GFP aggresomes in COS-7 cells 85 Figure 3.12 The localisation of GFP-tagged WT, K296E and P23H rod opsin in SH-SH5Y cells 86 Figure 3.13 Ubiquitin co-localises with GFP-tagged K296E and P23H rod opsin aggregates in SH-SY5Y cells 87 Figure 4.1 The glycoforms of WT rod opsin expressed in COS-7 cells 99 Figure 4.2 Comparison of DM soluble and DM insoluble WT and mutant opsin species 101 Figure 4.3 The glycoforms of DM soluble P23H opsin in COS-7 cells 102 Figure 4.4 TM prevents the degradation of mutant rod opsin in COS-7 cells 104 Figure 4.5 TM induces the accumulation of mutant rod opsin in the ER of COS-7 cells 105 Figure 4.6 TM induces the accumulation of mutant rod opsin in the ER of SH-SY5Y cells 107 Figure 5.1 The effects MG-132 on the steady state levels of WT, K296E and P23H rod opsin in COS-7 cells 116 Figure 5.2 Endo H digestion of mutant rod opsin glycoforms that accumulated following treatment with MG-132 117 Figure 5.3 The influence of MG-132 on rod opsin aggresome numbers in COS-7 cells 119 Figure 5.4 Effects of MG-132 on the targeting of WT rod opsin to the plasma membrane 121 Figure 5.5 MG-132 induces P23H rod opsin to accumulate in the ER of COS-7 cells 122 Figure 5.6 MG-132 induces the accumulation of K296E rod opsin in the ER of COS-7 cells 123 Figure 5.7 MG-132 induces GFP-tagged K296E and P23H mutant rod opsin to accumulate in the ER of COS-7 cell 124

8 Figure 5.8 In SH-SY5Y cells mutant rod opsin accumulates in the ER following incubation with MG-132 126 Figure 6.1 The effect of 9-c/s-retinal on the processing of P23H rod opsin in COS-7 cells 135 Figure 6.2 Endo H and PNGase F digestion of P23H rod opsin from cells incubated with 9-c/s-retinal 136 Figure 6.3 9-c/s-retinal promotes the translocation of P23H rod opsin to the cell surface in COS-7 cells 137 Figure 6.4 Low temperature incubation of COS-7 cells expressing the WT and P23H rod opsin genes 138 Figure 7.1 Detection of Gaddi 53 by RT-PCR in SH-SY5Y cells incubated with TM 146 Figure 7.2 Detection of the Gaddi 53 transcript in SH-SY5Y cells expressing K296E rhodopsin using RT-PCR 147 Figure 7.3 Gaddi 53 protein is present in SH-SY5Y cells following treatment with TM 148 Figure 7.4 Detection of Gaddi 53 protein in SH-SY5Y cells expressing GFP-tagged WT and mutant rod opsin 148 Figure 7.5 Tunicamycin induces an increase in BiP protein levels in COS-7 cells 150 Figure 7.6 The influence of GFP-tagged WT and mutant rod opsin on BiP protein levels in COS-7 cells 150

List of Tables: Table 1.1 Missense/nonsense mutations in rhodopsin causing RP 19 Table 1.2 Deletions in rhodopsin that cause adRP 21 Table 1.3 Nucleotide substitutions at splice sites in rhodopsin causing adRP 21 Table 1.4 Mutations in rhodopsin causing CSNB 21 Table 2.1 Antibody titres used in immunofluorescence labeling of COS-7 and SH-SY5Y cells 62 Table 2.2 Absorbance and Emmision spectra of fluorophores used in immunofluorescence labeling of cells 63 Abbreviations: adRP autosomal dominant RP arRP autosomal recessive RP APP amyloid precursor protein Ap P-amyloid ALS Amyotrophic Lateral Sclerosis AD Alzheimer’s disease p-COP coatamer protein BFA Brefeldin A BiP Immunoglobulin heavy chain binding protein BSA Bovine Serum Albumin cGMP cyclic guanosine 3', 5'-monophosphate GMP CNX Calnexin CRT Calreticulin CSNB congenital stationary night blindness DMEM Dulbecco’s Modified Eagles Medium DM n-Dodecyl-p-D-Maltoside DNA deoxyribosenucleic acid DAP I 4", 6' diamidino 2-phenylindole dihydrochloride DMSO dimethylsulfoxide DRPLA Dentatorubral-Pallidoluysian Atrophy EDTA Ethylenediaminetetraacetic acid Endo H Endoglycosidase H ER endoplasmic reticulum ERG electroretinogram ERAD ER associated degradation ERSR ER stress response element PBS Foetal Bovine Serum F lic Fluoroscein isothiocyanate FRET fluorescence resonance energy transfer GFP green fluorescent protein Grp94 glucose regulated protein 94 GlcNAc N-acetylglucosamine Gl glucosidase I Gil glucosidase II

10 HD Huntington’s disease MRP Horse Radish Peroxidase LB Luria-Bertani mAb monoclonal anti-body Man mannose Man I Mannosidase I Mann II Mannosidase II MG-132 Cbz-leu-leu-leucinal MTOC microtubule organizing center PAGE polyacrylamide gel electophoresis PNGase F Peptide: N glycosidase F PD Parkinson’s disease PBS Phoshate Buffered Saline PCR polymerase chain reaction QC Quality Control retGO retinal guanylyl cyclase Rhi Drosophila rhodopsin ROS rod outer segment RP Retinitis Pigmentosa SCA spinocerebellar ataxias SBMA Spinal and Bulbar Muscular Atrophy SOD superoxide dismutase SDS sodium dodecyl sulphate TM Tunicamycin TAB Tris Acetate EDTA UGGT uridine diphosphate (UDP)-glucose:glycoprotein glucosyltransferase UPR unfolded protein response UPS Ubiquitin Proteasome System

11 Chapter 1 : Introduction

1.1 Structure of the Retina

Visual perception begins in the retina, a highly specialised sensory organ, which is an intricate part of the CNS. Light entering the eye is projected onto the retina at the back of the eye and converted into electrical signals by photoreceptor cells. These electrical signals are sent through the optic nerve to higher centers in the brain for further processing that is necessary for perception.

The retina lies in front of the pigment epithelium that lines the back of the eye. Cells in the pigment epithelium contain the pigment, melanin, which absorbs any light not captured by the retina, preventing scattering of light back onto the retina. As shown in figure 1.1 the photoreceptors are in direct contact with the pigment epithelium. One interesting consequence of this arrangement is that light must travel through the layers of other retinal neurons before striking the photoreceptors. To allow light to reach the photoreceptors without light being absorbed or greatly scattered, the proximal neural layers of the retina are unmyelinated and therefore relatively transparent.

The human retina contains two types of photoreceptors, rods and cones. Cones are responsible for day vision and rods mediate night vision; rods function in dim light that is present at dusk or night, when most stimuli are too weak to excite cones. Rods are exquisitely sensitive to light; a single photon can evoke a detectable response in a rod cell and many rods synapse on the same target interneuron, which is known as the bipolar cell (Fig 1.1). All vertebrate retinas are composed of three layers of nerve cell bodies and two layers of synapses. The outer nuclear layer contains cell bodies of the rods and cones, the inner nuclear layer contains cell bodies of the bipolar, horizontal and amacrine cells and the ganglion cell layer contains cell bodies of ganglion cells and displaced amacrine cells. Dividing these nerve cell layers are two neuropils where synaptic contacts occur. The first area of neuropil is the outer plexiform layer (OPL) where connections between rods and cones, and vertically running bipolar cells and horizontally oriented horizontal cells occur. The second neuropil of the retina, is the inner plexiform layer (IPL), and it functions as a relay station for the vertical-information-carrying nerve cells, the bipolar cells, to connect to ganglion cells. In addition, different varieties of horizontally-

12 A. B.

_ Cvtopl09mic op lntraPi$C0l

Mitochondrion

Endoplssmic

/. V\ SynBptk

Figure 1.1 Simple organisation of the retina. A) The three nuclear layers of the retina contain three major functional classes of neurons: photoreceptors (rods and cones) in the outer nuclear layer, inter-neurons (bipolar cells, horizontal cells, amacrine cells) in the inner nuclear layer and ganglion cells in the ganglion cell layer. Photoreceptor, bipolar cells, and horizontal cells make synaptic connections with each other in the outerplexiform layer. The bipolar, amacrine, and ganglion cells make contact in the inner plexiform layer. Bipolar cells bridge the two plexiform layers, and information flows vertically from photoreceptors to bipolar cells to ganglion cells. Information also flows laterally mediated by horizontal cells in the outer plexifom layer and amacrine cells in the inner plexiform layer (from http://retina.umh.es/Webvlsion/) . B) Shematic diagram of a rod cell (from Stryer, 1988). The outer segment is the light responsive organelle containing many membranous discs in which rhodopsin is embedded. Disc morphogenesis occurs at the base of the cilium.

13 and vertically-directed amacrine cells, interact in further networks to influence and integrate the ganglion cell signals. It is at the culmination of all this neural processing in the inner plexiform layer that the message concerning the visual image is transmitted to the brain along the optic nerve.

The rod cell consists of an outer segment, which is the specialised region of the rod cell responsible for photoreception, containing rhodopsin embedded in discs (Fig. 1.1b). The outer segment contains a stack of about a 1000 dis(%, which are closed, flattened sacs about 160 A thick, which are densely packed with rhodopsin. A slender cilium joins the outer and the inner segment, containing the nucleus and many mitochondria and at the base of the rod cell is a synaptic terminal, which makes contact with retinal inter-neurons. The basal body at the base of the connecting cilium also serves as the microtubule organising center (MTOC) that organises the cytoplasmic microtubules of the inner segment (Joshi, 1993). New discs are added at the base of the outer segment at the cilium by a process of out pouching of the photoreceptor cell plasma membrane at this point (Steinberg et al., 1980). The stacks of discs containing visual pigment molecules in the outer segments of the photoreceptors are constantly renewed. At the same time old discs travel up the outer segment and are pinched off at the tips and engulfed by the apical processes of the pigment epithelium (LaVail at a/., 1994). These discarded discs become known as phagosomes in the pigment epithelial cells, which are then broken down (LaVail at a/., 1994). Photoreceptor outer segment discs are phagocytosed by the pigment epithelium in a diurnal cycle.

1.2 Retinitis Pigmentosa

Neurons in the retina are susceptible to degeneration, as are many other neurons in the central nervous system. Retinitis Pigmentosa (RP) is a heterogeneous group of retinal dystrophies, defined by progressive loss of photoreceptor function and viability, resulting in reduced or non-recordable electroretinogram amplitudes (ERG), constriction and gradual loss of the visual field and loss of visual acuity. The primary defect underlying RP effects the function of the rod photoreceptor, and subsequently, mostly unknown molecular and cellular mechanisms trigger the apoptotic degeneration of these photoreceptor cells. In classic RP as rod cell death progresses, cone cell viability is compromised with severe loss of vision being the end result. As the disease progresses the retina exhibits typical features of RP, such as scattered pigmentary deposits, attenuated retinal vessels and a pale appearance of the optic disk (Fig. 1.2).

14 RP represents the most prevalent cause of visual handicap among those of working age, while Leber congenital amaurosis (LCA), a congenital form of RP is the most prevalent cause of visual handicap in children (Boughman and Fishman, 1983). RP affects 50,000- 100,000 people in the United States and about 1.5 million world wide (Berson 1996) and fifty percent of these cases are sporadic, the remainder showing dominant, recessive, X- linked or digenic hereditary patterns (Kajiwara et a/., 1994). RP not only represents a disease in isolation but can also form a component of certain syndromes such as Usher’s syndrome (RP and deafness), Kearns-Sayre syndrome and the Bardet-BiedI syndrome (Dryja etal., 1995).

One of the first steps towards the discovery of RP-causing genes was made in 1984 and involved the localisation of a gene for X-linked RP to the short arm of chromosome X (Bhattacharya et al., 1984). In 1989 the first locus to co-segregate with autosomal dominant RP (adRP) was found on the long arm of chromosome 3 and the disease was subsequently found to be a result of mutations in the rhodopsin gene (Dryjaet a/., 1990a). A second locus for adRP was identified on 6q (Farrar et a/., 1991). Further investigations revealed the presence of mutations within the gene encoding RDS-peripheiin, a component of the photoreceptor outer segment disc membranes (Farrar et a!., 1991, Kajiwara et al., 1991). The discovery of these genes was just the beginning in the search for the diverse genetic basis of RP and now a decade later mutations in a total of 21 genes have now been identified, seven in Leber congenital amaurosis, four in Usher’s syndrome and three in Bardet-BiedI syndrome. In addition 24 disease loci have been mapped for RP or for syndromes incorporating RP but these genes remain to be isolated. Research over the last decade has revealed an extensive genetic heterogeneity and a complex underlying molecular pathology in retinal degeneration. For a comprehensive description and referencing of the genes involved in retinal degeneration please refer to

W W W .sph.uth.tmc.edu/RetNet/disease.htm or www.retina-international.com.

There is a large diversity in the function of the gene products that have been implicated in RP and of those genes, six encode proteins of the phototransduction cascade, which include rhodopsin, the subunits of rod cyclic GMP phosphodiesterase, the subunit of the rod cyclic GMP-gated channel and arrestin. Photoreceptor outer segment disc membranes contain essential structural proteins and mutations in two of these genes have been reported to cause RP; RDS-peripherin and Rom 1 (Kajiwaraet al., 1994). Mutations in two other genes encoding structural components of rod cells and hair cells of the inner ear; harmonin (Verpy et al.,2000) and myosin Vila (Weil et al., 1995) have been described

15 in Usher syndromes types 1A and 1B. Mutations in a gene encoding a protein with homology to co-factor C (involved in the folding of tubulin) co-segragate with X-linked RP (RP2)(Schwahn et al., 1998). Another gene responsible for X-linked RP (RP3) is the RPGR gene (retinitis pigmentosa GTPase regulator) (Meindletal., 1996)

Mutations in genes causing RP, Ushers syndrome or LCA have been discovered in the genes encoding proteins of the extracellular matrix and those involved in cell adhesion. For example arRP causing mutations in CRB1 an orthologue of the Drosophila crumbs gene (den Hollander et a/,.2001)'usherin’ in Usher’s type lia (Eudy et al., 1998), and a cadherin like protein in Usher’s type ID (Bolz etal., 2001).

An essential component of photoreceptor physiology is the retinoid cycle and mutations within five genes encoding proteins of this cycle have also been implicated in RP or LCA (Farrar etal., 2002).

Retinal pigment epithelial cells are responsible for the phagocytosis of outer segment disc membranes, which are constantly being shed. Mutations have been identified in Mertk which encodes a receptor tyrosine kinase in the RCS rat which shows abnormal disc phagocytosis (Gal etal., 2000) and recessive RP in man.

Three transcription factors have also been implicated in the aetiology of retinal degeneration, which include CRX, NRL and TULP1. CRX is thought to be involved in the regulation of photoreceptor specific genes (Chen et al., 1997) and can act individually or synergistically with NRL to mediate photoreceptor specific expression of rhodopsin by binding to the promoter region of that gene (Mitton et al., 2000). TULP1 is an orthologue of the murine tub gene and is thought to be a transcription factor (Hagstrom etal., 1998).

Recent findings have shown that genes involved in pre-mRNA splicing are involved in the pathogensis of RP (McKie et al., 2001, Vithana et al., 2001, Charkarova et al., 2002). Although these genes are expressed in all tissues, it only causes pathology in photoreceptor cells, once again emphasizing the functional diversity of the gene products implicated in human retinal degeneration.

16 #

Figure 1.2 Fundus photographs of a normal human retina (left) panel and RP retina (right) panel Typical features in a RP retina are shown; pigment epithelial thinning, pale optic disc, attenuated retinal vessels and the classical bone spicule' intraretinal pigmentary deposits (Farraret al., 2002).

17 1.2.1 Rhodopsin RP

The first indication that mutations in the rhodopsin gene might be responsible for some cases of RP came from linkage analysis in a large Irish family with adRP (McWilliam etal., 1989). In this family the disease was shown to co-seg regate with that region of the long arm of chromosome 3 to which the rhodopsin gene had been mapped. Based on this study, a number of researchers concentrated on looking for mutations in the rhodopsin gene in patients with adRP. Subsequently Dryja and co-workers (1990) reported a C to A mutation in codon 23 of rhodopsin, resulting in a Pro-to-His substitution, which co- seg ragated with the disease. Approximately 25% of all cases of adRP in the U.K., Europe and the U.S.A. (Ingleheam etal., 1992, Bunge etal., 1993) are caused by mutations in the rhodopsin gene, most of which result in single amino acid substitutions. The Pro-23-His rhodopsin mutation is not commonly found outside the U S A (Farrar et al., 1990) but is the most frequent mutation in the U S A. and was found in 12% of patients with adRP. Since 1990 more than 100 mutations in rhodopsin have been identified which cause disease (see OMIM, www3.ncbi.nlm.nih.gov/0min/).

Rhodopsin is a seven transmembrane G-protein coupled receptor found in the outer segment discs of the rod photoreceptor and is the light sensing protein within these specialised neurons. The ligand of rhodopsin, 11-c/s-retinal, unlike other ligands of G- protein coupled receptors is covalently attached to lysine 296 (as described later). The mutations so far identified in adRP are not localised to any specific domain of rhodopsin but are found throughout the coding region of the gene. The majority of the mutations implicated in adRP are single base pair substitutions. Of these, all but the nonsense mutations at codons 64 and 344 and the splice acceptor site mutations in intron 4, are missense mutations, responsible for the substitution a single amino acid in the polypeptide (Table 1.1). The remaining mutations consist of deletions (Table 1.2) and an insertion of 150 base pairs responsible for adRP has been discovered (Al-Maghthehet al., 1994a) as well as a duplication of four base pairs after nucleotide 998 in exon 5 of rhodopsin disrupting the normal reading (Bareil etal., 1999).

Certain amino acids are known to have essential roles in determining the structure and function of the rhodopsin protein and mutations in these particular codons cause adRP. For instance, three mutant rhodopsins that are unable to bind 11-c/s-retinal due to a substitution at Lysine 296 are found in patients with adRP (Table 1). As described later. In vitro experiments revealed that Lysine 296 mutants could activate transducin (Robinson et

18 al., 1992, Rim and Oprian 1995). Other important residues in the rhodopsin protein are asparagine residues 2 and 15 the sites of N-linked glycosylation, which are within the tripartite consensus sequence (Asn-X-Ser/Thr) for N-linked glycosylation. A single base pair mutation has been described at codon 17 (threonine-17 of the consensus sequence) which cosegregates with adRP (Sung et a/., 1991a). Autosomal dominant RP mutations disrupting the disulphide bond which forms between two cysteines at codons 110 and 187 and is important for the attainment of the correct tertiary structure of the rhodopsin protein (Karnik and khorana, 1990) have also been described in patients with the disease (Table 1). Although the majority of mutations in the rhodopsin gene are responsible for adRP two mutations have been described which lead to RP only in the homozygous state (Table 1.1). Furthermore, mutations that are causative for a condition known as congenital stationary night blindness (CSNB) have been reported (Table 1.4). CSNB is a separate, clinically and genetically heterogeneous disease, which is not accompanied by retinal dystrophy and is considered to be a purely functional defect. Three heterozygous missense mutations in rhodopsin have been described which are responsible for CSNB and in vitro analysis of these proteins indicates that they are active in G-protein signaling in the absence of 11-c/s-retinal (Sieving etal., 1992, Dryja etal., 1993, Rao etal., 1994)

Table 1.1 Missense/nonsense mutations in Rhodopsin causing RP

Codon Amino acid substitution Reference 4 Thr-Lys Bunge etal., 1993 15 Asn-Ser Kranich etal., 1993 17 Thr-Met Sung etal., 1991a 23 Pro-Hls/Leu/Ala Dryja etal., 1990a, Dryja etal., 1991, Oh etal., 2000 28 Gin-Hls Bunge etal., 1993 40 Leu-Arg Al-Maghtheh etal., 1994b 44 Met-Thr Reig etal., 1994 45 Phe-Leu Sung etal., 1991a 46 Leu-Arg Rodriguez ef a/., 1993 51 Gly-Vai/Arg/Ala Dryja etal., 1991, Dryja, 1992, Macke etal., 1993, 53 Pro-Arg ingleheam etal., 1992 58 Thr-Arg Dryja etal., 1990b 64 GIn-stop Macke etal., 1993, 87 Val-Asp Sung etal., 1991a 89 Gly-Asp Sung etal., 1991a 106 Giy-Trp/Arg Sung etal., 1991a, Ingleheam etal., 1992 109 Gly-Arg Goliath etal., 1998 110 Cys-Tyr/Phe Dryja, 1992, Fuchs etal., 1994, 114 Gly-Asp/Val Vaithinathan etal., 1994, Dryja etal., 2000 125 Leu-Arg Dryja etal., 1991 127 Ser-Phe Souied etal., 1994

19 131 Leu-Pro/Arg Fuchs etal., 1994, Souied etal., 1994 135 Arg-Leu/T rp/Gly/Pro Sung et al., 1991a, Sung et al., 1991a, Bunge et al., 1993, Gal etal., 1997, 136 Tyr-stop Sanchez ef a/., 1996 137 Val-Met Ayuso etal., 1996 140 Cys-Ser Macke etal., 1993 150“ Glu-Lys Kumaramanickavel etal., 1994 164 Ala-GluA/al Vaithinathan etal., 1994, Fuchs etal., 1994 167 Cys-Arg Dryja etal., 1991 170 Pro-Arg Sohocki ef al., 2001 171 Pro-Leu/Ser/GIn Dryja ef al., 1991, Vaithinathan ef al., 1994, Antinolo ef al., 1994 174 Gly-Ser Gal etal., 1997 178 Tyr-Cys/Asn Sung etal., 1991a, Souied etal., 1994, 180 Pro-Ala Gal etal., 1997 181 Glu-Lys Dryja etal., 1991 184 GIn-Pro Dryja etal., 2000 182 Gly-Ser Sheffield etal., 1991 185 Cys-Arg Sohocki etal., 2001 186 Ser-Pro/Trp Dryja etal., 1991, Gal etal., 1997 187 Cys-Tyr Richards etal., 1995 188 Gly-Arg/Glu Dryja etal., 1991, Macke etal., 1993 190 Asp- Asn/Gly/Tyr Keen ef al., 1991, Sung ef al., 1991a, Fishman ef al., 1992 207 Met-Arg Farrar ef a/., 1992 209 Val-Met Macke etal., 1993 210 Pro-Leu Sohocki etal., 2001 211 His-Pro/Arg Keen etal., 1991, Macke etal., 1993, 216 Met- Lys/Arg al-Maghtheh etal., 1994b, Haim etal., 1996, 220 Phe-Cys Bunge etal., 1993 222 Cys-Arg Bunge etal., 1993 249’ Glu-stop Rosenfeld etal., 1992 267 Pro-Leu/Arg Sheffield etal., 1991, Souied etal., 1994 270 Ser-Arg Gal etal., 1997 296 Lys-Glu/Met/Asn Keen etal., 1991, Gal etal., 1997, Sohocki etal., 2001 297 Ser-Arg Souied etal., 1994 328 Leu-Pro Gal etal., 1997 342 Thr-Met Gal etal., 1997 344 GIn-stop Sung etal., 1991a 345 Val-Met/Leu Dryja etal., 1991, Vaithinnathan etal., 1994 346 Ala-Pro Borrego etal., 1996 347 Pro-Leu/Ser/Arg/Ala/GIn/ Dryja etal., 1990b, Dryja etal., 1990b, Gal etal., 1991, Thr Macke ef al., 1995, Vaithinathan ef al., 1994, Gal ef al., 1997 349 stop Bessant etal., 1999 these mutations lead to autosomal recessive RP

20 Table 1.2 Deletions In the Rhodopsin gene that cause adRP Codon a.a. Deletion Reference 68-71 Del Leu-Arg-Thr-Pro Keen etal., 1991 255/256 Del lie Ingleheam etal., 1991 264 Del Cys Vaithinathan etal., 1994 318* Del Leu Sohocki etal., 2001 332-338 6 a.a. del and Gal etal., 1997 frameshift 340 Del Thr and frameshift Horn etal., 1992 341-443 2 a.a. del and Horn etal., 1992 frameshift 340-348 8 a.a. deletion Restango etal., 1993 in an isolated individual with RP

Table 1.3 Nucleotide substitutions at spiice sites causing adRP

Donor/acceptor Location Substitution Reference ds (Intron 4) +1 G-T Macke etal., 1993 as (Intron 4) -1 G-A Bell et al., 1994 as (Intron 4) -1 G-T Reig et al., 1996

Table 1.4 Mutations in the rhodopsin gene causing CSNB

Codon a.a. substitution Reference 90 Gly-Asp Sieving etal., 1992 94 Thr-lle Al-Jandal etal., 1999 292 Ala-Glu Dryja etal., 1993

21 1.3 Rhodopsin Structure and Function

1.3.1 Structure:

Rhodopsin consists of rod opsin apoprotein, an integral membrane protein of 348 amino acids that folds into seven transmembrane helices and ligand 11-c/s-retinal (forming rhodopsin, in this thesis I refer to the receptor without ligand as rod opsin and the gene as rhodopsin). The seven transmembrane a-helices vary in length from 19 to 34 residues and tilt at various angles with respect to the membrane surface, in addition one other a- helix is found in the cytoplasmic domain of the protein (Fig. 1.3). A network of hydrogen bonds and hydrophobic interactions form in the middle of the transmembrane region constraining rhodopsin in the inactive state. Inter-helical hydrogen bonds are mediated by the side chain of asparagine-55 (Asn-55) in TM 1 to aspartate-83 (Asp-83) in TM 2 and to the peptide carbonyl group of alanine-299 (Ala-299) of TM 7 (Palczewski et al., 2000). Asp-83 is in turn connected via a water molecule to the peptide carbonyl group of glycine- 120 (Gly-120) in TM 3. Another region mediating constraints on the heptahelical bundle includes Asn-78 of TM 2 which forms hydrogen bonds with the OH groups of serine-127 (ser-127) of TM 3, and threonine-160 (Thr-160), tryptophan-161 (Trp-161) of TM 4. Rhodopsin is modified by two palmitoyl groups at cysteine-322 and cysteine-323 in the cytoplasmic tail, anchoring this segment of the protein to the lipid membrane (Fig. 1.3) Ovchinnikov et al., 1988, Papac at a/., 1992) and a disulfide bond is formed between the intradiscal cysteines, Cys-110 and Cys-187 which is essential for the formation of functional rhodopsin (Fig. 1.3) (Karnik and Khorana, 1990).

The transmembrane helices of opsin form a pocket for the binding of 11-c/s-retinal and a space is created between helices 3,4,5,6 and 7 that is the presumptive binding site for retinal. The chromophore 11-c/s-retinal is attached covalently to opsin, through the formation of a Schiff base linkage (Fig 1.4) between the aldehyde moiety of retinal and the 8-amine of lysine 296 in the middle of TM 7 (Bownds, 1967, Wang etal., 1980). The Schiff base is protonated (Oseroff and Callender, 1974) and the positive charge on the nitrogen is stabilised through an electrostatic interaction with the side chain of glutamate-113 on TM3, the so called Schiff base counterion (Zhukovsky and Oprian, 1989, Sakmar et al., 1989, Nathans, 1990) thus forming a salt bridge between TM3 and TM7. Glutamate-113 is highly conserved among all known vertebrate visual pigments reflecting the importance of this residue in ligand binding.

22 84 ' " 85 E2 «4 f

C3 340

Figure 1.3 Secondary structure diagram of bovine rhodopsin. Rhodopsin’s polypeptide chain traverses the lipid bilayer seven times as transmembrane a-helical segments HI to H7. Amino acid residues are shown in single letter code. The amino terminal tail and intradiscal domain is towards the top, and the carboxy-terminal tail and cytoplasmic domain is toward the bottom. Retinal is covalently linked to lysine-296 in H7. An essential disulphide bond links Cys-110 and Cys-187. Two residues in H8, Cys- 322 and Cys-323 are palmitoylated and oligosaccharide chains are attached at Asn-2 and Asn-15. Hydrophilic loop sequences are exposed to the cytoplasmic surface (C1-C3) and to the intradiscal or extracellular surface (E1-E3). Diagram is from Menon et al., 2001.

23 A.

H CH

11-c/s-Retinal

H CH

CH Ali-frans-retinal

HjC CH 3 ÇH3 CH,OH

AII-fraMs-retinol (Vitamin A)

/ T. H , R— G + HpN— (CH2)4— Opsin R— C ^ N — (CH;),,— Opsin 4- H^O H H 11-c/s-Retinal Lysine side chain Protonated Schiff base

Figure 1.4 A) Structure of 11-c/s-retinal, aii-tracs-retina! and all-fra/7s-retinol (Vitmain A). B) Formation of a Schiff-base linkage between 11-c/s-retinal and Lys-296. (From Stryer, 1988)

24 Conserved residues on TM helices 3, 5, and 6 largely define the retinal binding pocket and there are a number of hydrophobic interactions around the chromophore. Photoaffinity labelling, mutational analysis and modeling show that the (3-ionone ring of retinal associates with TM 3, 5 and 6 particularly TM 6 Tryptophan-265 and Tyrosine-268 (Han et al., 1997). The 09 methyl group of 11-c/s-retinal appears to associate with TM 3 Glycine- 121 (Han et a!., 1997). In addition to forming a salt bridge between TM 3 and TM 7 a number of hydrogen bonds form between 11-c/s-retinal and amino acid residues lining the binding pocket of the ligand and the recent crystal structure of rhodopsin with attached chromophore has revealed that at least 16 amino acids are within 4.5A of 11-c/s-retinal.

The responsiveness of rhodopsin to light depends on the presence of 11-c/s-retinal, which is a very effective chromophore. 11-c/s-retinal gives rhodopsin a broad absorption maximum in the visible region of the spectrum with a peak at 500 nm (Wald, 1968) (Fig. 1.5).

tight dark I

300 400 600 Wavelength (nm)

Figure 1.5 The absorbance spectrum of bovine rhodopsin. The 280 nm peak represents the protein component of rhodopsin and the broad absorbance spectrum with a X^ax. of 500 nm represents rhodopsin (dark). After exposure to light, rhodopsin is converted to metarhodopsin II (meta-ll) with a X^ax value of 380 nm (light) characteristic of an unprotonated Schiff base imine (from Menon eta!., 2001)

Bovine rhodopsin was first shown to be a glycoprotein in 1968 (Heller 1968) and is glycosylated at asparagine residues 2 and 15 (Hargrave, 1977). From studies of the

25 oligosaccharide structure of rhodopsin from several species [cattle (Fukuda et a/., 1979, Liang ef a/., 1979), frog (Duffin etaL, 1993), human (Fujita at ai, 1994) and rat (Endo at ai, 1996) it has been demonstrated that the major oligosaccharide from each had the same structure:

Man a1 6 Man (31— ► 4 GlcNAc pi — ^4 GlcNAc 3 GlcNAc p i—^2 Man a1

In each of these species, two sites of glycosylation per mole of rhodopsin have been observed; Asn-2 and Asn-15. However, in bovine rhodopsin a small number of Asn-2 and

Asn-15 residues had N- linked oligosaccharides with Man4 and Mans (Fukuda at ai, 1979). Additional oligosaccharides of lesser abundance containing galactose and sialic acid have also been observed in frog, human and rat rhodopsins.

1.3.2 Function of rhodopsin

In 1958 George Wald and co-workers showed that light isomerises the 11-c/s-retinal group of rhodopsin to all-frans retinal. The cis-trans isomérisation of retinal (bleaching) and the consequent conformational changes in rhodopsin are the initial events in visual transduction. The Schiff-base linkage of retinal moves approximately 5 A in relation to the ring portion of the chromophore. The isomérisation of retinal takes place within a few picoseconds of the absorption of a photon, yielding a rhodopsin intermediate called bathorhodopsin, which contains a strained all-trans form of the chromophore. The protein conformation of rhodopsin continues to change as reflected in the formation of a series of transient intermediates with distinctive spectral properties (Fig. 1.6). The Schiff base linkage becomes deprotonated in the transition from metarhodopsin I to II which takes about a millisecond. Metarhodopsin II, called activated rhodopsin (R*), binds to and activates transducin (Emeis at ai, 1982), triggering an enzymatic cascade. The unprotonated Schiff base in metarhodopsin II is hydrolysed to yield opsin and aW-trans- retinal, which diffuses away from the protein. Isomérisation of 11-c/s-retinal results in conformational changes of the rhodopsin protein yielding activated rhodopsin. Work from several laboratories produced a model of the structural changes that occur during receptor

26 activation and proposed that the helical bundle opens at its cytoplasmic side, exposing various regions for interaction with transducin (reviewed by Meng and Bourne, 2001). The evidence indicates that there is activation-induced separation of TM 3 and TM 6 with an

hv Rhodopsin (SOOnm) ps j Bathorhodopsin (543 nm) ns I Lumirhodopsin (497 nm)

jlS j Metarhodopsin I (480 nm) ms I Metarhodopsin II (380 nm) (R*) S i Opsin + All-frans-retinal (380 nm)

Figure 1.6 Intermediates in the photolysis of rhodopsin. The wavelength of the absorption maximum of each species and the time constant of each transition are given. R* is activated rhodopsin that is able to catalyse guanine nucleotide exchange by transducin. outward movement (away from the helical bundle) of the cytoplasmic ends of TM 3, TM6 and TM7 (Meng and Bourne, 2001). A clockwise rotation of TM 6 about its helical axis viewed from the cytoplasmic side may also occur (Meng and Bourne, 2001).

1.4 Phototransduction in rod cells

The absorption of light by rhodopsin in rods triggers a cascade of events that leads to a change in ionic fluxes across the plasma membrane of these cells and a consequent change in membrane potential (Yau, 1994). A key molecule in the cascade is the cyclic guanosine 3', 5'-monophosphate GMP (cGMP) which acts as a second messenger. Cyclic cGMP controls ionic fluxes by opening a specialised species of ion channels, the cGMP- gated ion channel, which allow an inward current carried largely by Na^ ions to flow into the cell. In darkness the level of cGMP in a photoreceptor is high, and two currents predominate in the cell: an inward current that flows through cGMP-gated channels, which

27 are confined to the photoreceptor outer segment, and an outward K* current that flows through non-gated selective channels, which are confined to the inner segment. The intracellular concentrations of Na* and are maintained at a steady level by high levels of a sodium-potassium pump in the inner segment of the rod cell, which pumps Na* out and in to the cell. The outward current carried by the K* channels tends to hyperpolarise the photoreceptor toward the equilibrium potential for K* (approximately -70 mV). The steady inward current known as the dark current tends to depolarise the photoreceptor. Therefore in darkness the photoreceptor’s membrane potential is around -40 mV, significantly more depolarised than that of most neurons (Yau, 1994).

The activation of rhodopsin by light leads to a reduction in the cytosolic concentration of the second messenger cGMP (Yau, 1994). Two enzymes control the concentration of cGMP. It is synthesised from GTP by retinal guanylyl cyclase (retGC) and it is broken down to 5'-cGMP by cGMP phosphodiesterase, a protein associated with the disc membrane. The concentration of cGMP in rod outer segments is high and its concentration drops upon illumination. This drop in concentration is brought about by a specific phosphodiesterase for cGMP that is activated by a cascade triggered by light. Activation of rhodopsin by light leads to the hydrolysis of 10® cGMP molecules in one second. This amplification is achieved through the action of the G-protein transducin (T). Transducin a member of the family of signal transducing G-proteins that is rod cell specific has three subunits Ta, Tp and Ty. Like other G-proteins the activation of transducin requires the characteristic interaction with guanine nucleotides and the T» subunK binds a molecule of GDP tightly. Upon interaction of transducin with activated rhodopsin in the disc membrane the T« subunit exchanges GDP for GTP and now becomes active (Fig 1.X). Free T«-GTP then activates cGMP phophodiesterase by binding to and removing the inhibitory y subunit, thus releasing the catalytic a and p subunits in an active form which catalyse the breakdown of cGMP to 5’-GMP. This light evoked decrease in cGMP results in the closure of cGMP-gated channels in the rod cell by binding directly to the cytoplasmic face of the channel; a channel becomes activated by the cooperative binding of at least three molecules of cGMP. When light reduces the level of cGMP, closing cGMP-gated channels, the inward current that flows through these channels is reduced, and the cell membrane becomes hyperpolarised. As with other a subunits of G-proteins, a GTPase activity is intrinsic to the a subunit of transducin, which slowly converts light induced Ta- GTP back to Ta-GDP, which results in the inactivation of cGMP phosphodiesterase. However, the intrinsic GTPase activity of Ta is much slower than the rate of termination of

28 phototransduction, but a protein has been isolated which stimulates the GTPase activity of Ta and belongs to gene family called regulators of G protein signaling (RGS) and is termed RGS9 (He et al., 1998). Visual transduction can be shut down by a rod cell enzyme rhodopsin kinase, which phosphorylates the C-terminus of light activated rhodopsin, which is less able to activate transducin. At high ambient light the level of opsin phosphorylation is such that the protein arrestin binds to opsin totally blocking activation of transducin and causing a shut down in rod-cell activity.

1.5 Rhodopsin processing in vivo

1.5.1 Role of N-linked glycosylation

Kaushal and co-workers (1994) investigated the role of rhodopsin N-linked glycosylation and discovered that glycosylation is not required for the folding of rhodopsin to the inactive state structure but that unglycosylated metarhodopsin II photoproduct was unstable and had an abnormally short lifetime, rendering it defective in signal transduction. In the same report the wild type bovine rhodopsin gene was expressed in cell culture in the presence of the N-linked glycosylation inhibitor Tunicamycin (TM). The resulting non-glycosylated rod opsin was palmitoylated, translocated to the cell surface and bound 11-c/s-retinal to form rhodopsin with characteristic spectral properties. However, the non-glycosylated rhodopsin showed diminished light dependent activation of transducin (Gy) in vitro. Furthermore, amino acid substitutions at or near Asn-2 had little effect on rod opsin cell surface expression, spectral properties, or Gy activation. In contrast, substitutions at Asn- 15 or Lysine-16 formed opsins that were not transported to the cell surface but retained in the ER and did not associate with 11-c/s-retinal, suggesting these mutant opsins were misfolded. A mutation in the glycosylation consensus sequence T17M that is associated with adRP activated Gy poorly (Kaushal eta!., 1994).

The major rhodopsin (Rhi) in Drosophila has two consensus sites for N-linked glycosylation, at asparagine residue 20 (Asn-20) and asparagine 196 (Asn-196) but Rhi is exclusively glycosylated at Asn-20 (Katanosaka et al 1998). Insights into the role of N- linked glycosylation in Rhi maturation came from the work by OTousa, 1992 and Webel et al., 2000. These investigators revealed that substitution of N20 with isoleucine (N20I) in Rhi prevented translocation of the protein to the rhabdomere, which is the equivalent of the mammalian rod outer segment. Furthermore, mutant N20I Rhi accumulated in the ER and disrupted the normal maturation and translocation of WT Rhi (Webel et al., 2000).

29 The authors concluded that in Drosophila, glycosylation at Asn-20 of Rh1 is essential for its maturation and translocation to the rhabdomere (Webeleta!., 2000).

1.5.2 The role of ligand In rhodopsin biogenesis

The retinal ligand of rhodopsin (Rh1) in Drosophila is 11-c/s-3-hydroxyretinaI (Taninmura et al., 1986) and is essential for the maturation of protein. In the absence of ligand immature glycosylayed Rhi cannot progress to the mature non-glycosylated form and is rapidly degraded suggesting that 11-c/s-3-hydroxyretinal is required for the correct folding and processing of the protein (Ozaki et al., 1993). In mammalian rods 11-c/s-retinal is added soon after synthesis of opsin apoprotein in the ER (St Juleset al., 1989) and it is unknown whether opsin requires the binding of 11-c/s-retinal to translocate to the rod outer segment. However, 11-c/s-retinal is important in rod and cone function and a lack of vitamin A produces an initial decrease in rod sensitivity and eventually leads to the loss of both rods and cone outer segments (Dowling and Gibbons 1960, Hayes, 1974, Kemp et al., 1989). It has been proposed by Fain and Usman (1993) that photoreceptor degeneration observed in vitamin A deficiency is similar to degeneration that occurs in continuous exposure to light (Noell and Albrect, 1971). Fain and Usman (1993) suggest that free opsin, not bound to ligand, present during vitamin A deprivation excites the visual transduction cascade, which would produce a constant ‘equivalent light’ that triggers photoreceptor degeneration. However, the hypothesis that continuous activation of the visual pathway triggers photoreceptor degeneration is less likely, given the observation that mutations in cGMP phosphodiesterase which do result in the continues activation of the visual transduction cascade cause CSNB (Gal et al., 1994) rather than photoreceptor degeneration. Mutations in other genes of the visual transduction cascade can also cause CSNB and not RP; for instance, missense mutations in the a-subunit of rod transducin (Dryja et ai, 1996).

1.5.3 Molecular chaperones and rhodopsin biogenesis

The Drosophila protein Nina A is essential for the biogenesis and maturation of Rhi forming a stable and highly specific complex with the receptor (Stamnes et ai, 1991, Baker et ai, 1994). Nina A (with a central cyclophilin homology domain) is the only known molecular chaperone that has been identified to have a role in rhodopsin biogenesis and maturation, and a mammalian orthologue of Nina A has not yet been identified. However Ferreira and co-workers identified a putative retina specific cyclophilin from a bovine retina

30 cDNA library which was predominantly expressed in cones (Ferreira et al., 1995). Of the two isoforms discovered type II is identical to RanBP2 (Ran-binding protein 2) which contains a cyclophilin domain that can form a complex with bovine long and medium- wavelength (red/green) sensitive opsin (Ferreiraat a!., 1996). Cyclophilins induce cis- trans isomérisation of peptidyl-prolyl bonds in proteins and Nina A and RanBP 2 may have this particular role in the folding of rhodopsin and red/green opsin respectively (for further information see Chappie at a!., 2001).

1.6 Phenotypic behavior of rhodopsin RP mutants.

1.6.1 In vitro studies

To increase our understanding of the molecular mechanisms of rhodopsin RP the biochemical differences between wild-type (WT) and the adRP mutants of rhodopsin were investigated using cell culture models. The seminal work by Sung and co-workers in 1991 and 1993 revealed differences in the biochemical properties of the RP mutants of human rhodopsin expressed in cultured cells (Sung at a/., 1991b, Sung at a/., 1993); which included variations in the ability of mutants to regenerate with 11-c/s-retinal to form functional rhodopsin and also variations in yield and cell surface expression of the receptor. The correct tertiary structure of the mutant opsins was assessed by the ability to bind 11-c/s-retinal and produce pigments with the normal WT rhodopsin absorbance spectra, protein yield and cellular localisation. Two distinct classes of mutants emerged form this work: class I mutants resemble WT in yield, regeneration with 11-c/s-retinal and cellular localisation i.e. plasma membrane staining. The majority of rhodopsin mutations in class I lie in the C-terminus of the protein and do not effect the folding of the protein (eg. Q344ter, V345M, P347S). Class II mutants gave lower yields, regenerated with 11-c/s- retinal to variable extents or not at all, and show diminished plasma membrane expression. A further distinction was made within class II mutants between those mutants that showed predominantly intracellular localisation (Class lia) and those that also show significant cell surface localisation (Class lib). Rhodopsin mutations characterised as class II were found in the intradiscal domain, the transmembrane domains and the cytoplasmic domain of the protein. The analysis of rhodopsin adRP mutations in cell culture by Sung at a!., 1991b and Sung at a!., 1993 and other investigators (Kaushal and Khorana, 1994) has revealed that the majority of rhodospin mutations result in a protein that cannot attain its native conformation. In addition to these mutants three particular mutations that are responsible for adRP result in amino acid substitution of Lysine-296

31 (Keen et al., 1991, Gal et al., 1997, Sohocki et al., 2001) the site of attachment of 11-c/s- retinal, thus preventing its binding to opsin. K296E rod opsin purified from cultured cells was found to activate transducin (from bovine retinas) in the absence of light with activities comparable to that of light activated WT rhodopsin, revealing a novel phenotype for this particular mutant (Robinson et ai, 1992). However, K296E rhodopsin is inactive in vivo as this mutant protein was found in a stable complex with arrestin and was not, as previously thought, constitutively active (Li et ai, 1995).

1.6.2 Transgenic models of Rhodopsin RP

Research in to the phenotypic differences of rhodopsin RP mutants has continued using transgenic mouse technology to produce mouse models of retinal degeneration. These models revealed that RP rhodopsin transgene expression in mice lead to photoreceptor cell death by apoptosis (Portera-Cailliau et al., 1994) and in addition Drosophila mutant rhodopsin causes photoreceptor degeneration by apoptosis (Davidson and Steller. 1998). Transgenic mice have been created with a number of rhodospin RP mutants including class I and class II mutants. Of the class I mutants, mice expressing, P347S (Huang et ai, 1993, Weiss et ai, 1994, Li et ai, 1996), Q344ter (Sung et ai, 1994) and swine expressing P347L (Petters et ai, 1997, Tso et ai, 1997, Li et ai, 1998) were engineered. Of the class II mutants P23H (Olsson et ai, 1992, Roof et ai, 1994) K296E (Li et al 1995) and T17M (Li et ai, 1998) were expressed in mice. There appears to be no early developmental defect involving the retina in RP mouse models but retinal abnormalities (including abnormal ERGs, shorter than nonrial rod outer segments and rod and cone cell loss) are detected approximately 10 days after birth.

A number of Class I adRP mutants of rhodopsin are found within the rhodopsin C-terminal sequence QVAPA (QVSPA in frog and fish) which are the last 5 amino acids of the polypeptide and is highly conserved among different species. Amino acid substitutions of proline and valine along with a Q344 termination in this conserved sequence have been associated with adRP (Table 1). The carboxy-terminus of rhodopsin together with the palmitoylation of cysteines 322/323 are important for efficient rhodopsin transport and sorting to the outer segement (Tam et ai, 2000). The wild type carboxy-terminus binds to Tctex-1 protein, a widely expressed dynein light chain which is abundant in photoreceptor inner segments and is thought to guide the transport of rhodopsin laden post Golgi vesicles to the rod outer segment (Tai et ai, 1999). However, palmitoylation of cysteine residues 322 and 323 which provide a point of membrane attachment is also required for

32 efficient targeting of rhodopsin to outer segments in addition to the distal amino acids of the carboxy-terminus (Tam et al., 2000). The C-terminal RP mutants of rhodopsin just described fail to bind to Tctex-1 and in transgenic animals these mutants accumulate in the cell body of rod cells and show defective intracellular transport. For example, one class I mutant (Q344ter) showed inefficient outer segment localisation of the mutant opsin but not of the endogenous wild type opsin (Sung et al., 1994) and a second mutant (P347S) induces accumulation of vesicles at the base of the outer segment (Li et a/., 1996).

Of particular relevance to my work are the transgenic mice expressing the adRP mutants of rhodopsin P23H and K296E. The P23H transgenic mice revealed that P23H rod opsin accumulates in the inner segments and mislocalises to the synapses of rod cells at the outer plexiform layer but also translocates to the rod outer segment (Olsson et a/., 1992, Roof et a!., 1994). Furthermore increased expression of the P23H transgene led to a more rapid degeneration of the retina suggesting a gene dosage effect (Olssonet a/., 1992). K296E rhodopsin constitutively activates transducin in vitro (Robinson et ai., 1992) and causes RP in vivo (Keen et ai., 1991, Vaithinathan et ai., 1994). It has been suggested that this mutant causes disease by constitutive activation of the phototransduction cascade. Support for this hypothesis not only comes from the finding that K296E rhodopsin constitutively activates transducin in vitro but also from observations that constant light exposure damages photoreceptor cells (Lavail, 1980, Rapp and Williams, 1980) and opsin apoprotein without ligand was reported to activate phototransduction without light, albeit at six orders of magnitude lower efficiency (Fain and Cornwall, 1993). Surprisingly however, K296E in transgenic mice was found to be inactivated by phosphorylation and the binding of arrestin (Li et ai., 1995) thus showing that the K296E mutant does not cause photoreceptor degeneration by continuous activation of phototransduction.

Two classes of disease expression were found in patients with different mutations in the rhodopsin gene and disease type was found to be allele specific (Cideciyan et ai., 1998). Class A mutants (R135G, R135L, R135W, V345L and P347L) led to severely abnormal rod function across the retina early in life, in comparison class B mutants (T17M, P23H, T58R, V87D) were compatible with normal rods in adult life in some retinal regions or throughout the retina and there was a slow stereotypical disease sequence (Cideciyan et ai., 1998). Class A families reported onset of night blindness in early life where as class B families reported little or no night vision symptoms and rod function was relatively

33 preserved. Some class A disease mutants (V345L and P347L) have been described as class I mutants in vitro i.e. attain the native conformation and bind 11-c/s-retinal. However, three mutants in class A disease type involve R135G, R135L, R135W (arginine-135 is highly conserved in members of the rhodopsin GPCR family) and in vitro studies of this mutant showed abnormal folding and poor regeneration with 11-c/s-retinal (Sung et al., 1991, Sung et al., 1993). Most class B disease mutants show folding abnormalities in vitro. Thus the relationship between in vitro classification of rhodopsin mutants and the human disease is obviously complex.

1.7 N-linked glycosylation and ER quality control

1.7.1 N-linked Glycosylation

The majority of polypeptides which enter the secretory pathway at the ER are glycosylated as the nascent polypeptide chain extends in to the ER during translation. A branched oligosaccharide group containing three glucose, nine mannose and two N- acetylglucosamine molecules (core unit) (Fig. 1.7) is added to asparagine residues found within the tripeptide recognition sequence Asn-X-Ser/Thr (X is any amino acid except proline) of the growing polypeptide within the ER.

This core oligosaccharide is linked to dolichol a long chain polyisoprenoid lipid found within the ER membrane and long enough to span a phospholipid bilayer membrane four or five times and acts as a carrier of the oligosaccharide in the ER (Abeijon and Hirschberg, 1992). The initial steps involving the linking of sugar groups to dolichol involve a complex set of reactions utilising membrane attached enzymes of the ER, in which two N- acetylglucosamine, and five mannose residues, in the form of nucleoside sugars synthesised in the cytosol, are added one at a time to dolichol phosphate on the cytosolic face of the ER (Fig. 1.8) (Abeijon and Hirschberg, 1992). Tunicamycin, an analogue of UDP-N-acetylglucosamine, blocks the first enzyme of this pathway, and thus inhibits the synthesis of all N-linked oligosaccharides (Duksin and Mahoney, 1982). The now formed dolichol pyrophosphoryl oligosaccharide is flipped into the lumen of the ER where the remaining four mannose and all three glucose residues are added (Abeijon and Hirschberg, 1992). The core oligosaccharide is transferred en bloc by an ER enzyme, oligosaccharide-protein transferase, from the dolichol carrier to an asparagine residue on the nascent polypeptide. Once the core oligosaccharide has been added to growing polypeptides, almost immediately, the terminal glucose and mannose residues are

34 removed by ER glucosidases and mannosidases. Further modifications to the N-linked oligosaccharides take place in the Golgi, where different processing pathways lead to complex, high mannose, and hybrid oligosaccharides respectively. These pathways involve a sequential set of enzymatic reactions that add and remove specific sugar residues from the oligosaccharide as glycoproteins move through the c/s-, the medial-, and the frans-Golgi vesicles. The primary amino acid sequence of a protein and protein conformation determines whether an oligosaccharide is attached to a particular Asn-X- Ser/Thr sequence in the ER, since not all sequences are utilised. For instance the rapid folding of a segment of polypeptide containing a tripeptide sequence may restrict access of the oligosaccharide transferase enzyme and prevent the transfer of an oligosaccharide to it. Likewise the conformation of a segment of a protein may determine whether in the Golgi, a particular N-linked oligosaccharide becomes complex, hybrid or high mannose.

An important function of N-linked oligosaccharides in eukaryotic cells is the promotion of protein folding of newly synthesised polypeptides in the ER (Helenius, 1994). When N- linked glycosylation of nascent polypeptides is prevented, a frequent occurrence is the production of misfolded proteins that fail to reach their native conformational state (Varki, 1993, Helenius 1994). A well studied example of glycosylation dependent protein folding is the vesicular stomatitis virus (VSV) membrane protein, which normally translocates to the plasma membrane in cultured cells but was retained within the ER upon inhibition of the activity of ER glucosidases (Hammond and Helenius 1994). Not all proteins require added oligosaccharide motifs to facilitate their folding and this dependence varies between proteins and physiological context e.g. non-glycosylated rod opsin can attain its native conformation with characteristic spectral properties. Furthermore, N-linked oligosaccharides are not usually essential for maintaining the overall folded stucture once a glycoprotein has folded (Helenius 1994).

35 ■4 Glucosidase I J

[ Glucosklase II J

W J2)

—fMannosidase 11

fMannostdase 111

1.4) I GlcNAc I IP""' I GlcNAc I

— Asn—

Figure 1.7 Diagramatic representation of the N-linked core oligosaccharide added to growing polypetides within the ER lumen. The oligosaccharide chain contains 14 saccharide units in a branched configuration. Terminal glucose and mannose residues are removed in the ER by glucosidases and mannosidases. GlcNAc is N-acetylglucosamine, Man is mannose and Glc is glucose (from Ellgaard et al., 1999).

UMP UDP ■ UOP 5GDP-# 5 GOP

I . A/.acetylglucosamine

Mannose

Glucose

Figure 1.8 Diagrammatic representation of the biosynthesis of the dolichol pyrophosphoryl oligosaccharide precursor of N- linked oligosacharides. Two N-acetylglucosamine and five mannose residues from UDP sugars are added one at a time to a dolichol phosphate on the cytosolic face of the ER. Tunicamycin blocks the first enzyme in this pathway and thus inhibits the synthesis of all N-linked oligosaccharides. The dolichol pyro phosphoryl oligosacchaide is flipped to the lumenal face where the remaining 4 mannose and all 3 glucose residues are added (from Abeijon and Hirschberg, 1992)

36 1.7.2 ER Quality Control

The ER is the site of synthesis and folding of proteins destined for entry into the secretory pathway. The folding of proteins in the ER is facilitated by a number of molecular chaperones whose chaperone activity require the hydrolysis of ATP, such as BiP (immunoglobulin heavy chain binding protein also known as Glucose-regulated protein Grp78, Haas and WabI, 1983) of the Hsp70 family of molecular chaperones and Grp94 (Sorger and Pelham, 1987) of the Hsp90 family of molecular chaperones. The members of the Hsp70 chaperones family are highly conserved ATPases with an ability to couple ATP hydrolysis to the binding and release of substrate proteins with regulatory imput from auxiliary proteins known as co-chaperones (Agashe and Marti 2000). The Hsp70s are thought to have a major role in facilitating protein folding and protect nascent and newly synthesised polypeptides against aggregation. Hsp70s have two functional domains, an N-terminal ATPase domain and a C-terminal peptide binding domain which contains a peptide binding cleft and an a-helical extension that acts as a lid covering the peptide binding site (Agashe and Marti 2000). Msp70s bind short hydrophobic peptides of approximately 7 amino acids in length. The binding of ATP by Msp70 opens the peptide binding cleft and hydrolysis of ATP by the N-terminal domain causes the cleft to close thereby retaining the bound peptide of a given polypeptide, which is then released on exchange of ADP for ATP at the N-terminal domain of Msp70. The net result of this reaction cycle is the binding and release of an unfolded polypeptide. Folding of the polypeptide occurs only after release from the chaperone, which interacts with hydrophobic regions, which are destined to be buried in the hydrophobic core of the folded protein. A number of proteins regulate Msp70 ATPase activity and nucleotide exchange including Bag-1, Msp40 and Mip, thereby regulating the binding and release of hydrophobic polypetides (Agashe and Marti 2000). Other chaperones which facilitate folding of proteins include the lectin chaperones, calnexin (CNX) (David et al., 1993) and calreticulin (CRT) (Peterson at a/., 1995) and proteins which catalyse disulfide bond formation (protein disulfide isomerase (PDI), Erp72, Erp57 etc).

Misfolded and incompletely assembled proteins are common side products of protein synthesis in the ER. An elaborate mechanism within the ER consisting of molecular chaperones stringently checks the fidelity of protein folding, retains misfolded or unassembled proteins and prevents their entry into the secretory pathway, eventually targeting these polypeptides for ER associated degradation (ERAD). This process of conformation dependent molecular sorting of newly synthesised proteins has been termed

37 quality control (Hurtley and Helenius, 1989). An important component of the quality control mechanism of the ER involves two homologous lectin chaperones; CNX, an integral membrane protein and CRT present in the lumen of the ER. These chaperones specifically associate with oligosaccharide moieties of glycoproteins after they have been trimmed by glucosidases I and II, generating the monoglucosylated form

GIUilVlangGlcNAc 2. The resulting monoglucosylated glycoproteins can now associate with CNX and CRT. If there is a glycoslylation site within the first fifty amino acids of a nascent glycoprotein, interaction with CNX and CRT begins cotranslationally (Molinari and Helenius 2000). Also, Erp57, a thiol oxidoreductase homologue of protein disulfide isomerase (Oliver et a/., 1997) forms a complex with both of these lectin chaperones and short-lived disulphide bonds occur between Erp57 and cysteines in CNX and CRT bound glycoproteins (Molinari and Helenius 1999) which serve to facilitate correctly paired disulfide bonds within the bound glycoprotein (Huppa and Ploegh, 1998).

Association of glycoproteins with CNX and CRT involves a binding and release cycle driven by the opposite actions of two soluble ER enzymes, uridine diphosphate (UDP)- glucose:glycoprotein glucosyltransferase (UGGT) and glucosidase II (Hammond et al., 1994, Trombetta et a/., 1996, Parodi, 2000). In the first step of the cycle, the association of glycoprotein substrate with either CNX or CRT is terminated if the glucose residue of the oligosaccharide moiety is cleaved by glucosidase II. After release, if the glycoprotein remains unfolded/misfolded, it is recognised and reglucosylated by UGGT (Helenius and Aebi 2001). This allows the unfolded/misfolded glycoprotein to reassociate with CNX/CRT for another round of the refolding reaction (Fig. 1.9). Glycoproteins may go through this re-glucosylation cycle many times and in order to escape the cycle, a glycoprotein must refold so that it is no longer recognised by UGGT, and remain unglycosylated so as not to rebind CNX/CRT. The folded glycoprotein interacts with ER resident mannosidases I and

II which trims mannose residues generating Man8 -7GlcNAc 2 glycoprotein which is free to leave the ER because it has no glucose residue that could mediate its binding to CNX/CRT (Liu et a/., 1999). In a model proposed by Liu et a/., 1999, persistently unfolded/misfolded glycoproteins are marked for degradation by the action of mannosidase I, which removes the terminal mannose residue generating the

GICiMan8 GlcNAc 2 moiety. After reglucosylation by UGGT the protein rebinds CNX/CRT and because the GICiMan8 GlcNAc 2 form is a poor substrate for glucosidase II, as compared to the GICiManglcNAc 2 form (Grinna and Robbins, 1980), the glycoprotein remains associated with CNX/CRT. The abolished dissociation of CNX and CRT with misfolded/unfolded glycoprotein initiates the entrance of the substrate into the ER

38 associated degradation pathway (ERAD). This involves the retrograde transport of the misfolded protein out of the ER followed by ubiquitin dependent degradation by the proteasome (section 1.8). The identity of the ER lectins involved in the ERAD pathway remain unknown but a recently discovered novel ER a-mannosidase-like protein (EDEM) which binds Man 8 GlcNAc 2 glycoproteins is thought to be directly involved in the targeting of misfolded glycoproteins for degradation (Hosokawa etaL, 2001).

Misfolded proteins within the ER which are destined for degradation are first exported from the ER lumen or ER membrane into the cytosol so as to gain access to the cytosolic proteasomal machinery (Plemper and Wolf, 1999). This mechanism is known as retrotranslocation and it is believed that proteins are exported from the ER via the Sec61 complex or translocon (Plemper and Wolf, 1999). The molecular chaperone Grp78 (BiP) in addition to facilitating protein folding in the ER is also thought to regulate the conformation of the translocon pore (Hamman etaL, 1998).

39 GD Ribosome GICgMang Glc^Mang Mang Man9

c Man 1 ) ( Man I )

Gl&i Mang Man, Man I

Golgi

Cytosol

Proteasome

Figure 1.9 Calnexin/Calreticulin mediated degradation of misfolded glycoproteins. Glucosidase I and II (GI/GII) trim the first two glucose residues from newly synthesised glycoproteins which then associate with calnexin (CNX) or calreticulin. The substrate glycoprotein is released from CNX when Gll removes the terminal glucose of the substrate. If misfolded the glycoprotein is reglucosylated by the folding sensor uridine diphosphate (UDP)-glucose:glycoprotein glucosyltransferase (UGGT, but GT in this diagram) and the unfolded glycoprotein rebinds CNX. Associated with CNX is ERp57 (not shown) which catalyses the formation of any disulphide bonds within the protein. If the protein is terminally misfolded, it is marked for degradation by the action of mannosidase I (Man I). Because Glc^MangGlcNAc^ is a poor substrate for Gll compared to the Glc^MangGlcNACg oligosaccharide the protein is not rapidly released from CNX and is thus targeted for ERAD (Figure from Ellgaard et al., 1999).

40 1.8 The Ubiquitin Proteasome System (UPS)

1.8.1 The proteasome

The proteasome is widely recognised as the central enzyme complex of non-lysosomal protein degradation. It is an essential component of an ATP-dependent proteolytic pathway catalysing the rapid degradation of many enzymes, transcriptional regulators and critical regulatory proteins (DeMartino and Slaughter, 1999). The 268 proteasome complex is a large supramolecular complex, composed of a core proteinase known as the 208 proteasome and two regulatory complexes. The 208 complex comprises four stacked rings that enclose a central chamber where protyolysis occurs (DeMartino and Slaughter, 1999). The 208 catalytic core of the proteasome is built up of seven different a and seven different p-subunits arranged as a cylindrical complex in four stacked rings. Degradation of proteins requires co-operation of the 208 core protease with the 198 cap (also described as PA700), a structure that can bind to both ends of the barrel, resulting in the 268 proteasome. The 198 cap contains at least 20 subunits, some of which are involved in the recognition of ubiquitinated proteins and the removal of ubiquitin from these targeted proteins. It also possesses an ATPase activity that helps to unfold and feed the substrate into the chamber of the 208 core that harbours the proteolytically active sites. Further complexity of proteasome architecture arises from the additional 118 regulatory complex which can associate with the 208 proteasome in the absence of ATP. The 118 complex is a complex of alternating subunits, alpha and beta and is a ring shaped particle which like the 198, caps the proteasome by binding alpha-rings at both or either ends (DeMartino and 8laughter,1999).

1.8.2 Ubiquitin and proteasomal degradation

The vast majority of known protein substrates of the proteasome must be modified by the covalent attachment of a polyubiquitin chain, which acts as a substrate-targeting and recognition signal for the proteasome (Weissman, 2001). The polyubiquitination of polypeptides targeted for degradation is accomplished by the sequential action of three types of enzyme: an ATP-dependent ubiquitin-activating enzyme known as E l, a ubiqultin- conjugating enzyme known as E2 and thirdly a ubiquitin-protein ligase E3. The action of these enzymes results in the covalent attachment of the C terminus of ubiquitin to the e- amino group of a lysine residue of the target protein (Weissman, 2001). The ubiquitin

41 Chain grows as additional ubiquitin molecules are conjugated to lysine 48 of the already attached, preceding ubiquitin. Firstly, E1 a ubiquitin activating ezyme forms a thiol-ester bond with the carboxy-terminal glycine of ubiquitin, a process requiring ATP. Next, a ubiquitin-carrier enzyme E2 accepts ubiquitin from E1 which involves a trans-thiol reaction. In the last step a ubiquitin ligase E3 catalyses the transfer of ubiquitin from the E2 enzyme to the 8-amino group of a lysine residue on the protein substrate. Ubiquitin ligases (E3) represent two distinct families, with conserved protein domains: E3s with a HECT domain (Homologous to E6-AP carboxyl terminus, E6-AP being the founder member of this family) and E3s with a RING finger domain. HECT domain E3s form thiol-ester intermediates with ubiquitin as part of the reaction leading to ubiquitination of substrates. Members of the other class, E3s with RING finger domains, are believed to mediate the direct transfer of ubiquitin from E2 to substrate (Weissman, 2001).

Many ER lumenal and integral membrane proteins destined for degradation by the proteasome are known to undergo polyubiquitination during their retrotranslocation from the ER into the cytosol and this covalent modification is required for degradation. Expression of mutant ubiquitin (lacking lysine 48) in yeast (Hiller et al., 1996, Biederer et a/., 1996) or mammalian cells (Ward et a!., 1995, Yu and Kopito,1999) results in the stabilisation of proteins which were destined for proteasomal degradation. Furthermore mutations in many components of the ubiqitination process inhibit the proteolysis of proteins. For example, in a temperature sensitive mutant of the ubiquitin activating (El) enzyme, degradation of CFTR (Ward et a/,. 1995) and the a-subunit of the T cell receptor (Yu and Kopito, 1999) were prevented in mammalian cultured cells.

The precise role of polyubiquitination of ERAD substrates in retrotranslocation is not fully understood. In a number of substrates studied, including both lumenal and transmembrane proteins, polyubiquitination of these proteins was necessary to promote retrotranslocation. Given that degradation substrates accumulate in the ER when polyubiquitination is prevented (de Virgilio et a/., 1998, Kikkert et a/., 2001, Biederer et a/., 1997, Jarosch et ai 2002) it seems possible that this modification is required for retrotranslocation and not just for degradation by the proteasome. However, at least partial translocation of lumenal polypeptides must occur before the polyubiquitination of the substrate can occur. Once the substrate is accessible, polyubiquitination may serve as a ratcheting mechanism to move extra polypeptide segments into the cytosol; the polypetide would be prevented from moving back through the Sec61 channel by ubiquitin

42 molecules attached to the retrotranslocated polypeptide. Another possibility is that a polyubiquitin chain acts as a recognition signal for downstream components such as the proteasome or the recently identified Cdc48/p97 complex (Jarosch et al., 2002), which may be actively involved in the movement of the polypeptide chain.

Polyubiquitination alone may not be sufficient to for retotranslocation of polypeptides into the cytosol for proteasomal degradation. When polyubiquitination is carried out in a pemeabilised cell system in the presence of an ATP analogue, which cannot be hydrolysed, polyubiquitinated proteins remain associated with the ER membrane indicating that an ATP dependent step is required for release into the cytosol (Tsaiet a/., 2002). As the 19S complex of the proteasome can recognise polyubiquitinated chains, and has six ATPases this may, as has been suggested by Mayeret a/., 1998 actually pull the polypeptide chain through the Sec61 complex in the ER and feed the chain into the proteolytic cavity of the proteasome. However inhibition of the proteolytic activity of proteasomes can lead to the accumulation of non-degraded substrates in the cytosol rather than the ER (Ward et a/., 1995, Huppa and Ploegh, 1997, Yang et a/., 1998) and mutation of the 19S regulatory particle does not prevent retro-translocation of soluble proteins from the yeast ER (Jarosch et a!., 2002).

Recently a member of the AAA family of ATPases, Cdc48 in yeast and p97 in mammals was shown to play a role in the retrotranslocation of polyubiquitinated polypeptides (Yeet a/., 2001, Jarosch et a/., 2002, Bays et a/., 2001, Rabinovich et a/., 2002). Cdc48/p97 forms a stable complex with two other co-factors, Ufdl and Np14, in both yeast and mammals and are both required for its function in the ERAD pathway (Meyeret a/., 2000 Hitchcock et a/., 2001). In yeast mutants of cdc48, ufdl or npl4, the retrotranslocation and degradation of several ER lumenal, misfolded proteins and a short lived transmembrane protein was inhibited (Bays eta!., 2001, Braun eta!., 2002, Rabinovich et a!., 2002). Also polyubiquitinated polypeptide chains accumulated at the ER membrane in npl4 and ufd1 mutants (Jarosch eta!., 2001, Bays eta!., 2001) which implies that ubiquitination precedes retrotranslocation which requires the Cdc48-Np14-Ufd1 complex.

43 1.9 Neurodegenerative Diseases and Proteinaceous Intracellular Inclusions

1.9.1 Neurodegenerative Diseases

The function and viability of the neuronal cell may be threatened by accumulation of abnormally folded proteins, which can arise as a consequence of mutations in polypeptides preventing the normal folding or association with other subunits or stabilising factors (Sherman and Goldberg, 2001). In a variety of neurodegenerative diseases, unfolded polypeptides accumulate in neurons as insoluble inclusions and appear to have a role in disease pathogenesis (Sherman and Goldberg, 2001). Intracellular inclusions of misfolded proteins are characteristic features of many neurological diseases, such as Amyotrophic Lateral Sclerosis (ALS), Alzheimer’s disease (AD), Parkinson’s disease (PD), and several hereditary diseases caused by expansions of polyglutamine tracts (eg. Huntington’s disease (HD), spinocerebellar ataxias (SCA) and Spinal and Bulbar Muscular Atrophy, SBMA) (Kaytor and Warren, 1999, Sherman and Goldberg, 2001). In spinocerebellar ataxia type 1 (SCA1), an expanded polyglutamine tract is found within ataxin-1, and the protein forms ubiquitin positive nuclear inclusions in transgenic SCA1 mice and in transfected cells (Skinner et al., 1997). Similarly polyglutamine tracts are found in ataxin-3 and ataxin-7, the gene products responsible for spinocerebellar ataxia type 3 (SCA3) and type 7 (SCA7) respectively and these proteins also accumulate in ubiquitinated aggregates in the nuclei of affected neurons (Paulson eta!., 1997, Holmberg et a/., 1998). SCA6 is caused by a polyglutamine repeat expansion within the alA voltage-dependent calcium channel and forms cytoplasmic inclusions in Purkinje cells which are not ubiquitinated (Ishikawa et a/., 1999). Expanded polyglutamine tracts are also present in the HD gene product, huntingtin, in atrophin-1 responsible for Dentatorubral-Pallidoluysian Atrophy (DRPLA) and in the androgen receptor responsible for SBMA (Kaytor and Warren, 1999). These proteins all form ubiquitinated inclusions in neurons and cultured cells (Kaytor and Warren, 1999).

In addition to the polyglutamine diseases, several other major neurodegenerative diseases are also associated with an accumulation of proteinaceous inclusions. For example, ubiquitinated intracellular inclusion bodies (referred to as Lewy bodies) are found in the brains of patients with Parkinson’s disease. A major component of Lewy bodies is a- synuclein and missense mutations in this gene have been found in dominant cases of PD (Polymeropoulos et ai, 1997). Mutations in genes of the UPS have been shown to co-

44 segregate with hereditary forms of PD, including mutations in the gene parkin, which encodes a ubiquitin-protein ligase (E3) (Kitada et al., 1998, Tanaka at a/., 1998) and mutations in the ubiquitin carboxy-terminal hydrolase LI (Leroy at a/., 1998), which removes ubiquitin from peptide conjugates. Another neurological disease that is associated with proteinaceous inclusions is ALS characterised by the death of motor neurons. Some hereditary forms of ALS are due to missense mutations in the superoxide dismutase (SOD) gene (Rosen at a/., 1993). Mutant SOD forms cytoplasmic inclusions in motor neurons of transgenic mice expressing the mutant SOD transgene and in cultured cells (Johnston atal., 2000).

Alzheimer’s Disease (AD), the most common cause of senile dementia, appears superficially to be different from the above neurodegenerative diseases mentioned above, because this condition is associated with extracellular amyloid plaques. In addition Alzheimer’s is further characterised by the presence of cytoplasmic neurofibrillary tangles made up of highly phosphorylated and ubiquitinated forms of the microtubule associated factor tau (Morishima-Kawashima at a/., 1993). The major component of amyloid plaques is (3-amyloid (A(3), a 39-43 amino acid peptide derived from the larger amyloid precursor protein (APR) (Selkoe, 1998) which is cleaved by both p and y-secretase, generating Ap. The accumulation of Ap is considered to be a central part of AD pathogenesis and the Ap peptide of 42 amino acids (AP42) is highly amyloidogenic with mutations in presenilin 1 and

2 (associated with early onset AD) increasing the production and secretion of the Ap ,2 peptide (Gething, 2000)

1.9.2 Intracellular Proteinaceous Inclusions

The pathology and death of specific neuronal populations in many neurodegenerative disorders is associated with the accumulation of specific abnormal polypeptides. Those polypeptides, which are associated with neurodegenerative disease and form intracellular inclusions, appear to share a common mechanism of inclusion formation and provoke similar neuronal cell responses (Kopito, 2000).

The overall efficiency of protein folding is related to the proportion of intermediates that assume a native conformation but intermediates can also adopt misfolded ‘off-pathway’ conformations that are prone to oligomerise into non-native aggregates (Wetzel, 1994). Protein aggregates are oligomeric complexes of misfolded protein that arise from non-

45 native interactions among kineticaily trapped intermediates of protein folding or assembly (Wetzel, 1994). Thus, within cellular compartments where protein synthesis and folding occurs, there is an abundance of molecular chaperones that help to minimise the formation of ‘off-pathway’ intermediates during protein folding (for review Agashe and HartI, 2000).

If a cells capacity to refold or degrade abnormal proteins is exceeded, as can occur under experimental conditions in vitro the abnormally folded proteins accumulate and tend to form aggregates (Kopito, 2000, Sherman and Goldberg, 2001). Protein aggregates are defined by their poor solubility in aqueous or detergent solvents, abnormal intracellular or extracellular location and non-native secondary structure. These protein aggregates can be either structured (e.g. amyloid) or amorphous and tend to be insoluble and stable under physiological conditions (Kopito, 2000). Protein aggregation is widely viewed as non specific coagulation of incompletely folded or partially denatured polypeptides, driven by interaction among inappropriately exposed hydrophobic surfaces. According to this view, production of misfolded or denatured polypeptides has been suggested to be detrimental to cell survival, due to their ability to coaggregate with and thereby trap unrelated cellular proteins that may transiently display complementary surfaces (Trojanowski et a/., 2000, Perutz et al., 1994). However, refolding studies of chemically denatured polypeptides suggest that protein aggregation in vitro is due to specific intermolecular interactions among defined domains within structured folding intermediates (Wetzelet a/., 1996, Speed et a!., 1996). Evidence of this specificity is found in the seeding behavior of amyloidogenic proteins (Jarrett and Lansbury, 1992) and in the selectivity of aggregate formation by model proteins (Speed et a/., 1996). Although these studies suggest specificity of purified denatured proteins in dilute solutions few studies have addressed the mechanism and specificity of protein aggregation in vivo. In cells aggregated proteins are usually sequestered in discrete structures called inclusion bodies. Bacterial inclusion bodies are often highly enriched in a single aggregated protein species, suggesting that misfolded proteins do not co-aggregate with cellular proteins. In contrast inclusion bodies in mammalian cells are complex structures that contain many proteins, including molecular chaperones, components of the ubiquitin proteasome system, centrosomal material and cytoskeletal proteins. Inclusion bodies are also enriched in proteins involved in cell signaling, cell division and apoptosis, suggesting that coaggregation of misfolded damaged or mutant proteins with normal cellular proteins could explain both the presence of multiple proteins in inclusion bodies and the toxicity associated with protein aggregation in many neurodegenerative diseases (Trojanowskiet a/., 2000, Steffan et al., 2000,

46 Nucifora et ai, 2001). However, a significant finding by Rajan and co-workers has revealed that aggregation prone proteins expressed in the same cell do not co-aggregate even though they can be found within the same inclusion body. Co-expression of two aggregation prone proteins, AF508CFTR and mutant P23H rhodopsin, in cell culture lead to the formation of a AF508CFTR/ P23H opsin inclusion body. However, even though the two proteins in the inclusion were shown to co-localise in the cell using immunofluorescence microscopy, fluorescence resonance energy transfer (FRET) analysis revealed that the two different proteins did not in fact co-aggregate as they were more than 100 A apart (Rajan at al., 2001). Thus, non-specific aggregation between hydrophobic proteins does not occur and supports the view that protein aggregation is highly specific.

1.9.3 Aggresomes

Aggregated proteins can form an inclusion body in a discrete pericentriolar region, near the microtubule organizing centre (MTOC) and such an inclusion has been described as an aggresome (Johnston at a/., 1998). Formation of an aggresome is a multistep process, which involves retrograde transport of smaller aggregates on microtubules from the cell's periphery toward the MTOC (Garcia-Mata at a/., 1999). Depolymerisation of microtubules does not lead to the rapid dispersal of an aggresome, suggesing it is a not dynamic structure. Aggresome formation is blocked by drugs that depolymerise microtubules and by the expression of dynactin protein p50 dynamitin (Garcia-Mata atal., 1999), suggesting a role for the dynein retrograde transport on microtubules in aggresome formation. These studies, therefore, do not support a diffusion-based model for inclusion formation but confirm the idea that an aggresome is an inclusion body formed by the retrograde transport of aggregated protein on microtubules.

Using immunocytochemical analysis, several publications indicate that in addition to a major aggregated proteins species, aggresomes are enriched in molecular chaperones, including Hsc70, the Hsp40 proteins Hdji and Hdj2, and the chaperonin TriC/TCP (Garcia- Mata at al., 1999, Wigley at al., 1999). Components of the ubiquitin proteasome system (UPS) have been reported in aggresomes (Wigley et al., 1999, Wojcik at al., 1996, Wojcik, 1997, Fabunmi at al., 2000, Anton at al., 1999) but ubiquitin localisation to aggresomes is inconsistent. Aggresomes which formed in cells expressing AF508CFTR (Johnston at al.,

47 1998, Wigely et al., 1999) exon 1 of huntington with an expanded polyglutamine repeat (Waelter etaL, 2001) and peripheral myelin protein 22 (PMP22) reacted with antibodies to ubiquitin but not the aggresomes of the cytoplasmic proteins GFP250 (Garcia-Mata et al., 1999) and mutant SOD (Johnston etaL, 2000).

In cells harboring aggresomes, reports have consistently shown that intermediate filament proteins are displaced from their normal distribution (Johnston et a/., 1998). Type III intermediate filaments such as vimentin and glial fibrillary acidic protein (GFP) normally display an extended cytoplasmic distribution comprising 8-10 nm filaments (Herrmann and Aebi 2000). Vimentin is radically redistributed in cells with aggresomes and forms a cage like structure wrapped around the surface of the inclusion (Johnston etaL, 1998).

Heterologous expression of mutant genes in cultured cells can lead to the formation of aggresomes but the accumulation of protein aggregates at a juxta nuclear position is not unprecedented. For example, in mammalian cells stressed by heat shock (Vidair et a/., 1996) or viral infection (Laszio etaL, 1991, Brown etaL, 1994), Hsp70 accumulated at the centrosome, along with an increase in pericentriolar protein (Vidair et a!., 1996).

Futhermore, Wôjcik and co-workers reported the formation of inclusions near the MTOC consisting of electron dense material when Hela cells (Wôjcik et al., 1996) and other cells

(Wôjcik, 1997) were treated with proteasome inhibitors. The formation of these proteasome inhibitor induced inclusions were blocked with nocodazole suggesting that they are aggresomes-and reacted with antibodies to proteasomal subunits and ubiquitin,

Wôjcik termed this area the'proteolysis centre'. Other investigators have supported the idea that proteasomal degradation is associated with the centrosome. For instance, Wigley and co-workers, using immunofluorescence microscopy, showed that ubiquitin, proteasomal subunits and heatshock proteins colocalised with the pericentriolar matrix protein y-tubulin even in the absence of proteasome inhibitors. The authors went on to show that proteasomes that associated with centrosome-enriched preparations are catalytically active. Given that only approximately 1% of proteasomes in unstressed cells are associated with the centrosome (Fabunmi et al., 1999) it is doubtful that this region accounts for the major proportion of proteasome mediated proteolysis, as components of the UPS are associated with other cellular organelles, including the ER and the nucleus (Palmer et al., 1996). Therefore components of the UPS system may be recruited to sites only were they are needed rather than centralising proteasomal activity to the MTOC.

48 Exceeding the cell’s capacity to degrade misfolded proteins can lead to formation of aggregates, which are then recruited to the MTOC along microtubules where they form aggresomes. This mechanism removes aggregated proteins from the cytosol to a central region of the cell and prevents the interference of aggregated proteins in cellular functions, except MT based transport. In addition, delivery of aggregated protein to a centralised region might facilitate their disposal by an autophagic route. Autophagy, the term given to the process where cytoplasmic components such as ribosomes or intracellular organelles are destroyed, may be a mechanism for ridding the cell of aggresomes. Autophagosomes are membrane bound organelles, which eventually mature into lysosomes and have been shown to be present in cells over expressing aggregation prone proteins (Kegel et al., 2000). Furthermore, recent work by Ravikumar et a/., 2002 has revealed that aggregates of proteins with polyglutamine and polyalanine expansions are degraded by an autophagic route. It has been speculated that concentrating protein aggregates at the MTOC may increase the efficiency of capture into autophagic structures, but constant delivery of aggregated protein to this region of the cell may even overwhelm this process.

Neuronal cytoplasmic aggregates are present as a single inclusion or in low copy numbers and are nearly always found with displaced or abnormally phosphorylated neurofilaments (Mayer et a/., 1989). A characteristic of Pick’s disease is the redistribution of GFAR into Rosenthal fibres surrounding inclusion bodies (Mayer et a/., 1989). In many neurodegenerative dieseases neurofilaments, the major cytoskeletal component in axons, surround cytoplasmic inclusions. Neurofilaments were found to surround hyaline inclusion bodies in amyotrophic lateral sclerosis and Lewy bodies in Parkinson’s disease (Mayeret ai., 1989). Intracellular inclusions can occur within cell bodies or within cell processes. The comparison of aggresomes with disease associated inclusion bodies is complicated by the differences in the organisation of microtubules and the MTOC in post-mitotic neurons and dividing cells (Baas, 1998). However, mutant gene products associated with autosomal dominant neurodegenerative disease (e.g. superoxide dismutase-linked familial ALS) are targeted to the MTOC in a microtubule dependent manner to form aggresomes when expressed in cultured cells (Johnston et a!., 2000). The transport of aggregated protein via retrograde transport on microtubules may also be a property found in neurons.

Aggresome formation leads to the disruption of the normal architecture of the Golgi apparatus and partial disorganisation of microtubule arrays (Garcia-Mata et a/., 1999).

49 The Golgi apparatus is a dynamic organelle whose organisation depends on continuous microtubule vesicular traffic (Lippincott-Schwartz 1998). Disruption of the Golgi by aggresomes is unlikely to reflect a gross defect in microtubule based trafficking because other forms of microtubule based vesicular movement such as ER-Golgi transport, appear to be unaffected (Garcia-Mata et al., 1999). However, fragmentation of the Golgi apparatus might be due to displacement of Golgi cisternae by the accumulation of protein aggregates in and around the MTOC. In fact, displacement and fragmentation of the Golgi apparatus was observed in cultured cells over expressing aggregation-prone proteins or cells treated with proteasome inhibitors (Garcia-Mata at a!., 1999), in neurons from Alzhiemer’s disease brains, in spinal cords from human ALS, in a transgenic mouse model of ALS expressing mutant superoxide dismutase (Gonatas at a!., 1998) and a transgenic mouse model of adRP expressing P23H rhodopsin (Olsson at a!., 1992).

1.10 ER Stress and the Unfolded Protein Response

Proteins entering the secretory pathway fold within the ER, and to support efficient folding the ER maintains an environment enriched in molecular chaperones. A quality control system in the ER ensures that terminally misfolded proteins are destined for degradation by the proteasome. Certain situations will increase the levels of misfolded proteins in the ER, including diminished protein glycosylation caused by nutrient/glucose deprivation, disturbances in calcium homeostasis and redox status of the ER, increased demand for protein folding due to elevated synthesis of secretory proteins and finally the expression of mutant proteins which are misfolded. The cell detects and responds to an increase in the levels of misfolded proteins within the ER by inducing or upregulating the transcription of genes involved in protein folding i.e. molecular chaperones, genes involved in ER associated degradation and certain transcription factors, notably Gaddi 53. This signaling pathway from the ER to the nucleus is termed the unfolded protein response (UPR).

In yeast Irelp plays an essential role in the initiation of the UPR and is a type I transmembrane protein kinase whose N-terminal domain is located in the ER lumen; the serine/threonine specific protein kinase and endonuclease domains of Irelp are situated on the cytoplasmic side of the ER membrane (for review Kaufman, 1999). In unstressed conditions Irelp maintains a mainly monomeric and inactive state by binding to the ER chaperone KAR2 (the yeast homologue of BiP). Upon ER stress KAR2 (BiP) binds to hydrophobic surfaces that are exposed on unfolded proteins thus depleting KAR2 that is bound to Irelp. In the absence of KAR2 Irelp dimerises and frans-autophosphorylates.

50 leading to activation of the endonuclease activity present in its C-terminal domain. Activated Irelp splices out an intron from HAC1 mRNA that encodes the basic leucine zipper (bZIP)-type transcription factor H ad p. The cleaved 6' and 3' halves of the spliced HAC1 mRNA are ligated and translation of the mature HAC1 mRNA now follows. H ad binds to a sequence motif termed the unfolded protein response element (UPRE) as a dimer in the promoters of responsive genes. As well as activating the transcription of ER resident chaperone genes, H ad activates the transcription of genes that are involved in ER associated degradation and phospholipid biosynthesis (Kaufman, 1999). The mammalian genome contains two orthologues of yeast Irelp: IREIa, which is ubiquitously expressed, and IREip whose expression is restricted to epithelial cells of the intestinal tract. In a similar fashion to yeast Irelp, mammalian IRE1 dimerises and trans- autophosphorylates, activating its endoribonuclease activity. It cleaves the mRNA that encodes the bZIP transcription factor X-box binding protein (XBP1) to remove a small intron which changes the translational reading frame of XBP1 to yield a potent transcriptional activator that interacts with the mammalian ER stress response element (ERSE) in the promoters of responsive genes (Kaufmanetal., 2002).

Another mammalian ERSE-binding protein was identified; activating transcription factor 6 (ATF-6a and ATF-6p) which is an ER transmembrane protein and on activation of the UPR ATF-6a and ATF-6p move to the Golgi where they are cleaved by site 1 protease (SIP) and site 2 protease (S2P) to generate cytosolic bZIP containing fragments that migrate to the nucleus (Yoshida et a/., 1998, Kaufman at a/., 2002). In the nucleus both forms of ATF-6 with the CCAAT-binding factor (CBF) bind to the 3' half-site of the ERSE in the promoter regions of UPR-responsive genes (such as BiP and Gp94) to activate transcription (Kaufman etal., 2002).

The UPR not only results in the transcription of genes involved in this adaptive mechanism but also results in the attenuation of protein synthesis; a response that is mediated by the phosphorylation of the a-subunit of eukaryotic translation initiation factor 2 (elF2a). Upon activation of the UPR, a transmembrane, ER localised, elF2-a kinase known as PERK, oligomerises and frans-autophosphorylates. Activated PERK then phosphorylates elF2-a, resulting in a down regulation of overall protein synthesis. It has been suggested that BiP is a negative regulator of the pathways just described because BiP binds to the ER lumenal domains of IRE1 and PERK preventing their dimérisation and interacts with ATF-6 to prevent its transport to the Golgi compartment (Kaufman et a/., 2002). Thus when

51 misfolded proteins accumulate in the ER, BiP binds to exposed hydrophobic surfaces of accumulating proteins thereby competitively disrupting the interaction with the transducing proteins IRE1, ATF-6 and PERK.

Although the UPR is an adaptive response to disruption of ER homeostasis chronic ER stress can lead to apoptosis (Patil and Walter, 2001). Gaddi53, also known as Chop, is a leucine zipper transcription factor that is present at low levels under normal conditions but is robustly expressed in response to stress (Carlson et al., 1993, Wang et al., 1996). Following induction of the UPR the kinetics of Gaddi 53 induction parallel those seen for BiP (Wang et al., 1996). Gaddi 53 was originally identified based on its induction following treatment of cells with growth arresting and DNA damaging agents (Luethyet al., 1990), though induced expression of the gene has also been linked to perturbation of homeostasis in the ER. Some researchers have reported that induction of Gaddi 53 correlates with the onset of apoptosis (Friedman, 1996, Eymin et al., 1997). Furthermore, cell death in response to tunicamycin, which induces an UPR, was delayed in embryonic fibroblasts derived from Gaddi 53 knock out mice (Zinszner et al., 1998). However a clear mechanistic relationship between expression of Gaddi53 and cell death remains to be established, but recently elevated Gaddi53 expression was shown to down regulate the expression of the anti-apoptotic protein Bcl-2, deplete cells of glutathione and increase production of reactive oxygen species (McCullough et al., 2001). ER stress can also activate other signals, which lead to apoptosis such as murine caspase-12, localised on the cytoplasmic face of the ER and the c-Jun N-terminal kinase (JNK) pathway (Ferri and Kroemer, 2001). JNKs regulate gene expression through the phosphorylation and activation of transcription factors such as cJUN or ATF2 and ER stress mediated activation of JNKs is mediated through IRE1 and TRAF2 an adaptor protein that couples tumor necrosis factor receptor activation to JNK activation (Ferri and Kroemer, 2001). The human orthologue of murine caspase-12 has not yet been discovered but the closest homologue is caspase 4. Caspase-12 is associated with the cytosolic side of the ER and is specifically involved in apoptosis that results from stress in the ER (Nakagawa et al., 2000). In fact embryonic fibroblasts from caspase-12 knockout mice are resistant to ER stress induced apoptosis (Nakagawa et al., 2000). Caspase-12 like most other members of the caspase family requires cleavage of the prodomain to activate its proapoptotic form and is actually cleaved by m-calpain a Ca2+ responsive cytosolic cysteine protease (Nakagawa and Yuan, 2000). Thus, the accumulation of misfolded proteins in the ER inducing ER stress can be directly connected to the caspase activation casacade, resulting in apoptosis.

52 1.11 Aims of the Thesis: i) To examine in detail the intracellular location of mutant rod opsin using immunofluorescence microscopy, which led to the finding, that mutant opsin formed what appeared to be an aggresome. Thus, further investigations were carried out to determine if these opsin inclusions showed the characteristic hallmarks of aggresomes. Also the influence of mutant opsin on the processing of WT rod opsin in culture cells was investigated. ii) The role of N-linked glycosylation in the processing of WT opsin has been studied but the function of N-linked glycosylation in mutant rod opsin processing is unclear. Therefore the drug Tunicamycin was used to investigate whether N-linked glycosylation had any effect on the processing and degradation of mutant opsin in cultured cells. Furthermore, the role of the proteasome in the degradation of mutant opsin was determined. iii) 11-c/s-retinal is covalently attached to opsin and fits into a binding pocket created by the transmembrane domains of the receptor. The ligand creates a salt bridge between transmembrane 3 and 7 of opsin and 11-c/s-retinal also forms a number of hydrogen bonds with amino acids lining the binding pocket, thus stabilising the protein. Therefore, opsin ligand may stabilise mutant opsin and facilitate folding. Thus the effects of ligand on the processing of mutant opsin in cultured cells were examined. iv) Given that the mutant opsins in this study are misfolded and accumulate in the ER and in aggresomes, I determined whether mutant opsins can ellicit an unfolded protein response in cultured cells.

53 Chapter 2: Materials and Methods

2.1 Reagents

Tunicamycin (containing homologues A.B.C, and D), 9-c/s-retinai, DAP! for nuclear staining, and Protease Inhibitor Cocktail in DMSO for mammalian cell extracts were purchased from Sigma. The proteasome inhibitor MG-132 (Z-leu-leu-leu CHO) was purchased from Biomol Research Labs. PA 19462 USA. The GFP expression vector pEGFP-N1 was from Clontech. The glycosidases Endo H and PNGase F were from New England Biolabs. All restriction enzymes and T4 DNA ligase were purchased from Promega. Primary antibodies: mAb 1D4 to rhodopsin was a gift from Dr R Molday (University of British Columbia) and subsequently purchased from the National Cell Culture Centre, Minneapolis, Minnesota, USA; anti-p-COP (clone maD) and anti-Hsc70 (clone BRM22) were from Sigma; rabbit anti-Grp78 (BiP) and rabbit anti-Gaddi53 were from Santa Cruz Biotech; rabbit polyclonal anti-calnexin was from StressGen Biotechnologies; mouse monoclonal to vimentin was from Dako Ltd; Secondary antibodies: Goat anti-mouse Cy3 (Amersham Pharmacia Biotech); Donkey anti-rabbit Cy3 and Donkey anti-goat Cy3 were from Jackson Immuno Research Laboratories; goat anti mouse conjugated to Horseradish Peroxidase was from Pierce. Mammalian expression vectors encoding AF508CFTR-GFP and Hise-c-myc-tagged ubiquitin were a generous gift from Dr. R. Kopito (Stanford, USA).

2.2 Molecular Biology

2.2.1 Site directed mutagenesis by overlap extension using PGR

Bovine WT rhodopsin and K296E rhodopsin in pMT3 (Figue 2.1) were a gift from Dr D. Oprian (Brandeis). Rhodopsin was cloned into the Not I and Eco RI sites of pMT3. P23H rhodopsin in vector pMT3 was made by PCR mediated site directed mutagenesis using the WT in pMT3 as a template. The following primer combinations were used: For PCR reaction i) rhodopsin forward primer 5’GGATCCGAATTCCACCATGAACGGTACCGAA 3’, and P23H reverse primer encoding the P23H mutation (P23H B): 5’ GCCTCGAAGTGGCTGCGCACCAC 3’: For PCR reaction ii) P23H forward primer encoding the P23H mutation (P23H A) 5’GCAGCCACTTCGAGGCTCCGCAG 3’ and rhodopsin reverse primer 5'GCGGCCGCTTAGGCAGGCGCCACTTG 3’. The reaction

54 products from i) and ii) were gel purified, and used as a template in the final PCR reaction, which was amplified using rhodopsin forward primer and reverse rhodospin primer to generate the full length P23H rhodopsin gene. This was cloned into pGEM-T, digested with Eco RI and Not I and cloned into pMT3 (conditions for restriction digests, ligations and transformations are described in 2.2.3 and 2.2.4).

5000

amp'

MLP 1000 Not I < pMT3 'EcoRI (5.1 kb)

AAAA

2000 3000

Figure 2.1 The pMT3 mammalian expression vector

pMT3 vector is a transient expression vector in which the adenovirus major late promoter (Ad MLP;MLP) is coupled to the adenovirus tripartite leader (TPL) which contains a 5’ splice site and part of a mouse immunoglobulin gene that contains a 3’ splice site. The intervening sequence (IVS) is followed by two cloning sites{Not I and E go RI), the dihydrofolate reductase gene {dhfr), the SV40 polyadenylation site(polyA; AAAA), and the adenovirus VA| gene. The SV40 origin (SV40ori) and enhancer are also present on the vector.

The PCR reactions and conditions used to generate P23H rhodopsin are shown below; 1x advantage polymerase mix: Glycerol 1% Tris-HCL(pH7.5) 0.8 mM KCL 1 mM (NH4)2S04 0.5 mM EDTA 2pM (3-mercaptoethanol 0.1 mM Thesit 0.005%

55 The advantage polymerase mix contained KlenTaq-1 DNA polymerase (a 5’-exo-minus, N- terminal deletion of Taq DNA polymerase) and Taqstart antibody (1.1 pg/|xl) and a second DNA polymerase to provide proof reading activity. lOx PCR reaction buffer: Tricine-KOH (pH 9.2 at 25 0) 400 mM KOAc 150 mM Mg(0Ac)2 35 mM Bovine Serum Albumin 37.5 ^ig/ml The conditions for PCR using rhodopsin forward primer and P23HB primer are listed below: PMT3- WT ,0.5 ^ig

10 X reaction buffer 5 |il dNTPs (lOmM) 0.2 mM Forward primer 25 pmoles Reverse primer (P23H B) 25 pmoles

50 X polymerase mix 1 pl ddHzO to 50 pi Total reaction volume 50 pi Cycling parameters:

1 ) 94 °C 60 seconds X 1 2) 94 °C 30 seconds 65 °C 30 seconds X 10 cycles 72 °C 45 seconds This generated a 988 bp product and 10 pi of the PCR product was run on a 1% agarose gel, excised and then gel purified.

The conditions for PCR using the reverse rhodopsin primer and P23HA primer are listed below: pMT3- WT 0.5 pg

10 X reaction buffer 5 pi dNTPs (lOmM) 0.2 mM Forward primer 25 pmoles Reverse primer (P23H A) 25 pmoles

50 X polymerase mix 1 Hi ddHzO to 50 pi Total reaction volume 50 pi

56 Cycling parameters:

1) 94 °C 60 seconds X 1 2) 94 °C 30 seconds 65 °C 30 seconds X 10 cycles 72 °C 45 seconds This reaction generated a 93 bp PCR product and 10 was run on a 1% agarose gels excised and gel purified.

Equimolar amounts of the two PCR gel purified fragments from the above reactions were pooled and added to the PCR mix as listed below:

1 kb.p. PCR product 50 ng 100 b.p. PCR product 300 ng

10 X reaction buffer 5 |il dNTPs 0.2 mM Forward primer 25 pmoles Reverse primer 25 pmoles

50 X polymerase mix 1 pi ddHzO to 50 pi Total reaction volume 50 pi

Cycling Parameters: 1) 94°C 60 seconds x 1 2) 94°C 30 seconds 65°C 30 seconds x 26 72°C 45 seconds 3) 72°C 10 minutes x 1

2.2.2 Cloning of WT, K296E and P23H rhodopsin into pEGFP-N1

In order to generate GFP-tagged opsin I cloned rhodopsin into the BamH I and Age I site of pEGFP-N1, by introducing BamH I and Age I restriction sites into rhodopsin using PCR. The 5 ‘ end of each primer consists of the restriction site to be introduced preceded by three or four extra nucleotides. The restriction site sequences (bold) are followed by 14 bases that are homologous to the template DNA. The forward primer has the restriction

57 site for BamH I and the reverse primer has the restriction site for Age I and lacks the stop codon:

Forward primer: 5' GGATCCGAATTCCACCATGAACGGTACCGAA 3’ Reverse primer (no stop codon): 5' GACCGGTGCAGGCGCCACTTGGCTGGTCTC 3' pMT3 WT/K296E 0.5 Jig 10 X reaction buffer 5 \i\ dNTPs 0.2 mM Forward Primer 25 pmoles Reverse Primer 25 pmoles

50 X polymerase mix 1 W ddHzO to 50 pi Total reaction volume 50 pi

Cycling Parameters: 1) 94°C 60 seconds x1 2) 94°C 30 seconds 65°C 30 seconds x12 72°C 45 seconds 3) 72°C 10 minutes x1

WT, P23H and K296E opsin PCR products were ligated into pGEM-T, digested with Age I and BamH I restriction enzymes, gel purified and were then cloned in frame with GFP into the BamHI/Agel site of pEGFP-N1, such that the GFP sequence was fused to the C- termini of the opsin. All construct sequences were confirmed by sequencing.

2.2.3 Restriction Enzyme Digests

All restriction digests using the restriction enzymes Not I, EcoR I, Age I and BamH I were carried out using multicore restriction enzyme buffer (Promega): 1 X buffer is 25 mM Tris- Acetate (pH7.5 @ 37°C), 100 mM potassium acetate, 10 mM magnesium acetate, 1 mM DTT. DNA samples were mixed with 6x blue orange loading dye, which contained orange G, bromophenol blue and xylene cyanol and were loaded onto a 1% w/v agarose gel made up in 1 X Tris-Acetate (TAE) buffer containing 40 mM Tris-acetate and 1 mM EDTA. DNA

58 was visualised on a U.V. transilluminator and appropriate bands were excised using a scalpel. DNA was recovered using a gel purification kit (Geneclean II Kit).

2.2.4 Ligations and Transformations

Ligation reactions using a ratio of vector DNA to insert DNA of 1:3 were carried out using 3 units of 14 DNA ligase in the presence of 1 x 14 DNA ligase buffer (10 X T4 DNA ligase buffer: 300 mM Tris-HCI (pH7.8), 100 mM MgCb, 100 mM DTT and 10 mM ATP) in a total volume of 10 jj.1. The reaction mixtures were incubated at 4°C for 18 hours. For transformations, 5 \i\ of the ligation reaction was added to 50 pi of competent JM109 E.coli cells (Yanisch-Perron.C., et al., 1995) heat shocked at 42°C for 90 seconds, made up to 1.5 mis with pre-warmed LB (37°C) and incubated at 37°C for 90 minutes. Then 200 pi of the cell suspension was plated onto agar plates containing 100 pg/ml of ampicilin which were then incubated for 18 hours at 37°C.

2.2.5 RT-PCR of GADDI 53:

RNA was isolated from SH-SY5Y cells expressing WT and mutant GFP-tagged opsin using an RNA extraction kit from Qiagen according to the manufacturer’s instructions. RNA (1 pg) that was isolated was added to the reverse transcriptase reaction mix:

RNA 1 pg RT buffer x10 2 pi AMV RT 15 u

Oligo (dT)i5 0.5 pg

MgCl2 (25 mM) 4 pi (5 mM) DNTP (lOmM) 2 pi (1 mM) RNAsin 0.5 pi ddHzO to 20 pi Total reaction volume 20 pi

The reaction mix was incubated at 42°C for 15 mins and then heated at 99°C for 5 mins, so as to inactivate the reverse transcriptase.

59 The PCR reaction for GADD153 using the forward primer: 5' GAG ACT GTC AGC TGG GAG CT 3'and reverse primer: 5' CAT TCT CTT CAG CTA GCT GTG C 3' was set up as follows: cDNA 2 pi of RT reaction 10 X reaction buffer 5 |il dNTPs (10 mM) 1 pl Forward Primer 0.25 pmoles Reverse Primer 0.25 pmoles

50 X polymerase mix 1 pl ddHzO to 50 |il Total reaction volume 50 jil

Cycling Parameters: 94°C 1 minute x1

94°C 30 seconds 68°C 30 seconds x 25 cycles 72°C 30 seconds

10 |Lil of the PCR product was run on a 1% agarose gel stained with ethidium bromide (the 1 X TAE running buffer contained ethidium bromide at a concentration of 1 pg/ml) and placed on a U.V. transilluminator. An image of the gel was captured using a digital camera and the band intensities of the PCR products were analysed.using Kodak Digital Science 1D image analysis software

2.2.6 Mini preparations of piasmid DNA

E.coli (strain JM109) were grown in Luria-Bertani Medium (LB), which contained lOg tryptone per litre, 5g yeast extract per litre and lOg NaCI per litre. Plasmid DNA was extracted from the cells using a GFX Micro Plasmid Prep Kit (Amersham Pharmacia Biotech) according to the manufactureres instructions: 5 mis of LB were inoculated and incubated for 18 hours at 37°C and then 1.5 ml aliquots were placed into 1.5 ml microcentrifuge tubes and centrifuged for 30 seconds to pellet the cells. The supernatant was removed and the pellets were resuspended in 150 pi of solution I (lOOmM Tris-HCI, 10 mM EDTA, 400 pg/ml Rnase I), followed by the addition of 150 pi of Solution II (1 M NaOH, 5.3% SDS) and incubated at 22°C for two minutes. Then 300 pi of solution III was

60 added to the cell lysates, and mixed by inverting the 1.5 ml tubes 10 times. The lysates were centrifuged at 17,500g for 5 minutes at 22°C. The supernatant was transferred to a GFX column (containing a glass fibre matrix) and incubated for 1 minute at 22°C. The GFX columns were centrifuged for 30 seconds at 17,500g and then washed by adding 400 pi of wash buffer (10 mM Tris-HCI (pH 8), 1 mM EDTA, 80% ethanol). The GFX columns were then centrifuged for 60 seconds at 17,500g to remove the wash buffer. The DNA was eluted from the columns by the addition of 50 pi of sterile ddH20 and then the columns were centrifuged for 1 minute to collect the DNA solution in sterile 1.5 ml centrifuge tubes.

2.3 Cell Culture

2.3.1 Transfection of COS-7 and SH-SY5Y cells

COS-7 and SH-SY5Y cells were grown in DMEM/NUT.MIX.F12 with Glutamax-I + 10% heat inactivated FBS and 50 pg/ml gentamicin with an atmosphere of 5% CO2 at 37°C. Twenty four hours after seeding glass 8 well chamber slides with 2 x lO'* cells per well, the cells were transfected with 0.2 pg DNA per well with 2 pi of plus reagent and 1 pi of lipofectamine according to the manufacturers instructions. For 35 mm dishes the transfection mix was scaled up and cells were transfected 24 hours after seeding 35 mm dishes with 10® cells.

2.3.2 Drug treatments of COS-7 and SH-SY5Y cells

For Golgi disruption experiments, brefeldin A (BFA) was added to the cells (from a stock of 10 mg/ml) at a final concentration of 20 pg/ml for 15 mins prior to fixation. For proteasome inhibition, cells were treated with 5 pM MG-132 in DMEM/NUT.MIX.F12 with Glutamax-I + 10% heat inactivated FBS for 8 hours or 16 hours. A/-linked glycosylation was inhibited by the addition of TM to a final concentration of 0.8 pg/ml in DMEM/NUT.MIX.F12 with Glutamax-I + 10% heat inactivated FBS for 8 hours or 16 hours .

2.4 Immunofluorescence and Electron Microscopy

2.4.1 Immunofluorescence microscopy For vimentin, Hsc70, (3-COP and c-myc ubiquitin staining, COS-7 cells were fixed in 100% methanol at -20“C for 6 mins then washed twice in PBS. For calnexin and BiP staining

61 cells were fixed in 3% formaldehyde/0.1% gluteraldehyde in 0.08 M sodium cacodylate- HCI buffer, pH 7.4 for 20 mins followed by 2 five min washes in PBS and one 5 min wash in 50 mM NH4CI in PBS, followed by one rinse in PBS. Cells were permeabilised in 0.1% Triton X-100 and 0.05% SDS for 4 mins followed by three 5-minute washes in PBS. SH- SY5Y cells were fixed in 100% methanol at -20®C for 6 mins prior to immunolabeling with anti-Grp94, anti-BiP and anti- c-myc (9E10).

Following fixation and permeabilisation of COS-7 and SH-SY5Y cells, blocking agent made up of PBS +10% appropriate normal serum was added to the cells for 1 hour at 22°C, prior to incubation with antibodies. The primary antibodies were incubated for 1 hour at 22°C, followed by three 5 minute washes in PBS, followed by a 1 hour incubation at 22°C with conjugated secondary antibodies and then washed for 5 minutes, 4 times in PBS. To counterstain nuclei, DAPI was used diluted to 1:5000 in PBS and added for 5 mins in the third wash. The titres of primary and secondary (2^°) antibodies used and the appropriate animal serum are tabulated below:

Table 2.1 Antibody titres used in immunofluorescence labeling of COS-7 and SH-SY5Y cells Antibody to Host Titre Host Fluorophore Titre 2"" species species 2"^ Bip Goat 1 40 Donkey Cy3 1 200 Grp94 Rat 1 50 Goat FITC 1 50 Hsc70 Mouse 1 50 Goat Cy3 1 200 Calnexin Rabbit 1 100 Donkey Cy3 1 200 P-COP Mouse 1 50 Goat Cy3 1 500 Vimentin Mouse 1 10 Goat Cy3 1 500 c-myc Mouse Neat Goat Cy3 1 500 Hybridoma Supernatant Rod opsin Mouse 1.33 ^ig/ml Goat Cy3 1:200 Rod opsin Mouse 2.5 pg/ml conjugated to FITC

62 Table 2.2 Absorbance and Emmision spectra of fluorophores used in immunofluorescence labeling of cells Fluorophore Absorption Emission Maximum Observed colour Maximum (nm) (nm) Fluorescein (FITC) 494 518 Green Indocarbocyanine 550 570 Red (Cy3) DAPI Blue EGFP 488 507 Green

For dual labelling, opsin was visualised with mAb 1D4 conjugated to FITC, at a concentration of 2.5 p.g/ml in PBS + 10% normal mouse serum and added after primary and secondary antibodies had bound, and only after three 5 minute washes in PBS. Immunofluorescence was visualised with a Zeiss laser scanning confocal microscope. Excitation wavelengths: FITC/GFP: 488 nm, Cy3: 543 nm, DAPI: 364 nm Filter sets used: FITC/GFP: 505-530 nm, Cy3: 560 nm, DAPI: 475-525 nm.

2.4.2 Electron Microscopy

COS-7 cells were transfected with the plasmid pEGFP-P23H and 24 hours later the cells were given to Dr. P. Munro for processing for EM. The cells were fixed overnight in Karnovsky’s fixative (3% glutaraldehyde, 1% paraformaldehyde in 0.08 M sodium cacodylate-HCI buffer, pH7.4) then were rinsed 3 times in PBS and incubated in a 1% aqueous osmium tetroxide solution for 1h at RT. Following 3 rinses in distilled water, cells were dehydrated by 10 minute immersions in 50%, 70%, 90% and 4 x100% ethanol. Finally, the wells containing dehydrated cells were filled with araldite resin and placed in a 60°C oven to harden overnight. Sections for examination by transmission electron microscopy were cut using a Leica ultracut S microtome and diamond knife, mounted on copper grids and contrasted lead citrate only. Images were viewed using a JEOL 1010 TEM operating at 80kV and photographs recorded onto Kodak electron image film.

63 2.5 Biochemistry

2.5.1 Preparation of Cell Extracts Twenty four hours after transfection of COS-7 cells, the transfected cells were treated with or without 5 p.M MG-132 for 16 hrs or 0.8 )ig/ml Tunicamycin for 16 hrs in DMEM/NUT.MIX.F12 with Glutamax-I + 10% FBS without antibiotics. Then, the cells were washed twice in 4®C PBS (137 mM NaCI, 2.7 mM KCI, 10 mM Na2HP0 4 , 2 mM KH2PO4) and incubated in 290 |il PBS/1 % n-Dodecyl-p-D-Maltoside plus 10 pi of protease inhibitor cocktail and 166 pM MG-132 for 30 mins on ice. Cell lysates were transferred to 1.5 ml microcentifuge tubes on ice and homogenised by passage through a 23G needle 5 times. 200 pi of cell lysate was centrifuged at 17,500 g for 15 mins at 4°C. Pellets were solublised in 50 pi 10 mM Tris-HCI pH 7.5, 1% SDS + 2 pi protease inhibitor cocktail for 5 mins at 22°C. Then 150 pi of RIPA buffer (50 mM Tris-HCI pH 8, 150 mM NaCI, 1 mM EDTA, 1% NP-40, 0.1% SDS, 0.05% sodium deoxycholate) was added and the pellets were then sonicated for 5 seconds three times. For total cell lysate preparation from COS- 7 and SH-SY5Y cells: Cells growing in 35 mm dishes were lysed in 290 pi RIPA buffer + 10 pi of protease inhibitor cocktail (Sigma) at 4°C. Cell lysates were transferred to a 1.5 ml centrifuge tube and sonicated three times for 5 seconds. The protein concentration of whole cell lysates was determined using the Bio-Rad protein assay, based on the method of Bradford and involves the addition of an acidic dye (Coomassie Brilliant Blue G-250) to the protein solution and subsequent measurement at 595 nm with a spectrophotometer. A standard curve using BSA (2pg- lOpg) was established and protein concentrations of whole cell lysates were determined from the graph.

2.5.2 SDS-Polyacrylamide Gel Electrophoresis

For SDS-PAGE an equal volume of 2 x modified Laemmli sample buffer (1 x sample buffer 0.125 M Tris-HCI, pH 6.8, 5% SDS, 15% glycerol, 10% 2-mercaptoethanol, 0.012% bromophenol blue) was added to the DM soluble and DM insoluble fraction of cell lysates and run on a 10% polyacrylamide gel, 7 cm (L) x 8 cm (W). The stacking gel was made up of 4% polyacrylamide in 0.375 M Tris-HCI (pH 6.8) with 0.1% SDS. The resolving gel was made up of 10% polyacrylamide in 0.125 M Tris-HCI (pH 8.8) containing acrylamide:bisacrylamide in a ratio of 30:1. The Tris-glycine running buffer contained 25 mM Tris-HCI, 250 mM Glycine and 0.1% SDS. Bio-Rad low range molecular weight markers were used as molecular weight standards.

64 2.5.3 Western Blotting

Proteins were transferred from 10% polyacrylamide gels to nitrocellulose using a semi dry electrophoretic transfer cell (Bio-Rad). The polyacrylamidie gel and nitrocellulose membrane were sandwiched between filter paper soaked in electrotransfer buffer. The electrotransfer buffer was made up of 25 mM Tris, 192 mM glycine and 20% methanol. Electrotransfer of proteins was conducted for 15 mins at 15 V. For immunodetection of opsin on the nitrocellulose blots, mAb 1D4 was used at a concentration of 0.5 jig/ml in PBS + 5% w/v Marvel™ + 0.1%v/v Tween-20 for 1 hour followed by one fifteen minute wash and three 5 minute washes in PBS + 0.1% v/v Tween-20. Next the blot was incubated with goat anti-mouse HRP, 1:50,000 in PBS + 5%w/v Marvel™, 0.1%v/v Tween- 20 for 1 hour and washed once for 15 minutes followed by four 5 minute washes in PBS + 0.1%v/v Tween-20. For the detection of Grp78 (COS-7 cell lysates) and Gaddi53 (SH- SY5Y cell lysates) the antibody titres for anti-Grp78 and anti-Gaddi 53 were 1:200 in PBS + 5%w/v Marvel™, 0.1%v/v Tween-20 and the secondary goat anti-rabbit HRP was used at 1:5000 in PBS + 5%w/v Marvel™, 0.1%v/v Tween-20. Incubations and washes were performed as described above. The chemiluminescent detection reagent ECL+ Plus (Amersham Pharmacia Biotech) was used to detect immobilised antigens, according to the manufacturer’s instructions.

2.5.4 Deglycosylation of rod opsin

DM soluble cell lysate (15 |ig) from transfected cells expressing rhodopsin was digested with Peptide: N-glycosidase F (PNGase F) or Endoglycosidase H (Endo H) (1500 units) for 2 hours at 37°C in 1 x G7 (PNGase) or 1 x G5 (Endo H) buffer before resolving by SDS-PAGE. Unit definition: one unit is defined as the amount of enzyme (PNGase F or Endo H) required to remove > 95% of the carbohydrate from 10 p,g of denatured RNase B in a 10 111 reaction in 1 hour at 37°C

PNGase F reaction: PNGase F purified from Flavobacterium meningosepticum was supplied in 20 mM Tris-HCI

(pH 7.5 @25 C), 50 mM NaCI, 5 mM Na2EDTA and 50% glycerol. lOx G7 Buffer: 0.5 M sodium phosphate (pH 7.5 @ 25°C)

65 DM soluble lysate 15 îg 10xG7 Buffer 2^1 PNGase F (500 U/îl) 3 [i\ (1500 U) ddHzO to 20 îl Total reaction volume 20 p.1 Reaction was incubated at 37°C for 2 hours.

Endo H reaction: Endo H cloned from Streptomyces plicatus and over expressed in E.coli was supplied in

20 mM Tris-HCL (pH7.5) 50 mM NaCI and 5 mM Na2EDTA. 10xG5 buffer: 0.5M Sodium Citrate (pH5.5)

DM soluble lysate 15 |ig 10xG5 Buffer 2 pi Endo H (500 U/pl) 3 pi (1500 U) ddH20 to 20 pi Total reaction volume 20 pi Reaction was incubated at 37®C for 2 hours.

66 Chapter 3: Intracellular inclusions of RP mutants of rhodopsin

3.1 Introduction

Cell culture models have improved our understanding of expanded poly-glutamine repeat diseases [Huntington’s, the Spinocerebellar Ataxias, Spinal and Bulbar Muscular Atrophy (SBMA)] and other neurodegenerative disorders such as Familial Amyotrophic lateral sclerosis (missense mutations in Cu,Zn-superoxide dismutase gene) and hereditary forms of Parkinson’s disease (dominant missense mutations in a-Synuclein). Phenomena such as protein aggregation and ubiquitination of protein aggregates have been studied usingin vitro experiments and have contributed to our understanding of the cellular mechanisms of aggregate formation and the role of the UPS in the proteolysis of misfolded proteins.

Investigation of the expression of RP mutants of rhodopsin in cell culture, revealed that mutations in rhodopsin led to the aberrant folding of the protein, resulting in ER retention of misfolded rod opsin, rapid degradation and failure to associate with 11-c/s-retinal to form the functional photopigment (Sung et a/., 1991b, Sung at a!., 1993, Kaushal and Khorana, 1994). To further our understanding of the phenotypic properties of RP mutants of rhodopsin, I established a cell culture model of rod opsin processing using African green monkey kidney cells (COS-7) and the human neuroblastoma cell line SH-SY5Y. The two RP mutants of rhodopsin used in this study where Lys-296-Glu and Pro-23-His. These two missense mutations are found within different regions of the polypeptide. The P23H mutation lies within the intradiscal domain and K296E lies within the seventh helical transmembrane domain of the rod opsin protein. These mutant proteins share some similarities in their phenotypic behaviour; they are both misfolded proteins, albeit a certain proportion of these molecules can attain the native conformation (Sung at a/., 1991b, Kaushal and Khorana, 1994). However, Lysine-296 is the amino acid residue to which 11- c/s-retinal forms a covalent attachment, and replacing this positively charged lysine with either a neutral or negatively charged amino acid results in a constitutively active form of rod opsin in vitro (Robinson at a/., 1992).

K296E and P23H mutant rhodopsin genes translate into proteins which cannot attain the native conformational state (Sung at a/., 1991b, Kaushal and Khorana 1994) and in eukaryotic cells, the ER quality control mechanism targets misfolded proteins for degradation by the UPS, preventing the accumulation of potentially toxic aggregation prone proteins. If the rate of synthesis of a given protein exceeds the rates of folding or

67 degradation, a proportion of the protein molecules will be unable to proceed along the secretory pathway and inevitably accumulate and possibly form aggregates.

In this chapter I show that K296E and P23H mutant rod opsin has a propensity to form aggregates in cultured cells leading to the formation of aggresomes which are ubiquitinated. Furthermore, the data show that mutant rod opsin can influence the normal processing of WT rod opsin in COS-7 cells.

3.2 Results

3.2.1 Generation of P23H Rhodopsin and GFP tagged rhodopsin constructs

WT and K296E rhodopsin in the vector pMTS, were kindly provided by Dr. D. Oprian. In order to add another mutant to my investigations, the P23H mutation was produced by PCR mediated mutagenesis. The disease causing mutation in rhodopsin, P23H, was found to be the commonest cause of rhodopsin RP in the USA in a study by Sohocki et al., 2001 and is found in the intradiscal domain of the protein. P23H rhodopsin in the vector pMT3 was created by site directed mutagenesis by overlap extension using PCR (Ho et a/., 1989) with the WT gene in pMT3 as a template. In the first PCR reaction rhodopsin forward primer 5’ GGATCCGAATTCCACCATGAACGGTACCGAA 3’, and P23H reverse primer: 5’ GCCTCGAAGTGGCTGCGCACCAC 3’ were used to generate a 988 b.p. product shown in figure 3.1a. Another PCR reaction using P23H site forward primer 5’ GCAGCCACTTCGAGGCTCCGCAG 3' and rhodopsin reverse primer 5’ GCGGCCGCTTAGGCAGGCGCCACTTG 3’ was performed and generated a 93 b.p. product as shown in figure 3.1a. Reaction products, from both reactions were gel purified, added together (equal molar ratios) and used in the final PCR reaction step using rhodopsin forward primer 5’ GGATCCGAATTCCACCATGAACGGTACCGAA 3’ and reverse rhodospin primer 5’ GCGGCCGCTTAGGCAGGCGCCACTTG 3’, to generate the full length P23H rhodopsin gene (Fig. 3.1b). This was then TA cloned into pGEM-T, sequenced, then digested with Eco RI and Nof I and cloned into pMT3.

Rhodopsin-GFP constructs were generated, as these would be important tools required for the investigation of rod opsin trafficking in living cells and reduce the problem of anti body cross reactivity encountered in dual immunofluorescence labeling. Also these constructs were useful in co-expression studies with GFP tagged WT opsin and mutant untagged opsin.

68 Rhodopsin-GFP: WT, P23H and K296E rhodopsin were cloned in frame with GFP into the Bam Hi/Age I site of pEGFP-N1 following PCR mutagenesis (wt and mutant rhodopsin in pMT3 as template) to remove the stop codon and include restriction sites, such that the GFP sequence was fused to the C-terminus of rhodopsin. The PGR primers used in this reaction were: forward primer incorporating a Bam HI site 5’ GGATCCGAATTCCACCATGAACGGTACCGAA 3', and a reverse primer where the stop codon has been removed: 5’ GCGGCCGCTTAGGCAGGCGCCACTTG 3’ incorporating an Age I site. The PCR products were resolved on a 1% agarose gel (Fig. 3.2a for WT and K296E rhodopsin and Figure 3.2d for P23H rhodopsin) and gel purified. The PCR products were then TA cloned into pGEM-T, digested with Bam HI and Age I (Fig. 3.2b and 3.2e) gel purified and then cloned into pEGFP-N1 (Fig. 3.2c and 3.2f shows digested pEGFP-N1 constructs). All construct sequences in pGEM-T were confirmed by sequencing.

3.2.2 The localisation of WT and mutant rod opsin in COS-7 ceils using immunofluorescence microscopy

The intracellular localisation of mutant rod opsin in COS-7 cells, was determined using immunofluorescence microscopy, with antibodies to the ER and Golgi. As described in the materials and methods, transfected cells were fixed 24 hours later in PBS/3.7% formaldehyde. The ER was stained using an antibody to the chaperone BiP, which is present in the lumen of the ER (Shiu and Pastan, 1979). Rod opsin was detected with 2 pg/ml FITC conjugated mouse mAb 1D4, which detects an epitope in the C-terminus of the protein (Hodges et al., 1988). WT rod opsin translocated to the plasma membrane of the cells and faint staining was seen in the ER (Fig. 3.3) with opsin strong immunofluorescence co-localising with the Golgi marker (Fig. 3.4). K296E rod opsin accumulated in the ER as shown by co-localisation with BiP (Fig. 3.3) and also showed faint plasma membrane staining (Fig. 3.3). In some cells K296E rod opsin formed a bright, round structure, located at a juxtanuclear position (Fig. 3.3). P23H rod opsin also accumulated in the ER and in a proportion of transfected cells formed a bright round structure at a juxtanuclear position (Fig. 3.3) but plasma membrane staining was very faint.

69 B. P23H P23H P23H

Figure 3.1 Construction of P23H rhodopsin by PCR mutagenesis

A) The plasmid pMTS-WT (*) was used as template to generate two PCR products overlapping the P23H mutation, a 988 b.p. product (arrow) and a 93 b.p. product (arrow head). B) The PCR products were gel purified, pooled together and used in a second PCR reaction to generate full length P23H rhodopsin gene (arrow). C) The P23H PCR product was then T-A cloned into pGEM-T and digested with EcoRI and Not I (arrow indicates P23H). D) P23H EcoR I/Not I was then cloned into the EcoR I/Not I site of pMT3 and digested with EcoR I/Not I restriction enzymes (arrow indicates P23H). Molecular weight 1 kb DNA markers are shown on the left and products were resolved on a 1% agarose gel.

70 B. C.

Figure 3. 2 Cloning of WT, K296E and P23H rhodopsin into pEGFP-N1

A) pMT3-Wr and pMT3-K296E were used as template (arrow head) to generate PCR products of WT and K296E rhodopsin (arrow) with restriction sites Age I and Bam HI and no stop codon. PCR products were then T-A cloned into pGEM-T. B) pGEMT-WT and pGEMT-K296E clones were digested with Age I and Bam HI, releasing WT and K296E rhodopsin (arrow). C) WT and K296E rhodopsin were cloned into the Age I/Bam HI site of pEGFP-N1 and digested with Age 1/Bam HI restriction enzymes. Arrow indicates full length WT and K296E rhodopsin. D) pGEMT-P23H was used as a template (arrow head) to generate a PCR product of P23H (arrow) with restriction sites Age 1 and Bam HI and no stop codon which was then T-A cloned into pGEM-T. E) pGEMT-P23H was digested with Age 1 and Bam HI restriction enzymes to release P23H rhodopsin (arrow). F) P23H was then cloned into the Age I/Bam HI site of pEGFP-N1 and then digested with Age I/Bam HI restriction enzymes, releasing P23H (arrow). Gels were 1% agarose and molecular weight DNA 1 kb markers are shown on the left.

71 P23H rod opsin

Figure 3.3 The ER localisation of WT and mutant rod opsins in COS-7cells

COS-7 cells were transfected with the plasmids pMT3-WT, pMT3-K296E and pMT3-P23H. WT rod opsin (green) targeted to the plasma membrane (arrow heads) and the Golgi (arrows). K296E rod opsin and P23H rod opsin accumulated in the ER as shown by co-localisation with BiP (merge) and formed a round, bright, fluorescent, inclusion at a juxta nuclear position (arrows). K296E rod opsin was also detected at the plasma membrane (arrow head). Nuclei were stained with DAPI. Images are z-sections and scale bars represent 10 pim.

72 The juxtanuclear bright round mutant opsin structures may represent mutant opsin accumulating in the Golgi (Sung et al., 1993), therefore antibodies to the Golgi protein p- COP were used to label the Golgi in transfected cells. p-COP is one of the proteins that coats non-clathrin coated vesicles which bud from the ER and is involved in membrane traffic between the ER and Golgi and between different membranes of the Golgi (Klumperman, 2000). The Golgi appeared as a discrete organelle in a juxtanuclear position when detected with antibodies to p-COP (Figure. 3.4). WT rod opsin, as expected, was present in the secretory pathway in the Golgi as shown by co-localisation with p-COP (Figure 3.4). P23H rod opsin inclusions, however, disrupted the normal morphology of the Golgi as shown by a very noticeable reduction and scattering of p-COP staining (Figure 3.4) which implied that the Golgi apparatus had been disrupted in these cells. Phenomena of this kind have been described in the literature for the CFTR mutant, AF508-CFTR (Johnston et a!., 1998) and this suggested that mutant rod opsin formed intracellular inclusions known as aggresomes and did not accumulate in the Golgi apparatus. To corroborate these data the drug brefeldin A (BFA), which disperses the Golgi and induces p-COP to rapidly relocate to the cytosol (Lippincott-Schwartz et a!., 1989), was used to show that mutant rod opsin aggresomes did not represent rod opsin which had accumulated in the Golgi. COS-7 cells were transfected with pEGFP-WT, pEGFP-K296E, pEGFP-P23H and pMT3-P23H. Twenty four hours later the cells were incubated with 20 pg/ml BFA for 15 minutes prior to fixation in -20°C, 100% methanol. As shown in figure 3.5, BFA dispersed p-COP from its discrete location and dispersed WT rod opsin-GFP that was present in the Golgi. Mutant rod opsin, however, remained in a juxtanuclear position (Fig. 3.5) showing that mutant rod opsin that had accumulated in this region was not in the Golgi.

3.2.3 Mutant rod opsin inclusions co-localise with c-myc-tagged ubiquitin in COS-7 cells

Ubiquitin has been shown to be a component of some aggresomes, e.g. AF508CFTR and mutant presenilin-1 (Johnston et a!., 1998) and propeptide mutants of surfactant protein C (Kabore et a!., 2000), but not all aggresomes are ubiquitinated (Garcia-Mataet a!., 1999). Experiments were carried out to establish whether rod opsin inclusions were ubiquitinated. COS-7 cells were co-transfected with a plasmid encoding c-myc-tagged ubiquitin and the plasmids: pEGFP-WT, pEGFP-K296E, pEGFP-P23H, and pMT3-P23H, (at a vector ratio

73 Figure 3.4 inclusions of mutant rod opsin disrupt the Goigi apparatus in COS-7 ceiis.

COS-7 cells were transfected with plasmids pMT3-WT and pMT3- P23H. For comparison, cells were transfected with plasmids pEGFP-WT, pEGFP-K296E and pEGFP-P23H. WT rod opsin (green) and GFP-tagged WT rod opsin (WT-GFP upper panels) accumulated in the Golgi (red) as shown by co-localisation with p - COP. In contrast, the Golgi was disrupted in cells that expressed P23H rod opsin (P23H rod opsin) and GFP-tagged K296E and P23H rod opsin (K296E-GFP, P23H-GFP, lower panels) as shown by weak and scattered p-COP fluorescence (arrows). Nuclei were counter stained with DAPI. Images are z-sections and scale bars represent 10 pm. ■ ■ H ■

P23H rod opsin

P23H-GFP

K296E-GFP

74 K296E-GFP |‘.-COP Merge 4^

+BFA +BFA

Figure 3.5 Dispersal of the Golgi with BFA does not disrupt mutant rod opsin inclusions.

COS-7 cells were transfected with the plasmids pEGFP-WT PEGFP-K296E, pEGFP-P23H and pMT3-P23H. The Golgi (red) dispersed after BFA treatment. WT rod opsin-GFP (WT-GFP) that was present within the Golgi also dispersed following BFA treatment. P23H rod opsin (P23H rod opsin) and P23H rod opsin- GFP (P23H-GFP) inclusions remained intact as did K296E rod opsin-GFP (K296E-GFP, arrows) and showed no dispersal following BFA treatment. Nuclei were stained with DAPI. Images are z-sections and scale bars represent 10 ^m.

75 of 1:1). The cells were fixed, 24 hours after transfection, in -20°C 100% methanol and c- myc-tagged ubiquitin was detected with neat hybridoma supernatant from clone 9E10 followed by anti-mouse Cy3. Rod opsin was detected with FITC conjugated mAb 1D4 or GFP fluorescence. Normally c-myc-tagged ubiquitin was found in the cytosol and nucleus in cells transfected only with the plasmid encoding c-myc ubiquitin (Fig. 3.6 last panel) and in cells expressing WT rhodopsin-GFP (Fig. 3.6 first row of panels). Furthermore, c-myc- tagged ubiquitin did not co-localise with Golgi or plasma membrane WT rod opsin-GFP (Fig. 3.6 first row of panels). In contrast, untagged P23H and K296E rod opsin juxtanuclear inclusions co-localised with c-myc-tagged ubiquitin as did the GFP-tagged P23H rod opsin mutant (Fig. 3.6). Ubiquitin was recruited to mutant opsin inclusions thus removing ubiquitin from its normal cellular location.

3.2.4 Mutant rod opsin inclusions alter the normal distribution of vimentin Another feature of aggresomes is their ability to redistribute the intermediate filament protein vimentin to the aggresomal region forming a cage like structure around it (Johnston et al., 1998, Garcia-Mata et al., 1999). The affect of rod opsin inclusions on vimentin distribution was determined using indirect immunofluorescence on cells expressing GFP- tagged and untagged mutant rod opsin in COS-7 cells. Normally, vimentin was seen beneath the plasma membrane and around the nucleus as in cells expressing WT rod opsin-GFP or in non-transfected cells (Fig. 3.7). Normal vimentin distribution, however, was disrupted in cells with mutant rod opsin inclusions and now formed a ring around the inclusion (Fig. 3.7). The above data (Figs 3.4-3.7) suggest that mutant rod opsin inclusions share characteristic features of aggresomes

3.2.5 Cytosolic chaperones are recruited to mutant rod opsin aggresomes Since aggresomes contain misfolded proteins destined for degradation, it is conceivable that such polypeptides could be complexed with specific molecular chaperones. To determine if distinct classes of molecular chaperones are associated with mutant rod opsin aggresomes, indirect immunofluorescence was used to assess the localisation of the abundant cytosolic chaperone Hsc70 in cells with aggresomes. Hsc70 is distributed throughout the cytosol in non-transfected cells and cells expressing WT rhodopsin-GFP, but Hsc70 showed strong co-localisation with P23H rod opsin at the periphery of the aggresome (Fig. 3.8). The antibody used in this experiment is a mouse monoclonal antibody (clone BRM-22), which recognises constitutively expressed (Hsc70) and inducible forms (Hsp70) of Hsp70.

76 Figure 3.6 Mutant rod opsin inciusions are ubiquitinated in COS-7 ceiis.

COS-7 cells were co-transfected with a plasmid encoding c-myc- ubiquitin and plasmids pEGFP-WT, pEGFP-P23H and pMT3-K296E, pMT3-P23H. Normally c-myc ubiquitin (red) was found distributed within the cytosol (upper panel WT rod opsin-GFP and lower panel labelled c-myc Ub only), but in cells co-transfected with c-myc ubiquitin and PMT3-K296E (K296E rod opsin) or pEGFP-P23H (P23H-GFP) or pMT3-P23H (P23H rod opsin) the tagged ubiquitin was recruited to mutant rod opsin inclusions (arrows) and co-localised with mutant rod opsin (merge). Nuclei were counter stained with DAPI. Images are z- sections and scale bars represent 10 jim. V\^T-GFP c-myc Ub

c-myc Ub only

77 Figure 3.7 Mutant rod opsin inciusions disrupt the normai vimentin network in COS-7 ceiis

COS-7 cells were transfected with the plasmids pMT3-P23H (P23H rod opsin) and pMT3-K296E (K296E rod opsin). For comparison cells were also transfected with plasmids pEGFP-WT (WT-GFP) and pEGFP-P23H (P23H-GFP). In transfected cells expressing mutant rhodopsin genes, vimentin no longer formed the normal distribution (as in non-transfected cells) but now formed a ring around mutant opsin inclusions (arrows). Nuclei were counter stained with DAPI. Images are z-sections and scale bars represent 10 pm. WT-GFP vimentin

78 w M sp/c/0 ■

P23H-GFP H s p / C / 0

Figure 3.8 Mutant rod opsin-GFP aggresomes recruit molecular chaperones

COS-7 cells were transfected with the plasmids pEGFP-WT and pEGFP-P23H. WT rod opsin-GFP (WT-GFP) did not show co- localisation with the cytosolic molecular chaperone Hsp/c70 (red). In contrast, Hsp/c70 associated with P23H rod opsin-GFP aggresomes (P23H-GFP) (arrows) as shown by co-localisation of Hsp/c70 with P23H rod opsin-GFP (merge). Nuclei were counter stained with DAPI. Images are z-sections and scale bar represents 10 pim.

79 3.2.6 The ultrastructure of P23H rod opsin-GFP aggresomes

Electron microscopy (performed by Dr. P. Munro) of COS-7 cells transfected with pEGFP- P23H confirmed the localisation of the P23H-GFP inclusion close to the microtubule organising centre (MTOC) and the disruption of the Golgi and intermediate filament networks (Fig. 3.9). The juxta-nuclear structure was also surrounded by mitochondria. Therefore, P23H rod opsin inclusions display most of the ultrastructural hallmarks of aggresomes. However, the membrane content of the inclusions was much higher than that observed in aggresomes formed from other proteins (Heathet a/., 2001, Johnston at a/., 1998, Waelter et ai, 2001) and in places the structure appears to be surrounded by membranes (Fig. 3.9).

3.2.7 Comparison of the numbers of WT and mutant rod opsin aggresomes in COS-7 cells over time

In order to compare aggresome numbers between WT rod opsin and mutant rod opsin over time, COS-7 cells were transfected with pEGFP-WT, pEGFP-K296E and pEGFP- P23H. At various time points after transfection (16 hours, 24 hours and 48 hours) the cells were fixed in PBS/3.7% formaldehyde and numbers of aggresomes were counted. I scored an aggresome only if it was round and bigger than the Golgi. A total of 150-200 transfected cells were counted from different fields of view in duplicate wells for each mutant at each time point. The results are graphically represented in figure 3.10. WT rod opsin-GFP aggresome numbers were low, approximately 5 percent of transfected cells harboured an aggresome, 16 hours post transfection and remained unchanged at 24 hours post transfection. In contrast to WT opsin-GFP, aggresome numbers were higher in cells expressing GFP-tagged mutant rhodopsin. At 16 hours post transfection, approximately 11 percent of cells expressing K296E rhodopsin-GFP contained an aggresome and for P23H rhodopsin-GFP, the percentage was approximately 20 percent. These experiments revealed a particular trend in mutant opsin aggresome formation. As time progressed numbers of mutant aggresome numbers increased and by 48 hours post transfection mutant aggresome numbers had doubled. Numbers of aggresomes of untagged opsin were similar over the same time points.

80 Figure 3.9 The ultrastructure of P23H rod opsin-GFP aggresomes

COS-7 cells were transfected with the plasmid pEGFP-P23H and 24 hours later were processed for electron microscopy as described in materials and methods, chapter 2. Electron micrograph of a COS-7 cell with a P23H rod opsin-GFP aggresome. i) The aggresome (A) forms near the nucleus (N) and centrosome (C) and is surrounded by mitochondria (M) ii) At higher magnification the intermediate filaments (IF) that surround the aggresome can be seen iii) The aggresome also has a high membrane content and in places appears to be surrounded by membrane (arrowheads). Scale bar 1 pm

81 60 50 1 i | 40 1 II:: 1 0 - ^ QJ 16 24 , 24 48

WT-GFP K296E-GFP P23H-GFP

Figure 3.10 The numbers of mutant rod opsin-GFP aggresomes in COS-7 cells increase with time

COS-7 cells were transfected with pEGFP-WT, pEGFP-K296E and pEGFP-P23H and the cells were fixed at the time points indicated (16 hours, 24 hours and 48 hours) and aggresomes were counted from duplicate wells. Indicated above the individual bars are the number of fields of view from which aggresomes were scored, along with error bars (SEM).

82 3.2.8 The influence of P23H rod opsin on the number of WT rod opsin-GFP aggresomes in COS-7 ceiis

Dominant mutations in the rhodopsin gene, Nina-E, of Drosophila cause autosomal dominant retinal degeneration and exert a dominant negative effect on the biosynthesis of the normal rhodopsin protein (Colley et ai, 1995). Therefore, the effects of mutant rod opsin on WT rod opsin aggresome numbers in COS-7 cells were investigated. COS-7 cells were co-transfected with pEGFP-WT encoding wild-type rhodopsin-GFP and pMTS- P23H encoding untagged P23H rhodopsin in a vector ratio of 1:10. The control for this experiment involved the co-transfection of COS-7 cells with the plasmids pEGFP-WT and pMT3-WT in a vector ratio of 1:10. To make a clearer distinction between normal trafficking of WT rod opsin-GFP through the Golgi and WT rod opsin-GFP aggresomes, the cells were treated with 20 pg/ml BFA for 15 mins prior to fixation and determination of aggresome numbers. An independent observer (Dr M. Cheetham), who was unaware of the transfection status of the cells, counted aggresomes from a total of four different experiments. Co-transfection of WT rhodopsin-GFP with an excess of WT untagged rhodopsin did not lead to an increase in aggresome formation above that seen with the GFP-tagged protein alone. In contrast, co-transfecting WT rhodopsin-GFP with untagged P23H rhodopsin led to an increase in the incidence of wild-type rod opsin aggresomes (Fig. 3.11). In the presence of untagged P23H rod opsin, the percentage of cells with WT rod opsin-GFP aggresomes approached that of untagged P23H rod opsin alone (35%- 45% at 24 hour).

3.2.9 The localisation of WT and mutant GFP-tagged rod opsin in SH-SY5Y cells using immunofluorescence microscopy

Rhodopsin is a rod cell specific GPCR and a rod cell is essentially a neuron, albeit a highly specialised one. Therefore to compare the phenotypic behavior and cellular localisation of GFP-tagged WT and mutant rod opsin in COS-7 cells with human neuronal cells I carried out studies using the human neuroblastoma cell line SH-SY5Y. These cells were transfected with plasmids pEGFP-WT, pEGFP-K296E, pEGFP-P23H and fixed in -20°C 100% methanol, 24 hours after transfection. The ER was labeled with an antibody to the ER resident chaperone BiP, followed by donkey anti-goat Cy3. WT rod opsin-GFP targeted to the plasma membrane (Fig. 3.12). K296E rod opsin-GFP, in the majority of transfected cells, formed numerous bright, intracellular aggregates, which did not co- localise with BiP and appeared to be in the cytosol (Fig. 3.12). Furthermore, faint K296E

83 rod opsin-GFP plasma membrane staining was observed. P23H rod opsin-GFP accumulated in the ER as shown by co-localisation with BiP, but also formed numerous small, bright, green spots that appeared to be in the cytosol (Fig. 3.12).

3.2.10 Mutant rod opsin-GFP aggregates sequester ubiquitin in SH-SY5Y cells

To determine if these intracellular aggregates were ubiquitinated, SH-SY5Y cells were co transfected with an equal vector ratio of plasmid encoding c-myc-tagged ubiquitin and plasmids: pEGFP-WT, pEGFP-K296E and pEGFP-P23H. The transfected cells were fixed 24 hours after transfection in -20'"C, 100% methanol, c-myc-tagged ubiquitin was detected with neat anti-c-myc hybridoma supernatant (9E10) followed by goat anti-mouse Cy3. Plasma membrane WT rod opsin-GFP showed no co-localisation with c-myc-tagged ubiquitin (Fig. 3.13). Epitope tagged ubiquitin in SH-SY5Y cells expressing WT opsin- GFP was found in the nucleus and the cytosol. In contrast, c-myc-tagged ubiquitin co localised with K296E-GFP and P23H-GFP aggregates (Fig. 3.13) and the staining pattern of c-myc ubiquitin was different to the normal pattern found in cells expressing WT rhodopsin-GFP. Interestingly, in a small percentage of cells a single ubiquitinated K296E- GFP aggregate was found in a juxta-nuclear position suggesting that multiple aggregates may coalesce to form an aggresome (Fig. 3.13).

84 A.

+BFA

W T-G FP /W T rod opsin WT-GFP/P23H rod opsin

WT-GFP/ WT-GFP/ WT OPSIN P23H-OPSIN

Figure 3.11 P23H rod opsin increases the numbers of WT rod opsin-GFP aggresomes in COS-7 cells

A. Co-transfection of GFP tagged WT opsin in pEGFP with untagged WT opsin in pMT3 (WT-GFP/WT-opsin) at a vector ratio of 1:10. This showed no increase in aggresome formation over the WT-GFP alone, whereas co-transfection of GFP tagged WT opsin in pEGFP with untagged P23H opsin in pMT3 (WT-GFP/P23H-opsin) at a vector ratio of 1:10 increased the percentage of WT-opsin aggresomes, as judged by GFP fluorescence. B. Quantification of co-transfection WT rod opsin-GFP aggresome formation, 24 hours after transfection. The percentage of cells with WT-GFP fluorescing aggresomes was counted double blind from 4 separate experiments, error bars ± 2 s.d.; BFA (20 mg/ml for 15 minutes prior to fixation) was added to cells in both experiments.

85 WT-GFP

K296E-GFP

P23H-GFP Merge

Figure 3.12 The localisation of GFP-tagged WT, K296E and P23H rod opsin in SH-SH5Y cells.

SH-SY5Y cells were transfected with the plasmids pEGFP-WT, pEGFP- K296E and pEGFP-P23H. The ER was detected using an antibody to the ER chaperone BiP (red). WT rod opsin-GFP (WT-GFP) translocated to the plasma membrane (arrow heads) and was not detected in the ER. K296E rod opsin-GFP (K296E-GFP) formed bright inclusions which did not co-localise with BiP (red). P23H rod opsin-GFP (P23H-GFP) showed partial co-localisation with BiP (merge) and formed several bright spots that did not appear to co-localise with BiP (arrows). Images are z-sections and scale bars are 10 pim.

86 Figure 3.13 Ubiquitin co-iocalises with GFP-tagged K296E and P23H rod opsin aggregates in SH-SY5Y ceiis.

SH-SY5Y cells were co-transfected with a plasmid encoding c-myc tagged ubiquitin and plasmids: pEGFP-WT, pEGFP-K296E and pEGFP-P23H. WT rod opsin-GFP (WT-GFP) translocated to the plasma membrane (arrow heads). The c-myc-tagged ubiquitin (c-myc Ub, red) was found in the cytosol and the nucleus. In cells expressing K296E rhodopsin-GFP (K296E-GFP) c-myc-tagged ubiquitin (c-myc Ub) was recruited to intra cellular K296E rod opsin-GFP (arrows) showing co-localisation. P23H rod opsin-GFP (P23H-GFP) that formed aggregates also co-localised with c- myc-tagged ubiquitin (arrows). Images are z-sections and scale bars represent 10 pm. K296E-GFP c-myc Ub

1 1I

P23H-GFP c-myc Ub

87 3.3 Discussion

The work described in this chapter has revealed that when RP mutants of the rhodopsin gene are expressed in COS-7 and SH-SY5Y cells the mutant rod opsins have a propensity to form intracellular ubiquitinated inclusions. Mutant rod opsin inclusions found in 008-7 cells show many features in common with aggresomes: they are found at or near the MTOC, are ubiquitinated, disrupt the Golgi apparatus, are encased by vimentin, and sequester molecular chaperones. The co-localisation of ubiquitin with opsin aggresomes suggests that opsin is ubiquitinated and the data in the following chapters hint at this possibility. However, just recently a report by llling et al 2002 has confirmed that mutant opsin is indeed ubiquitinated.

There appears to be a difference in the incidence of aggresome formation between the mutants K296E and P23H in COS-7 cells. The percentage of rod opsin expressing COS-7 cells with aggresomes was higher for the P23H mutant. P23H accumulates in the ER and forms aggresomes, but plasma membrane staining was very faint, in agreement with Sung at a!., 1991b. K296E accumulated in the ER, formed aggresomes, and showed plasma membrane staining greater than P23H but the percentage of cells with aggresomes was lower than that for the P23H mutant at 16 and 24 hours. The difference in aggresome numbers between the two mutants in COS-7 cells may be a reflection of the ease with which P23H folding intermediates assume misfolded off-pathway conformations in comparison to the K296E mutant. WT opsin also formed aggresomes but only in appoximately 5% of transfected cells and may be a result of over expression of the protein in these particular cells. Interestingly transgenic mice over expressing WT opsin experience retinal degeneration (Ollson at a!., 1992). P23H opsin has a propensity to form more aggresomes than the K296E mutant, but whether this has any bearing on toxicity to cultured cells or rod cells is unknown at present. Patients with the K296E mutation exhibit severe retinal degeneration, in contrast to the milder form of the disease shown by patients with the P23H mutation (Gal at a!., 1997). In K296E heterozygous transgenic mice, arrestin is bound to K296E rhodopsin (Li at a/., 1995) and as a result a proportion of constitutively active K296E may escape inactivation as the levels of arrestin available to bind K296E or even wild type rhodopsin may be depleted. This perhaps, would in turn, contribute to the rod cell’s vulnerability to light damage. It has been shown that the retinas of arrestin homozygous knockout mice (where, rhodopsin cannot be deactivated) are particularly susceptible to light induced degeneration which is not dependent on transducin i.e. the phototransduction cascade (Mao at a!., 2002). Thus, in vivo, K296E rhodopsin may be toxic in two ways; firstly as a result of aggregation as with P23H rhodopsin and 88 secondly by depleting the rod cell of arrestin, preventing inactivation of WT and K296E rhodopsin. Furthermore, in D.melanogaster photoreceptor degeneration has been associated with the formation of rhodopsin-arrestin complexes and blocking the formation of these complexes retards degeneration (Alloway et ai, 200).

Mutant opsin aggresomes appear spontaneously in cultured cells and increase over time. Misfolded integral membrane proteins (including opsin) are retrotranslocated from the ER membrane to the cytoplasm for degradation by the ubiquitin proteasome pathway (Chapter 5 and Saliba at ai, 2002, llling at ai, 2002). Retrotranslocated opsin is a highly hydrophobic protein and may be more prone to aggregation in the aqueous environment of the cytosol. The proportion of retrotranslocated opsin that is degraded compared to that which aggregates may be determined by the relative rates of proteasomal degradation and aggregation. To degrade misfolded opsin effectively, proteasomes must degrade opsin before it aggregates. Thus it appears that the rate of proteasomal degradation of mutant opsin is slower than aggregation, given the high incidence of aggresomes in COS-7 cells. Also as the copy number of the pMT3 plasmids increases over time (pMT3 contains the SV40 origin of replication) opsin levels increase, which probably exceed the cell’s capacity to fold or degrade mutant opsin leading to an increase in frequency of aggresomes over time.

K296E-GFP and P23H-GFP form multiple aggregates in SH-SY5Y cells that most likely eventually coalesce into aggresomes given that some cells harboured a single K296E- GFP inclusion at a juxta nuclear position. It appears that the coalescence of individual K296E-GFP and P23H-GFP rod opsin aggregates into aggresomes in SH-SY5Y cells takes longer than in COS-7 cells. It is possible that aggresomes in SH-SY5Y cells would disrupt the neurofilament network, (neurofilaments belong to a class of intermediate filaments which are specifically found in neurons) but awaits further investigation.

Rod opsin aggresomes are electron dense structures circled by vimentin and by mitochondria. Within rod opsin aggresomes membranous material was seen and may reflect the highly hydrophobic nature of the transmembrane domains of rod opsin, which may sequester lipids as well as chaperones. In addition ER resident chaperones BiP (lumenal) and Calnexin (transmembrane) are occasionally associated with rod opsin aggresomes forming a ring around the inclusion body implying that ER membranes are disrupted by large rod opsin aggresomes. BiP was also shown to associate with the periphery of Huntington aggresomes (Waelter et al., 2001). The observation of electron dense opsin aggregates in vivo has been mostly inconclusive. Electron dense material 89 was described in transgenic Drosophila heterozygous for mutant NinaE alleles (Kurada and O'Tousa, 1995) but the authors did not present data showing that these inclusions consisted of Rhi. In mammals, electron dense material present at a juxtanuclear position that disrupted the Golgi apparatus was described in the rods of transgenic mice heterozygous for P23H mutant rhodopsin (Olsson et a/., 1992). However, once again no data were presented confirming that these inclusions contained rod opsin and it remains unknown whether mutant rod opsin forms intracellular aggregates in rod photoreceptors. Electron dense material immunoreactive for rod opsin, however, was discovered in the transgenic pig heterozygous for P347L rhodopsin (Li et a/., 1998). This mutation does not alter folding, as P347L rod opsin expressed in cell culture targets to the plasma membrane, is expressed at WT levels and associates with 11-c/s-retinal to form functional receptor (Kaushal and Khorana 1994). The P347L mutation prevents the interaction of Tctex-1, a dynein light chain, with the cytoplasmic tail of rod opsin. As a result P347L rhodopsin transport to the rod outer segment is disrupted leading to the mislocalisation of the mutant protein to the cell body plasma membrane (Li et ai, 1998). The detection of mutant rod opsin aggregates in rod photoreceptors needs further investigation and it is of interest that a rod cell has a specialised MTOC for the connecting cilia, presenting the possibility that opsin aggregates could coalesce at these sites.

In Drosophila retinal degeneration is caused by dominant mutations in the NinaE gene encoding rhodopsin (Rhi) and these mutants exert a dominant negative affect on the normal maturation of WT Rhi (Colley et ai, 1995, Kurada and O'Tousa 1995). In flies heterozygous for the dominant NinaE mutant allele, WT Rhi protein levels are dramatically reduced. The mutant protein prevents the normal maturation of WT Rhi and as a result accumulates in the ER and does not target to the rhabdomere. Also the levels of WT rhodopsin in the outersegments of transgenic mice heterozygous for a rhodopsin triple mutant (V20G, P23H and P27L) were reduced (Wu et ai, 1998). Co-transfection studies, revealed that P23H rod opsin expression increased the proportion of WT rod opsin present in aggresomes. It appears that P23H rod opsin recruits the WT protein to intracellular inclusions but one can only speculate as to how this might happen. Since WT rod opsin present in aggresomes has been retrotranslocated from the ER it is perhaps likely that disruption of the normal processing of WT rod opsin by the mutant protein is an event which occurs prior to retrotranslocation i.e. in the ER. In Drosophila, WT rhodopsin initially forms a dimer which eventually leads to a momomeric form, as the protein matures and translocates through the secretory pathway (Kurada et ai, 1998). Also mutant Rhi dimerises but does not mature to a monomeric form but is rapidly degraded and mammalian mutant rhodopsin forms dimers almost exclusively in vivo when expressed in 90 homozygous rhodopsin knockout mice (Frederick et al., 2001). As opsin aggregation is highly specific and opsin does not co-aggregate with different proteins (Rajan et a!., 2001) it is possible that in COS-7 cells or indeed in vivo there may be an association of misfolded mutant rod opsin with folding intermediates of WT rod opsin increasing the proportion of WT folding intermediates that form off-pathway products. Presumably WT opsin and mutant opsin co-localise in aggresomes, and fluorescence resonance energy transfer (FRET) analysis would reveal whether they also co-aggregate. Alternatively the influence of mutant opsin on WT processing may be indirect, as mutant opsin may compromise the folding environment of the ER, perhaps by depleting the levels of available chaperones, thus having a general effect on the folding and processing of cellular proteins in the secretory pathway. This may be investigated by the co-expression of AF508CFTR-GFP and WT rhodopsin and would reveal the effects of AF508CFTR-GFP aggresomes on WT opsin processing. Mutant rod opsin may also influence the steady state levels of the wild type protein as seen in Drosophila. Time permitting, the influence of mutant rod opsin on the steady state levels of WT protein could have been assessed in COS-7 cells by co transfection studies using the plasmids encoding GFP-tagged WT and untagged P23H rhodopsin followed by immunoblotting with an antibody to GFP. This would reveal whether mutant opsin reduces the steady state levels of the WT protein perhaps by promoting its ER associated degradation.

Aggresome numbers in COS-7 cells differed between mutants and immunofluorescence data from the studies using SH-SY5Y cells revealed a difference in the processing of K296E and P23H rod opsin-GFP mutants. P23H rod opsin-GFP in SH-SY5Y cells showed ER staining and small aggregates scattered throughout the cell. ER fluorescence of K296E rod opsin-GFP was undetectable in most cells and aggregates were much larger and fewer in number, once again demonstrating the different properties of these two mutants. Also, there appears to be a difference in the processing of mutant opsin between the two cell lines used. There may be various reasons for this but one reason relating to the levels of molecular chaperones between the two cell lines is an attractive idea.

The expression of misfolded proteins is toxic to neurons in vitro and in vivo and the mechanisms by which misfolded proteins lead to cell death is still unclear (Sherman and Goldberg, 2001, Taylor et al., 2002). Future experiments investigating the toxicity of mutant rod opsin to SH-SY5Y cells and COS-7 cells may be performed by monitoring cell death of transfected cells over different time points, perhaps using the TUNEL assay, or counting numbers of remaining transfected cells. In fact my preliminary findings revealed

91 that 48 hours after transfection, many SH-SY5Y cells expressing GFP-tagged mutant rod opsins were either rounded up or floating in the medium. In contrast a greater number of SH-SY5Y cells expressing WT rhodopsin-GFP or GFP had a normal morphology and remained attached to the culture dish. These findings have not yet been quantified, but experiments in the laboratory of Dr. M. Cheetham are underway, and initial reports confirm my preliminary findings (Dr. S. Wilkins, personal communication).

Furthermore it remains to be established, whether rod opsin aggresomes are toxicper se, or whether toxicity is due to the continued expression of the mutant rhodopsin gene, and not necessarily a result of rod opsin aggresomes. However, mutant rod opsin aggresomes severely impair the UPS in cultured cells (Dr. R. Kopito, personal communication and llling et al., 2002) as do other protein aggresomes (Bence et a!., 2001). It has been suggested that there is a link between UPS dysfunction and neurodegeneration as revealed by genetic evidence linking mutations in the UPS to several neurodegenerative diseases and animal models of neurodegeneration (as described in the introduction). The toxicity of opsin aggregates may therefore be due to the impairment of the UPS. Protein aggregates have recently been shown to be degraded by an autophagic route in cultured cells (Ravikumar et a!., 2002) and aggresome formation might be a means of concentrating protein aggregates at the MTOC increasing the efficiency of capture into autophagic vesicles. Thus the targeting of smaller aggregates to the centrosome in a dynein- dependent retrograde transport on microtubules may serve a protective role (Sisodiaet a/., 1998) as interruption of inclusion formation by mutant polyglutamine using microtubule destabilising drugs resulted in enhanced toxicity in yeast (Muchowski et a!.. 2002). Also transgenic mice expressing ataxin-1 with an expanded polyglutamine repeat showed greater toxicity when it was not sequestered into nuclear inclusions (Cummings et a!., 1999). Recent studies support the hypothesis that pre-fibrillar intermediates (protofibrils) and not mature amyloid fibrils which make up inclusions, may be the key toxic species in neurodegenerative diseases such as AD, PD and HD (Muchowski, 2002). it Is unknown whether misfolded opsin forms ordered fibrillar structures (amyloid), but it is possible that only certain opsin intermediates may be toxic (i.e. smaller aggregates or pre-aggregation intermediates) rather than opsin aggresomes.

Rod opsin aggresomes may be stable structures, which are resistant to further degradation or may be subjected to degradation via an autophagic route. Therefore, to investigate further the stability of opsin aggresomes and autophagic degradation futurein vitro experiments might involve the expression of mutant rhodopsin genes under the control of an inducible promoter, where opsin expression could be regulated. In fact, the 92 first year of my Ph.D. was spent generating stable SH-SY5Y cells expressing WT, G90D and K296E under the control of a tetracycline regulated promoter. I established SH-SY5Y cells stably expressing WT, K296E and G90D rhodopsin. Rhodopsin was cloned into the tetracycline regulated retroviral vector pBPSTR-1 (Paulus et al., 1996) which was then used to generate SH-SY5Y clones. This vector is composed of two parts: (a) the E.coli derived tetracycline resistance operon regulatory elements (tetO) embedded within a minimal cytomegalovirus promoter and (b) the gene encoding a hybrid protein, the tetracycline repressor protein (tTA) fused to the herpes simplex virus transactivator protein (VP6). The vector also contains the gene for puromycin resistance, which allows for the positive selection of transfected cells. The opsin gene was inserted downstream of the tetO-minimal CMV promoter, thus making the expression of the gene dependent on tTA which binds to the tetO sequences and recruits cellular transcription factors through its VP16 domain and gene expression is inhibited by the addition of tetracycline. However, tight regulation of rhodopsin expression was not achieved using this system and many of the clones isolated were not clonal, as many cells within in a given clone were not expressing rhodopsin. However, only one clone expressing WT rod opsin was tightly regulated by tetracycline. Other clones (WT and G90D), which were isolated, were found to continuously express rod opsin. Only one K296E clone was isolated and this was responsive to tetracycline but approximately only 1% of the cells in this clone were expressing K296E as assessed by immunofluorescence microscopy. On removal of tetracycline from the medium for 48 hours the immunofluorescence signal was weak compared to clones expressing WT and G90D opsin. Weak plasma membrane staining of K296E rod opsin was observed and no intracellular aggregates were detected. The difficulty in growing and isolating K296E clones may be related to the toxicity of this mutant protein given that tight regulation of gene expression was not achieved using tetracycline in these cells. The G90D mutant protein (causing CSNB rather than retinal degeneration) is not toxic to rod cells and this may be the reason why more clones were isolated expressing this mutant.

Molecular chaperones are typically found in the neuronal inclusions associated with many neurodegenerative diseases (Welch and Gambetti, 1998, Sherman and Goldberg, 2001). The molecular chaperone Hsp70 was found to be associated with rod opsin aggresomes implying that rod opsin is misfolded in these inclusions. A number of recent reports have revealed that molecular chaperones suppress polyglutamine-mediated neurodegeneration in fruit fly and mouse models (Warrick et a/., 1999, Jana et al., 2000, Cummings et al., 2000). Further more, molecular chaperones (Hsp70 acting synergistically with Hsp40) enhanced the solubility of polyQ proteins in Drosophila, suggesting that these chaperones 93 mediate suppression of polyQ neurotoxicity by increasing the solubility of polyQ proteins (Chan et al., 2000). It may be possible to manipulate mutant rod opsin processing by altering the levels of molecular chaperones in this cell culture model. Co-expression of mutant rhodopsin with Hsp70 or with Hsp40 (a co-chaperone of Hsp70) could possibly increase misfolded opsin solubility and influence mutant opsin aggregation or degradation as was shown for other intracellular protein aggregates (Bailey et al., 2002). In fact expression of K296E rhodopsin in SH-SY5Y cells over expressing HSJ1a (a neuron specific chaperone, which is a member of the DnaJ family of molecular chaperones which all have a region known as the J domain that stimulates the ATPase activity of Hsp70) appeared to increase the proportion of K296E rod opsin at the plasma membrane (M.E.Cheetham unpublished results).

Mutant opsin accumulates in the ER where a quality control mechanism exists ensuring the detection and retention of terminally misfolded proteins. Important components of ER quality control are the lectin chaperones CNX and CRT, which associate with misfolded glycoproteins. Prolonged association of misfolded proteins with CNX or CRT leads to the subsequent retrotranslocation and degradation of these misfolded polypeptides. In the next chapter I explore the role of N-linked oligosaccharides in the ER associated degradation of mutant opsin.

94 Chapter 4: N-linked giycosylation of rod opsin and ER associated degradation

4.1 Introduction

An important post-translational event occurring during the biogenesis of glycoproteins is the addition of N-linked oligosaccharides within the lumen of the endoplasmic reticulum. N-linked oligosaccharides ensure the correct folding and cellular targeting of many glycoproteins (Paulson 1989, Helenius, 1994). Bovine rod opsin is N-linked glycosylated at asparagine residues 2 and 15 and the oligosaccharide groups are modified as the protein translocates through the ER and Golgi, eventually reaching the rod outer segment or the cell membrane in cell culture (Chapter 1).

One of the ways that has been used to determine the role of N-linked oligosacchharides in protein biogenesis is the use of pharmacological agents, which inhibit this process, such as the antibiotic tunicamycin (TM). Isolated from Stmptomyces lysosuperficus, TM is a hydrophobic analogue of UDP-N-acetylglucosamine, which blocks the addition of N- acetylglucosamine to dolichol phosphate, the first step in the formation of the core oligosaccharide (as described in chapter 1 ). TM has been used in a number of biological systems to determine the role of carbohydrate moieties in the cellular functions of glycoproteins, including rod opsin.

The function of the carbohydrate groups of rod opsin in the folding, biological activity and cellular targeting of the protein has been studied in a number of experiments using TM. Intravitreal injection of TM in the leopard frog {Rana pipiens) resulted in retinal abnormalities, similar to those seen in some retinal dystrophies, in which there is degeneration of both rods and cones but without apparent pathological effects on other retinal neurons (Fliesler et al., 1984). Further experiments on the isolated retinas from Xenopus laevis and Rina pipiens, revealed that TM prevented giycosylation of rod opsin, without perturbing the synthesis and transport of the protein through the rod cell body to the apical plasma membrane of the inner segment (Fliesleret a!., 1985, Fliesler and Basinger, 1985). Disc assembly, however, was severely disrupted and resulted in an accumulation of rod opsin and membranous material in the intersegmental space (Ulshafer et a/., 1986). The oligosaccharide groups of rod opsin, may not be required; however, for the incorporation into ROS disc membranes in mammalian retinas. For example.

95 example, unglycosylated rod opsin has been isolated from rod outer segments from bovine retinas treated with TM (Plantneret al., 1980).

Drosophila rhodopsin Rhi has two consensus sites for N-linked glycosylation, at asparagine residue 20 (Asn-20) and asparagine 196 (Asn-196) but Rhi is exclusively glycosylated at Asn-20 (Katanosaka et al 1998). Insights into the role of N-linked glycosylation in Rhi maturation came from the work by O'Tousa, 1992 and Webel et al., 2000. These investigators revealed that substitution of N20 with isoleucine (N20I) in Rhi prevented translocation of the protein to the rhabdomere, which is the equivalent of the mammalian rod outer segment. Furthermore, mutant N20I Rhi accumulated in the ER and disrupted the normal maturation and translocation of WT Rhi (Webel et al., 2000). The authors concluded that in Drosophila, glycosylation at Asn-20 of Rhi is essential for its maturation and translocation to the rhabdomere (Webel et al., 2000). In contrast, the folding and trafficking of rhodopsin in cultured mammalian cells is not dependent on N- linked glycosylation (Kaushal et ai, 1994)

The evidence from in vivo and in vitro studies, therefore, suggests that the N-linked oligosaccharide groups of vertebrate wild type rod opsin are not essential for the folding of the protein or translocation through the secretory pathway. Nevertheless, N-linked oligosaccharide groups of rod opsin appear to be required for disc assembly in vivo (Ulshafer et ai, 1986) and possibly transducin activation in vitro (Kaushal et ai, 1994).

Not only is N-linked glycosylation essential for the folding, maturation and subcellular distributuion of many glycoproteins but plays a pivotal role in the targeting of terminally misfolded proteins for ERAD as discussed in chapter 1. As the involvement of the oligosaccharide groups of rod opsin in the ERAD of mutant rod opsin is unclear, I examined the role of N-linked oligosaccharides in the degradation of mutant opsins in cell culture.

96 4.2 Results

4.2.1 The glycoforms of wild type rod opsin processed in COS-7 cells

To investigate rod opsin processing in cell culture and examine further the biochemical properties of mutant rod opsin, I established the necessary conditions to examine different rod opsin species by western blotting. COS-7 cells were transfected with the plasmid pMT3-WT and 24 hours later cells were lysed in PBS containing 1% n-Dodecyl-p-D- Maltoside (DM) a non-denaturing detergent which affords maximal stability for folded opsin (Cha et al., 2000). Cell lysates were partitioned into DM soluble and DM insoluble fractions by centrifugation as described in chapter 2. Fractionated lysates equivalent to 10 jig of total cell lysate were resolved on a 10% SDS-PAGE gel followed by immunoblotting with 1D4. Western blots revealed that wild type rod opsin, in the DM soluble fraction, migrated as a series of bands with a sharp leading edge at «40 kDa (Fig. 4.1a) in agreement with Kaushal and Khorana (1994). Rod opsin g ly coforms migrated from «40 kDa up to the opsin dimer (Fig. 4.1a and Fig. 4.1c) and represent different glycosylated species because these glycoforms disappeared after deglycosylation with the glycosidase; Peptide N-Glycosidase F (PNGase F). Deglycosylated rod opsin migrated as a 33 kDa species (Figure 4.1a) and a small amount of rod opsin migrating at approximately 60 kDa probably represents opsin dimer. In addition expression of the WT rhodopsin gene in COS-7 cells in the presence of TM (0.8 pg/ml) for 16 hours, also yielded a 33 kDa unglycosylated species (Figure 4.1b). This observation agrees with the previous findings of Oprian (1987), Kaushal and Khorana (1994) and Heyman and Subramaniam (1997). PNGase F is an amidase that cleaves between the innermost N-acetyglucoseamine and asparagine residue of high mannose, hybrid and complex oligosaccharides from N-linked glycoproteins, essentially cleaving all N-linked oligosaccharides from glycoproteins:

PNGase F X-Man ^ Man-GlcNAc-GlcNA5-Asn-i

X-Man [GlcNAc: N-acetyglucosamine, Man: Mannose, X = H or sugar(s)]

97 A 33 kDa rod opsin species was present in the DM insoluble fraction and a band at approximately «60 kDa probably representing rod opsin dimer, along with a high molecular weight rod opsin smear which extended to the top of the resolving gel and into the stacking gel (Fig. 4.1a).

The N-linked oligosaccharide chains of rod opsin are modified as the protein translocates through the secretory pathway. In the ER, mannose groups are removed and other sugar groups are added in the Golgi (as described in chapter 1) eventually yielding mature glycosylated rod opsin, which translocates to the cell surface. Mature rod opsin isolated from COS cells has been reported to migrate as a species of «40 kDa on an SDS-PAGE gel (Doi et al., 1990, Kaushal and Khorana 1994). Opsin species residing in the ER can be distinguished from Golgi and post Golgi species by digesting opsin with the enzyme endoglycosidase H (Endo H). This enzyme cleaves only high mannose oligosaccharides or hybrid structures from N-linked glycoproteins. Endo H therefore will digest only ER resident opsin (which are high mannose glycoforms) to the core polypeptide, albeit with one attached N-acetylgucosamine residue remaining at each consensus sequence: Endo H

(Man)n-Man ^ ^ Man-GlcNAc-GlcNAc-Asn- X -M a n ^ / Y

n — 2-150, X — (Man)i_2, Y — H In hybrid structures: n = 2, X and/or Y = AcNeu-Gal-GlcNAc

In order to distinguish between immature glycosylated ER resident rod opsin and mature post Golgi rod opsin, the DM soluble fraction from COS-7 cells expressing the WT rhodopsin gene, was incubated with Endo H, followed by SDS-PAGE and western blotting. The blot in Figure 4.1c, shows that the majority of the WT rod opsin glycoforms are post ER as they remain insensitive to Endo H. A very faint band of 33 kDa appeared, however, following Endo H treatment and revealed the presence of high mannose, ER resident, rod opsin species.

98 A. PNGase p ^ + PNGase F TM —- + — s s s

112 — 81 — ** 49.9 —

* u 36.2 — B 29.9 —

c. PNGase F — — + Endo H — + — S s s

113 — 92 — ** 52.3 —

9 35 — W 28.9 —

Figure 4.1 The glycoforms of WT rod opsin expressed in COS-7 cells.

COS-7 cells were transfected with the plasmid pMT3-WT and 24 hours later were lysed in PBS+1% DM. DM soluble (8) and DM insoluble (P) fractions equivalent to 10 pig of whole cell lysate (T) were resolved on a 10% SDS- PAGE gel followed by immunoblotting with 1D4. A) WT rod opsin in the DM soluble (8) fraction migrates as a series of species from «40 kDa (*) to «86 kDa, which migrate as a 33 kDa (arrow) species and opsin dimer (**) following deglycosylation with PNGase F. A 33 kDa species (arrow) and dimer (**) are present in the DM insoluble (P) fraction, plus a high molecular weight smear which stretches to the top of the resolving gel and into the stacking gel (arrow head). B) The DM soluble (8) WT rod opsin glycoforms also migrate as a 33 kDa species (arrow) following a 16 hour incubation with 0.8 [xg/ml tunicamycin. 0) A faint deglycosylated opsin band appears following Endo H treatment (arrow). For comparison PNGase F treated DM soluble WT rod opsin was loaded on to the same gel, showing deglycosylated opsin (arrow) and dimer(**). Molecular weight markers in kDa are shown on the left.

99 4.2.2 The different glycoforms of WT, K296E and P23H rod opsin in COS-7 cells

In order to compare the different WT rod opsin species present in COS-7 cells with mutant rod opsin species, COS-7 cells were transfected with pMT3-WT, pMT3-K296E and pMT3- P23H and 24 hours later the cells were lysed in PBS+1% DM. Transfection efficiencies for all the vectors used were approximately 50% as assessed by immunofluorescence microscopy. The lysates were fractionated into DM soluble and DM insoluble fractions by centrifugation as (described in chapter 2). Equal volumes of DM soluble and DM insoluble fractions corresponding to 20 pg of whole cell lysate were analysed by SDS-PAGE and immunoblotting with 1D4. K296E and P23H rod opsin was present in COS-7 cells at lower levels than the wild type protein (Fig. 4.2) which was a consistent finding agreeing with previous reports (Sung et al., 1991b, Kaushal and Khorana, 1994). My immunocytochemical observations (chapter 3) also revealed less mutant opsin plasma membrane staining compared to WT. K296E and P23H rod opsins in the DM soluble fraction appear to have a similar glycosylation pattern to wild type (Fig. 4.2) with the presence of a «40 kDa band. The amount of K296E and P23H rod opsin in the DM soluble fraction is greater than that in the DM insoluble fraction. Non-glycosylated 33 kDa K296E and P23H rod opsin was present in the soluble and insoluble fractions and a small amount of dimer was present in the insoluble fraction (Fig. 4.2). Furthermore, a high molecular weight smear of mutant rod opsin was observed in the DM insoluble fraction which reached to the top of the resolving gel (Fig 4.2) and into the stacking gel. An immunoreactive species migrating faster than the 33 kDa species was detected in the K296E and P23H rod opsin blots (Fig. 4.2.). This species has also been observed in published blots of mutant rod opsins (Kaushal and Khorana 1994). A non-specific immunoreactive species was detected in non-transfected COS-7 lysates, which migrated more slowly than the opsin dimer.

To determine if the P23H rod opsin glycoforms in the DM soluble fraction represented the mature glycosylated species, the DM soluble fraction was incubated with Endo H. The «40 kDa P23H rod opsin species is partially insensitive to Endo H, suggesting that a percentage of P23H rod opsin translocated to the Golgi and plasma membrane (Fig. 4.3) but there is a noticeable reduction in the intensity of this 40 kDa band following digestion with Endo H. PNGase F treatment of DM soluble P23H rod opsin produces the expected 33 kDa species, however, there is an increase in the intensity of the high

100 W K296E P23H COS-7 S P s P i P T

112 “ “

81 —

49 9 —

36.2 — 29.9 —

Figure 4.2. Comparison of DM soluble and DM insoluble WT and mutant opsin species

COS-7 cells were transfected with the plasmids pMT-3WT, pMT3-K296E and pMT3-P23H and 24 hours later the cells were lysed in PBS+1% DM. A volume of the soluble (S) and insoluble (P) fraction equivalent to 20 pg of total cell lysate was resolved on a 10% SDS-PAGE gel. The K296E and P23H opsin banding pattern appears to be similar to that of the WT protein with an 33 kDa species (arrow head) » 40 kDa species (arrow), plus higher molecular weight forms in the DM soluble faction. A K296E and P23H opsin species migrating faster than the 33 kDa species (*) was seen in the DM soluble fraction. The protein levels of the mutants were consistently lower than that of WT (WT>K296E>P23H). Whole cell lysate (T) from control non- transfected COS-7 cells (20 pg) was used to show specific immunoreactivity. A non-specific immunoreactive species was observed in non-transfected cell lysates which migrated more slowly than opsin dimer (**). Molecular weight markers in kDa are shown on the left.

101 Endo H PNGase F —

113—

Figure 4.3 The glycoforms of DM soluble P23H opsin in COS-7 cells

The DM soluble (8) fraction (20 pig) was treated with Endo H or PNGase F. The P23H rod opsin doublet migrating at «40 kDa t>ecomes one band which migrated at «40 kDa after incubation with Endo H. This band was less intense following Endo H treatment, leaving a faint band representing an Endo H resistant species (arrow). This showed that a species migrating at « 40 kDa (arrow) is mature glycosylated P23H opsin and migrated close to an Endo H sensitive P23H immature species. Incubation with PNGase F resulted in deglycosylated P23H rod opsin (arrow head). A high molecular weight smear from the 52.3 kDa marker to the top of the resolving gel was present in all lanes. Molecular weight markers in kDa are shown on the left.

102 molecular weight smear (Fig. 4.3) suggesting that deglycosylated P23H rod opsin has a greater propensity to form high molecular weight oligomers under these in vitro conditions

4.2.3 N-linked glycosylation of mutant rod opsins is required for ERAD in COS-7 cells

The role of N-linked glycosylation in WT and mutant rod opsin processing and quality control in COS-7 cells was examined using TM. As expected, treatment with 0.6 pg/ml TM for 16 hours prevented the formation of the mature di-glycoslyated form of WT and mutant rod opsins and resulted in the appearance of non-glycosylated 33 kDa rod opsin (Fig. 4.4). TM did not effect the transport of the wild type protein to the cell surface (after an 8 hour and 16 hour incubation) (Fig. 4 5) in agreement with the findings of Kaushal et a/.,1994 In addition, the steady state levels of WT rod opsin appeared to remain unchanged following TM treatment (Fig. 4.4). In contrast, the degradation of K296E and P23H rod opsins was inhibited by TM showing a striking increase in the steady state levels of the 33 kDa mutant rod opsin species (Fig. 4.4).

To determine the intracellular location of mutant rod opsin in COS-7 cells, which had accumulated as a result of the incubation with TM, immunofluorescence microscopy was used. COS-7 cells were transfected with pMT3-WT, pMT3-K296E and pMT3-P23H and 24 hours later the cells were incubated with 0.8 pg/ml TM for 8 hours and 16 hours. K296E and P23H rod opsin in untreated cells showed faint ER staining and some cells harbored aggresomes (Fig. 4.5). K296E was also detected at the plasma membrane (Fig. 4.5). Immunofluorescence detection of K296E and P23H rod opsins in COS-7 cells following an 8 hour incubation with TM, revealed a striking increase in intracellular mutant rod opsin fluorescence in all transfected cells (Fig. 4.5). Mutant rod opsin which had accumulated as a result of TM treatment was detected at the level of the ER as shown by the co-localisation with the ER integral membrane chaperone calnexin (Fig. 4.5). A 16 hour incubation of transfected cells with TM also appeared to result in the accumulation of mutant rod opsin in the ER but also caused the cells to round up which made it difficult to asses the intracellular location of mutant rod opsin.

103 TM — + + — + T T T T T

110 90 * 51.2 —

36.2 — 29.9 —

WT rod opsin K296E rod opsin P23H rod opsin

Figure 4.4 TM prevents the degradation of mutant rod opsin in COS-7 cells

COS-7 cells were transfected with pMT3-WT, pMT3-K296E, and pMT3-P23H and 24 hours later the cells were treated for 16 hours with 0.8 p.g/ml of TM. Whole cell lysates (T) (20 ^g) were resolved on a 10% SDS-PAGE gel, then analysed by immunoblotting with mAb 1D4. As expected WT and mutant rod opsin after TM treatment (as indicated) now migrated as a 33 kDa species (arrow) with the appearance of rod opsin dimer (*). TM had no noticeable effect on WT rod opsin steady state levels but a striking increase in mutant rod opsin steady state levels occurred following incubation with TM. Blot exposures were adjusted to give equivalent band intensity between WT and mutants. Molecular weight markers in kDa are shown on the left.

104 Figure 4.5 TM Induces the accumulation of mutant rod opsin in the ER of COS-7 ceils

COS-7 cells were transfected with the plasmids pMT3-WT, pMT3-K296E and pMT3-P23H and 24 hours later the cells were incubated with 0.8 )ng/ml TM for 8 hours. Rod opsin was detected with mAb 1D4-FITC and the ER was labelled using a rabbit antibody to calnexin followed by anti-rabbit Cy3. TM did not disrupt the translocation of WT rod opsin (green) through the Golgi to the plasma membrane. K296E and P23H rod opsin (green) showed a faint fluorescent signal in the ER of untreated cells. In contrast TM caused a striking accumulation of mutant rod opsin in the ER (panels labelled +TM) as shown by co-localisation with calnexin (red). Nuclei were counter stained with DAPI. Images are z-sections and scale bars are 10 ^m WT rod opsin C ahexin

WT rod opsin Calnexin

■ 1 ■1 1 I 1 1 1 um ■ 1 ■ 1 11

105 4.2.4 Tunicamycin induced the accumulation of mutant rod opsins in the ER of SH-SY6Y cells

The role of N-linked glycosylation in the of mutant opsins using TM was initially assessed in human neuroblastoma cells (SH-SY5Ys) prior to experiments with COS-7 cells. However, since transfection efficiency was low in SH-SY5Y cells, experiments using COS-7 cells were used follovdng my preliminary findings in SH-SY5Y ..ells. TM (0.8 pg/ml) was added to SH-SY5Y cells 24 hours after transfection with the plasmids pMTS-WT, pMT3'K29ôE and pMf3-F^23H and cells were fixed in 100% methanol at -20 "C prior to labeling. An 8 hour or 16 hour incubation of SH-SY5Y cells expressing the \AvT rhodopsin gene, with TM (0.8 ug/ml) did not perturb the translocation of WT rod opsin to the coll surface and no ER VfT rod opsin was detected (Fig 4.6) In untreated SH-SY6Y cells untagged K296E and P23H rod opsin showed faint plasma membrane staining and ER opsin fluorescence was observed in some cells as revealed by co-localisation with the ER resident chaperone Grp94 (of the Hsp90 family of molecular chaperones). In addition mutant opsin accumulated inside the cell forming a speckled fluorescence-staining pattern which did not co-localise with Grp94 (Fig. 4.6). Incubation of SH-SY5Y cells expressing the K296E or P23H rhodopsin gene with TM resulted in a striking increase in the mutant rod opsin fluorescence signal, which extended through out the cell to the plasma membrane and co-localised with Grp94 (Fig. 4.6).

106 Figure 4.6 TM induces the accumuiation of mutant rod opsin in the ERofSH-SYSYcelis

SH-SY5Y cells were transfected with the plasmids pMT3-WT, pMT3- K296E and pMT3-P23H and 24 hours later the cells were treated with 0.8 [ig/ml TM for 8 hour. A) WT rod rod opsin was detected with 1D4-F1TC. The ER was labelled with goat polyclonal antibody to the ER resident chaperone BiP followed by donkey anti-goat Cy3. TM did not influence the translocation of WT rod opsin (green) to the cell surface (panels labelled +TM). Rod opsin was seen at the cell surface (arrow head) before and after incubation with TM. B) Mutant rod opsin was detected with 1D4 followed by goat anti-mouse Cy3. The ER was labelled with a rat mAb to the ER chaperone Grp94 followed by goat anti-rat-FlTC. In untreated cells intracellular K296E and P23H rod opsin (red) gave a scattered fluorescent signal. In addition faint plasma membrane staining was seen (arrow heads). TM induced a striking increase in the mutant rod opsin fluorescent signal which partially co-localised with Grp94 (panels labelled +TM). Images are z-sections and scale bars represent 10 |xm. B.

« tf

’M erge ' ; .« L

K296E rod opsin

107 4.3 Discussion

The different WT and mutant rod opsin species detected by immunoblotting reveal the heterogeneity of the oligosacharide chains of rod opsin and is indicative of oligosaccharide processing as the protein translocates through the ER and Golgi. Using the enzyme Endo H it was possible to distinguish betvæen immature ER resident rod opsin and post ER mature glycosylated rod opsin. The majority of P23H rod opsin in the DM soluble fraction is the immature glycoforms in the ER membrane as these species were sensitive to Endo h and only a small proportion is mature glycosylated P23H rod opsin. Daglycosylating P23H rod opsin with PNGase F promotes aggregation as shown in figure 4.3. Even W1 rod opsin tends to aggregate, but to a lesser extent, following digestion with PNGase F The glycans of glycoproteins have a stabilising effect on proteins and increase their solubility (imperiali and O' Connor 1999) and may explain why glycosylated P23H rod opsin is more soluble than the unglycosylated form

The fractionation of cell lysates into soluble and insoluble fractions was not done in a stringent detergent, rather a DM based buffer was used similar to the buffer that is used in opsin reconstitution studies (where the protein is kept in a folded state ready for retinal binding and spectral studies). DM allows for maximal opsin stability (Cha et al., 2000) but is too mild to provide a definitive fractionation between ‘soluble protein’ and aggregated protein. WT-opsin in the ‘insoluble’ fraction contains unglycosylated opsin and probably also contains folding intermediates that would not normally be regarded as aggregated but aggregate post-lysis as they are not stable in the DM buffer. When opsin is separated by SDS-P.AGE, a series of bands is often generated corresponding to dimer, trimer and higher molecular weight oligomers, but rhodopsin is thought to be a monomer in disk membranes (Cone 1972; Downer 1985). This generation of a multiplicity of bands is considered to be a unique characteristic of the protein (Sung et a!., 1991b, Hargrave, 1995) but is also a nuisance when trying to determine the different opsin species present in COS-7 cells. Using more stringent detergents like 1% Triton X-100 to solubilise opsin in cells, increased the amounts of high molecular weight oligomers, so the use of the non-denaturing detergent DM was continued in my experiments. Also, boiling exacerbated the problem of aggregation, therefore samples were incubated at 22°C for 15 mins in sample buffer, prior to loading on a SDS-PAGE gel.

108 There is less mutant opsin in the insoluble fraction than the soluble fraction and one would expect to see more in the insoluble fraction if mutant opsin is present in aggresomes. The insoluble fraction of the mutant opsins (in particular P23H) contain less opsin, relative to the soluble fraction, probably because the opsin present in aggresomes was not solubilised effectively post-fractionation and never enters the SDS-PAGE gel. Indeed the blots reveal evidence for this, as mutant opsin was detected in the well at the entry point of the stacking gel. Since opsin can form high molecular weight oligomers in denaturing detergents, distinguishing between opsin that is present as aggregated oligomers within cells and those oligomers which have formed as a result of detergent solubilisation is not straightforward. However, just recently it has been shown, by co-immunoprecipitation and western blotting, that ubiquitinated oligomeric mutant opsin is in fact present within cultured human embryonic kidney (HEK) 293 cells (lilinget al., 2002).

Rlegarding rod opsin N-linked glycosylation and ER.4D, the work in this chapter has demonstrated that N-linked glycosylation of rod opsin appears to be essential for the targeting of the mutant protein for degradation and implies that calnexin or calreticulin play a role in its degradation. The mechanisms of ER quality control of proteins are slowly beginning to be unraveled and involves the cycling of glyco-proteins between the lectin chaperones CNX/CRT and a folding sensor (UDP)-glucose:glycoprotein glucosyltransferase (Helenius and Aebi 2001). The folding of normal rod opsin appears to be independent of this lectin quality control mechanism as unglycosylated rod opsin targets to the plasma membrane in cell culture and forms photopigment with characteristic spectral properties (Kaushal et a/., 1994). In contrast, inhibiting the N-linked glycosylation of mutant rod opsin, prevents its degradation where it accumulates in the ER, thereby revealing a quality control mechanism that is not lectin dependent. Detection and retention of misfolded rod opsin may be mediated by other ER resident chaperones, such as Bip and Grp94. In fact, misfolded deletion mutants of rod opsin (which were specifically generated to study the role of the intradiscal domain of rod opsin in the folding of the protein and are not RP mutants) co-immunoprecipitated with BiP and Grp94 in cell culture (Anukanth and Khorana, 1994). An additional example of the retention of misfolded polytopic glycoproteins by ER chaperones other than calnexin or calreticulin is that of the low-density lipoprotein LDL receptor (Jorgensen at a!., 2000). The mutant LDL receptor (which causes familial hypercholesterolemia) was retained within the ER by the chaperone BiP and no association of the mutant receptor with CNX or CRT was observed (Jorgensen at a/., 2000). Thus it appears that polypeptides can interact with different

109 chaperone systems within the ER. Molinari and Helenius (2000) have discovered that growing polypeptides within the ER associate with CNX and CRT, without prior interaction with BiP only if glycans are present within 50 residues of the NH2 terminus of the nascent protein. In addition, if calnexin association with nascent polypeptides is prevented, for example with castanospermine, then the BiP chaperone system will associate with the grcwing polypeptide. Rod opsin is glycosylated at asparagine residues 2 and 15, so according to the model proposed by Molinari and Helenius (2000), it is conceivable that CNX/CRT may be the first chaperone system to interact with the nascent rod opsin polypeptide. However, the data in this chapter suggest that other chaperone systems can assist opsin folding because Wh rod opsin targets to the plasma membrane in tlie presence of TM. Also rod opsin reconstituted with 11-c/s-retinal from ceils incubated with TM forms rhodopsin with characteristic spectral properties (Kaushal et ai, 1994) showing that unglycosylated rhodopsin can attain the riative conformation.

It is emerging from several studies on the degradation of misfolded proteins, that mannose trimming in the ER plays an essential role in ERAD. For instance, the ERAD of yeast prepro-a-factor expressed in mammalian cells (Su et ai, 1993), the Hong Kong null form of ai-antitrypsin (Liu et a/., 1997, Liu et ai, 1999), the yeast CRY (Jacob et a /1998), the 5-subunit of the T-cell antigen receptor (Yang et ai, 1998), the cog thyroglobulin mutant and y-carboxylation-deficient protein C (Tokunaga et ai, 2000) could also be hindered by inhibiting the activity of ER mannosidase- I. A model of the mechanisms of protein targeting for degradation is emerging in which mannose trimming of glycoproteins associated with calnexin or calreticulin prolongs the association with these lectin chaperones, which is thought to trigger the proteasomal degradation of these substrates. For example proteasomal degradation of misfolded ai-antitrypsin (AAT) required the physical interaction with calnexin, but also, mannose trimming was responsible for mediating the onset of misfolded AAT degradation (Liuet ai, 1997, Liu et ai, 1999). Also the major histocompatability complex (MHC) class I heavy chain, which is rapidly degraded by the proteasome if not associated with p2 microglobulin, is dependent on calnexin interaction for its proteasomal degradation (Wilson et ai, 2000) and removal of the N-linked glycosylation site in MHC class I heavy chain rendered the protein stable to degradation. The inhibition of mannose I trimming of core glycosylated MHC I, using deoxymannojirimycin also had the same effect (Wilson et ai, 2000. However, there may be alternative mechanisms, as studies on the cog thyroglobin mutant revealed a calnexin independent mechanism for ERAD, as castanospermine, an inhibitor of glucosidases I

110 and II, thereby blocking lectin-like binding to calnexin, had no effect on the degradation of this particular protein. The role of mannose trimming in ERAD of misfolded opsin requires further investigation but my preliminary findings suggest that mannosidase I may play a part. SH-SY5Y cells expressing K296E opsin v/ere incubated for 8 hours in deoxymannojirimycin, then fixed and processed for immunofluorescense microscopy. An increase in the intracellular K296E fluorescent signal was observed following incubation with deoxymanojirimycin, suggosting that the action of mannosidase I may be involved in ERAD of mutant opsin.

The involvement of CNX/CRT in targeting of terminally misfolded opsin for degradation could be assessed by co-irnmuncprecipitation studies which would help to clarify whether CNX/CRT associate with mutant rod opsin and perhaps establish the role of other potential chaperones involved in targeting of the protein for degradation such as the recently discovered EDEM (Hosokawa et al., 2001). Any interaction with CNX/CRT would probably be dependent on an asparagine linked monoglucosylated oligosaccharide motif and TM presumably would prevent rod opsin association with these lectin chaperones, consequently preventing the targeting of misfolded opsin for degradation. If this hypothesis were correct the drug castanospermine, which prevents the formation of monoglucoslyated oligosaccharides, would also hinder rod opsin association with lectin chaperones thus inhibiting the degradation of misfolded rod opsin. Recent studies suggest that the translocon, the Sec61p complex, participate in the retrograde translocation of misfolded proteins from the ER to the cytosol, for degradation by the proteasome (Johnson and Haigh 2000). The exact mechanisms by which retrotranslocation substrates are recognised by and targeted to the translocon remain unclear. It is conceivable that an opsin CNX/CRT interaction or even an interaction with EDEM may be required to donate misfolded opsin to the translocon and degradation machinery, but requires further investigation.

TM prevents the N-linked glycosylation of all glycoproteins within the secretory pathway and is obviously not specific for rod opsin. To study the specific effects of glycosylation on ERAD of misfolded opsin the substitution of both asparagine residues may be required (Kaushal at a/., 1994). As this particular double mutant is misfolded (Kaushal at a/., 1994) it would be of use to specifically study the requirements of N-linked glycosylation in ERAD of misfolded rod opsin without disturbing the glycosylation of other proteins involved in this mechanism. N-linked glycosylation appears not to be essential for the folding of WT rod

111 opsin but by contrast the maturation of mutant Drosophila Rh1 (N20I), which cannot be glycosylated, is impaired and the unglycosylated protein is retained within the ER (O’Tousa, 1992, Webel et a!., 2000). The authors conclude that glycosylation at Asn-20 is required for folding and targeting of Rhi inDrosophila, but changes in the primary amino acid sequence may have an influence on Rh1 folding, resulting in ER retention of the protein, as a result of quality control. This appears to be the case for mammalian rod opsin, as amino acid substitutions in the tripepotide consensus sequence for N-linked glycosylation of bovine rod opsin (specifically N15Q) lead to misfolding of the protein and therefore E R retention (Kaushal et al., 1994).

Finally as with other misfolded glycoproteins N-linked oligosaccharide groups of rod opsin seemi to play an important part in the targeting of terminally misfolded miutant rod opsin for degradation. As mutant rod opsin accumulates at the level of the ER following inhibition of N-linked glycosylation with TM it is possible that the proteolytic processes within the cell required for the degradation of misfolded rod opsin involves the ubiquitin proteasome system (UPS). In the next chapter the role of the proteasome in the degradation of misfolded rod opsin is explored.

112 Chapter 5: Degradation of mutant opsin is dependent on the proteasome

5.1 Introduction

An important function of the ER is to ensure that polypeptides, which are inserted into the ER membrane or imported into the lumen, fold into their native conformation before reaching their final destination. Proteins, which fail to fold or assemble with other subunits, are detected by an ER quality control system composed of folding sensors and molecular chaperones and targeted for degradation by the proteasome located in the cytosol. It is believed that misfolded or unassembled proteins are, retrotranslocated via the SEC61 translocon to the cytosol prior to or in concert with degradation by the proteasome which recognises and degrades ubiquitinated proteins (Plemper and Wolf 1999, Jarosch et a/.,2002).

Heterologous expression of RP mutants of rhodopsin in cell culture has revealed that the majority of these mutations cause misfolding of the rod opsin protein (Sung at a/., 1991b and Sung at a/., 1993, Kaushal and Khorhana 1994). It has been suggested that as a result of misfolding these mutants are expressed at much lower levels compared to those of the WT protein, presumably due to the accelerated degradation of these mutants by the cellular proteolytic machinery. Indeed, pulse chase analysis has revealed that P23H rod opsin had a much shorter half-life than WT rod opsin in cell culture (Sung at a/., 1991b) and the exact pathway of degradation was unknown.

A number of pharmacological agents have been developed that can inhibit the proteolytic activity of the proteasome, and have been used to investigate the degradation of misfolded proteins or unassembled subunits of protein complexes in cell culture models. In this study the peptide aldehyde Cbz-leu-leu-leucinal (MG-132) which acts as a substrate analogue and is a potent transition state inhibitor of the chymotrypsin-like activity of the proteasome (Rock at a/., 1994, Lee and Goldberg, 1996) was used to investigate the degradation pathway of mutant rod opsin in COS-7 and SH-SY5Y cells. In addition, the effects of inhibiting the proteolytic activity of the proteasome on opsin aggresome formation were investigated.

113 5.2 Results

5.2.1 MG-132 inhibits the degradation of mutant rod opsin in COS-7 cells

The low steady state level of mutant rod opsin in cultured cells compared to the WT protein (Sung et a,l 1991b, Kaushal and Khorana 1994) may be due to rapid degradation. 1 hereioie, to increase our understanding of the degradation pathway of mutant lod cpsin, an in vitro system using COS-7 cells and SH-SY5Y cells was employed COS-7 cells were transfected with pMT3-WT, pM13-K296E and pl\/lT3-P23H and 24 hours later the cells were incubated In complete medium (without antibiotics) containing 6 pM of MG-132 for i6 hours. After veli lysis in FT3S + T:b DM, the lysates were centrifuged to separate the DM soluble and DM insoluble fractions and were resolved on a 10% SDS-PAGE gel, followed by immunoblotting with the mAb 1D4 Figure 6.1a illustrates the effects of MG- 132 on the levels of WT rod opsin protein after 16 hours. Steady state levels of WT rod opsin were unaffected by MG-132. However, MG-132 treatment of cells expressing K296E and P23H rod opsin, resulted in a noticeable increase in the levels of all the mutant opsin species in both the DM soluble and DM insoluble fractions (Figure 5.1b and 5.1c respectively). Mutant opsin species in the DM soluble fraction migrating at %40 kDa increased after incubation with MG-132 and are most likely to be ER resident immature glycosylated species (see below Fig. 5.2).

In the DM insoluble fraction the non-glycosylated 33kDa species (*) of both mutants increased after MG-132 treatment and may be a degradation intermediate as glycoproteins are deglycosylated prior to proteasomal degradation. It is also clear that following MG-132 treatment the amount of high molecular weight oligomers of mutant ospin, migrating from the rod opsin dimer (**) (Figure 5.1b and 5.1c) to the top of the resolving gel, increased.

The DM soluble fraction from P23H and K296E expressing cells was digested with Endo H in order to determine if these species (migrating at »40 kDa) are ER resident immature glycosylated species. Figure 5.2 shows a western blot of Endo H digested DM soluble fractions from WT, K296E and P23H expressing cells incubated with MG-132. A decrease in the intensity of this band can be seen after Endo H treatment, but there is not a concomitant increase in the «33 kDa species. In fact all forms of the protein migrating within this range (40-33 kDa) have decreased in intensity and rod opsin could now be

114 detected in the stacking gel (a phenomenon described in chapter 4). Thus it appears that following MG-132 treatment a proportion of mutant opsin receptors accumulate in the ER as shown by Endo H sensitivity.

115 A. MG-132

2 1 8 -

43.8 —

B. C. MG 132 “ - MG-.3; 4 4- S P S S P S

Figure 5.1 The effects of MG-132 on the steady state levels of WT, K296E and P23H rod opsin in COS-7 cells.

Equal volumes of DM soluble (S) and DM insoluble (P) fractions corresponding to 10 pg of whole cell lysates were resolved on a 10% SDS-PAGE gel followed by immunoblotting with mAb 1D4. A) Immunoblot of WT rod opsin +MG132; as indicated B) Immunoblot of K296E and 0) P23H rod opsin +MG-132; as indicated. The arrows points to the «40 kDa glycoform. In the DM insoluble (P) fraction the non glycosylated «33 kDa species is indicated (arrow head) and (*) indicates rod opsin dimer. Blot exposures were adjusted to give equivalent band intensity between WT and mutants. Molecular weight markers are shown on the left in kDa.

116 '/VT K296E P23H Endo H

113 —

Figure 5.2 Endo H digestion of mutant rod opsin glycoforms that accumulated following treatment with MG-132

10 pig of the DM soluble (S) fraction from lysates of COS-7 cells expressing WT, K296E and P23H rhodopsin genes in the presence of MG-123, were incubated with Endo H and resolved on a 10% SDS-PAGE gel in parallel with 10 pg of undigested DM soluble fraction, followed by immunoblotting with mAb 1D4. Arrow head: WT and mutant rod opsin glycoforms migrating at «40 kDa. Arrow: WT and mutant rod opsins in the stacking gel following Endo H digestion. Molecular weight markers are shown on the left (kDa).

117 5.2.2 The influence of MG-132 on mutant rod opsin aggresome numbers inhibition of proteasomal activity in cells expressing AF508CFTR-GFP lead to a large increase in aggresome numbers (Johnston et al., 1998). In order to determine if MG-132 led to and increase in the incidence of opsin aggresomes, COS-7 cells expressing GFP- tagged WT and mutant rhodopsin genes were examined using immunofluorescence microscopy after incubation with MG-132. Aggresomes were still visible following an 8 hour and 16 hour incubation with 5 p.M MG-132. Aggresome numbers were determined following an 8-hour incubation with MG-132, 24 hours after transfection. After a 16 hour incubation with MG-132 it was difficult to assess aggresome numbers due to MG-132 related cell death. An independent observer (Dr. M. Cheetham) unaware of the transfection status or treatment status of the cells counted aggresomes, and the results are represented graphically in figure 5.3. These experiments revealed that incubation of cells with MG-132, induced a small increase in GFP-tagged P23H opsin aggresome numbers and negligible change in GFP-tagged WT rod opsin aggresome numbers. In contrast, AF508CFTR-GFP aggresome numbers increased dramatically following incubation with MG-132.

118 I 100 1

O)i O) 80 < 60 (A o ■ D 40 B 0 1 c 20 (0 o 0 □ I WT-GFP WT-GFP P23H-GFP P23H-GFP AF508CFTR AF508CFTR GFP GFP

MG-132 — + —

Figure 5.3 The influence of MG-132 on rod opsin aggresome numbers in COS-7 cells

COS-7 cells were transfected with the plasmids pEGFP-WT, pEGFP-P23H and pEGFP-AF508CFTR and 24 hours later the cells were incubated with 5 piM MG-132 for 8 hours. Cells expressing mutant CFTR showed a very noticeable increase in aggresome numbers after an 8 hour incubation with MG-132, approximately 100% of transfected cells now formed AF508CFTR-GFP aggresomes compared to 32%, prior to MG-132 treatment. Cells expressing P23H rhodopsin-GFP showed a small increase in aggresome numbers following incubation with MG-132 and WT-GFP showed a negligible increase.

119 5.2.3 Detection of WT and mutant rod opsins in COS-7 cells following MG-132 treatment using immunofiuorescence microscopy.

Even though mutant opsin aggresome numbers did not dramatically increase after treatment with MG-132, unlike AF508CFTR-GFP, mutant opsin did accumulate in a peri nuclear region suggestive of the ER. To determine whether K296E and P23H rod opsin accumulated in the ER following incubation with MG-132, COS-7 cells were dual labeled for opsin and an ER resident chaperone, calnexin as shown in figure 5.5 and figure 5.6. Following an 8-hour incubation with MG-132 the trafficking of WT rod opsin appeared to be unaffected and was present at the plasma membrane (Fig. 5.4). Mutant ER rod opsin fluorescence was weak in untreated cells but following incubation with MG-132 a stronger mutant opsin immunofluorescence signal was observed in the ER as indicated by co localisation with calnexin (Fig. 5.5 and 5.6). In some cells MG-132 also induced calnexin to relocate to an area within the cell from which mutant rod opsin was excluded (Figs 5.5 and 5.6). In contrast to mutant opsin the majority of AF508CFTR-GFP in the presence of MG-132 was found in aggresomes (Fig. 5.5). In some cells calnexin was observed at the surface of AF508CFTR-GFP aggresomes (Fig. 5.5).

The effects of MG-132 on the processing of GFP-tagged WT and mutant rod opsins were also investigated using immunofluorescence microscopy. These experiments were performed to asses whether the GFP-tagged rod opsin behaved in the same way as untagged rod opsin following inhibition of proteasomal activity. MG-132 caused an increase in GFP-tagged mutant rod opsin accumulation in the ER as shown by co localisation with the ER luminal chaperone BiP (Fig 5.7). MG-132 did not appear to influence the trafficking of the GFP-tagged WT rod opsin, as plasma membrane and Golgi staining could still be observed (Fig 5.7).

120 WT rod opsin Cahexin

Figure 5.4 Effects of MG-132 on the targeting of WT rod opsin to the plasma membrane.

COS-7 cells were transfected with the plasmid pMT3-V\nr, and 24 hours later the cells were treated with 5 p,M MG-132 for 8 hours. Rod opsin was labelled with 1D4-FITC and the ER was labelled with an anti-body to calnexin (red). WT rod opsin (green) translocated through the Golgi to the cell surface (arrow head) of untreated cells. MG-132 did not alter WT rod opsin (panels labelled +MG-132) translocation to the cell surface (arrow heads). Nuclei were counter stained with DAPI. Images are z-sections and scale bars are 10 ^im.

121 Figure 5.5 MG-132 induces P23H rod opsin to accumulate in the ER of COS-7 ceils

COS-7 cells were transfected with the plasmid pMT3-P23H and 24 hours later were incubated with 5 |iM MG-132 for 8 hours. P23H rod opsin was labelled with 1D4-FITC and the ER was labelled with an anti body to calnexin (red). A) Untreated cells: P23H rod opsin (green) accumulated in the ER showing faint immunofluorescence and formed a characteristic aggresome (arrow). B) Two different fields: MG-132 lead to brighter ER staining of P23H rod opsin which co-localised with calnexin. Also, there was an area within the cell which stained for calnexin but from which mutant rod opsin was excluded (arrow). 0) AF508CFTR-GFP expressed in COS-7 cells + MG-132 (5 ^iM). The majority of AF508CFTR-GFP transfected cells contained aggresomes after incubation with MG-132. The nuclei were counter stained with DAPI. Images are single z-sections and scale bars are 10 ^im. y

P23H rod opsin Calnexin Merae

■ H H

122 K296E rod opsin Calnexin

+MG-132 +MG-132 +MG-132

K296E rod opsin Calnexin

Figure 5.6 MG-132 induces the accumulation of K296E rod opsin in the ER of COS-7 cells

COS-7 cells were transfected with the plasmid pMT3-K296E and 24 hours later the cells were incubated with 5 pM MG-132 for 8 hours. K296E rod opsin was labelled with 1D4-FITC and the ER was detected using anti bodies to calnexin (red). A) Untreated cells: K296E rod opsin (green) accumulated in the ER showing faint immunofluorescence and formed characteristic aggresomes (arrow). B) MG-132 led to brighter ER staining of K296E rod opsin which co-localised with calnexin. Also, there was an area within the cell which stained for calnexin but from which K296E rod opsin was excluded (arrow). The nuclei were counter stained with DAPI. Images are single Z-sections and scale bars are 10 pM.

123 Figure 5.7 MG-132 induces GFP-tagged K296E and P23H mutant rod opsin to accumulate in the ER of COS-7 cells.

0 0 8-7 cells were transfected with the vectors pEGFP-WT, pEGFP-K296E, and pEGFP-P23H and 24 hours later the cells were incubated with 5 p,M MG-132 for 8 hours. In these experiments the ER was labelled with an anti-body to BiP (red). A) MG-132 did not disrupt the translocation of WT rod opsin-GFP (WT- GFP) to the cell surface. B) and 0) K296E rod opsin-GFP (K296E-GFP) and P23H rod opsin-GFP (P23H-GFP) respectively, show faint ER staining and bright aggresomes. After incubation with MG-132 (images labelled +MG-132) mutant rod opsin-GFP now shows brighter ER staining. Nuclei were counter stained with DAPI. Images are z-sections and scale bars are 10 jim. B.

K2S6F-GFP

MG-132 t

K296E-GFP

P23H-GFP

+MG-132 MG-132 MG-132

P23H-GFP

124 5.2.4 The influence of MG-132 on mutant opsin degradation in SH-SY5Y cells

As in the previous chapters WT and mutant rod opsin processing was also investigated in the neuronal cell line SH-SY5Y. In chapter 4 I had shown that TM prevented the degradation of mutant opsin which appeared to accumulate in the ER of SH-SY5Y cells and COS-7 cells. Inhibiting the proteolytic activity of the proteasome may have the same effect as tunicamycin on mutant rod opsin causing accumulation in the ER. Therefore, SH-SY5Y cells were transfected with the plasmids pMT-3WT, pMT3-K296E and pMT3- P23H and 24 hours later were treated with a range of concentrations of MG-132 (10-50 ^M) for 8 hours and then fixed in -20°C 100% methanol and processed for immunofluorescence microscopy. In untreated cells faint plasma membrane staining of K296E and P23H rod opsin could be seen. Also K296E and P23H rod opsin formed faint spots within the cell, which did not co-localise with Grp94. Incubation of the cells with MG- 132 (20 pM), however, resulted in a very different staining pattern of mutant rod opsin to that of untreated cells (Fig.5.8). Bright intracellular staining of mutant opsin was now seen which showed partial co-localisation with Grp94 (Fig.5.8).

125 A.

Merge

+MG-132 +MG-132 +MG-132

P23H rod opsin Merge

Figure 5.8 In SH-SY5Y cells mutant rod opsin accumulates in the ER following incubation with MG-132.

SH-SY5Y cells were transfected with the plasmids pMT3-K296E and pMT3-P23H and 24 hours later cells were incubated in 20 pM MG-132 for 8 hours. Rod opsin (red) was labelled with mAb 1D4 and anti mouse Cy3. The ER was labelled with antibodies to the ER resident chaperone Grp94. A) K296E (K296E rod opsin) shows faint cell surface staining. Also K296E rod opsin and P23H rod opsin (P23H rod opsin) intracellular staining can be seen that dose not co-localise with Grp94. B) SH-SY5Y cells incubated with MG-132 for 8 hours. K296E and P23H rod opsin accumulate in the ER and give a brighter fluorescence signal, showing partial co-localisation with Grp94. Images are z-sections and scale bars represent 10 pm.

126 5.3 Discussion

The work described in this chapter has revealed that the proteolytic activity of the proteasome is required for the degradation of mutant opsin. Furthermore, MG-132 did not appear to induce the accumulation of mutant rod opsin within aggresomes but in the ER.

The inhibition of the proteolytic activity of proteasomes caused an increase in the DM insoluble fraction of high molecular weight mutant opsin species. These may represent polyubiquitinated rod opsin species that have accumulated, a phenomenon which has been reported for a number of ubiquitinated proteins (Wardet al., 1995, Halaban et al., 1997, Yu and Kopito, 1999, Xiong et al., 1999). A number of studies have shown that lumenal and transmembrane proteins destined for ERAD accumulate as polyubiquitinated species following inhibition of the proteolytic activity of the proteasome in yeast and mammalian cells. For example, polytopic transmembrane proteins such as AF508CFTR (Ward et al., 1995) and the human 5 opioid receptor (Petaja-Repo et al., 2001) accumulate in the cytoplasm as polyubiquitinated species following acute proteasome inhibition. In addition, proteasome inhibition results in the cytosolic accumulation of ubiquitinated type I membrane proteins such as unassembled a subunits of the T cell receptor (TCRa) (Yang et al., 1998) and MHC class I heavy chains targeted for degradation in human cytomegalovirus-infected cells (Wiertz et al., 1996). However, as rod opsin is known to form high molecular weight oligomers on denaturing SDS-PAGE gels, which appears to be an intrinsic property of the protein, it remains to be determined if these high molecular weight rod opsin species are indeed ubiquitinated. MG-132 induces an accumulation of mutant misfolded opsin, which is probably unstable in DM and may form high molecular weight oligomers which partition to the DM insoluble fraction, but this does not rule out the presence of ubiquitinated high molecular weight species. To determine if MG-132 induced the accumulation of polyubiquitinated mutant opsin, I immunoprecipitated GFP-tagged WT and mutant opsin from cells treated with MG-132 using anti-GFP, followed by immunoblotting with anti-c-myc (9E10). These data were inconclusive and required further changes to the experiment (time permitting) to optimise conditions for solubility of insoluble opsin present in cell lysate pellets. However, it has recently been shown that ubiquitinated mutant opsin does indeed accumulate following treatment of HEK 293 cells with MG-132 (llling et al., 2002).

127 The detection of AF508CFTR-GFP fluorescence in COS-7 cells is very low and it initially appeared that only a small percentage of cells were transfected. However, this was misleading because following incubation with MG-132 the percentage of cells that had actually been transfected was much greater as revealed by the presence of AF508CFTR- GFP aggresomes. This is a consequence of AF508CFTR-GFP being rapidly degraded in COS-7 cells. MG-132 did not lead to a substantial increase in rod opsin aggresome numbers, which would be expected if retrotranslocation and degradation were uncoupled, but lead to a noticeable increase in opsin immunofluorescence in the ER. This seems to suggest that mutant rod opsin targeted for proteasomal degradation remains in the ER membrane or remains associated with the ER when the proteolytic activity of the proteasome is compromised. Interestingly, COS-7 cells expressing P23H rod opsin and c- myc tagged ubiquitin in MG-132 treated cells co-localised in a perinuclear region of the cell (data not shown) which may suggest that the cytoplasmic domains of mutant opsin are ubiquitinated at the ER prior to retrotranslocation. Therefore mutant rod opsin, targeted for degradation, may require a functional proteasome for retrotranslocation prior to degradation, or degradation may occur in concert with dislocation as seen for some transmembrane proteins (Mayer et a/., 1998, Xiong at a/., 1999, Plemper et al., 1998). Some reports have suggested that the proteasome itself is involved in the dislocation of transmembrane proteins (Mayer at ai, 1998). Also novel ER associated proteins have been recently described which are involved in the dislocation of proteins from the ER (Jarosch at a/., 2002).

Immunoblotting of cell lysates from cells treated with MG-132 revealed that the steady state levels of all mutant rod opsin species increased. A mutant rod opsin species of 33 kDa, accumulated following the addition of MG-132, and partitioned to the detergent insoluble fraction. This 33 kDa species may represent a rod opsin intermediate in the proteasomal degradation pathway, as it is known that proteins destined for proteasomal degradation are deglycosylated prior to proteolysis (Wiertz et al., 1996, Kopito 1997). In fact, N-glycanases which could be involved in oligosaccharide removal from targeted proteins have been found in the cytosol (Suzuki at si, 1998) and in the ER (Suzuki et al., 1997, Weng and Spiro, 1997) but it remains unclear as to where this deglycosylated rod opsin species accumulated following incubation with MG 132. This species may remain in the ER membrane however, as increased ER rod opsin immunofluorescence was observed and numbers of mutant rod opsin aggresomes did not increase greatly. It remains to be discovered if rod opsin is deglycosylated prior to retrotranslocation but this

128 phenomenon has been described for misfolded human 5 opioid receptors (a G-protein coupled receptor) targeted for proteasomal degradation (Petâjâ-Repo et al., 2001).

The membrane bound ER chaperone, calnexin and its lumenal homologue calreticulin relocates to a juxtanuclear position within the ER upon incubation of cells with proteasome inhibitors along with protein substrates targeted for proteasomal degradation (Kamhi- Nesher at al. 2001). This region is dependent on microtubules as it can be disrupted with nocodazole and is thought to be the site of retrotranslocation of proteins targeted for proteasomal degradation. In the experiments reported in this chapter, calnexin did appear to relocate to a juxtanuclear position following incubation with MG-132, but mutant rod opsin was excluded from this region. Alternatively, these calnexin positive structures may represent aggresomes formed from other cellular proteins following incubation with MG- 132 unlike mutant opsin, which is retained in the ER and excluded from these structures.

The proteasomal degradation of misfolded rod opsin targeted for degradation, may involve retrotranslocation through the SEC61 translocon, but the complex mechanisms of retrotranslocation of this polytopic membrane protein prior to proteasomal degradation of aberrantly folded rod opsin remain to be unravelled. For instance is ubiquitination of rod opsin a prerequisite for dislocation, as some studies have revealed that ubiquitination of transmembrane proteins targeted for proteasomal degradation require prior ubiquitination for retrotranslocation (Yu and Kopito, 1999, Shamu et al.,2001. Ye et al., 2001).

Misfolded mutant rod opsin is targeted for proteasomal degradation as a consequence of ER quality control, which recognises and retains misfolded proteins within the ER, preventing their translocation through the secretory pathway. Preventing mutant opsin forming off-pathway conformera would inevitably lead to the abrogation of ER retention of mutant opsin, and increase in mutant opsin at the plasma membrane. In the next chapter I explore the effects of opsin ligand and low temperature on the folding of mutant opsin in COS-7 cells.

129 Chapter 6: The influence of ligand and temperature on the processing and trafficking of mutant rod opsin in vitro

6.1 Introduction

The acquisition of the native conformation and export from the ER represent limiting factors for the expression of polytopic membrane proteins (Kopito, 2000). RP mutants of rhodopsin generate misfolded receptors that are retained in the ER and degraded by the proteasome in cultured cells (Sung et al., 1991b, Saliba at a!., 2002 and chapter 5) and also have a propensity to form aggresomes (Saliba at a/., 2002, chapter 3). The processing and degradation of RP rhodopsin mutants in vivo could be envisaged to present a heavy burden to cellular proteolytic processes and may exceed the rod cell’s capacity to process such an abundance of misfolded protein. By increasing the proportion of misfolded rod opsin proteins that attain the native conformational state and proceed along the secretory pathway, it may be possible to reduce the burden on the proteolytic processes of rod cells and prevent accumulation of misfolded protein. A number of reports now suggest that protein folding defects can be corrected with a number of small molecules (such as ligands and chemical chaperones) and thus restore impaired protein trafficking (Sato at ai., 1995, Welch and Brown, 1996, Galbiati at ai., 2000, Kagan at a/.,2000, Petaja-Repo et al., 2002) which has become an important experimental and therapeutic goal.

In Drosophiia the Rhi ligand, 11-c/s-3-hydroxy retinal, is required for the efficient processing and maturation of the receptor (Ozaki at ai., 1993). In the case of mammalian rhodopsin, vitamin A deprivation lead to a decrease in rhodopsin in the rod outer segment (Dowling and Wald ,1958, Engbretson and Witkovsky, 1978) and furthermore a lack of vitamin A produced a decrease in rod sensitivity and eventually lead to the loss of both rods and cones (Dowling and Gibbons, 1960, Hayes, 1974, Kemp et al., 1989). Vitamin A supplements included in the diets of T17M rhodopsin transgenic mice preserved the ERG amplitude and delayed retinal degeneration (Liat ai., 1998). In the same study, the authors reported that 11-c/s-retinal or 9-c/s-retinal stabilised T17M mutant rod opsin and increased the proportion of mutant rod opsin reaching the plasma membrane in cell culture (Li at ai., 1998); therefore, ligand appears to facilitate opsin folding.

130 Low temperatures can also facilitate protein folding. Expression of genes with disease causing mutations, which give rise to protein products that fail to achieve their native conformational state, in cultured cells at 30 ®C rescued the folding abnormalities. For example, the folding and trafficking defects of AF508-CFTR (Denning et al., 1992), Limb- girdle Muscular Dystrophy mutants of caveolin-3 (Galbiati et al., 2000) and mutant HERO K" channel (responsible for hereditary long QT syndrome, Kagan at a/., 2000) in cultured cells were overcome by an incubation temperature of 30°C. At this temperature, a proportion of these mutant proteins, which normally accumulate in the ER, escaped ER quality control and translocated to the plasma membrane.

The instability of a described set of mutant proteins can be overcome by small molecules (e.g. ligands) and low temperature. Therefore, in this chapter I have investigated the effects of ligand and the lowering of cell culture growing temperature on P23H rod opsin folding by examining the trafficking of this mutant protein to the plasma membrane in COS-7 cells.

6.2 Results

6.2.1 The Influence of 9-c/s-retinal on P23H rod opsin processing and trafficking in COS-7 cells

Initially, 11-c/s-retinal (a generous gift from Dr. Rosalie Crouch) was used in studies examining the role of the molecular chaperone Hsp90 in rod opsin trafficking in SH-SY5Y cells, stably expressing the WT rhodopsin gene. The drug geldanamycin, which inhibits the activity of Hsp90, caused WT rod opsin to accumulate in the Golgi. Therefore I decided to investigate the effects of 11-c/s-retinal on the geldanamycin induced accumulation of WT rod opsin in the Golgi. Unfortunately, a new batch of geldanamycin failed to produce the same effects on opsin trafficking seen with the previous batch of drug. In addition geldanamycin from different companies failed to reproduce the effects on opsin trafficking, which I had initially observed, as did other Hsp90 inhibitors. Therefore, these experiments had to be abandoned. By the time the experiments using COS-7 cells and P23H rod opsin were performed, the stock of 11-c/s-retinal from Dr. Rosalie Crouch which was dissolved in ethanol had been stored at -70®C for more than a year; so 9-c/s-retinal, commercially available from Sigma was used to investigate the role of ligand on P23H rod opsin processing in cell culture.

131 COS-7 cells were transfected with the plasmids, pMT3-WT, pMT3-K296E and pMT3-P23H and 24 hours later the cells were incubated in the presence 9-c/s-retinal (made up from a 20 mM stock in 100% ethanol) for 16 hours. As ethanol can induce upregulation of the Hsc70 gene and increase levels of the Hsc70 protein at ethanol concentrations of 50-100 mM (Miles et al., 1991), which in itself may influence the processing of misfolded rod opsin, control cells were incubated in 0.1% ethanol (17 mM). Retinal was added in the dark under the light from a Kodak type 2-safelight filter. Following a 16-hour incubation with 9-cis-retinal (20 ^M, as suggested in previous reports Oprian et a/., 1987) the cells were lysed in PBS+1%DM and cell lysates were partitioned into DM soluble and DM insoluble fractions by centrifugation. Volumes of DM soluble and DM insoluble fractions equivalent to 10 )ig of total cell lysate were resolved on a 10% SDS-PAGE gel and immunoblotted with mAb 1D4.

The steady state levels of WT rod opsin were not influenced by the presence of ligand in the culture medium as shown in Figure 6.1a. In contrast 9-c/s-retinal induced a noticeable increase in the steady state levels of all P23H rod opsin species (Fig. 6.1b) where as ethanol (0.1%) had no effect (Fig. 6.1c). In particular, a species migrating at «40 kDa increased following addition of ligand as well as an increase in the intensity of a band representing a species that migrated faster than the 33 kDa non-glycosylated species. In order to verify whether P23H species migrating at «40 kDa were Endo H resistant, 10 ^ig of the DM soluble fraction from cells that expressed the P23H rhodopsin gene in the presence of 9-c/s-retinal was digested with Endo H. For comparison, the DM soluble fraction (10 p.g) from untreated control cells expressing P23H rhodopsin was digested with Endo H. Figure 6.2 shows an Endo H resistant form, which remained following digestion and was of greater intensity than the equivalent species from untreated cells (compare lanes 1-3 and 4-6 of Fig. 6.2). Thus, 9-c/s-retinal, appears to increase the proportion of Endo H resistant (i.e. post ER) P23H rod opsin, suggesting that an increased proportion of P23H rod opsin escaped quality control retention and continued through the secretory pathway.

To rule out the possibility that the effects of 9-c/s-retinal on P23H rod opsin processing and degradation are indirect, I examined the effects of ligand on the processing and degradation of K296E rod opsin, which cannot bind retinal. Ligand did not alter the steady state levels of K296E rod opsin (Figure 6.Id).

132 Even though 9-c/s-retinal facilitates the maturation of P23H rod opsin to a post-ER compartment, these data do not establish that ‘rescued’ rod opsin actually translocates to the plasma membrane. Therefore, immunofluorescence microscopy was used to test for plasma membrane localisation of P23H rod opsin in the presence of ligand. COS-7 cells expressing the P23H rhodopsin gene were incubated for 12 hours in the presence of 20 )iM 9-c/s-retinal, twenty-four hours after transfection. P23H rod opsin in untreated 008-7 cells accumulated in the ER, formed aggresomes and showed very faint plasma membrane staining (Fig. 6.3). In the presence of 9-c/s-retinal P23H rod opsin showed increased immunofluorescence staining at the cell surface (Fig. 6.3.) even in cells with aggresomes. Given that 9-c/s-retinal appears to facilitate protein folding and therefore may have some bearing on P23H rod opsin aggregation, numbers of aggresomes were counted following addition of ligand (for 12 hours) twenty four hours after transfection and revealed that aggresome numbers remained unchanged.

6.2.2 The influence of a low temperature incubation on the processing of P23H rod opsin in COS-7 cells

Low temperature incubation of cultured cells facilitates the folding of some polytopic membrane mutant proteins (Sato et a/., 1996, Galbiati at al., 2000, Kagan at al., 2000); thus, cells expressing the P23H rhodopsin gene were grown at 30°C to investigate the effects of low temperature incubation on P23H rod opsin processing and maturation. An increase in the proportion of P23H rod opsin receptors reaching the native conformational state would presumably be reflected in an increase in the proportion of the mutant protein escaping ER quality control retention and consequently translocating to the plasma membrane. COS-7 cells were transfected with the plasmids pMT3-WT and pMT3-P23H and after 24 hours at 37®C the cells were incubated for an additional 16 hours at 30^0. Control cells were cultured for the same amount of time at 37°C. Cells were lysed in PBS+ 1% DM, after a 16 hour incubation at 30®C and partitioned into DM soluble and DM insoluble fractions. Volumes of soluble and insoluble fractions equivalent to 10 ^g of whole cell lysate were loaded onto a 10% SDS-PAGE gel. WT and P23H rod opsin levels were assessed by immunobloting with 1D4. Figure 6.4a shows a slight increase in WT rod opsin steady state levels observed after a 30°C incubation. In contrast, the steady state levels of P23H rod opsin increased noticeably following incubation at 30®C, which was reflected by an increase in the amounts of P23H rod opsin in the DM soluble and DM

133 insoluble fraction (Fig. 6.4b). In particular, the intensity of a doublet migrating at «40 kDa in the soluble fraction increased and from previous experiments, the slower migrating species most likely represents an Endo H resistant glycoform.

134 A. B. WT rod opsin P23H rod opsin

9-c/S'Retlnal — — + + 9-c/s-Retlnal — — + + S P S P S P S P

112 — 112 — 81 81 — mm## $ 49.9 — 49 9 — 1 ‘ 36.2 — 36.2 — 29 9 — 29.9 —

^ P23H rod opsin ^ K296E rod opsin

+EtOH —- — + + 9-c/s-Retinal — — + + S P S P S P S P

Figure 6.1 The effect of 9-c/s-retinal on the processing of P23H rod opsin in COS-7 cells

COS-7 cells were transfected with the plasmids pMT3-WT, pMT3-K296E and pMT3-P23H. Twenty four hours later the transfected cells were incubated with 20 pM 9-c/s-retinal for 16 hours. Equal amounts of DM soluble (8) and DM insoluble (P) fractions (corresponding to 10 pig of total cell lysate) were resolved on a 10% SDS-PAGE gel and immunoblotted with mAb 1D4. A) Soluble (S) and insoluble (P) fraction from cells expressing the WT rhodopsin gene in the presence of 9-c/s-retinal as indicated. B) Soluble (S) and insoluble (P) fraction from cells expressing the P23H rhodopsin gene in the presence of 9-c/s-retinal as indicated. The P23H rod opsin ^ 0 kDa species (arrow head), the 33 kDa species (arrow) and a species migrating faster than the 33kDa species (*) are indicated. C) Soluble (S) and insoluble (P) fraction from cells expressing the P23H rhodopsin gene in the presence of 0.1% ethanol for 16 hours, as indicated. D) Soluble (S) and insoluble (P) fraction from cells expressing the K296E rhodopsin gene in the presence of 9-c/s- retinal as indicated. Molecular weight markers in kDa are shown on the left.

135 P23H rod opsin WT rod opsin

9-c/s-Retinal — — — + + +

PNGase F — — + — — + — — +

Endo H — + — — + —

113 — 92 —

5 2 .3 —

3 5 ------

2 8 . 9 ------

1 2345 6789

Figure 6.2 Endo H and PNGase F digestion of P23H rod opsin from cells incubated with 9-c/s-retinal

The DM soluble fraction (10 pg) of cell lysates from cells that expressed the WT and P23H rhodopsin genes in the presence or absence of 9-c/s-retinal were incubated with PNGase F and Endo H and resolved on a 10% SDS-PAGE gel and immunoblotted with 1D4. Lanes 1-3 Endo H and PNGase F digestion of P23H from untreated cells as indicated. Lanes 4-6 Endo H and PNGase F digestion of P23H from cells treated with 9-c/s-retinal as indicated. Lanes 7-9 Endo-H and PNGase F digestion of WT rod opsin from untreated cells as indicated; arrow head points to the «40 kDa glycoform and arrow to the 33 kDa deglycosylated rod opsin species. Molecular weight markers in kDa are shown on the left.

136 A.

P23H rod opsin

B. + 9-c/s-retinal

P23H rod opsin

+ 9-c/s-retinal

P23H rod opsin

Figure 6.3 9-c/s-retinal promotes the translocation of P23H rod opsin to the cell surface in COS-7 cells.

COS-7 cells were transfected with the plasmid pMT3-P23H and 24 hours later the cells were incubated in the presence of 20 pM 9-c/s- retinal for 12 hours in the dark. Rod opsin was detected with 104- FITC. A) Untreated cells: P23H rod opsin formed aggresomes and faint ER staining was observed. B) In contrast P23H rod opsin in the presence of 9-c/s-retinal (panels labelled + 9-c/s-retinal two different fields of view) showed strong immunofluorescence staining at the plasma membrane (arrow heads). However aggresomes were still present for this time period of incubation with ligand (lower panel, arrow). Images are z-sections and scale bars represent 10 ^m

137 NM m

B. 37X + + - - 30°C - - + + S P S P

52.3 —

Figure 6.4 Low temperature incubation of COS-7 cells expressing the WT and P23H rhodopsin genes

COS-7 cells were transfected with the plasmids pMT3-WT and pMT3- P23H. Twenty four hours later the cells were incubated at 30°C for 16 hours. A volume of DM soluble (8) and DM insoluble (P) fraction equivalent to 20 |ig of total cell lysate was loaded onto a 10% SDS- PAGE gel and immunblotted with 1D4. A) Soluble (S) and insoluble (P) fraction from cells expressing the WT rhodopsin gene at 37°C and 30°C as indicated. B) Soluble (8) and insoluble (P) fraction from cells expressing the P23H rhodopsin gene at 37°C and 30°C as indicated. The P23H rod opsin species migrating at «40 kDa (arrow) and the 33 kDa species (arrow head) are indicated. Molecular weight markers in kDa are shown on the left.

138 6.3 Discussion

In this chapter strategies for increasing the proportion of P23H rod opsin that targets to the plasma membrane (suggestive of the correct folding of the protein) have been investigated. This work revealed that the opsin ligand, 9-c/s-retinal increased the proportion of P23H rod opsin that targeted to the plasma membrane and increased the steady state levels of the mutant receptor. In addition, conditions that facilitate protein folding such as low temperatures (i.e. 30°C) appear to increase the proportion of P23H rod opsin that escapes ER quality control retention, and forms the mature glycosylated species.

The precise molecular mechanism by which retinal could enhance stabilisation of newly synthesised rod opsin is unclear. As has been suggested for chemical chaperones the most likely mechanism by which 11-c/s-retinal and 9-c/s-retinal could mediate their action is stabilisation of mutant rod opsin in the native or intermediate state of its folding pathway. Ligand may maintain immature P23H rod opsin in a maturation competent-state, either by minimising reactions that are on the off pathway or by enhancing reactions that are on the folding pathway. Retinal most likely enhances conformational stabilisation of mutant rod opsin by promoting the stable packing of its transmembrane (TM) a-helices helices. Two critical amino acids required for ligand binding in rhodopsin are lysine-296 (Lys-296) in transmembrane a-helix VII (TM7); the site of attachment of retinal to the protein through a protonated Schiff base linkage and glutamic acid-113 (Glu-113) in transmembrane a-helix III (TM3); the counterion to the protonated Schiff base (Strader et al., 1994). 11-c/s-retinal constrains rhodopsin in an inactive conformation by a salt bridge between TM3 Glu-113 and TM7 Lys-296 (Robinson et a/., 1992). Thus, it is reasonable to speculate that binding of ligand to the newly synthesised receptor might stabilise the labile protein by inducing conformational constraints within the a-helical bundle. Consequently, binding of ligand may alter the equilibrium in favour of the stable correctly folded protein, ultimately leading to an increase in the fraction of mutant rhodopsin escaping ER quality control retention and consequently to an increase in receptor numbers at the cell surface.

The use of ligands to stabilise other GPCR protein conformations has just recently been described. A number of membrane permeable ô opioid receptor ligands facilitate maturation and ER export of wild type and mutant 8 opioid receptors retained in the ER (Petâjâ-Repo et al., 2002); the authors have described these ligands as pharmacological

139 chaperones. A different example of receptor protein stabilisation by ligands is that of Transthyretin (TTR) and ligand thyroxine. Transthyretin, is composed of four identical p- sheet-rich subunits and binds and transports thyroxine and retinol binding protein (Blake et al., 1978). In the absence of thyroxine TTR, in vitro, undergoes dissociation (linked to a conformational change) to form the monomeric intermediate, which self-assembles into amyloid (Miroy at a/., 1996). Thyroxine, therefore stabilises the tetramer of TTR against dissociation and the subsequent conformational changes required for amyloid fibril formation. Dominant mutations in the TTR gene lead to Familial Amyloid Polyneuropathy (FAR) where extracellular amyloid deposits of TTR are found mostly in peripheral neurones. Mutations in TTR destabilise the tetrameric form of TTR and favour the monomeric amyloidogenic form (Colon at a/., 1996). Therefore it is conceivable that thyroxine may stabilise the tetrameric form and prevent amyloidosis, suggestive of a potential therapy for this disease (Miroyat a/., 1996).

Rod opsin ligands appear to facilitate the folding of the T17M (Li at a/., 1998) and P23H (this chapter) mutant opsin in cell culture models. 9-c/s-retinal increased the proportion of Endo H resistant P23H rod opsin and appeared to increase cell surface immunofluorescence staining of the mutant receptor. The effects of ligand on P23H opsin folding appear to be direct, as the processing of K296E rod opsin appeared unchanged following incubation with 9-c/s-retinal. The proportion of misfolded rod opsin that escapes ER quality control retention may be dependent on the concentration of ligand used. This increase could be assayed quantitatively by pulse chasing, metabolically labeled cells expressing the P23H rhodopsin gene followed by immunoprécipitation of these receptors. Immunoprecipitated P23H rod opsin would then be analysed by SDS-PAGE and band intensities quantified using a phosphorimager. Alternatively examination of the spectral properties of P23H rod ospin isolated from COS-7 cells incubated with ligand would give an indication of the amount and correct native conformation of the mutant protein. In my experiments retinal was added to cells dissolved in ethanol, which would pose a problem when using concentrations of ligand higher than 20 )iM due to ethanol induced upregulation of Hsc70 (Miles at a/., 1991). Thus, an alternative delivery system for the ligand may need to be employed to eliminate ethanol from these experiments: for instance, Li and co-workers (1998) used liposomes as a carrier for delivery of retinals to cells in culture. By facilitating P23H rod opsin folding with the addition of ligand, it is conceivable that aggregation of the mutant receptor could be prevented or at least reduced. In the

140 experiments I performed, ligand was added twenty-four hours post transfection and approximately 35% of transfected cells already harbored aggresomes at this time point. Twenty-four hours after transfection, no noticeable change in aggresome numbers was observed following a 12 hour incubation of the cells with ligand. This implies that P23H rod opsin aggresomes are stable structures and cannot be solublised by 9-c/s-retinal. As the covalent attachment of 11-c/s-retinal to mammalian rod opsin is an early postranslational event (St. Jules et a/., 1989), the addition of ligand shortly after transfection, prior to the formation of P23H rod opsin aggresomes, may prevent or reduce the formation of mutant rod opsin aggresomes.

There are a number of compounds that exist which have been described as chemical chaperones’ due to their ability to facilitate the folding of misfolded proteins (Sato et al., 1996, Kagan et a!., 2000, Galbiati et a/., 2000). For instance glycerol and other polyhydric alcohols are known to stabilise protein conformation (Gekko et a/., 1981), increase the rate of in vitro protein folding (Sawano et a/., 1992) and increase the kinetics of oligomeric assembly (Shelanski et ai., 1973). As retinal appears to stabilise RP mutants of rhodopsin, like wise chemical chaperones might be able to stabilise P23H rod opsin in a native state and reduce aggregation of the receptor.

Low temperatures can also facilitate the folding of misfolded proteins glycoproteins; for example AF508CFTR, mutant HERG \C channel and mutant caveolin-3 are stabilised at reduced temperature in cultured cells. It appears that low temperature increases the proportion of mature glycosylated P23H rod opsin, suggesting that the folding defect has been overcome by incubation at this temperature. The data in this chapter also show that low temperature increases the steady state levels of P23H rod opsin, and suggests that the mutant protein is stabilised. Alternatively this increase may be a result of a decrease in the activity of the UPS system at this particular temperature which may be assessed with the use of a short lived green fluorescent protein as described by Bence et ai., 2001 in which the fluorescence signal gives a measure of the activity of the UPS. In addition, an increase in high molecular weight oligomers of P23H rod opsin in the DM insoluble fraction was observed at 30“C, which appeared as a smear on the blot and reached the stacking gel. These high molecular weight oligomers may be a result of detergent solubilisation and probably do not represent an increase in opsin aggregates within the cell as discussed in previous chapters. The influence of low temperatures on P23H rod opsin aggresome numbers remains to be determined. Experiments should incorporate a

141 30°C incubation soon after transfection (i.e. prior to aggresome formation) and a 30°C incubation following aggresome formation.

Mutant rod opsin exerts a dominant negative effect on the processing of WT rod opsin in Drosophila (Kurada and OTousa 1995, Colley et si, 1995, Kurada ef a/., 1998) and in vitro (chapter 3, Saliba et a!., 2002). It may be possible to reverse the dominant negative influences of mutant rod opsin and rescue WT rod opsin from ERAD and recruitment to mutant rod opsin aggresomes (chapter 3) by the use of 11-c/s-retinal, chemical chaperones or low temperatures. In chapter 3, I showed that mutant rod opsin can increase WT rod opsin aggresome numbers in COS-7 cells. Repetition of this experiment under various conditions that facilitate protein folding may reveal a way of minimising the dominant phenotype of P23H rod opsin.

The accumulation of misfolded proteins in the ER can lead to an ER stress response known as the unfolded protein response and ER stress can lead to apoptosis. An abundance of mutant rod opsin in the ER of rod cells and a compromised UPS could lead to an ER stress response and it may be important and beneficial to rod cell survival to alleviate ER stress. If it is possible to increase the proportion of mutant rod opsin, which attains the native conformational state with 11 -c/s-retinal or chemical chaperones, to such an extent as to reduce ER stress and also reduce the proteolytic demands on the UPS system, then this may mitigate mutant opsin toxicity. In the next chapter I investigate whether mutant rod opsin can evoke an unfolded protein response in cultured cells.

142 Chapter 7: RP mutants of rhodopsin and ER stress in cultured cells

7.1 Introduction

RP rod opsin mutant proteins which are misfolded accumulate in the ER of cultured cells (Chapter 3, Sung et al., 1991b, Sung at a!., 1993) and in the ER of rod cells (Frederick at a!., 2001). Accumulation of misfolded proteins within the ER leads to increased transcription of genes whose protein products are involved in alleviating the build up of misfolded polypeptides in the ER; the unfolded protein response (DPR). The UPR was first discovered over ten years ago with the observation that mammalian cells respond to the accumulation of misfolded proteins by increasing the production of the heat shock proteins BiP and Grp94 (Kozutsumi at a!., 1988). Since then many genes involved in the UPR have been characterised and signal-transducing molecules, which play pivotal roles in transmitting a signal from the stressed ER to the nucleus have been, discovered (as described in the introduction). Even though the UPR is an adaptive response, accumulation of misfolded proteins in the ER may exceed the ER s capacity to process such an abundance of polypeptides, which consequently can lead to death of the cell (Patil and Walter, 2001). One of the many genes which is upregulated in the UPR is Gaddi53, a leucine zipper transcription factor (Kaufman, 1999) which has been suggested to play a role in ER stress induced apoptosis (McCullough at a!., 2001).

As mutant rod opsin accumulates in the ER of cultured cells and forms aggresomes (Chapter 3, Saliba at a!., 2002, llling at a!., 2002) which impair the UPS (llling at a/., 2002) it is conceivable that accumulated mutant opsin may trigger an unfolded protein response as a means of coping with this sudden demand on the folding environment of the ER and an increased burden on ERAD. Furthermore as ER stress can lead to apoptosis, this may suggest mechanisms by which mutant opsin induces apoptosis in cultured cells and rod cells. Therefore to determine whether mutant opsin induces an UPR, RP rhodopsin genes were expressed in cultured cells and the detection of Gaddi 53 and BiP upregulation was used as an indicator of ER stress.

143 7.2 Results

7.2.1 Detection of the Gaddi 53 transcript in SH-SY5Y ceils expressing the K296E rhodopsin gene

Gaddi 53 is one of the many genes induced as a result of ER stress and therefore I used the induction of the Gaddi 53 gene as an indicator of ER stress in SH-SY5Y cells expressing the K296E rhodopsin gene. TM has been reported to induce the expression of Gaddi53 (Wang et a/.,1996), therefore, SH-SY5Y cells were incubated with 10 pg/ml for 16 hours to induce ER stress. To determine whether the Gaddi 53 transcript was present in SH-SY5Y under normal growth conditions or induced as a result of ER stress, RT-PCR was used to detect the Gaddi 53 transcript using the conditions described in the materials and methods. RNA isolated from untreated cells and cells treated with TM, was subjected to RT-PCR as described in the materials and methods and 1 |ig of RNA was reverse transcribed. Next, 2 pi of the reverse transcription reaction was used in the PGR reaction to detect the Gaddi 53 transcript. The Gaddi 53 transcript was not detected in SH-SY5Y cells under normal growth conditions at PGR cycle number 20, but was detected in cells treated with TM (Fig.7.1). Under normal growth conditions, however, the Gaddi 53 transcript was detectable at higher PGR cycle numbers (i.e. 25). As the Gaddi53 transcript is not detectable at cycle numbers 20-25 under normal growth conditions I investigated, using RT-PGR, whether the expression of the K296E rhodopsin gene in SH- SY5Y cells leads to an ER stress response, as determined by the upregulation of the Gaddi53 transcript. SH-SY5Y cells were transfected with pMT3-K296E and control cells were mock transfected with a p-galactosidase expression vector. RNA from cells was isolated 48 hours after transfection and 1 pg of RNA was reverse transcribed. In order to compare the band intensities of PGR products on an agarose gel over the linear range of the PGR reaction, six separate reactions containing 2 pi of the reverse transcritption reaction were set up. Six reactions for cells transfected with pMT3-K296E, six separate reactions for mock transfected cells and six separate reactions for cells incubated with TM. The PGR reaction was halted at each cycle number from 24 to 30. PGR products (10 pi) were resolved on a 1% agarose gel and placed on a U.V. transilluminator. Band intensities of the PGR products were measured using Kodak Digital Science 1D image analysis software. A photograph of the gel and graphical representation of the band intensities are shown in figure 7.2a and 7.2b respectively. To determine if equal amounts of cDNA from pMT3-K296E transfected cells and mock transfected cells had been used,

144 the p-actin transcript was used as an internal control for equal loading. The p-actin transcript was amplified using PCR and band intensities were determined from the linear range of the PCR reaction (Fig. 7.2A).

7.2.2 Detection of the Gaddi 53 protein in SH-SY5Y cells expressing mutant rhodopsin-GFP.

A much higher transfection efficiency of SH-SY5Y cells was achieved with the rhodopsin- GFP expression vectors (approximately 10%) than the pMTS vectors (approximately 1% of cells). Also by using the rhodopsin-GFP vectors transfection efficiency could be determined prior to cell lysis. TM induces the expression of the Gaddi 53 gene, therefore to determine if the Gaddi 53 transcript is translated into polypeptide, SH-SY5Y cells were incubated with a range of TM concentrations, 10-40 ug/ml for 16 hours (Fig. 7.3). SH- SY5Y cells were lysed in RIPA buffer and 30 ug of whole cell lysates were resolved on a 10% SDS-PAGE gel followed by immunoblotting with anti-Gaddi53. The Gaddi53 protein was present in cells treated with TM and migrated as an « 33 kDa species, but in untreated cells the Gaddi53 protein was not detected (Fig.7.3). Furthermore, increased concentrations of TM appeared to have no influence on the further induction of Gaddi 53.

To asses whether the expression of mutant rod opsin in SH-SY5Y cells induces an ER stress response, which would consequently lead to the expression of Gaddi 53, SH-SY5Y cells were transfected with pEGFP-WT, pEGFP-K296E, pEGFP-P23H. Transfection efficiencies were approximately equal. The presence of Gaddi53 protein was assayed using western blotting. Transfected cells were lysed in RIPA buffer 48 hours after ttransfection according to the procedure described in chapter 2. Control cells were treated with TM (10 ^ig/ml) for 16 hours. Whole cell lysates (80 pg) were resolved on a 10 % SDS-PAGE gel followed by immunoblotting with anti-Gaddi 53. An immunoreactive species was present in lysates from cells expressing WT, K296E, and P23H rhodopsin- GFP, which migrated with a mobility closely matching that of Gaddi53 from TM treated cells (Fig. 7.4) but gave a very faint signal. This band is absent from non-transfected cells (Fig. 7.4).

145 TM

Figure 7.1 Detection of Gaddi 53 by RT-PCR in SH-SY5Y cells incubated with TM

SH-SY5Y cells were treated with 10 ^ig/ml TM for 16 hours (as indicated) and the Gaddi 53 transcript was detected using RT-PCR. In untreated cells the transcript was not detected at cycle number 20. In TM treated cells the Gaddi53 PCR product was detected. DNA molecular weight markers are shown on the left.

146 Figure 7.2 Detection of the Gaddi 53 transcript in SH-SY5Y cells expressing K296E rhodopsin using RT-PCR.

SH-SY5Y cells were transfected with pMT3-K296E or mock transfected with a plasmid encoding p-galactosidase and 48 hours later RNA was isolated from the transfected cells and subjected to RT-PCR using primers to Gaddi53. As a positive control SH-SY5Y cells were incubated with TM (lOpg/ml) for 16 hours. A) i) Photograph of PCR products from reactions which were terminated at cycles 25-30 (as indicated) and were loaded on a 1% agarose gel. ii) Photograph of p-actin RT-PCR product from mock transfected cells and cells transfected with pMT3-K296E. The mean intensities of the p-actin bands are indicated. B) The mean band intensities of the Gaddi 53 PCR products were measured and represented graphically. A. Mock Transfection pMT3-K296E +TM

25 26 27 28 29 30 25 26 27 28 29 30 25 26 27 28 29 30

Mean Band Intensity tp-actin)

PMT3-K296E; 79.10 Mock : 86.03

P-actin

140

120

100

B 80 c , Mock transfection

■ O 60 I K296E

CÛ 40 .+TM c s 20

PCR cycle number

147 TM (fig/ml) 0 10 20 40

34.9—

29.9—

Figure 7.3 The Gaddi53 protein is present in SH-SY5Y cells following treatment with TM

SH-SY5Y cells were incubated with 0,10,20 and 40 jiQ/ml of TM (as indicated) for 16 hours and 30 pg of whole cell lysates were resolved on a 10% SDS-PAGE gel and immunoblotted with anti- Gaddi 53. In untreated cells Gaddi53 protein was not detected, whereas in cells incubated with TM (at the concentrations indicated) Gaddi53 was present (arrow). (*) indicates a non specific immunoreactive species Molecular weight markers in kDa are shown on the left.

SH-SY5Y WT-GFP K296E-GFP P23H-GFP +TM only 36.2

29.9

Figure 7.4 Detection of Gaddi 53 protein in SH-SY5Y cells expressing GFP-tagged WT and mutant rhodopsin.

SH-SY5Y cells were transfected with pEGFP-WT, pEGFP-K296E and pEGFP-P23H and 72 hours later the cells were lysed in RIPA buffer. Control non-transfected cells were incubated with TM (10 pg/ml for 16 hours) (+TM). Cell lysate from non-transfected cells not treated with TM is indicated (SH-SY5Y only). Whole cell lysates (80 pg) were resolved on a 10% SDS-PAGE gel and immunoblotted with anti- Gaddi 53. A band closely matching the mobility of Gaddi 53 in lane 4 (arrow) was present in all lysates, except non-transfected, untreated cells (lane 5). (*) indicates a non-specific immunoreactive species. Molecular weight markers in kDa are shown on the left.

148 7.2.3 The influence of GFP-tagged WT and GFP-tagged mutant rod opsin on BiP protein ievels in COS-7 cells.

The levels of the ER resident chaperone BiP (Grp78) increase dramatically following ER stress in cell culture models and have been reported to increase five fold in cells expressing mutant proteins (Kozutsumi et al., 1988). Due to the high transfection efficiency obtained with the rhodopsin-GFP expression vectors (consistently approximately 60%) in COS-7 cells, this system was used to assess the influence of WT and mutant rhodopsin-GFP expression on BiP levels. The detection of Gaddi53 protein was not used as an indicator of ER stress as this was detected under normal growth conditions in 0 0 8- 7 cells, but Gaddi 53 protein levels increased after incubation with TM in these cells. A 50% transfection efficiency would translate into a 2.5 fold increase in BiP levels overall assuming that BiP levels increase by 5 fold upon ER stress. In order to demonstrate BiP induction in COS-7 cells following ER stress, the cells were incubated with TM (10 ^ig/ml) for 16 hours. Treated cells were lysed in RIPA buffer and increasing amounts of whole cell lysates (1.25 jig -10 pg) were resolved on a 10% SDS-PAGE gel followed by immunoblotting with anti-BiP. In cells that were treated with TM, the levels of BiP protein increased as shown in figure 7.5. Comparing the lanes loaded with 1.25 p.g of whole cell lysates from TM treated and untreated cells, a clear difference in band intensity can be seen (Fig. 7.5).

To assess the influence of mutant rhodopsin-GFP gene expression in COS-7 cells on the levels of BiP, cells were transfected with pEGFP-WT, pEGFP-K296E and pEGFP-P23H. Transfection efficiency was determined (WT: 60%, K296E: 55% and P23H: 60%) and cells were lysed 48 hours after transfection in RIPA buffer as described in chapter 2. Whole cell lysates (2 pg) were resolved on a 10% SDS-PAGE gel followed by immunoblotting with anti-BiP. The levels of BiP appeared to increase in cells expressing the mutant rhodopsin-GFP genes in comparison to cells expressing WT rhodopsin-GFP (Fig.7.6). The level of BiP appeared to be equal in cells expressing K296E and P23H but was greater than the level of BiP in cells expressing WT-GFP only (P23H=K296E>WT).

149 +TM — — — — + + + + lOpg 5 ug 2.5 Jig 1.25 fig 1.25 fig 2.5 fig 5 fig 10 fig 81 —

49.9—

Figure 7.5 Tunicamycin induces an increase in BIP protein levels in COS-7 cells.

COS-7 cells were incubcited with 10 ag/ml TM for 16 hours and lysed in RIPA buffer. Different amounts of whole cell lysates froni TM treated and untreated cells (as indicated) were resolved on a 10% SDS-PAGE gel followed by immunoblotting with anti-BiP. Molecular weight markers in kDa are shown on the left.

COS-7 WT-GFP K296E-GFP P23H-GFP 112-—

Figure 7.6 The influence of GFP-tagged WT and mutant rod opsin on BiP protein levels in COS-7 cells

COS-7 cells were transfected with pEGFP-WT, pEGFP-K296E, pEGFP-P23H and cells were lysed in RIPA buffer 48 hours later. Total cell lysates (2.5 pg) were loaded on a 10% SDS-PAGE gel and immunoblotted with anti-BiP. Control non-transfected cells were treated with TM (+TTWI) or left untreated (COS-7 only). Molecular weight markers in kDa are shown on the left.

150 7.3 Discussion

The data presented in this chapter are suggestive of a mutant opsin induced UPR. RT- PCR was used to determine if the expression of Gaddi 53 was induced or if transcription was upregulated in cells expressing mutant rod opsin. If Gaddi 53 was not expressed at low levels under normal growth conditions in SH-SY5Y cells it would have been a simple case of just determining whether the Gaddi 53 transcript is induced in the presence of mutant rod opsin. However, as the Gaddi53 transcript was detected under normal growth conditions albeit at high PCR cycle numbers, a quantitative RT-PCR method would have to be employed to give an accurate measurement of Gaddi53 upregulation. This would require the quantification of the RNA target using a method known as competitive RT-PCR (Clementi et al., 1993, Siebert and Larrick, 1992) in which the target RNA (in this case Gaddi53) is co-amplified with serial dilutions of a synthetic RNA standard which is shorter than the RNA target and shares the same primer sequences. Then a calibration curve is constructed relating the logarithm of the ratio of both PCR products to the logarithm of the initial amount of RNA standard added. Thus if both amplicons are amplified with the same efficiency the amount of initial RNA target can be obtained from the point on the curve where the amounts of target and standard are equal. As this would be time consuming and complex, an alternative approach was used: detection of the Gaddi 53 protein.

Under normal growth conditions the Gaddi 53 protein was not detected, but was detected after the treatment of cells with TM. A protein migrating with a similar mobility to Gaddi53 appears in SH-SY5Y cells expressing WT and mutant GFP tagged rhodopsin for 48 hours, which was absent from non-transfected cells. This protein appears to migrate slightly slower than Gaddi 53 from cells treated with TM, which is a much thicker band on the blot which may give the appearance of a species with a different mobility. In addition, given the low transfection efficiency of SH-SY5Y cells (consistently approximately 10%) the Gaddi 53 protein is not particularly abundant making detection difficult. One way to get around this would be to immunoprecipitate Gaddi 53 from lysed cells, thus concentrating the protein. In addition after 48 hours of mutant GFP-tagged rod opsin of expression, SH-SY5Y cells floated into the medium, and thus would make the detection of Gaddi 53 protein in the remaining cells by western blotting more difficult. Time permitting, future experiments should involve the detection of Gaddi 53 after expression of rod opsin in SH-SY5Y cells at different time periods, possibly starting at 18 hours post transfection then 24 hours, then 48 hours.

151 The levels of BiP increased in COS-7 cells after a 16 hour incubation with TM, which causes the accumulation of misfolded proteins in the ER, demonstrating an intact unfolded protein response in these cells. To determine if the presence of mutant rod opsin results in ER stress, the levels of BiP were assessed in transfected cells using western blotting. The data suggest that the levels of BiP increase following 48 hours of expression of mutant rhodopsin genes, whereas the expression of WT rhodopsin appeared to have a less noticeable effect on BiP levels. Quantifying the increase in BiP levels using western blotting may be achieved by using pre-flashed X-ray film and measuring the band intensity with a densitometer. Given that the presence of mutant rod opsin appears to evoke a UPR it would be of interest to establish any causal relationship between rod opsin aggregates and the induction of the stress response. Rod opsin aggresomes impair the activity of the UPS (Dr. R. Kopito, personal communication and llling et a/.,2002) which consequently may cause more opsin to accumulate in the ER, as a functional proteasome is required for efficient retrotranslocation (chapter 5). Futhermore, inhibiting the proteolytic activity of the proteasome with pharmacological agents leads to an ER stress response (Bush et al., 1997). Similarly, aggresomes could also induce ER stress as a result of inhibiting the proteolytic activity of proteasomes.

Stress in the ER can lead to apoptosis via the activation of caspase-12 (Nakagawa et a!., 2000) and the expression of mutant opsin in cultured cells appears to induce a stress response in the ER. As mentioned in chapter 3, the expression of mutant rhodopsin genes is toxic to COS-7 and SH-SY5Y cells, and may be due to an apoptotic signal emanating from the ER as a result of ER stress. Caspase-12 may be a candidate, as it has been recently shown that protein aggregates in cultured cells lead to the activation of caspase-12, which was closely correlated with programmed cell death in these cells (Kouroku eta!., 2002)

A link between ER stress and mutant opsin aggresomes may be investigated further by the expression of a dominant negative form of IRE1 lacking the C-terminal effector domain as described by Wang et a!., 1988. This dominant negative form of IRE1 prevents the induction of BiP and Gaddi53 in response to ER stress (Wang eta!., 1998). Thus, the co expression of dominant negative IRE1 with mutant opsin in COS-7 cells may prevent the upregulation of BiP and Gaddi53, preventing an unfolded protein response. If opsin

152 aggresomes mediate their toxicity by inducing ER stress, it is conceivable that over expression of the dominant negative form of IRE1 may mitigate mutant opsin toxicity.

153 General Discussion The K296E and the P23H RP mutants of rhodopsin encode a protein that cannot attain a native conformation and consequently has a propensity to form intracellular ubiquitinated aggregates in cultured cells. The majority of mutations in the rhodopsin gene result in misfolding and a loss of function of the protein, as shown by accumulation in the ER and the inability of mutant rod opsin to efficiently bind 11-c/s-retinal to form rhodopsin with characteristic spectral properties (Sung et a/., 1991b, 1993, Kaushal and Khorana 1994). However, it appears that the RP disease phenotype is not a consequence of haplo- insufficiency as heterozygous knockout mice display no retinal degeneration, but minor structural disorganisation of rod inner and outer segments (Humphries et al., 1997). My work (chapter 3, Saliba et a/., 2002) and the work of other investigators (Sung et a/., 1991b, Sung et a/., 1993, llling et a/., 2002) suggest that mutations in the rhodopsin gene which lead to programmed rod cell death (Chang et a/., 1993) can be described as toxic gain of function mutations. In fact mutant rod opsin forms aggregates in cultured cells (chapter 3) and has a disruptive influence on the processing of the WT protein recruiting it to aggresomes (Chapter 3, Saliba et a/., 2002). This dominant negative behaviour of mutant rod opsin is also reflected in vivo; Drosophila mutant Rhi disrupts the normal processing and targeting of the WT protein without influencing the processing of other membrane proteins destined for the rhabdomere (Kurada and OTousa 1995). Furthermore rats expressing the P23H transgene have shorter rod outer segments than the normal animals and this may be explained as a result of the dominant effects of P23H rod opsin on the processing of the WT protein (Machida et al., 2000).

RP mutations in the rhodopsin gene that generate misfolded opsin are found through out the coding sequence. The position of the amino acid substitution within the polypeptide appears to influence the ease with which rod opsin forms off-pathway (off the folding pathway) conformers. As off pathway conformers can aggregate there may be a correlation between the site of an amino acid substitution within the rod opsin protein and the propensity to form aggregates. For instance K296E rod opsin formed aggresomes at a slower rate compared to the P23H mutant and a fraction of K296E rod opsin targeted to the plasma membrane unlike the P23H mutant. These data suggest that the kinetics of folding for these two mutants differ and consequently so does the frequency of ubiquitinated rod opsin aggresomes. The presence of rod opsin aggregates in photoreceptor cells needs further corroboration but mutant rod opsin does aggregate in cultured cells (chapter 3). The mechanisms by which protein aggregates in

154 many neurodegenerative diseases might be related to a toxic gain of function are still unclear. However, given the presence of elevated levels of ubiquitinated intracellular protein deposits in nearly all sporadic and hereditary neurodegenerative diseases, it has been suggested that there is a link between a defective UPS and pathogenesis (Mayeret al., 1989). This has been consolidated by genetic evidence linking mutations in the UPS to several neurodegenerative diseases and models (Kitadaet a!., 1998, Saiogh et a/., 1999, Cummings et a/., 1999, Fernandez-Funez et a/., 2000). For example, a form of early onset Parkinson's disease is caused by a missense mutation in the ubiquitin carboxy-terminal hydrolase LI (UCH-L1) gene (Leroy et a/., 1998). Uchll hydrolyses bonds between degraded end products and ubiquitin to generate free monomeric ubiquitin. The SCA1 phenotype was recently found to be accelerated by the loss of function of an E3 ligase (Cummings et a/., 1999) and in a Drosophila SCA1 model the disease was exacerbated by mutations in genes involved in the UPS (Fernadez-Funezet al., 2000). A recent report by Bence et al., 2001 revealed that protein aggregates (such as AF508CFTR and expanded polyglutamine repeat aggregates) in cultured cells almost completely inhibited the proteolytic activity of the proteasome showing that there is a specific relation between protein aggregation and the function of the UPS. Furthermore proteasome activity is also impaired in cultured cells harbouring P23H rod opsin aggresomes (Dr. R. Kopito, personal communication, llling et al., 2002).

Rod photoreceptor cell death in animal models of rhodopsin RP is considered to be an apoptotic process (Chang et ai, 1993, Portera-Cailliau et ai, 1994) but it is unknown what triggers apoptosis and in what cellular compartment these apoptotic signals originate. The expression of mutant rhodopsin gene in cultured cells has revealed that the mutant protein accumulates in the ER and forms stable aggregates. The presence of rod opsin aggregates presumably would compromise the UPS leading to further accumulation of rod opsin in the ER and in aggregates. Thus, it is conceivable that the accumulation of misfolded rod opsin in the ER together with a UPS compromised by rod opsin aggregates would trigger an unfolded protein response, as the data in chapter 7 suggests. This is an adaptive response of the ER to the presence of misfolded proteins but can eventually lead to apoptosis of the cell experiencing ER stress (Kaufman 1999, Ferri and Kroemer 2001). Recently an apoptotic signal emanating from the ER as a consequence of stress within this organelle mediated by caspase-12 has been observed (Nakagawaet ai, 2000, Yoneda et ai, 2001). Furthermore, caspase-12 is activated by intracellular polyglutamine protein aggregates and this activation was closely associated with apoptotic cell death

155 (Kouroku et al., 2002). In rod cells an apoptotic signal may be initiated at the ER as a consequence of the accumulation of mutant rod opsin in the ER and an impaired UPS. This is perhaps reflected by the influence of gene dosage on mutant rhodopsin retinal degeneration which was accelerated in rats expressing high levels of the P23H transgene (Olsson at a/., 1992) and may be related to the increased burden on the folding environment of the ER and the UPS.

Thus it would seem important to alleviate the demands on ER processing and the proteolytic load on the UPS by reducing or abolishing the expression of mutant rhodopsin. This has been successively achieved through the delivery of allele specific ribozymes using recombinant adeno-associated virus (rAAV) vectors in P23H transgenic rats which led to the delay of retinal degeneration (Lewinat a!., 1998). Even ribozymes delivered to older transgenic rats with a distinct loss of photoreceptors rescued the retina from degeneration (LaVail at a/., 2000). This shows that the disease can be slowed once the mutant protein has been removed even after the disease process has begun and it appears that disease progression requires the continues expression of P23H rhodopsin. An alternative to gene therapy is the use of vitamin A supplements as a means of slowing retinal degeneration. Retinal degeneration of T17M rhodopsin transgenic mice was delayed by applying Vitamin A supplements to the diets of these experimental mice (Liat ai, 1998). In addition, oral vitamin A supplementation slowed the rate, on average, of retinal degeneration in adult patients with the common forms of RP (Berson at ai, 1993) but the genetic basis of these patient’s diseases were unknown. Given the ability of opsin ligand to promote ER export of P23H rod opsin in cell culture (chapter 6, Saliba at ai, 2002) which would likely reduce the burden on the ERAD process and may reduce rod opsin aggregation, this may be a viable option for the protection of rod cells from the destructive influence of misfolded rod opsin and offer an alternative treatment to patients with this particular mutation.

Molecular chaperones have been consistently shown to associate with proteinaceous intracellular inclusions in cellular models of neurodegenerative diseases and can give neuronal protection from the damaging effects of abnormal proteins (Sherman and Goldberg 2001). Therefore, a number of investigators have thought it prudent to investigate any potential role for molecular chaperones in reducing the cellular toxicity of terminally misfolded proteins. Interestingly, the elevated expression of Hsp70 in an animal model of polyglutamine repeats disorders, suppressed neuronal degeneration without

156 changing the frequency of polyglutamine inclusions (Cummings et al., 2001). In Drosophila models of human polyglutamine diseases over expression of Hsp70 and Hsp40 effectively protected neurons that expressed proteins with expanded polyglutamine repeats (Warrick at ai, 1999, Fernandez-Funez at ai, 2000, Kazemi-Esfarjani and Benzer, 2000) without having any effect on the frequency of inclusions but increased the solubility of the polyglutamine proteins (Chan at ai, 2000). In neuronal cultures, expression of Hsp70 and Hsp40 protected cells from apoptosis induced by expanded polyglutamine repeat androgen receptor (Kobayashi at ai, 2000). Over expression of Hsp70 and its co-chaperone Hsp40 increased the solubility of polyglutamine repeat androgen receptor aggregates formed in cell culture and furthermore these molecular chaperones accelerated the degradation of this aggregation prone protein (Bailey et al., 2002). Based on the findings that molecular chaperones have a protective influence on neurons expressing genes whose products produce aggregation prone proteins, it is possible that the over expression of molecular chaperones in rod cells expressing mutant rhodopsin may also serve a protective role. This possibility could be explored at first in cell culture models; assessing the influence of over expression of molecular chaperones (Hsp70 and/or Hsp40) on the solubility of mutant opsin, aggresome frequency and cell death.

Many neurodegenerative diseases are associated with an accumulation of misfolded protein ultimately leading to protein aggregates and this process appears to be intricately linked to the UPS. Similarly RP rhodopsin mutants aggregate presumably due to the failure of molecular chaperones to fold mutant rod opsin and also as a consequence of failed degradation by an over burdened UPS. The work reported in this thesis suggests that there are protein processing mechanisms in common with those of other toxic proteins associated with neurodegeneration. However, unlike other protein deposition’ diseases rhodopsin RP may not be refractory to treatment as the use of rod opsin ligands to overcome folding defects and perhaps promoting rod cell survival offers a promising mode of therapeutic intervention. Furthermore, investigation into the role of molecular chaperones in rod opsin folding, aggregation and cellular toxicity now appears to be an important experimental goal.

157 References:

Abeijon.C. and C.B.Hirschberg. 1992. Topography of glycosylation reactions in the endoplasmic reticulum. Trends Biochem. Sci. 17:32-36.

Agashe.V.R. and F.U.Hartl. 2000. Roles of molecular chaperones in cytoplasmic protein folding. Semin. Cell Dev. Biol. 11:15-25. al Jandal.N., G.J.Farrar, A.S.Kiang, M.M.Humphries, N.Bannon, J.B.Findlay, P.Humphries, and P.F.Kenna. 1999. A novel mutation within the rhodopsin gene (Thr-94- lle) causing autosomal dominant congenital stationary night blindness. Hum. Mutat. 13:75- 81.

Al Maghtheh,M., R.Y.Kim, A.Hardcastle, C.Inglehearn, and S.S.Bhattacharya. 1994a. A 150 bp insertion in the rhodopsin gene of an autosomal dominant retinitis pigmentosa family. Hum. Mol. Genet. 3:205-206.

Al-Maghtheh, M., C. Inglehearn, P. Lund, M. Jay, A. Bird, and S. Bhattacharya. 1994b. Two new rhodopsin transversion mutations (L40R; M216K) in families with autosomal dominant retinitis pigmentosa. Hum. Mutat. 3: 409-410

Al Maghtheh,M., E.N.Vithana, C.F.Inglehearn, T.Moore, A.C.Bird, and S.S.Bhattacharya. 1998. Segregation of a PRKCG mutation in two RP11 families. Am. J. Hum. Genet. 62:1248-1252.

Antinolo,G., B.Sanchez, S.Borrego, T.Rueda, P.Chaparro, and J.C.Cabeza. 1994. Identification of a new mutation at codon 171 of rhodopsin gene causing autosomal dominant retinitis pigmentosa. Hum. Mol. Genet. 3:1421.

Anton,L.C., U.Schubert, I.Bacik, M.F.Princiotta, P.A.Wearsch, J.Gibbs, P.M.Day, C.Realini, M.C.Rechsteiner, J.R.Bennink, and J.W.Yewdell. 1999. Intracellular localization of proteasomal degradation of a viral antigen.J. Cell Biol. 146:113-124.

Anukanth,A. and H.G.Khorana. 1994. Structure and function in rhodopsin. Requirements of a specific structure for the intradiscal domain. J. Biol. Chem. 269:19738-19744.

Ayuso,C., M.J.Trujillo, M.Robledo, C.Ramos, J.Benitez, F.Martin-Oses, T.Del Rio, and B.Garcia-Sandoval. 1996. Novel rhodopsin mutation in an autosomal dominant retinitis

158 pigmentosa family: phenotypic variation in both heterozygote and homozygote Val137Met mutant patients. Hum. Genet 98:51-54.

Baas.P.W. 1998. The role of motor proteins in establishing the microtubule arrays of axons and dendrites. J. Chem. Neuroanat. 14:175-180.

Back.J.F., D.Oakenfull, and M.B.Smith. 1979. Increased thermal stability of proteins in the presence of sugars and polyols. Biochemistry 18:5191-5196.

Bailey,C.K., I.F.Andriola, H.H.Kampinga, and D.E.Merry. 2002. Molecular chaperones enhance the degradation of expanded polyglutamine repeat androgen receptor in a cellular model of spinal and bulbar muscular atrophy. Hum. Mol. Genet. 11:515-523.

Baker,E.K., N.J.Colley, and C.S.Zuker. 1994. The cyclophilin homolog NinaA functions as a chaperone, forming a stable complex in vivo with its protein target rhodopsin. EMBO J. 13:4886-4895.

Bareil 0, Hamel 0, Pallares-Ruiz N, Arnaud B, Démaillé J, Claustres M. 1999. Molecular analysis of the rhodopsin gene in southern France: identification of the first duplication responsible for retinitis pigmentosa, c.998999ins4. Ophthalmic. Genet. 20:173-82

Bays,N.W., S.K.Wilhovsky, A.Goradia, K.Hodgkiss-Harlow, and R.Y.Hampton. 2001. HRD4/NPL4 is required for the proteasomal processing of ubiquitinated ER proteins. Mol. Biol. Ce//12:4114-4128.

Bell,C., C.AConverse, H.M.Hammer, A.Osborne, and N.E.Haites. 1994. Rhodopsin mutations in a Scottish retinitis pigmentosa population, including a novel splice site mutation in intron four. Br. J. Ophthalmol. 78:933-938.

Berson,E.L. 1996. Retinitis pigmentosa: unfolding its mystery. Proc. Natl. Acad. Sci. U. S. A 93:4526-4528.

Bessant, D.A., Khaliq, S., Hameed, A,. Anwar. K., Payne, A.M., Mehdi, S.Q., Bhattacharya, 8.8.1999. Severe autosomal dominant retinitis pigmentosa caused by a novel rhodopsin mutation (Ter349Glu). Mutations in brief no. 208. Online. Hum Mutat. 13:83

Bhattacharya,8.S., A.F.Wright, J.F.CIayton, W.H.Price, C.l.Phillips, C.M.McKeown, M.Jay, A.C.Bird, P.L.Pearson, E.M.Southern, and . 1984. Close genetic linkage between

159 X-linked retinitis pigmentosa and a restriction fragment length polymorphism identified by recombinant DNA probe L1.28. Nature 309:253-255.

Biederer.T-, C.Volkwein, and T.Sommer. 1996. Degradation of subunits of the SecBlp complex, an integral component of the ER membrane, by the ubiquitin-proteasome pathway. EMBO J. 15:2069-2076.

Biederer,T., C.Volkwein, and T.Sommer. 1997. Role of Cuelp in ubiquitination and degradation at the ER surface. Science 278:1806-1809.

Bolz,H., B.von Brederlow, A.Ramirez, E.G.Bryda, K.Kutsche, H.G.Nothwang, M.Seeliger, C.-S.del, M.C.Vila, O.P.Molina, A.Gal, and C.Kubisch. 2001. Mutation of CDH23, encoding a new member of the cadherin gene family, causes Usher syndrome type 1D.Nat. Genet 27:108-112.

Borrego,S., B.Sanchez, A.Ruiz, and G.Antinolo. 1996. Missense mutation A346P in the rhodopsin gene in one family with autosomal dominant retinitis pigmentosa.Hum. Mutat. 7:180-181.

Boughman,J.A. and G.A.Fishman. 1983. A genetic analysis of retinitis pigmentosa. Br. J. Ophthalmol. 67:449-454.

Bownds,D. 1967. Site of attachment of retinal in rhodopsin. Nature 216:1178-1181.

Braun,S., K.Matuschewski, M.Rape, S.Thoms, and S.Jentsch. 2002. Role of the ubiquitin- selective CDC48(UFD1/NPL4 )chaperone (segregase) in ERAD of 0LE1 and other substrates. EMBOJ. 21:615-621.

Brown,C.R., S.J.Doxsey, E.White, and W.J.Welch. 1994. Both viral (adenovirus E1B) and cellular (hsp 70, p53) components interact with centrosomes. J. Cell P/?ys/o/160:47-60.

Bunge,S., H.Wedemann, D.David, D.J.Terwilliger, L.I.van den Born, C.Aulehla-Scholz, C.Samanns, M.Horn, J.Ott, E.Schwinger, and . 1993. Molecular analysis and genetic mapping of the rhodopsin gene in families with autosomal dominant retinitis pigmentosa. Genomics 17:230-233.

Bush,K.T., A.L.Goldberg, and S.K Nigam. 1997. Proteasome inhibition leads to a heat- shock response, induction of endoplasmic reticulum chaperones, and thermotolerance. J. Biol. Chem. 272:9086-9092.

160 Carlson,s.G., T.W.Fawcett, J.D.Bartlett, M.Bernier, and N.J.Holbrook. 1993. Regulation of the C/EBP-related gene gaddISS by glucose deprivation. Mol. Cell Biol. 13:4736-4744.

Cha,K., P.J.Reeves, and H.G.Khorana. 2000. Structure and function in rhodopsin: destabilization of rhodopsin by the binding of an antibody at the N-terminal segment provides support for involvement of the latter in an intradiscal tertiary structure. Proc. Natl. Acad. Sci. U. S. A 97:3016-3021.

Chakarova,C.F., M.M.Hims, H.Bolz, L.Abu-Safieh, R.J.Patel, M.G.Papaioannou, C.F.Inglehearn, T.J.Keen, C.Willis, A.T.Moore, T.Rosenberg, A.R.Webster, A.C.Bird, A.Gal, D.Hunt, E.N.Vithana, and S.S.Bhattacharya. 2002. Mutations in HPRP3, a third member of pre-mRNA splicing factor genes, implicated in autosomal dominant retinitis pigmentosa. Hum. Mol. Genet. 11:87-92.

Chan,H.Y., J.M.Warrick, G.L.Gray-Board, H.L.Paulson, and N.M.Bonini. 2000. Mechanisms of chaperone suppression of polyglutamine disease: selectivity, synergy and modulation of protein solubility in Drosophila. Hum. Mol. Genet. 9:2811-2820.

Chang, G.Q., Y. Hao, F. Wong. 1993. Apoptosis: Final common pathway of photoeceptor death in rd, rds and rhodopsin mutant mice. Neuron 11: 595-605

Chapple,J.P., C.Grayson, A.J.Hardcastle, R.S.Saliba, S.J.van der, and M.E.Cheetham. 2001. Unfolding retinal dystrophies: a role for molecular chaperones? Trends Mol. Med 7:414-421.

Chen,J., J.G.Flannery, M.M.LaVail, R.H.Steinberg, J.Xu, and M l.Simon. 1996. bcl-2 overexpression reduces apoptotic photoreceptor cell death in three different retinal degenerations. Proc. Natl. Acad. Sci. U. S. A 93:7042-7047.

Clementi,M., S.Menzo, P.Bagnarelli, A.Manzin, A.Valenza, and P.E.Varaldo. 1993. Quantitative PCR and RT-PCR in virology. PCR Methods Appi. 2:191-196.

Colley,N.J., J.A Cassill, E.K.Baker, and C.S.Zuker. 1995. Defective intracellular transport is the molecular basis of rhodopsin- dependent dominant retinal degeneration.Proc. Natl. Acad. Sci. U. S. A 92:3070-3074.

Cone, R.A. 1972. Rotational diffusion of rhodopsin in the visual receptor membrane. Nat. New Biol. 236:39-43.

161 Cummings,C.J., E.Reinstein, Y.Sun, B.Antalffy, Y.Jiang, A.CIechanover, H.T.Orr, A.L.Beaudet, and H.Y.Zoghbi. 1999. Mutation of the E6-AP ubiquitin ligase reduces nuclear inclusion frequency while accelerating pciyglutamine-induced pathology in SCA1 mice. Neuron 24 :879-892.

Cummings, C.J, Sun, Y, Opa,l P., Antalffy, B., Mestril, R., Orr, H.T., Dillmann, W.H., Zcghbi, H.Y. 2001. Over-expression of inducible HSP70 chaperone suppresses neuropathology and improves motor function in SCA1 mice. Hum. Mol. Genet. 10:1511-8

David,V., F.Hochstenbach, S.Rajagopalan, and M.B.Brenner. 1993. Interaction with newly synthesized and retained proteins in the endoplasmic reticulum suggests a chaperone function for human integral membrane protein IP90 (calnexin). J. Biol. Chem. 268:9585- 9592.

Davidson,F.F. and H.Steller. 1998. Blocking apoptosis prevents blindness in Drosophila retinal degeneration mutants. Nature 391:587-591. de Virgilio,M., H.Weninger, and N.E.Ivessa. 1998. Ubiquitination is required for the retro translocation of a short-lived luminal endoplasmic reticulum glycoprotein to the cytosol for degradation by the proteasome.J. Biol. Chem. 273:9734-9743.

DeMartino,G.N. and C.A.SIaughter. 1999. The proteasome, a novel protease regulated by multiple mechanisms. J. Biol. Chem. 274:22123-22126.

Den Hollander,A.I., J.R.Heckenlively, L.I.van den Born, Y.J.de Kok, S.D.Velde-Visser, U.Kellner, B.Jurklies, M.J.van Schooneveld, A.BIankenagel, K.Rohrschneider, B.Wissinger, J.R.Cruysberg, A.F.Deutman, H.G.Brunner, E.Apfelstedt-Sylla, C.B.Hoyng, and F.P.Cremers. 2001. Leber congenital amaurosis and retinitis pigmentosa with Coats like exudative vasculopathy are associated with mutations in the crumbs homologue 1 (CRB1) gene. Am. J. Hum. Genet. 69:198-203.

Denning,G.M., M.P.Anderson, J.F.Amara, J.Marshall, A.E.Smith, and M.J.Welsh. 1992. Processing of mutant cystic fibrosis transmembrane conductance regulator is temperature-sensitive. Nature 358:761-764.

Doi,T., R.S.Molday, and H.G.Khorana. 1990. Role of the intradiscal domain in rhodopsin assembly and function. Proc. Natl. Acad. Sci. U. S. A 87:4991-4995.

162 Downer,N.W. 1985. Cross-linking of dark-adapted frog photoreceptor disk membranes. Evidence for monomeric rhodopsin. Biophys. J. 47:285-293.

Dowling,J.E. and Gibbons,I.R. 1960. The effect of vitamin A deficiency on the fine structure of the retina. In The Structure of the Eye. (Ed. Smelser, G.K.). Pg. 85-99. Academic Press: New York

Dowling, J.E. and G. Wald. 1958. Vitamin A deficiency and night blindness. Proc. Natl. Acad. Sci. U.S.A. 44: 648-61

Dryja,T.P., T.L.McGee, E.Reichel, L.B.Hahn, G.S.Cowley, D.W.Yandell, M.A.Sandberg, and E.L.Berson. 1990a. A point mutation of the rhodopsin gene in one form of retinitis pigmentosa. Nature 343:364-366.

Dryja,T.P., T.L.McGee, L.B.Hahn, G.S.Cowley, J.E.Olsson, E.Reichel, M.A.Sandberg, and E.L.Berson. 1990b. Mutations within the rhodopsin gene in patients with autosomal dominant retinitis pigmentosa. N. Engl. J. Med. 323:1302-1307.

Dryja,T.P., L.B.Hahn, G.S.Cowley, T.L.McGee, and E.L.Berson. 1991. Mutation spectrum of the rhodopsin gene among patients with autosomal dominant retinitis pigmentosa.Proc. Natl. Acad. Sci. U. S. A 88:9370-9374.

Dryja,T.P. 1992. Doyne Lecture. Rhodopsin and autosomal dominant retinitis pigmentosa. Eye 6 (Pt1):1-10.

Dryja,T.P., E.L.Berson, V.R.Rao, and D.D.Oprian. 1993. Heterozygous missense mutation in the rhodopsin gene as a cause of congenital stationary night blindness. Nat. Genet. 4:280-283.

Dryja,T.P. and E.L.Berson. 1995. Retinitis pigmentosa and allied diseases. Implications of genetic heterogeneity. Invest Ophthalmol. Vis. Sci. 36:1197-1200.

Dryja,T.P., L.B.Hahn, T.Reboul, and B.Arnaud. 1996. Missense mutation in the gene encoding the alpha subunit of rod transducin in the Nougaret form of congenital stationary night blindness. Nat. Genet. 13:358-360.

Dryja,T.P., J.A.McEvoy, T.L.McGee, and E.L.Berson. 2000. Novel rhodopsin mutations Gly114Val and Gln184Pro in dominant retinitis pigmentosa. Invest Ophthalmol. Vis. Sci. 41:3124-3127.

163 Duffin.K.L., G.W.Lange, J.K.Welply, R.FIorman, P.J.O'Brien, A.Dell, A.J.Reason, H.R.Morris, and S.J.FIiesler. 1993. Identification and oligosaccharide structure analysis of rhodopsin glycoforms containing galactose and sialic acid. Glycobiology 3:365-380.

Duksin,D. and W.C.Mahoney. 1982. Relationship of the structure and biological activity of the natural homologues of tunicamycin. J. Biol. Chem. 257:3105-3109.

Edington,B.V., S.A.Whelan, and L.E.Hightower. 1989. Inhibition of heat shock (stress) protein induction by deuterium oxide and glycerol: additional support for the abnormal protein hypothesis of induction. J. CellPhysiol^39:2^9-228.

Ellgaard,L., M.Molinari, and A.Helenius. 1999. Setting the standards: quality control in the secretory pathway. Science 286:1882-1888.

Emeis,D., H.Kuhn, J.Reichert, and K.P.Hofmann. 1982. Complex formation between metarhodopsin II and GTP-binding protein in bovine photoreceptor membranes leads to a shift of the photoproduct equilibrium . FEBS Lett. 143:29-34.

Endo,T., N.Niu, D.T.Organisciak, Y.Sato, A.Kobata, and E.L.Kean. 1996. Analysis of the oligosaccharide chains of rhodopsin from normal rats and those with hereditary retinal dystrophy. Exp. Eye Res. 63:395-405.

Engbretson,G.A. and P.Witkovsky. 1978. Rod sensitivity and visual pigment concentration in Xenopus. J. Gen. Physiol72:80^-8^9.

Eudy,J.D., M.D.Weston, S.Yao, DM. Hoover, H.L.Rehm, M.Ma-Edmonds, D.Yan, I.Ahmad, J.J.Cheng, C.Ayuso, C.Cremers, S.Davenport, C.Moller, C.B.Talmadge, K.W.Beisel, M.Tamayo, C.C.Morton, A.Swaroop, W.J.Kimberling, and J.Sumegi. 1998. Mutation of a gene encoding a protein with extracellular matrix motifs in Usher syndrome type lia. Sc/ence 280:1753-1757.

Eymin,B., L.Dubrez, M.AIIouche, and E.Solary. 1997. Increased gadd153 messenger RNA level is associated with apoptosis in human leukemic cells treated with etoposide. Cancer Res. 57:686-695.

Fabunmi,R.P., W.C.Wigley, P.J.Thomas, and G.N.DeMartino. 2000. Activity and regulation of the centrosome-associated proteasome. J. Biol. Chem. 275:409-413.

164 Fain, G.L. and Cornwall, M.C. 1993. Light and dark adaptation in vertebrate photoreceptors. In Contrast Sensitivity: from Receptors to Clinic. (Eds. Shapely, R and Lam, D.K.) Pg. 3-32. MIT Press: Cambridge. MA.

Fain,G.L. and J.E.Lisman. 1993. Photoreceptor degeneration in vitamin A deprivation and retinitis pigmentosa: the equivalent light hypothesis.Exp. Eye Res. 57:335-340.

Farrar,G.J., P.Kenna, R.Redmond, P.McWilliam, D.G.Bradley, M.M.Humphries, E.M.Sharp, C.F.Inglehearn, R.Bashir, M.Jay, and . 1990. Autosomal dominant retinitis pigmentosa: absence of the rhodopsin proline— histidine substitution (codon 23) in pedigrees from Europe. Am. J. Hum. Genet. 47:941-945.

Farrar,G.J., S.A.Jordan, P.Kenna, M.M.Humphries, R.Kumar-Singh, P.McWilliam, V.AIIamand, E.Sharp, and P.Humphries. 1991. Autosomal dominant retinitis pigmentosa: localization of a disease gene (RP6) to the short arm of chromosome 6. Genomics 11:870-874.

Farrar,G.J., J.B.Findlay, R.Kumar-Singh, P.Kenna, M.M.Humphries, E.Sharpe, and P.Humphries. 1992. Autosomal dominant retinitis pigmentosa: a novel mutation in the rhodopsin gene in the original 3q linked family.Hum. Mol. Genet. 1:769-771.

Farrar,G.J., P.F.Kenna, and P.Humphries. 2002. On the genetics of retinitis pigmentosa and on mutation-independent approaches to therapeutic intervention. EMBO J. 21:857- 864.

Ferreira,P.A., J.T.Hom, and W.L.Pak. 1995. Retina-specifically expressed novel subtypes of bovine cyclophilin. J. Biol. Chem. 270:23179-23188.

Ferreira,P.A., T.A.Nakayama, W.L.Pak, and G.H.Travis. 1996. Cyclophilin-related protein RanBP2 acts as chaperone for red/green opsin. Nature 383:637-640.

Ferri,K.F. and G.Kroemer. 2001. Organelle-specific initiation of cell death pathways. Nat. Cell Biol. 3:E255-E263.

Fishman,G.A., K.Vandenburgh, E.M.Stone, L.D.Gilbert, K.R.Alexander, and V.C.Sheffield. 1992. Ocular findings associated with rhodopsin gene codon 267 and codon 190 mutations in dominant retinitis pigmentosa. Arch. Ophthalmol. 110:1582-1588.

165 Fliesler.S.J., L.M.Rapp, and J.G.Hollyfield. 1984. Photoreceptor-specific degeneration caused by tunicamycin. Nature 311:575-577.

Fliesler.S.J. and S.F.Basinger. 1985. Tunicamycin blocks the incorporation of opsin into retinal rod outer segment membranes.Proc. Natl. Acad. Sol. U. S. A 82:1116-1120.

Fliesler.S.J., M.E.Rayborn, and J.G.Hollyfield. 1985. Membrane morphogenesis in retinal rod outer segments: inhibition by tunicamycin. J. Cell Biol. 100:574-587.

Frederick,J.M., N.V.Krasnoperova, K.Hoffmann, J.Church-Kopish, K.Ruther, K.Howes, J.Lem, and W.Baehr. 2001. Mutant rhodopsin transgene expression on a null background. Invest Ophthalmol. Vis. Soi. 42:826-833.

Friedman,A.D. 1996. GADDI 53/CHOP, a DMA damage-inducible protein, reduced CAAT/enhancer binding protein activities and increased apoptosis in 32D cl 3 myeloid cells. Cancer Res. 56:3250-3256.

Fuchs,S., H.Kranich, M.J.Denton, E.Zrenner, S.S.Bhattacharya, P.Humphries, and A.Gal. 1994. Three novel rhodopsin mutations (OilOF, L131P, A164V) in patients with autosomal dominant retinitis pigmentosa. Hum. Mol. Genet. 3:1203.

Fujita,S., T.Endo, J.Ju, E.L.Kean, and A.Kobata. 1994. Structural studies of the N-linked sugar chains of human rhodopsin. G/ycob/o/ogy 4:633-640.

Fukuda,M.N., D.S.Papermaster, and P.A.Hargrave. 1979. Rhodopsin carbohydrate. Structure of small oligosaccharides attached at two sites near the NH2 terminus. J. Biol. Chem. 254:8201-8207.

Gahmberg,C.G. and M.Tolvanen. 1996. Why mammalian cell surface proteins are glycoproteins. Trends Biochem. Sci. 21:308-311.

Gal,A., A.Artlich, M.Ludwig, G.Niemeyer, K.OIek, E.Schwinger, and A.Schinzel. 1991. Pro- 347-Arg mutation of the rhodopsin gene in autosomal dominant retinitis pigmentosa. Genomics 11:468-470.

Gal,A., U.Orth, W.Baehr, E.Schwinger, and T.Rosenberg. 1994. Heterozygous missense mutation in the rod cGMP phosphodiesterase beta- subunit gene in autosomal dominant stationary night blindness. Nat. Genet. 7 :64-68.

166 Gal, A., E. Apfelstedt-Sylla, A. R. Janecke, and E. Zrenner. 1997. Rhodopsin mutations in inherited retinal dystrophies and dysfunctions. Prog. Ret. Eye Res. 16:51-79

Gal,A., Y.Li, D.A.Thompson, J.Weir, U.Orth , S.G.Jacobson, E.Apfelstedt-Sylla, and D.Vollrath. 2000. Mutations in MERTK, the human orthologue of the RCS rat retinal dystrophy gene, cause retinitis pigmentosa. Nat. Genet. 26:270-271.

Galbiati,F., D.Volonte, C.Minetti, D.B.Bregman, and M.P.Lisanti. 2000. Limb-girdle muscular dystrophy (LGMD-1C) mutants of caveolin-3 undergo ubiquitination and proteasomal degradation. Treatment with proteasomal inhibitors blocks the dominant negative effect of LGMD-1C mutanta and rescues wild-type caveolin-3. J. Biol. Chem. 275:37702-37711.

Garcia-Mata,R., Z.Bebok, E.J.Sorscher, and E.S.Sztul. 1999. Characterization and dynamics of aggresome formation by a cytosolic GFP- chimera. J. Cell Biol. 146:1239- 1254.

Gekko,K. and S.Koga. 1983. Increased thermal stability of collagen in the presence of sugars and polyols. J. Biochem. (Tokyo) 94:199-205.

Gerlsma,S.Y. and E.R.Stuur. 1972. The effect of polyhydric and monohydric alcohols on the heat-induced reversible dénaturation of lysozyme and ribonuclease. Int. J. Pept. Protein Res. 4:377-383.

Goliath,R., S.Bardien, A.September, R.Martin, R.Ramesar, and J.Greenberg. 1998. Rhodopsin mutation G109R in a family with autosomal dominant retinitis pigmentosa. Hum. Mutat. Supp11:840-841.

Gonatas,N.K., J.O.Gonatas, and A Stieber. 1998. The involvement of the Golgi apparatus in the pathogenesis of amyotrophic lateral sclerosis, Alzheimer's disease, and ricin intoxication. Histochem. Cell Biol. 109:591-600.

Grinna,L.8. and P.W.Robbins. 1980. Substrate specificities of rat liver microsomal glucosidases which process glycoproteins. J. Biol. Chem. 255:2255-2258.

Haas,I.G. and M.Wabl. 1983. Immunoglobulin heavy chain binding protein. Nature 306:387-389.

167 Hagstrom.SA, MANorth, P.L.Nishina, E.L.Berson, and T.P.Dryja. 1998. Recessive mutations in the gene encoding the tubby-like protein TULP1 in patients with retinitis pigmentosa. Nat Genet 18:174-176.

Haim M, Grundmann K, Gal A, Rosenberg T. 1996. Novel rhodopsin mutation (M216R) in a Danish family with autosomal dominant retinitis pigmentosa.Opf?f/7a/m/c Genef. 17:193-7.

Halaban.R., E.Cheng, Y.Zhang, G.Moellmann, D.Hanlon, M.Michalak, V.Setaluri, and D.N.Hebert. 1997. Aberrant retention of tyrosinase in the endoplasmic reticulum mediates accelerated degradation of the enzyme and contributes to the dedifferentiated phenotype of amelanotic melanoma cells. Proc. Natl. Acad. Sci. U. S. A 94:6210-6215.

Hamman.B.D., L.M.Hendershot, and A.E.Johnson. 1998. BiP maintains the permeability barrier of the ER membrane by sealing the lumenal end of the translocon pore before and early in translocation. Cell 92:747-758.

Hammond,C. and A.Helenius. 1994. Folding of VSV G protein: sequential interaction with BiP and calnexin. Science 266:456-458.

Hammond,C., I.Braakman, and A.Helenius. 1994. Role of N-linked oligosaccharide recognition, glucose trimming, and calnexin in glycoprotein folding and quality control. Proc. Natl. Acad. Sci. U. S. A 91:913-917.

Han,M., M.Groesbeek, T.P.Sakmar, and S.O.Smith. 1997. The C9 methyl group of retinal interacts with glycine-121 in rhodopsin. Proc. Natl. Acad. Sci. U. S. A 94:13442-13447.

Hargrave,P.A. 1977. The amino-terminal tryptic peptide of bovine rhodopsin. A glycopeptide containing two sites of oligosaccharide attachment. Biochim. Biophys. Acta 492:83-94.

Hargrave, P.A. 1995. Future directions for Rhodopsin structure and function studies. Behavioral and Brain Sc/encesi 8:403-414.

Hayes,K.C. 1974. Retinal degeneration in monkeys induced by deficiencies of vitamin E or A. Invest Ophthalmol. 13:499-510.

He,W., C.W.Cowan, and T.G.Wensel. 1998. RGS9, a GTPase accelerator for phototransduction. Neuron 20:95-102.

Helenius,A. 1994. How N-linked oligosaccharides affect glycoprotein folding in the endoplasmic reticulum. Mol. Biol. Cell 5:253-265.

168 Helenius.A- and M.Aebi. 2001. Intracellular functions of N-linked glycans. Science 291:2364-2369.

Heller,J. 1968. Structure of visual pigments. I. Purification, molecular weight, and composition of bovine visual pigmentSOO. Biochemistry 7:2906-29^ 3.

Henle,K.J., J.W.Peck, and R.Higashikubo. 1983. Protection against heat-induced cell killing by polyols in vitro. Cancer Res. 43:1624-1627.

Herrmann,H. and U Aebi. 2000. Intermediate filaments and their associates: multi-talented structural elements specifying cytoarchitecture and cytodynamics. Curr. Opin. Cell Biol. 12:79-90.

Heymann,J.A. and S.Subramaniam. 1997. Expression, stability, and membrane integration of truncation mutants of bovine rhodopsin. Proc. Natl. Acad. Sci. U. S. A 94:4966-4971.

Hiller,M.M., A.Finger, M.Schweiger, and D.H.Wolf. 1996. ER degradation of a misfolded luminal protein by the cytosolic ubiquitin-proteasome pathway. Science 273:1725-1728.

Hitchcock,A.L., H.Krebber, S.Frietze, A.Lin, M.Latterich, and P.ASilver. 2001. The conserved npl4 protein complex mediates proteasome-dependent membrane-bound transcription factor activation. Mol. Biol. Cell 12:3226-3241.

Ho,S.N., H.D.Hunt, R.M.Horton, J.K.Pullen, and L.R.Pease. 1989. Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77:51-59.

Hodges,R.S., R.J.Heaton, J.M.Parker, L.Molday, and R.S.Molday. 1988. Antigen-antibody interaction. Synthetic peptides define linear antigenic determinants recognized by monoclonal antibodies directed to the cytoplasmic carboxyl terminus of rhodopsin. J. Biol. Chem. 263:11768-11775.

Holmberg,M., C.Duyckaerts, A.Durr, G.Cancel, I.Gourfinkel-An, P.Damier, B.Faucheux, Y.Trottier, E.G.Hirsch, Y.Agid, and A.Brice. 1998. Spinocerebellar ataxia type 7 (SCA7): a neurodegenerative disorder with neuronal intranuclear inclusions. Hum. Mol. Genet. 7:913-918.

Horn,M., P.Humphries, M.Kunisch, C.Marchese, E.Apfelstedt-Sylla, L.Fugi, E.Zrenner, P.Kenna, A.Gal, and J.Farrar. 1992. Deletions in exon 5 of the human rhodopsin gene

169 causing a shift in the reading frame and autosomal dominant retinitis pigmentosa.Hum. Genet. 90:255-257.

Hosokawa,N., I.Wada, K.Hasegawa, T.Yorihuzi, L.O.Tremblay, A.Herscovics, and K.Nagata. 2001. A novel ER alpha-mannosidase-like protein accelerates ER-associated degradation. EMBO Rep. 2:415-422.

Huang,P.C., A.E.Gaitan, Y.Hao, R.M.Petters, and F.Wong. 1993. Cellular interactions implicated in the mechanism of photoreceptor degeneration in transgenic mice expressing a mutant rhodopsin gene. Proc. Natl. Acad. Soi. U. S. A 90:8484-8488.

Huppa.J.B. and H.L.PIoegh. 1998. The eS-Sence of-SH in the ER. Ce//92:145-148.

Hurtley.S.M. and A Helenius. 1989. Protein oligomerization in the endoplasmic reticulum. Anna. Rev. Cell Biol. 5:277-307. llling.M.E., R.S.Rajan, N.F.Bence, and R.R.Kopito. 2002. A rhodopsin mutant linked to autosomal dominant retinitis pigmentosa is prone to aggregate and interacts with the ubiquitin proteasome system. J. Biol. Chem.

Imperiali.B. and S.E.O'Connor. 1999. Effect of N-linked glycosylation on glycopeptide and glycoprotein structure. Curr. Opin. Chem. Biol. 3:643-649.

Inglehearn.C.F., R.Bashir, D.H.Lester, M.Jay, A.C.Bird, and S.S.Bhattacharya. 1991. A 3- bp deletion in the rhodopsin gene in a family with autosomal dominant retinitis pigmentosa. Am. J. Hum. Genet. 48:26-30. lnglehearn,C.F., T.J.Keen, R.Bashir, M.Jay, F.Fitzke, AC.Bird, ACrombie, and S.Bhattacharya. 1992. A completed screen for mutations of the rhodopsin gene in a panel of patients with autosomal dominant retinitis pigmentosa.Hum. Mol. Genet. 1:41-45. lshikawa,K., M.Watanabe, K.Yoshizawa, T.Fujita, H.Iwamoto, T.Yoshizawa, K.Harada, K.Nakamagoe, Y.Komatsuzaki, A.Satoh, M.Doi, T.Ogata, I.Kanazawa, S.Shoji, and H.Mizusawa. 1999. Clinical, neuropathological, and molecular study in two families with spinocerebellar ataxia type 6 (SCA6). J. Neurol. Neurosurg. Psychiatry 67:86-89. lsono,K., T.Tanimura, Y.Oda, and Y.Tsukahara. 1988. Dependency on light and vitamin A derivatives of the biogenesis of 3- hydroxyretinal and visual pigment in the compound eyes of Drosophila melanogaster.J. Gen. Physiol 92:587-600.

170 Jakob,C.A., P.Burda, J.Roth, and M.Aebi. 1998. Degradation of misfolded endoplasmic reticulum glycoproteins in Saccharomyces cerevisiae is determined by a specific oligosaccharide structure. J. Cell Biol. 142:1223-1233.

Jana NR, Tanaka M, Wang G, Nukina N. 2000. Polyglutamine length-dependent interaction of Hsp40 and Hsp70 family chaperones with truncated N-terminal huntingtin: their role in suppression of aggregation and cellular toxicity. Hum Mol Genet 12:2009-18

Jarosch,E., R.Geiss-Friedlander, B.Meusser, J.Walter, and T.Sommer. 2002. Protein dislocation from the endoplasmic reticu. Traffic. 3:530-536.

Jarosch,E., C.Taxis, C.Volkwein, J.Bordallo, D.Finley, D.H.Wolf, and T.Sommer. 2002. Protein dislocation from the ER requires polyubiquitination and the AAA- ATPase Cdc48. Nat. Ce//B/o/. 4:134-139.

Jarrett,J.T. and P.T.Lansbury, Jr. 1992. Amyloid fibril formation requires a chemically discriminating nucléation event: studies of an amyloidogenic sequence from the bacterial protein OsmB. Biochemistry 31:12345-12352.

Joshi H.C. 1993. y-Tubulin: The Hub of Cellular Microtubule Assemblies. BioEssays 15: 637-643

Johnson,A.E. and N.G.Haigh. 2000. The ER translocon and retrotranslocation: is the shift into reverse manual or automatic? Ce//102:709-712.

Johnston,J.A., C.L.Ward, and R.R.Kopito. 1998. Aggresomes: a cellular response to misfolded proteins. J. Cell Biol. 143:1883-1898.

Jorgensen,M.M., O.N.Jensen, H.U.Holst, J.J.Hansen, T.J.Corydon, P.Bross, L.Bolund, and N.Gregersen. 2000. Grp78 is involved in retention of mutant low density lipoprotein receptor protein in the endoplasmic reticulum. J. Biol. Chem. 275:33861-33868.

Kabore,A.F., W.J.Wang, S.J.Russo, and M.F.Beers. 2001. Biosynthesis of surfactant protein 0: characterization of aggresome formation by EGFP chimeras containing propeptide mutants lacking conserved cysteine residues. J. Cell Sci. 114:293-302.

Kagan,A., Z.Yu, G.I.Fishman, and T.V.McDonald. 2000. The dominant negative LQT2 mutation A561V reduces wild-type HERG expression. J. Biol. Chem. 275:11241-11248.

171 Kajiwara.K., L.B.Hahn, S.Mukai, G.H.Travis, E.L.Berson, and T.P.Dryja. 1991. Mutations in the human retinal degeneration slow gene in autosomal dominant retinitis pigmentosa. Nature 354:480-483.

Kajiwara.K., E.L.Berson, and T.P.Dryja. 1994. Digenic retinitis pigmentosa due to mutations at the unlinked peripherin/RDS and R0M1 loci. Science 264:1604-1608.

Kamhi-Nesher,S., M.Shenkman, S.Tolchinsky, S.V.Fromm, R.Ehrlich, and G.Z.Lederkremer. 2001. A novel quality control compartment derived from the endoplasmic reticulum. Mol. Biol. Ce//12:1711-1723.

Karnik.S.S. and H.G.Khorana. 1990. Assembly of functional rhodopsin requires a disulfide bond between cysteine residues 110 and 187. J. Biol. Chem. 265:17520-17524.

Katanosaka.K., F.Tokunaga, S.Kawamura, and K.Ozaki. 1998. N-linked glycosylation of Drosophila rhodopsin occurs exclusively in the amino-terminal domain and functions in rhodopsin maturation. FEBS Lett. 424:149-154.

Kaufman,R.J. 1999. Stress signaling from the lumen of the endoplasmic reticulum: coordination of gene transcriptional and translational controls. Genes Dev. 13:1211-1233.

Kaufman,R.J., D.Scheuner, M.Schroder, X.Shen, K.Lee, C.Y.Liu, and S.M.Arnold. 2002. The unfolded protein response in nutrient sensing and differentiation.Nat. Rev. Mol. Cell Biol. 3:411-421.

Kaushal,S., K.D.Ridge, and H.G.Khorana. 1994. Structure and function in rhodopsin: the role of asparagine-linked glycosylation. Proc. Natl. Acad. Sci. U. S. A 91:4024-4028.

Kaushal,S. and H.G.Khorana. 1994. Structure and function in rhodopsin. 7. Point mutations associated with autosomal dominant retinitis pigmentosa. Biochemistry 33:6121-6128.

Kaytor,M.D. and S.T.Warren. 1999. Aberrant protein deposition and neurological disease. J. Biol. Chem. 274:37507-37510.

Keen,T.J., C.F.Inglehearn, D.H. Lester, R.Bashir, M.Jay, A.C.Bird, B.Jay, and S.S.Bhattacharya. 1991. Autosomal dominant retinitis pigmentosa: four new mutations in rhodopsin, one of them in the retinal attachment site. Genomics 11:199-205.

172 Kegel,K.B., M.Kim, E.Sapp, C.McIntyre, J.G.Castano, N.Aronin, and M.DiFiglia. 2000. Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation, and autophagy. J. Neurosci. 20:7268-7278.

Kemp,C.M., S.G.Jacobson, D.J.Faulkner, and R.W.Walt. 1988. Visual function and rhodopsin levels in humans with vitamin A deficiency. Exp. Eye Res. 46:185-197.

Kikkert,M., G.Hassink, M.Barel, C.Hirsch, F.J.Van der Wal, and E.Wiertz. 2001. Ubiquitination is essential for human cytomegalovirus US11-mediated dislocation of MHC class I molecules from the endoplasmic reticulum to the cytosol. Biochem. J. 358:369-377.

Kim,R.Y., M AI Maghtheh, F.W.Fitzke, G.B.Arden, M.Jay, S.S.Bhattacharya, and A.C.Bird. 1993. Dominant retinitis pigmentosa associated with two rhodopsin gene mutations. Leu- 40-Arg and an insertion disrupting the 5'-splice junction of exon 5. Arch. Ophthalmol. 111:1518-1524.

Kitada,T., S.Asakawa, N.Hattori, H.Matsumine, Y.Yamamura, S.Minoshima, M.Yokochi, Y.Mizuno, and N.Shimizu. 1998. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature 392:605-608.

Klumperman,J. 2000. Transport between ER and Golgi. Curr. Opin. Cell Biol. 12:445-449.

Kopito,R.R. 1997. ER quality control: the cytoplasmic connection. Ce//88:427-430.

Kopito,R.R. and D.Ron. 2000. Conformational disease. Nat. Cell Biol. 2:E207-E209.

Kouroku,Y., E.Fujita, A.Jimbo, T.Kikuchi, T.Yamagata, M.Y.Momoi, E.Kominami, K.Kuida, K.Sakamaki, S.Yonehara, and T.Momoi. 2002. Polyglutamine aggregates stimulate ER stress signals and caspase-12 activation. Hum. Mol. Genet. 11:1505-1515.

Kozutsumi,Y., M.Segal, K.Normington, M.J.Gething, and J.Sambrook. 1988. The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature 332:462-464.

Kranich,H., S.Bartkowski, M.J.Denton, S.Krey, P.Dickinson, C.Duvigneau, and A.Gal. 1993. Autosomal dominant 'sector' retinitis pigmentosa due to a point mutation predicting an Asn-15-Ser substitution of rhodopsin. Hum. Mol. Genet. 2:813-814.

173 KumaramanickaveI.G., M.Maw, M.J.Denton, S.John, C.R.Srikumari, U.Orth, R.Oehlmann, and A.Gal. 1994. Missense rhodopsin mutation in a family with recessive RP. Nat Genet.

8 :10- 11.

Kurada.P. and J.E.OTousa. 1995. Retinal degeneration caused by dominant rhodopsin mutations in Drosophila. Neuron 14:571-579.

Kurada.P., T.D.Tonini, M.A.Serikaku, J.P.Piccini, and J.E.OTousa. 1998. Rhodopsin maturation antagonized by dominant rhodopsin mutants. Vis. Neurosci. 15:693-700.

Laszlo.L., J.Tuckwell, T.Self, J.Lowe, M.Landon, S.Smith, J.N.Hawthorne, and R.J.Mayer. 1991. The latent membrane protein-1 in Epstein-Barr virus-transformed lymphoblastoid cells is found with ubiquitin-protein conjugates and heat-shock protein 70 in lysosomes oriented around the microtubule organizing centre. J. Pathol. 164:203-214.

LaVail, M. M. 1976 Rod outer segment disc shedding in rat retina: relationship to cyclic lighting. Science 194:1071-1074.

LaVail,M.M. 1980. Interaction of environmental light and eye pigmentation with inherited retinal degenerations. Vision Res. 20:1173-1177.

Lee,D.H. and A.L.Goldberg. 1996. Selective inhibitors of the proteasome-dependent and vacuolar pathways of protein degradation in Saccharomyces cerevisiae. J. Biol. Chem. 271:27280-27284.

Leroy,E., R.Boyer, G.Auburger, B.Leube, G.Ulm, E.Mezey, G.Harta, M.J.Brownstein, S.Jonnalagada, T.Chernova, A.Dehejia, C.Lavedan, T.Gasser, P.J.Steinbach, K.D.Wilkinson, and M.H.Polymeropoulos. 1998. The ubiquitin pathway in Parkinson's disease. Nature 395:451-452.

Li,G.C. and Z.Werb. 1982. Correlation between synthesis of heat shock proteins and development of thermotolerance in Chinese hamster fibroblasts. Proc. Natl. Acad. Sci. U. S. A 79:3218-3222.

Li,T., W.K.Franson, J.W.Gordon, E.L.Berson, and T.P.Dryja. 1995. Constitutive activation of phototransduction by K296E opsin is not a cause of photoreceptor degeneration. Proc. Natl. Acad. Sci. U. S. A 92:3551-3555.

174 Li,T., W.K.Snyder, J.E.OIsson, and T.P.Dryja. 1996. Transgenic mice carrying the dominant rhodopsin mutation P347S: evidence for defective vectorial transport of rhodopsin to the outer segments. Proc. Natl. Acad. Sci. U. S. A 93:14176-14181.

Li,T., M.A.Sandberg, B.S.Pawlyk, B.Rosner, K.C.Hayes, T.P.Dryja, and E.L.Berson. 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 95:11933-11938.

Li,Z.Y., F.Wong, J.H.Chang, D.E.Possin, Y.Hao, R.M.Petters, and A.M.Milam. 1998. Rhodopsin transgenic pigs as a model for human retinitis pigmentosa. Invest Ophthalmol. Vis. Sci. 39:808-819.

Liang,C.J., K.Yamashita, C.G.Muellenberg, H.Shichi, and A.Kobata. 1979. Structure of the carbohydrate moieties of bovine rhodopsin. J. Biol. Chem. 254:6414-6418.

Lin,P.S., L.Kwock, and K.Hefter. 1981. Protection of heat induced cytoxicity by glycerol. J. Cell Phys/o/108:439-443.

Lippincott-Schwartz,J., L.C.Yuan, J.S.Bonifacino, and R.D.KIausner . 1989. Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: evidence for membrane cycling from Golgi to ER. Cell 56:801-813.

Lippincott-Schwartz,J. 1998. Cytoskeletal proteins and Golgi dynamics. Curr. Cpin. Cell Biol. 10:52-59.

Liu,Y., P.Choudhury, C.M.Cabral, and R.N.Sifers. 1997. Intracellular disposal of incompletely folded human alpha 1-antitrypsin involves release from calnexin and post- translational trimming of asparagine-linked oligosaccharides. J. Biol. Chem. 272:7946- 7951.

Liu,Y., P.Choudhury, C.M.Cabral, and R.N.Sifers. 1999. Oligosaccharide modification in the early secretory pathway directs the selection of a misfolded glycoprotein for degradation by the proteasome .J. Biol. Chem. 274:5861-5867.

Luethy,J.D., J.Fargnoli, J.S.Park, A.J.Fornace, Jr., and N.J.Holbrook. 1990. Isolation and characterization of the hamster gadd153 gene. Activation of promoter activity by agents that damage DNA. J. Biol. Chem. 265:16521-16526.

175 Machida, S., M. Kondo, J.A Jamison, N.W. Khan, L.T. Kononen, T. Sugawara, R. A. Bush, and P. A. Sieving. 2000. P23H rhodopsin transgenic rat: correlation of retinal function with histopathology. Invest. Ophthalmol. Vis. Sci. 41:3200-3209.

Macke,J.P., C.M.Davenport, S.G.Jacobson, J.C.Hennessey, F.Gonzalez-Fernandez, B.P.Conway, J.Heckenlively, R.Palmer, I.H.Maumenee, P.Sieving. 1993. Identification of novel rhodopsin mutations responsible for retinitis pigmentosa: implications for the structure and function of rhodopsin. Am. J. Hum. Genet. 53:80-89.

Macke,J.P., J.C.Hennessey, and J.Nathans. 1995. Rhodopsin mutation proline347-to- alanine in a family with autosomal dominant retinitis pigmentosa indicates an important role for proline at position 347. Hum. Mol. Genet. 4:775-776.

Mayer,R.J., J.Lowe, G.Lennox, F.Doherty, and M.Landon. 1989. Intermediate filaments and ubiquitin: a new thread in the understanding of chronic neurodegenerative diseases. Prog. Clin. Biol. Res. 317:809-818.

Mayer,T.U., T.Braun, and S.Jentsch. 1998. Role of the proteasome in membrane extraction of a short-lived ER- transmembrane protein. EMBO J. 17:3251-3257.

McCullough,K.D., J.L.Martindale, L.O.KIotz, T.Y.Aw, and N.J.Holbrook. 2001. Gaddi53 sensitizes cells to endoplasmic reticulum stress by down- regulating Bcl2 and perturbing the cellular redox state. Mol. Cell Biol. 21:1249-1259.

McKie,A.B., J.C.McHale, T.J.Keen, E.E.Tarttelin, R.Goliath, J.J.Lith-Verhoeven, J.Greenberg, R.S.Ramesar, C.B.Hoyng, F.P.Cremers, D.A.Mackey, S.S.Bhattacharya, AC.Bird, A.F.Markham, and C.F.Inglehearn. 2001. Mutations in the pre-mRNA splicing factor gene PRPC8 in autosomal dominant retinitis pigmentosa (RP13). Hum. Mol. Genet. 10:1555-1562.

McWilliam,P., G.J.Farrar, P.Kenna, D.G.Bradley, M.M.Humphries, E.M.Sharp, D.J.McConnell, M.Lawler, D.Sheils, C.Ryan, and . 1989. Autosomal dominant retinitis pigmentosa (ADRP): localization of an ADRP gene to the long arm of chromosome 3. Genomics 5:619-622.

Meindl,A., K.Dry, K.Herrmann, F.Manson, A.Ciccodicola, A.Edgar, M.R.Carvalho, H.Achatz, H.Hellebrand, A.Lennon, C.Migliaccio, K.Porter, E.Zrenner, A.Bird, M.Jay, B.Lorenz, B.Wittwer, M.D'Urso, T.Meitinger, and A.Wright. 1996. A gene (RPGR) with

176 homology to the RCC1 guanine nucleotide exchange factor is mutated in X-linked retinitis pigmentosa (RP3). A/af. Genef. 13:35-42.

Meng.E.C. and H R.Bourne. 2001. Receptor activation: what does the rhodopsin structure tell us? Trends Pharmacol. Sci. 22:587-593.

Menon.S.T., M.Han, and T.P.Sakmar. 2001. Rhodopsin: structural basis of molecular physiology. Physiol Rev. 81:1659-1688.

Meyer,H.H., J.G.Shorter, J.Seemann, D.Pappin, and G.Warren. 2000. A complex of mammalian ufdl and npl4 links the AAA-ATPase, p97, to ubiquitin and nuclear transport pathways. EMBOJ. 19:2181-2192.

Miles, M.F., J.E. Diaz and V S. DeGuzman. 1991. Mechanisms of Neuronal Adaptation to Ethanol. J. Biol. Chem. 266:2409-2414

Miroy,G.J., Z.Lai, H.A.Lashuel, S.A.Peterson, C.Strang, and J.W.Kelly. 1996. Inhibiting transthyretin amyloid fibril formation via protein stabilization.Proc. Natl. Acad. Sci. U. S. A 93:15051-15056.

Mitton,K.P., P.K.Swain, S.Chen, S.Xu, D.J.Zack, and A.Swaroop. 2000. The leucine zipper of NRL interacts with the CRX homeodomain. A possible mechanism of transcriptional synergy in rhodopsin regulation. J. Bid. Chem. 275:29794-29799.

Molinari,M. and A Helenius. 2000. Chaperone selection during glycoprotein translocation into the endoplasmic reticulum. Science 288:331-333.

Morishima-Kawashima,M., M.Hasegawa , K.Takio, M.Suzuki, K.Titani, and Y.lhara. 1993. Ubiquitin is conjugated with amino-terminally processed tau in paired helical filaments. Neuron 10:1151-1160.

Muchowski,P.J., K.Ning, C.D'Souza-Schorey, and S.Fields. 2002. Requirement of an intact microtubule cytoskeleton for aggregation and inclusion body formation by a mutant huntingtin fragment. Proc. Natl. Acad. Sci. U. S. A 99:727-732.

Muchowski,P.J. 2002. Protein misfolding, amyloid formation, and neurodegeneration. A critical role for molecular chaperones? Neuron 35:9-12.

177 Nakagawa.T., H.Zhu, N.Morishima, E.Li, J.Xu, BAYankner, and J.Yuan. 2000. Caspase- 12 mediates endoplasmic-reticulum-specific apoptosis and cytotoxicity by amylojd-beta. A/afure 403:98-103.

Nakagawa.T. and J.Yuan. 2000. Cross-talk between two cysteine protease families. Activation of caspase-12 by calpain in apoptosis. J. Cell Biol. 150:887-894.

Nathans,J. 1990. Determinants of visual pigment absorbance: identification of the retinylidene Schiffs base counterion in bovine rhodopsin. Biochemistry 29:9746-9752.

Noell W.K. and Albrecht, R. 1971. Irreversible effects of visible light on the retina: role of vitamin A. Science 172: 76-9.

Nucifora.F.C., Jr., M.Sasaki, M.F.Peters, H.Huang, J.K.Cooper, M.Yamada, H.Takahashi, S.Tsuji, J.Troncoso , V.L.Dawson, T.M.Dawson, and C.A.Ross. 2001. Interference by huntingtin and atrophin-1 with cbp-mediated transcription leading to cellular toxicity. Sc/ence 291:2423-2428.

OTousa.J.E. 1992. Requirement of N-linked glycosylation site in Drosophila rhodopsin. Vis. Neurosci. 8:385-390.

Oh.K.T., R.G.Weleber, A.Lotery, D.M.Oh, A.M.Billingslea, and E.M.Stone. 2000. Description of a new mutation in rhodopsin, Pro23Ala, and comparison with electroretinographic and clinical characteristics of the Pro23His mutation. Arch. Ophthalmol. 118:1269-1276.

Oliver,J.D., F.J.Van der Wal, N.J.Bulleid, and S.High. 1997. Interaction of the thiol- dependent reductase ERp57 with nascent glycoproteins. Science 275:86-88.

Olsson,J.E., J.W.Gordon, B.S.Pawlyk, D.Roof, A. Hayes, R.S.Molday, S.Mukai, G.S.Cowley, E.L.Berson, and T.P.Dryja. 1992. Transgenic mice with a rhodopsin mutation (Pro23His): a mouse model of autosomal dominant retinitis pigmentosa. Neuron 9:815- 830.

Oprian,D.D., R.S.Molday, R.J.Kaufman, and H.G.Khorana. 1987. Expression of a synthetic bovine rhodopsin gene in monkey kidney cells. Proc. Natl. Acad. Sci. U. S. A 84:8874-8878.

178 Oseroff.A.R. and R.H.Callender. 1974. Resonance Raman spectroscopy of rhodopsin in retinal disk membranes. S/oc/?em/sfAy 13:4243-4248.

Ovchinnikov,Y., N.G.Abdulaev, and A.S.Bogachuk. 1988. Two adjacent cysteine residues in the C-terminal cytoplasmic fragment of bovine rhodopsin are palmitylated. FEBS Lett. 230 :1-5.

Ozaki.K., H.Nagatani, M.Ozaki, and F.Tokunaga. 1993. Maturation of major Drosophila rhodopsin, ninaE, requires chromophore 3- hydroxyretinal. Neuron 10:1113-1119.

Palczewski,K., T.Kumasaka, T.Hori, CA.Behnke, H.Motoshima, B.A.Fox, T.Le, I, D.C.Teller, T.Okada, R.E.Stenkamp, M.Yamamoto, and M.Miyano. 2000. Crystal structure of rhodopsin: A G protein-coupled receptor. Science 289:739-745.

Palmer,A., A.J.Rivett, S.Thomson, K.B.Hendil, G.W.Butcher, G.Fuertes, and E.Knecht. 1996. Subpopulations of proteasomes in rat liver nuclei, microsomes and cytosol. Biochem. J. 316 ( Pt 2):401-407.

Papac,D.I., K.R.Thornburg, E.E.Bullesbach, R.K.Crouch, and DR. Knapp. 1992. Palmitylation of a G-protein coupled receptor. Direct analysis by tandem mass spectrometry. J. Biol. Chem. 267:16889-16894.

Parodi,A.J. 2000. Protein glucosylation and its role in protein folding. Annu. Rev. Biochem. 69:69-93.

Patil,C. and P.Walter. 2001. Intracellular signaling from the endoplasmic reticulum to the nucleus: the unfolded protein response in yeast and mammals. Curr. Opin. Cell Biol. 13:349-355.

Paulson,H.L., M.K.Perez, Y.Trottier, J.Q.Trojanowski, S.H.Subramony, S.S.Das, P.Vig, J.L.Mandel, K H.Fischbeck, and R.N.Pittman. 1997. Intranuclear inclusions of expanded polyglutamine protein in spinocerebellar ataxia type 3. Neuron 19:333-344.

Paulus,W., I.Baur, F.M.Boyce, X.O.Breakefield, and S A.Reeves. 1996. Self-contained, tetracycline-regulated retroviral vector system for gene delivery to mammalian cells. J. Virol. 70:62-67.

179 Perutz.M.F., T Johnson, M.Suzuki, and J.T.FInch. 1994, Glutamine repeats as polar zippers: their possible role in inherited neurodegenerative diseases.Proc. Natl. Acad. Sci. U. S. >A 91:5355-5358.

Petaja-Repo,U.E., M.Hogue, A.Laperriere, S.Bhalla, P.Walker, and M.Bouvier. 2001. Newly synthesized human delta opioid receptors retained in the endoplasmic reticulum are retrotranslocated to the cytosol, deglycosylated, ubiquitinated, and degraded by the proteasome. J. Biol. Chem. 276:4416-4423.

Petaja-Repo,U.E., M.Hogue, S.Bhalla, A.Laperriere, J.P.Morello, and M.Bouvier. 2002. Ligands act as pharmacological chaperones and increase the efficiency of delta opioid receptor maturation. EMBOJ. 21:1628-1637.

Peterson,J.R., A.Ora, P.N.Van, and A.Helenius. 1995. Transient, lectin-like association of calreticulin with folding intermediates of cellular and viral glycoproteins. Mol. Biol. Cell 6:1173-1184.

Petters,R.M., C.A.AIexander, K.D.Wells, E.B.Collins, J.R.Sommer, M.R.BIanton, G.Rojas, Y.Hao, W.L.Flowers, E.Banin, A.V.Cideciyan, S.G.Jacobson, and F.Wong. 1997. Genetically engineered large animal model for studying cone photoreceptor survival and degeneration in retinitis pigmentosa. Nat. Biotechnol. 15:965-970.

Plantner,J.J., L.Poncz, and E.L.Kean. 1980. Effect of tunicamycin on the glycosylation of rhodopsin. Arch. Biochem. Biophys. 201:527-532.

Plemper,R.K. and D.H.Wolf. 1999. Retrograde protein translocation: ERADication of secretory proteins in health and disease. Trends Biochem. Sci. 24:266-270.

Polymeropoulos,M.H., C.Lavedan, E.Leroy, S.E.Ide, A.Dehejia, A.Dutra, B.Pike, H.Root, J.Rubenstein, R. Boyer, E.S.Stenroos, S.Chandrasekharappa, A.Athanassiadou, T.Papapetropoulos, W.G.Johnson, A.M.Lazzarini, R.C.Duvoisin, G.Di lorio, L.I.Golbe, and R.L.Nussbaum. 1997. Mutation in the alpha-synuclein gene identified in families with Parkinson's disease. Science 276:2045-2047.

Portera-Cailliau,C., C.H.Sung, J.Nathans, and R Adler. 1994. Apoptotic photoreceptor cell death in mouse models of retinitis pigmentosa. Proc. Natl. Acad. Sci. U. S. A 91:974-978.

180 Rabinovich,E., A Kerem, K.U.Frohlich, N.Diamant, and S.Bar-Nun. 2002. AAA-ATPase p97/Cdc48p, a cytosolic chaperone required for endoplasmic reticulum-associated protein degradation. Mol. Cell Biol. 22:626-634.

Rao,V.R., G.B.Cohen, and D.D.Oprian. 1994. Rhodopsin mutation G90D and a molecular mechanism for congenital night blindness. Nature 367:639-642.

Rapp,L.M. and T.P.Williams. 1980. The role of ocular pigmentation in protecting against retinal light damage. Vision Res. 20:1127-1131.

Ravikumar,B., R.Duden, and D.C.Rubinsztein. 2002. Aggregate-prone proteins with polyglutamine and polyalanine expansions are degraded by autophagy Hum.. Mol. Genet. 11:1107-1117.

Reig,C., J.Antich, E.Gean, B.Garcia-Sandoval, C.Ramos, C.Ayuso, and M.Carballo. 1994. Identification of a novel rhodopsin mutation (Met-44-Thr) in a simplex case of retinitis pigmentosa. Hum. Genet. 94:283-286.

Reig,C., A.I.Alvarez, I.Tejada, M.Molina, E.Arostegui, R.Martin, J.Antich, and M.Carballo. 1996. New mutation in the 3'-acceptor splice site of intron 4 in the rhodopsin gene associated with autosomal dominant retinitis pigmentosa in a Basque family.Hum. Mutat. 8:93-94.

Restagno,G., M.Maghtheh, S.Bhattacharya, M.Ferrone, S.Garnerone, R.Samuelly, and A.Carbonara. 1993. A large deletion at the 3' end of the rhodopsin gene in an Italian family with a diffuse form of autosomal dominant retinitis pigmentosa. Hum. Mol. Genet. 2:207- 208.

Richards,J.E., K M.Scott, and P.A.Sieving. 1995. Disruption of conserved rhodopsin disulfide bond by Cys187Tyr mutation causes early and severe autosomal dominant retinitis pigmentosa. Ophthalmology 102:669-677.

Rim,J. and D.D.Oprian. 1995. Constitutive activation of opsin: interaction of mutants with rhodopsin kinase and arrestin. S/ocf7em/sf/y 34:11938-11945.

Robinson,P.R., G.B.Cohen, E.A.Zhukovsky, and D.D.Oprian. 1992. Constitutively active mutants of rhodopsin. Neuron 9:719-725.

181 Rock,K L, C.Gramm, L.Rothstein, K CIark, R.Stein, L.Dick, D.Hwang, and A L.Goldberg. 1994. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78:761-771.

Rodriguez,J.A., C.A.Herrera, D.G.Birch, and S.P.Daiger. 1993. A leucine to arginine amino acid substitution at codon 46 of rhodopsin is responsible for a severe form of autosomal dominant retinitis pigmentosa. Hum. Mutat. 2:205-213.

Roof,D.J., M.Adamian, and A.Hayes. 1994. Rhodopsin accumulation at abnormal sites in retinas of mice with a human P23H rhodopsin transgene. Invest Ophthalmol. Vis. Sci. 35:4049-4062.

Rosen,D R., T.Siddique, D.Patterson, D.A.Figlewicz, P.Sapp, A.Hentati, D.Donaldson, J.Goto, J.P.O'Regan, H.X.Deng, and . 1993. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362:59-62.

Rosenfeld,P.J., G.S.Cowley, T.L.McGee, M.A.Sandberg, E.L.Berson, and T.P.Dryja. 1992. A null mutation in the rhodopsin gene causes rod photoreceptor dysfunction and autosomal recessive retinitis pigmentosa. Nat. Genet. 1:209-213.

Sakmar,T.P., R.R.Franke, and H.G.Khorana. 1989. Glutamic acid-113 serves as the retinylidene Schiff base counterion in bovine rhodopsin. Proc. Natl. Acad. Sci. U. S. A 86:8309-8313.

Saliba,R.S., P.M.Munro, P.J.Luthert, and M.E.Cheetham. 2002. The cellular fate of mutant rhodopsin: quality control, degradation and aggresome formation. J. Cell Sci. 115:2907- 2918.

Sanchez,B., S.Borreo, P.Chaparro, T.Rueda, F.Lopez, and G.Antinolo. 1996. A novel null mutation in the rhodopsin gene causing late onset autosomal dominant retinitis pigmentosa. Hum. Mutat. 7:180.

Sato,S., C.L.Ward, M.E.Krouse, J.J.Wine, and R.R.Kopito. 1996. Glycerol reverses the misfolding phenotype of the most common cystic fibrosis mutation. J. Biol. Chem. 271:635-638.

Sawano,H., Y.Koumoto, K.Ohta, Y.Sasaki, S.Segawa, and H.Tachibana. 1992. Efficient in vitro folding of the three-disulfide derivatives of hen lysozyme in the presence of glycerol. FEBS Lett. 303:11-14.

182 Schwahn.U., S.Lenzner, J.Dong, S.Fell, B.Hinzmann, G.van Duijnhoven, R.Kirschner, M.Hemberger, A.A.Bergen, T.Rosenberg, AJ.Pinckers , R.Fundele, A.Rosenthal, F.P.Cremers, H.H.Ropers, and W.Berger. 1998. Positional cloning of the gene for X- linked retinitis pigmentosa 2. Nat. Genet 19:327-332.

Selkoe,D.J. 1994. Cell biology of the amyloid beta-protein precursor and the mechanism of Alzheimer's disease. Annu. Rev. Cell Biol. 10:373-403.

Shamu,C.E., D.FIierman, H.L.PIoegh, T.A.Rapoport, and V.Chau. 2001. Polyubiquitination is required for US11-dependent movement of MHC class I heavy chain from endoplasmic reticulum into cytosol. Mol. Biol. Ce//12:2546-2555.

Sheffield,V.C., G.A.Fishman, J.S.Beck, A.E.Kimura, and E.M.Stone. 1991. Identification of novel rhodopsin mutations associated with retinitis pigmentosa by GC-clamped denaturing gradient gel electrophoresis. Am. J. Hum. Genet. 49:699-706.

Shelanski,M.L., F.Gaskin, and C.R.Cantor. 1973. Microtubule assembly in the absence of added nucleotides. Proc. Natl. Acad. Sci. U. S. A 70:765-768.

Sherman,M.Y. and AL.Goldberg. 2001. Cellular defenses against unfolded proteins: a cell biologist thinks about neurodegenerative diseases.Neuron 29:15-32.

Shiu,R.P. and I.H.Pastan. 1979. Properties and purification of a glucose-regulated protein from chick embryo fibroblasts. Biochim. Biophys. Acta 576:141-150.

Siebert,P.D. and J.W.Larrick. 1992. Competitive PCR. Nature 359:557-558.

Sieving,P.A, J.E. Richards, E.L. Bingham, and F. Naarendorp. 1992. Dominant congenital complete nyctalopia and Gly-90-Asp rhodopsin mutation. Invest. Ophthalmol. Vis. Sci. (Suppl.) 33: 1397

Sieving,P.A, J.E.Richards, F.Naarendorp, E.L.Bingham, K. Scott, and M. AI pern. 1995. Dark-light: model for nightblindness from the human rhodopsin Gly-90- >Asp mutation. Proc. Natl. Acad. Sci. U. S. A 92:880-884.

Sisodia,S.S. 1998. Nuclear inclusions in glutamine repeat disorders: are they pernicious, coincidental, or beneficial? Ce//95:1-4.

183 Skinner,P.J., B.T.Koshy, C.J.Cummings, lAKIement, K.Helin, AServadio, H.Y.Zoghbi, and H.T.Orr. 1997. Ataxin-1 with an expanded glutamine tract alters nuclear matrix- associated structures. Nature 389:971-974.

Sohocki,M.M., S.P.Daiger, S.J.Bowne, J.A.Rodriquez, H.Northrup, J.R.Heckenlively, D.G.Birch, H.Mintz-Hittner, R.S.Ruiz, R.A.Lewis, D.ASaperstein, and L.S.Sullivan. 2001. Prevalence of mutations causing retinitis pigmentosa and other inherited retinopathies. Hum. Mutat 17:42-51.

Sorger,P.K. and H.R.Pelham. 1987. The glucose-regulated protein grp94 is related to heat shock protein hsp90. J. Mol. Biol. 194:341-344.

Souied,E., S.Gerber, J.M.Rozet, D.Bonneau, J.L.Dufier, I.Ghazi, N.Philip, G.Soubrane, G.Coscas, A.Munnich, and . 1994. Five novel missense mutations of the rhodopsin gene in autosomal dominant retinitis pigmentosa. Hum. Mol. Genet. 3:1433-1434.

Speed,M.A., D.I.Wang, and J.King. 1996. Specific aggregation of partially folded polypeptide chains: the molecular basis of inclusion body composition. Nat. Biotechnol. 14:1283-1287.

St Jules,R.S., J.C.Wallingford, S.B.Smith, and P.J.O'Brien. 1989. Addition of the chromophore to rat rhodopsin is an early post- translational event. Exp. Eye Res. 48:653- 665.

Stamnes,M.A., B.H.Shieh, L.Chuman, G.L.Harris, and C.S.Zuker. 1991. The cyclophilin homolog ninaA is a tissue-specific integral membrane protein required for the proper synthesis of a subset of Drosophila rhodopsins. Cell 65:219-227.

Steffan,J.S., A. Kazantsev, O.Spasic-Boskovic, M.Greenwald, Y.Z.Zhu, H.Gohler, E.E.Wanker, G.P.Bates, D.E.Housman, and L.M.Thompson. 2000. The Huntington's disease protein interacts with p53 and CREB-binding protein and represses transcription. Proc. Natl. Acad. Sci. U. S. A 97:6763-6768.

Steinberg, R. H., Fisher, S. K. and Anderson, D. H. 1980. Disc morphogenesis in vertebrate photoreceptors. J. Comp. Neurol. 190:501-518.

Stryer L., 1988. Biochemistry. Freeman, New York.

184 Su,K., T.Stoller, J.Rocco, J.Zemsky, and R.Green. 1993. Pre-Golgi degradation of yeast prepro-alpha-factor expressed in a mammalian cell. Influence of cell type-specific oligosaccharide processing on intracellular fate. J. Biol. Chem. 268:14301-14309.

Sung.C.H., C.M. Davenport, J.C.Hennessey, I.H.Maumenee, S.G.Jacobson, J.R.Heckenlively, R.Nowakowski, G.Fishman, P.Gouras, and J.Nathans. 1991a. Rhodopsin mutations in autosomal dominant retinitis pigmentosa. Proc. Natl. Acad. Sci. U. S. A 88:6481-6485.

Sung,CH., B.G.Schneider, N.Agarwal, D.S.Papermaster, and J.Nathans. 1991b. Functional heterogeneity of mutant rhodopsins responsible for autosomal dominant retinitis pigmentosa. Proc. Natl. Acad. Sci. U. S. A 88:8840-8844.

Sung,C.H., C.M.Davenport, and J.Nathans. 1993. Rhodopsin mutations responsible for autosomal dominant retinitis pigmentosa. Clustering of functional classes along the polypeptide chain. J. Biol. Chem. 268:26645-26649.

Sung,C.H., C.Makino, D.Baylor, and J.Nathans. 1994. A rhodopsin gene mutation responsible for autosomal dominant retinitis pigmentosa results in a protein that is defective in localization to the photoreceptor outer segment. J. Neurosci. 14:5818-5833.

Suzuki,!., K.Kitajima, Y.Emori, Y.Inoue, and S.Inoue. 1997. Site-specific de-N- glycosylation of diglycosylated ovalbumin in hen oviduct by endogenous peptide: N- glycanase as a quality control system for newly synthesized proteins. Proc. Natl. Acad. Sci. U. S. A 94:6244-6249.

Suzuki,!., H.Park, K.Kitajima, and W.J.Lennarz. 1998. Peptides glycosylated in the endoplasmic reticulum of yeast are subsequently deglycosylated by a soluble peptide: N- glycanase activity. J. Biol. Chem. 273:21526-21530.

!ai,A.W., J.Z.Chuang, C.Bode, U.Wolfrum, and C.H.Sung. 1999. Rhodopsin's carboxy- terminal cytoplasmic tail acts as a membrane receptor for cytoplasmic dynein by binding to the dynein light chain !ctex-1. Cell 97:877-887.

!am,B.M., O.L.Moritz, L.B.Hurd, and D.S.Papermaster. 2000. Identification of an outer segment targeting signal in the COOH terminus of rhodopsin using transgenic Xenopus laevis. J. Cell Biol. 151:1369-1380.

185 Tanaka,K., T.Suzuki, and T.Chiba. 1998. The ligation systems for ubiquitin and ubiquitin- like proteins. Mo/. Ce//s 8:503-512.

Tanimura, T., Isono, K., and Tsukahara, Y. 1986. 3-Hydroxy-retinai as a chromophore of Drosophila melanogaster visual pigment analysed by high-pressure liquid chromatography. Photochem. Photobiol. 43, 225-228

Tokunaga,F., C.Brostrom, T.Koide, and P.Arvan. 2000. Endoplasmic reticulum (ER)- associated degradation of misfolded N-linked glycoproteins is suppressed upon inhibition of ER mannosidase I. J. Biol. Chem. 275:40757-40764.

Trojanowski,J.Q. and V.M.Lee. 2000. "Fatal attractions" of proteins. A comprehensive hypothetical mechanism underlying Alzheimer's disease and other neurodegenerative disorders. Ann. N. Y. Acad. Sci. 924:62-67.

Trombetta,E.S., J.F.Simons, and A.Helenius. 1996. Endoplasmic reticulum glucosidase II is composed of a catalytic subunit, conserved from yeast to mammals, and a tightly bound noncatalytic HDEL-containing subunit. J. Biol. Chem. 271:27509-27516.

Tsai,B., Y.Ye, and T.A.Rapoport. 2002. Retro-translocation of proteins from the endoplasmic reticulum into the cytosol. Nat. Rev. Mol. Cell Biol. 3:246-255.

Tso,M.O., W.W.Li, C.Zhang, T.T.Lam, Y.Hao, R.M.Petters, and F.Wong. 1997. A pathologic study of degeneration of the rod and cone populations of the rhodopsin Pro347Leu transgenic pigs. Trans. Am. Ophthalmol. Soc. 95:467-479.

Ulshafer,R.J., C.B.AIIen, and S.J.FIiesler. 1986. Tunicamycin-induced dysgenesis of retinal rod outer segment membranes. I. A scanning electron microscopy study. Invest Ophthalmol. Vis. Sci. 27:1587-1594.

Vaithinathan,R., E.L.Berson, and T.P.Dryja. 1994. Further screening of the rhodopsin gene in patients with autosomal dominant retinitis pigmentosa .Genomics 21:461-463.

Varki,A. 1993. Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3:97-130.

Verpy,E., M.Leibovici, I.Zwaenepoel, X.Z.Liu, A.Gal, N.Salem, A Mansour, S.BIanchard, I.Kobayashi, B.J.Keats, R.Slim, and C.Petit. 2000. A defect in harmonin, a PDZ domain-

186 containing protein expressed in the inner ear sensory hair cells, underlies Usher syndrome type 1C. Nat Genet 26:51-55.

Vidair.C.A., R.N.Huang, and S.J.Doxsey. 1996. Heat shock causes protein aggregation and reduced protein solubility at the centrosome and other cytoplasmic locations. Int. J. Hyperthermia 12:681-695.

Vithana,E.N., L.Abu-Safieh, M.J.AIIen, A.Carey, M.Papaioannou, C.Chakarova, M.AI Maghtheh, N.D.Ebenezer, C.Willis, A.T.Moore, A.C.Bird, D.M.Hunt, and S.S.Bhattacharya. 2001. A human homolog of yeast pre-mRNA splicing gene, PRP31, underlies autosomal dominant retinitis pigmentosa on chromosome 19q13.4 (RP11). Mol. Ce//8:375-381.

Waelter,S., A.Boeddrich, R.Lurz, E.Scherzinger, G.Lueder, H.Lehrach, and E.E.Wanker. 2001. Accumulation of mutant huntingtin fragments in aggresome-like inclusion bodies as a result of insufficient protein degradation. Mol. Biol. Ce//12:1393-1407.

Wald,G. 1968. Molecular basis of visual excitation. Science 162:230-239.

Wang,J.K., J.H.McDowell, and P.A.Hargrave. 1980. Site of attachment of 11-cis-retinal in bovine rhodopsin. Biochemistry 19:5111-5117.

Wang,X.Z., B.Lawson, J.W.Brewer, H.Zinszner, ASanjay, L.J.Mi, R.Boorstein , G.Kreibich, L.M.Hendershot, and D.Ron. 1996. Signals from the stressed endoplasmic reticulum induce C/EBP-homologous protein (CHOP/GADDI 53). Mol. Cell Biol. 16:4273- 4280.

Wang,X.Z., H.P.Harding, Y.Zhang, E.M.Jolicoeur, M.Kuroda, and D.Ron. 1998. Cloning of mammalian Irel reveals diversity in the ER stress responses.EMBO J. 17:5708-5717.

Ward,C.L., S.Omura, and R.R.Kopito. 1995. Degradation of CFTR by the ubiquitin- proteasome pathway. Ce//83:121-127.

Warrick,J.M., H.Y.Chan, G.L.Gray-Board, Y.Chai, H.L.Paulson, and N.M.Bonini. 1999. Suppression of polyglutamine-mediated neurodegeneration in Drosophila by the molecular chaperone HSP70. Nat. Genet. 23:425-428.

Webel,R., I.Menon, J.E.OTousa, and N.J.Colley. 2000. Role of asparagine-linked oligosaccharides in rhodopsin maturation and association with its molecular chaperone, NinaA. J. Biol. Chem. 275:24752-24759.

187 WeiI.D., S.BIanchard, J.Kaplan, P.Guilford, F.Gibson, J.Walsh, P.Mburu, A.Varela, J.Levilliers, M.D. Weston, and . 1995. Defective myosin VIIA gene responsible for Usher syndrome type 1B. Nature 374:60-61.

Weiss,E.R., Y.Hao, C D.Dickerson, S.Osawa, W.Shi, L.Zhang, and F.Wong. 1995. Altered cAMP levels in retinas from transgenic mice expressing a rhodopsin mutant. Biochem. Biophys. Res. Commun. 216:755-761.

Weissman,A.M. 2001. Themes and variations on ubiquitylation. Nat. Rev. Mol. Cell Biol. 2:169-178.

Welch,W.J. and C.R.Brown. 1996. Influence of molecular and chemical chaperones on protein folding. Cell Stress. Chaperones. 1:109-115.

Welch,W.J. and P.Gambetti. 1998. Chaperoning brain diseases. Nature 392:23-24.

Weng,S. and R.G.Spiro. 1997. Demonstration of a peptide:N-glycosidase in the endoplasmic reticulum of rat liver. Biochem. J. 322 ( Pt 2):655-661.

Wetzel,R. 1994. Mutations and off-pathway aggregation of proteins. Trends Biotechnol. 12:193-198.

Wetzel,R. 1996. For protein misassembly, it's the "I" decade. Ce//86:699-702.

Wiertz,E.J., T.R.Jones, L.Sun, M.Bogyo, H.J.Geuze, and H.L.PIoegh. 1996. The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84:769-779.

Wigley,W.C., R.P.Fabunmi, M.G.Lee, C.R.Marino, S.Muallem, G.N.DeMartino, and P.J.Thomas. 1999. Dynamic association of proteasomal machinery with the centrosome. J. Cell Biol. 145:481-490.

Wilson,C.M., M.R.Farmery, and N.J.Bulleid. 2000. Pivotal role of calnexin and mannose trimming in regulating the endoplasmic reticulum-associated degradation of major histocompatibility complex class I heavy chain. J. Biol. Chem. 275:21224-21232.

Wojcik,C., D.Schroeter, S.Wilk, J.Lamprecht, and N.Paweletz. 1996. Ubiquitin-mediated proteolysis centers in HeLa cells: indication from studies of an inhibitor of the chymotrypsin-like activity of the proteasome. Eur. J. Cell Biol. 71:311-318.

188 Wojcik.C. 1997. An inhibitor of the chymotrypsin-like activity of the proteasome (PS!) induces similar morphological changes in various cell lines. Folia Histochem. Cytobiol. 35:211-214.

Wu.T.H., T.D.Ting, T.I.Okajima, D.R.Pepperberg, Y.K.Ho, H.Ripps, and M.I.Naash. 1998. Opsin localization and rhodopsin photochemistry in a transgenic mouse model of retinitis pigmentosa. Neuroscience 87:709-717.

Xiong,X., E.Chong, and W.R.Skach. 1999. Evidence that endoplasmic reticulum (ER)- associated degradation of cystic fibrosis transmembrane conductance regulator is linked to retrograde translocation from the ER membrane. J. Biol. Chem. 274:2616-2624.

Yang,M., S.Omura, J.S.Bonifacino, and A.M.Weissman. 1998. Novel aspects of degradation of T cell receptor subunits from the endoplasmic reticulum (ER) in T cells: importance of oligosaccharide processing, ubiquitination, and proteasome-dependent removal from ER membranes. J. Exp. Med. 187:835-846.

Yau,K.W. 1994. Phototransduction mechanism in retinal rods and cones. The Friedenwald Lecture. Invest Ophthalmol. Vis. Sci. 35:9-32.

Ye,Y., H.H.Meyer, and T.A Rapoport. 2001. The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER into the cytosol. Nature 414:652-656.

Young, R. W. 1971. The renewal of rod and cone outer segments in the rhesus monkey. J. Cell Biol. 49: 303-318.

Yoshida,H., K.Haze, H.Yanagi, T.Yura, and K Mori. 1998. Identification of the cis-acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose- regulated proteins. Involvement of basic leucine zipper transcription factors. J. Biol. Chem. 273:33741-33749.

Yu,H. and R.R.Kopito. 1999. The role of multiubiquitination in dislocation and degradation of the alpha subunit of the T cell antigen receptor. J. Biol. Chem. 274:36852-36858.

Zhukovsky,E.A. and D.D.Oprian. 1989. Effect of carboxylic acid side chains on the absorption maximum of visual pigments. Science 246:928-930.

189 Zinszner.H., M.Kuroda, X.Wang, N.Batchvarova, R.T.Lightfoot, H.Remotti, J.L.Stevens, and D.Ron. 1998. CHOP is Implicated in programmed cell death in response to impaired function of the endoplasmic reticulum. Genes Dev. 12:982-995.

190 Research Article 2907

The cellular fate of mutant rhodopsin: quality control, degradation and aggresome formation

Richard S. Sallba\ Peter M. G. Munro^, Philip J. Luthert^ and Michael E. Cheetham^ * ^Division of Pathology and ^Electron Microscopy, Institute of Ophthalmology, University College London, London, UK ‘Author for correspondence (e-mail: [email protected])

Summary Mutations in the photopigment rhodopsin are the major proteasome is required for the efficient retrotranslocation cause of autosomal dominant retinitis pigmentosa. Theof the mutant protein. Inhibition of N-linked glycosylation majority of mutations in rhodopsin lead to misfolding with of tunicamycin leads to the selective retention of the the protein. Through the detailed examination of P23H andmutant protein within the ER and increases the steady state K296E mutant opsin processing in COS-7 cells, we have level of mutant opsin. Glycosylation, however, has no shown that the mutant protein does not accumulate in theinfluence on the biogenesis and targeting of wild-type opsin Golgi, as previously thought, instead it forms aggregates in cultured cells. This demonstrates that N-linked that have many of the characteristic features of anglycosylation is required for ER-associated degradation of aggresome. The aggregates form close to the centrosome the mutant protein but is not essential for the quality and lead to the dispersal of the Golgi apparatus. control of opsin folding. The addition of 9-cis-retinal to the Furthermore, these aggregates are ubiquitinated, recruit media increased the amount of P23H, but not K296E, that cellular chaperones and disrupt the intermediate filament was soluble and reached the plasma membrane. These data network. Mutant opsin expression can disrupt the show that rhodopsin autosomal dominant retinitis processing of normal opsin, as co-transfection revealed thatpigmentosa is similar to many other neurodegenerative the wild-type protein is recruited to mutant opsin diseases in which the formation of intracellular protein aggregates. The degradation of mutant opsin is dependent aggregates is central to disease pathogenesis, and they on the proteasome machinery. Unlike the situation withsuggest a mechanism for disease dominance. AF508-CFTR, proteasome inhibition does not lead to a marked increase in aggresome formation but increases theKey words: Rhodopsin, Aggresome, Proteasome, Chaperone, retention of the protein within the ER, suggesting that theGlycosylation, Retinitis pigmentosa

Introduction Sung et al, 1993). The mechanisms by which these misfolded Retinitis pigmentosa (RP) is the most common cause ofand misrouted proteins lead to photoreceptor death by inherited retinal degeneration and is clinically characterised by apoptosis (Portera Cailliau et al, 1994), however, remain night blindness and the loss of peripheral vision followed byunidentified. the loss of central vision. Mutations in the rod visual pigment,In transfected cells, rhodopsin with mutations in the rhodopsin were first described in 1990 (Dryja et al., 1990) and transmembrane, intradiscal or cytoplasmic domains fails to are now recognised as the most common cause of retinitistranslocate to the plasma membrane and accumulates within pigmentosa (RP), accounting for approximately 15% of the all cell, in what has been described as the ER and Golgi inherited retinal degenerations (Chappie et al., 2001) (OMIM (Kaushal and Khorana, 1994; Sung et al, 1991; Sung et al, 180380: http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id= 1993). These mutant proteins trapped within the cell cannot 180380). Over 100 mutations in rhodopsin have nowform the visual pigment with 11-cis-retinal (Kaushal and been described (RetNet: http://www.sph.uth.tmc.edu/RetNet/ Khorana, 1994; Sung et al., 1991; Sung et a l, 1993) and are home.htm). Although some of these mutations cause recessivefound in a complex with the ER-resident chaperones BiP and RP and congenital stationary night blindness (CSNB), the vastGrp94 (Anukanth and Khorana, 1994), supporting the notion majority are associated with autosomal dominant RP (ADRP).that they are incorrectly folded. It appears, therefore, that these Several of these mutations have been studied in detail at themutations result in a protein that cannot translocate to the cellular and transgenic animal level. These studies suggest that plasma membrane in cultured cells or to the outer segment in rhodopsin mutations can be divided into two major categoriesphotoreceptors because it is misfolded. on the basis of the mechanisms of pathogenesis. Mutations atThe failure of rhodopsin to translocate to the outer segment the C-terminus of rhodopsin interfere with the normal targetingper se does not appear to be enough to cause retinitis of the protein to the photoreceptor outer segment (Sung etpigmentosa, al., as heterozygous rhodopsin knockout mice display 1991; Sung et al., 1994; Tam et al, 2000), whereas mutations little photoreceptor cell death (Humphries et al, 1997). Rather, in the transmembrane, intradiscal or cytoplasmic domains it would appear that misfolded rhodopsin acquires a ‘gain of result in the misfolding of the protein (Kaushal and Khorana,function’ that leads to cell death. Recent insights into the fate 1994; Olsson et al, 1992; Roof et a l, 1994; Sung et a l, 1991; of misfolded proteins in mammalian cells has suggested that 2908 Journal of Cell Science 115(14)

there may be a common cellular response to unfolded proteins,primer 5' GCGGCCGCTTAGGCAGGCGCCACTTG 3: The reaction the formation of a specialised structure, the aggresomeproducts from (i) and (ii) were gel purified and used as a template in (Kopito, 2000). Aggresomes are thought to result from athe final PGR reaction, which was amplified using rhodopsin forward saturation of the normal proteolytic machinery by misfoldedprimer and reverse rhodospin primer to generate the full-length P23H proteins and accumulate as ubiquitinated inclusions near therhodopsin gene. This was cloned into pGEM-T, digested with EcoRl and N otl and cloned into pMT3. WT, P23H and K296E opsin were centrosome. Other characteristic features of aggresomes are the cloned in frame with GPP into the BamHl/Agel sites of pEGFP-Nl disruption of the Golgi and intermediate filament networks and following PGR mutagenesis to remove the stop codon and include the recruitment of cellular chaperones (Garcia-Mata et al.,restriction sites, such that the GPP sequence was fused to the G- 1999; Johnston et al., 1998; Kopito, 2000). termini of the opsin, following a similar methodology to an The presence of ubiquitinated, proteinacious inclusionsexperiment described previously (Ghapple et al., 2000). All construct within neurones is associated with many forms ofsequences were confirmed by sequencing. neurodegeneration, including Amyotrophic Lateral Sclerosis, Alzheimer’s disease, Parkinson’s disease and several Cell culture hereditary diseases caused by expansions of polyglutamine repeats (e.g., Huntington’s disease and the spinocerebellar GOS-7 cells were grown in DMEM/P12 with Glutamax-I+10% heat inactivated PBS and 50 pg/ml gentamicin with an atmosphere of 5% ataxias) (Alves-Rodrigues et al., 1998; Kaytor and Warren, GO2 at 37°G. 24 hours after seeding glass eight-well chamber slides 1999; Sherman and Goldberg, 2001). Studies of the inclusions with 2x10“^ cells per well, the cells were transfected with 0.2 gg DNA formed in some of these diseases and the pathogenetic basis ofper well with 2 |il of plus reagent and 1 pi of lipofectamine according mutations that lead to some neurodegenerations have suggestedto the manufacturers instructions. Por 35 mm dishes, the transfection that they share features with aggresomes (Johnston et al, 2000;mix was scaled up and cells were transfected 24 hours after seeding Waelter et al., 2001). Indeed, a recent study of P23H mutant with 10^ GOS-7 cells. Por Golgi disruption experiments, brefeldin A opsin expression in cell culture has shown that mutant opsin(BPA) was added to the cells to a final concentration of 20 pg/ml for aggregates colocalise, but do not co-aggregate, with other15 minutes prior to fixation. Por proteasome inhibition, cells were proteins that form aggresomes, such as AF508 CFTR andtreated with MG-132 at 5 pM for the indicated time and a parallel TCRa subunits (Rajan et al., 2001). In this study, we have transfection of AP508 GPTR-GPP was used as a positive control for aggresome formation via proteasome inhibition. N-linked further investigated the fate of mutant rhodopsin in vitro, glycosylation was inhibited by the addition of tunicamycin to a final identifying the formation of ubiquitinated mutant opsin concentration of 0.8 pg/ml for the indicated treatment time; the aggresomes and delineating mechanisms of mutant opsinefficiency of inhibition was confirmed by PNGase P treatment of quality control and degradation. The relevance to RP is furthersamples. The incidence of aggresome formation was estimated by highlighted by the observation that mutant opsin can recruitscoring the 400 transfected cells from four separate experiments for the normal protein to aggresomes. presence of aggresomes by an observer that was blind to the cell transfection status.

Materials and Methods Immunofluorescence Reagents The following fixation conditions were used for each antibody: for DAPI for nuclear staining, tunicamycin (containing homologues vimentin, Hsc70, P-GOP and c-myc ubiquitin staining, GOS-7 cells A,B,G, and D), 9-cis-retinal and Protease Inhibitor Gocktail in DMSO were fixed in 100% methanol at -20°G for 6 minutes then washed for mammalian cell extracts were purchased from Sigma. The twice in PBS; for BiP (Grp78) staining, cells were fixed in 3.7% proteasome inhibitor MG-132 (Z-leu-leu-leu GHO) was purchased formaldehyde in PBS for 20 minutes then washed three times in PBS from Biomol Research Labs. Lipofectamine and Plus reagent were for 5 minutes then permeabilised in 0.05% Triton X-100 for 5 minutes from Life Technologies. PNGase P and Endo H were from NEB. The followed by two 5 minute washes in PBS. Por calnexin staining, cells primary antibodies used were mAb 1D4 to rhodopsin, which was a were fixed in 3% formaldehyde/0.1% glutaraldehyde in 0.08 M gift from R. Molday (University of British Golumbia, Ganada) and sodium cacodylate-HGl buffer, pH 7.4 for 20 minutes followed by two subsequently purchased from the National Gell Gulture Gentre; anti- 5 minute washes in PBS and one 5 minute wash in 50 mM NH4GI in p-GOP (clone maD) and anti-Hsc70 (clone BRM22), which were PBS, followed by one rinse in PBS. Gel Is were permeabilised in 0.1% from Sigma; anti-Grp78 (BiP), which was from Santa Gruz Biotech; Triton X-100 and 0.05% SDS for 4 minutes followed by three 5 rabbit polyclonal anti-calnexin, which was from StressGen minute washes in PBS. Following fixation and permeabilisation, cells Biotechnologies; mouse monoclonal to vimentin from Dako Ltd; anti- were blocked for 1 hour at 22°G with 10% appropriate normal serum, ERGIG-53, which was a gift from H. P. Kauri (Basel). The secondary prior to incubation with antibodies. The primary antibodies were antibodies used were Goat anti-mouse Gy3 (Amersham Pharmacia incubated for 1 hour at 22°G, followed by three 5 minute washes in Biotech); donkey anti-rabbit Gy3 and donkey anti-goat Gy3, both of PBS, followed by a 1 hour incubation at 22°G with conjugated which were from Jackson Immuno Research Laboratories; goat anti secondary antibodies and then washed with four washes for five mouse antibodies conjugated to horseradish peroxidase were from minutes in PBS; DAPI was used at 1:5000 in PBS for 5 minutes in Pierce. The AP508 GFTR-GPP and c-myc-ubiquitin plasmids (Ward the third wash. The titres of antibodies used were: mouse anti-P-GOP et al., 1995) were gifts from R. R. Kopito (Stanford). Bovine WT 1:50 dilution in PBS+10% normal goat serum (NGS) followed by goat opsin and K296E opsin in pMT3 were gifts from D. Oprian anti-mouse Gy3 1:500 in PBS+10% NGS; mouse anti-vimentin 1:10 (Brandeis). P23H opsin in vector pMT3 was made by PGR-mediated dilution in PBS + 10% NGS followed by goat anti-mouse Gy3 1:500 site-directed mutagenesis using the WT in pMT3 as a template. in PBS+10% NGS; mouse anti-Hsc70(BRM22) 1:50 dilution in Briefly, the following primer combinations were used: PGR reaction PBS+10% NGS followed by goat anti-mouse Gy3 1:200 in PBS+10% (i) rhodopsin forward primer 5' GGATGGGAATTGGAGGATGA- NGS; anti-c-myc hybridoma supernatant, clone 9E10, was used neat AGGGTAGGGAA 3', P23H reverse primer: 5' GGGTGGAAGTGGG- followed by goat anti-mouse Gy3 1:500 in PBS+10% NGS; ERGIG- TGGGGAGGAG 3'; PGR reaction (ii) P23H site forward primer 5' 53 1:1000 dilution of mouse ascites in PBS+10% NGS followed by GGAGGGAGTTGGAGGGTGGGGAG 3' and rhodopsin reverse goat anti-mouse Gy3 1:500 in PBS+10% NGS; rabbit anti-calnexin Mutant opsin aggresome formation 2909

1:100 in PBS+10% normal donkey serum (NDS) followed by donkey and Khorana, 1994; Sung et al., 1991; Sung et al., 1993). To anti-rabbit Cy3 1:200 in PBS+10% NDS; goat anti-BiP 1:40 in investigate this further, we studied the fate of wild-type and PBS+10% NDS followed by donkey anti-goat Cy3 1:200 in mutant P23H and K296E opsin with and without a C-terminal PBS+10% NDS. For dual labelling, opsin was visualised with mAb green fluorescent protein (OFF) tag in COS-7 cells. COS-7 1D4 conjugated to FITC, at a concentration of 2.5 |ig/ml in PBS+10% cells have been used extensively to study opsin expression. The normal mouse serum and added after primary and secondary localisation of the opsins was investigated by confocal antibodies had bound and after three 5 minute washes in PBS. Immunofluorescence was visualised with a Zeiss LSM 510 laser immunofluorescence using the intrinsic GFP fluorescence and scanning confocal microscope. The following excitation/emission with opsin and Golgi protein antibody markers (Fig. 1). In the conditions were used in separate channels with either the x40 or x63 case of wild-type opsin, the majority of the protein is processed oil immersion objective: DAPI 364/475-525 nm; FITC/GFP 488/505- to the plasma membrane with some staining in the FR and 530 nm; Cy3 543/560 nm. Golgi network indicating the normal biogenesis of the protein (Fig. 1). By contrast, the mutant opsins P23H (Fig. 1) and K296F (data not shown) are not processed to the plasma Electron microscopy membrane but accumulate in the ER, and, in a significant Cells were fixed overnight in Karnovsky’s fixative (3% proportion of cells, the protein accumulates in a juxtanuclear glutaraldehyde, 1% paraformaldehyde in 0.08 M sodium cacodylate- region that shows a partial colocalisation with the Golgi and HCl buffer, pH 7.4) then rinsed three times in PBS and incubated in a 1% aqueous osmium tetroxide solution for 1 hour at room intermediate compartment markers p-COP (Fig. 1) and temperature. Following three rinses in distilled water, cells were FRGIC-53 (data not shown). However, the level of (3-COP dehydrated by 10 minute immersions in 50%, 70%, 90% and 4x100% staining is greatly reduced in the cells that have mutant opsin ethanol. Finally, the wells containing dehydrated cells were filled with aggregates (Fig. 1), suggesting that the Golgi is disrupted by araldite resin and placed in a 60°C oven to harden overnight. Sections the presence of the mutant protein. The addition of GFP to the for examination by transmission electron microscopy were cut using C-terminus of the protein does not alter the cellular fate of the a Leica ultracut S microtome and diamond knife, mounted on copper opsin (Fig. 1) but does appear to increase the rate of inclusion grids and contrasted with lead citrate. Images were viewed using a formation for both the wild-type and mutant proteins, such that JEOL 1010 TEM operating at 80 kV and photographs recorded onto a higher proportion of cells possess the juxtanuclear aggregates Kodak electron image film. (incidence of inclusion formation P23H-GFP>K296F- GFP>P23H>K296F>WT-GFP>WT). Preparation of cell extracts To confirm that the mutant protein was not present in the 24 hours after transfection, cells were treated with 5 p,M MG-132 for Golgi we used brefeldin A (BFA) to disrupt the Golgi 16 hours or 0.8 pg/ml Tunicamycin for 16 hours in DMEM/F12 with apparatus. Treatment with BFA dispersed the Golgi in all the Glutamax-1+10% FBS without antibiotics. Cells were washed twice in cells and dispersed the wild-type opsin trafficking through the 4°C PBS and incubated in 290 pi PBS/1 % n-Dodecyl-P-D-Maltoside Golgi but had no effect on the mutant opsin inclusion (Fig. 2). plus 10 pi of protease inhibitor cocktail and 166 pM MG-132 for 30 Treatment with BFA for periods up to 8 hours did not reduce minutes on ice. Cell lysates were transferred to 1.5 ml microcentifuge the incidence of aggregate formation, suggesting that transport tubes on ice and homogenised by passage through a 23G needle five of opsin from the FR to Golgi is not required for the formation times. 200 pi of cell lysate was centrifuged at 17,500 g for 15 minutes of these aggregates. Therefore, the mutant protein was not at 4°C. Pellets were solublised in 50 pi 10 mM Tris HCl pH 7.5, 1 % SDS+2 pi protease inhibitor cocktail for 5 minutes at 22°C. 150 pi of trapped within the Golgi and instead appeared to be associated RIPA buffer (50 mM Tris-HCl pH 8,150 mM NaCl, 1 mM EDTA, 1% with a distinct structure located close to the Golgi. NP-40, 0.1% SDS, 0.05% sodium deoxycholate) was added, and the pellets were then sonicated for three 5 second bursts. For SDS PAGE, an equal volume of 2x modified Laemmli sample buffer (Ixsample Mutant opsin forms ubiquitinated aggresome-iike buffer, 0.125 M Tris-HCl, pH 6.8, 5% SDS, 15% glycerol, 10% 2- structures mercaptoethanol 0.012% bromophenol blue) was added to the soluble As the mutant opsin appeared to form an intracellular inclusion and insoluble fraction. For immunoblotting, cell fractions were that had some of the characteristics of an aggresome (Kopito, normalised for total protein and separated by 10% SDS-PAGE and 2000), we investigated whether they shared other features with electroblotted onto nitrocellulose. For immunodetection of opsin, mAb aggresomes. Cotransfection of P23H opsin (Fig. 3) or K296F 1D4 was used at a concentration of 0.5 pg/ml and goat anti-mouse opsin (data not shown) with c-myc-tagged ubiquitin resulted in HRP (Pierce) was used at 1:50,000 in PBS+5% Marvel (Premier Brands), 0.1% Tween-20. The chemiluminescent detection reagent the recruitment of ubiquitin to the opsin inclusions, suggesting ECL Plus (Amersham Pharmacia Biotech) was used to detect that the structures are ubiquitinated. There was no association immobilised antigens according to the manufacturer’s instructions. The of ubiquitin with the wild-type protein, the ubiquitin retained electrophoretic mobihty of different opsin glycoforms was determined its diffuse pan-cellular staining pattern and did not colocalise empirically using PNGase F and Endo H; 15 pg protein was digested with the opsin at any stage of its biogenesis (data not shown). with PNGase F (1500 units) or Endo H (1500 units) for 2 hours at 37°C The presence of opsin inclusions also led to the disruption of in lxG7 (PNGase) or lxG5 (Endo H) buffer (NEB) before resolving the intermediate filament network as shown by vimentin by SDS-PAGE as described above. staining (Fig. 3). Vimentin is relocalised from its normal distribution to a bright ring surrounding the inclusion, a feature seen in many aggresome-like structures (Kopito, 2000). Results Examination of the colocalisation of molecular chaperones Mutant rhodopsin does not accumulate in the Golgi with inclusions demonstrated that the cytoplasmic chaperone Several studies have suggested that mutant opsins that cannotHsc70 (of the Hsp70 family) is recruited to the inclusions. By fold correctly are retained within the ER and Golgi (Kaushalcontrast, the FR-resident Hsp70, BiP, is predominantly 2910 Journal of Cell Science 115(14)

Fig. 1. Mutant opsin accumulates in a structure close to the Golgi apparatus. The cellular localisations of opsin in transiently transfected COS-7 cells expressing WT opsin and WT opsin with a C-terminal GFP tag (WT-GFP) were compared with mutant opsin (P23H-opsin) and GFP-tagged mutant opsin (P23H-GFP) by confocal immunofluorescence. The intracellular opsin colocalised with the Golgi marker P-COP; however, in cells with mutant opsin inclusions, the p-COP staining is dispersed. Bar, 10 pm. excluded from the inclusions but does occasionally localise at content of the structures, however, was much higher than that the periphery of the inclusion forming a ring around the observed in aggresomes (Heath et al., 2001; Johnston et al., inclusion (data not shown), a pattern that was also seen with 1998; Waelter et al., 2001) formed from other proteins, and in other ER markers. Collectively, these data suggest that the places the structure appears to be surrounded by membranes mutant opsin inclusions are aggresomes. (arrowheads in Fig. 4). This may correspond to the ‘ring’ of ER staining that was observed in some cells and could reflect the highly hydrophobic nature of the opsin transmembrane Ultrastructure of opsin aggresomes domains acting to recruit lipid vesicles as well as cellular Electron microscopy of COS-7 cells containing mutant opsins chaperones. confirmed the localisation of the aggregate close to the microtubule organising centre (MTOC) and the disruption of the Golgi and intermediate filament networks (Fig. 4). The Co-expression of wild-type and mutant opsin results in juxtanuclear structures were also surrounded by mitochondria. recruitment of the normal protein to the mutant opsin These mutant opsin structures, therefore, display most of the aggresome ultrastructural hallmarks of aggresomes. The membrane We investigated whether mutant opsin aggresome formation Mutant opsin aggresome formation 2911

WT-GFP -r BFA — W r g fe — m-COP-r BFA

Fig. 2. Mutant opsin does not accumulate in the Golgi. Cells expressing WT (WT-GFP) and mutant (P23H-GFP) opsin tagged with GFP were treated with BFA (20 pg/ml for 15 minutes prior to fixation). BFA dispersed the Golgi (P-COP) and the WT-GFP opsin that was trafficking through the Golgi but did not disperse the P23H-GFP opsin from its juxtanuclear position. Bar, 10 pm. P23H-GFP -H BFA P-COP -r BFA

could influence the processing of the normal wild-type protein hours). This suggests that when both wild-type and mutant in cell culture. COS-7 cells were co-transfected with the wild- proteins are expressed, as in ADRP, the mutant opsin type protein tagged at its C-terminus with GFP and untagged aggresome can recruit the normal wild-type protein. wild-type or untagged P23H mutant protein. To make a clearer distinction between normal trafficking through the Golgi and inclusion formation, the cells were treated with BFA prior to fixation and analysis of wild-type opsin-GFP localisation. Co-transfection of normal wild-type opsin-GFP with an excess of WT untagged opsin did not lead to an increase in inclusion formation above that seen with the GFP-tagged protein alone. By contrast, co-transfecting P23H-GFP Ubiquitin wild-type opsin-GFP with untagged P23FI opsin led to a significant increase in the incidence of wild-type opsin aggresomes (Fig. 5). In the presence of an excess of untagged P23H opsin, the incidence of cells with aggresomes containing the WT opsin-GFP approached the level of aggresome formation seen with untagged P23H opsin alone (between 35 and 50% at 24 P23H-GFP vim entin

Fig. 3. Mutant opsin inclusions have the characteristics of an aggresome. Mutant opsin (F23H-GFP) colocalised with c-myc- tagged ubiquitin when co-tranfected in COS-7 cells. The P23H-GFP opsin inclusions led to the disruption of the intermediate filament network (Vimentin). The cytoplasmic chaperone Hsc70 was recruited to the P23H-GFP inclusions. Bar, 10 pm. P23H-GFP 2912 Journal of Cell Science 115 (14)

Fig. 4. Ultraslructure of opsin aggresomes. Electron micrograph of a COS-7 cell with a P23H-GFP mutant aggresome. (i) The aggresome (A) forms near the nucleus (N) and centrosome (C) and is surrounded by mitochondria (M). (ii) At higher magnification the intermediate filaments (IF) that surround the aggresome can be seen, (iii) The aggresome also has a S'- high membrane content and in places appears to be surrounded by membrane (arrowheads). Bar, 1 pm.

Degradation of mutant opsin is via the ubiquitin- increase in the level of mutant protein, as judged by western proteasome pathway and the proteasome is required for blotting (Fig. 7). Following MG-132 treatment there was little efficient retrotranslocation change in the amount or the composition of the WT opsin in As aggresome formation is generally enhanced by treatment either the n-Dodecyl-(3-D-Maltoside (DM) soluble or insoluble with proteasome inhibitors, we examined the effect of the fractions. By contrast, the total amount of P23H and K296E opsin proteasome inhibitor M G -132 on opsin aggresome formation. is dramatically increased after incubation with the proteasome In contrast to the expression of AF508-CFTR where inhibitor (Fig. 7). In the case of the P23FI mutant, this increase proteasome inhibition leads to aggresome formation in 100% was predominantly in the insoluble fraction. DM solubilisation of transfected cells (Johnston et al., 1998), treatment with MG-132 for 8 or 14 hours did not have a marked effect on the incidence of aggresome formation by P23H, K296E or wild-type opsin as assessed by aggresome counts (data not shown). Morphologically, the most pronounced effect of proteasome inhibition was to increase the intensity of ER staining of mutant opsin as demonstrated by colocalisation with BiP (Fig. 6) and calnexin (data not shown). In untreated cells (Fig. I and Fig. 8, untreated), the level of mutant opsin i staining in the ER was relatively weak compared with the staining of the aggregates. Following treatment with MG-132, there was little change in the pattern of wild-type opsin staining but a dramatic increase in the intensity of ER staining WT-GFP/WT-opsin +BFA — |wT-GPt>/P23H-opsin +BFA ■ of the mutant opsins was observed in all cells (Fig. 6). The increase in mutant opsin ER staining correlated with an

Fig. 5. Co-transfection of normal and mutant opsin leads to the recruitment of the normal protein to the mutant opsin aggresome. (A) GFP-tagged WT opsin in pFGFP was co-transfected with untagged WT opsin in pMT3 (WT-GFP/WT-opsin) at a vector ratio m 20 of 1:10. Cells showed no increase in aggresome formation compared with WT-GFP alone, whereas co-transfection of GFP-tagged WT opsin in pFGFP with untagged P23H opsin in pMT3 (WT- GFP/P23H-opsin) at a vector ratio of 1:10 increased the percentage of WT opsin recruited to aggresomes, as judged by GFP fluorescence. Bar, 10 pm. (B) Quantification of co-transfection aggresome formation after 24 hours of opsin expression. The percentage of cells with WT-GFP fluorescence aggresomes was counted double blindly from four separate experiments (error bars±2 s.d.). BFA (20 pg/ml for 15 minutes prior to fixation) was added to both experiments to reduce confusion over WT opsin in the secretory WT-GFP/ WT-GFP/ pathway. WT-OPSIN P23H-OPSIN Mutant opsin aggresome formation 2913

MG-132 f MG-132 f MG-132 Fig. 6. Inhibition of the proteasome leads to an increase of mutant opsin ER staining. Confocal immunofluorescence shows that inhibition of the proteasome machinery with MG-132 (5 |iM for 8 hours) did not affect the processing of WT opsin. The treatment led to an increase in the level of P23H and K296E mutant opsin in the ER, as shown by colocalisation with the ER-resident Hsp70, BiP. Bar, 10 pm.

WT-ODSin — I Merae

+MG-132 -f-MG-132 +MG-132 retrotranslocation from the ER. Treatment with MG-132 led to an increase in the Endo-H-sensitive glycoforms of mutant opsin in the DM soluble and DM insoluble fractions (data not shown). In addition to the deglycosylated protein, the insoluble fraction consisted mainly of opsin dimer and high molecular weight species that migrated as a smear that P23H-opsin extended to the top of the gel, a feature

fMG-132 +MG-132 +MG-132 that is characteristic of ubiquitinated proteins (Ward et al., 1995). This smear may correspond to the variable number of ubiquitin molecules that are added to individual opsin molecules during the process of polyubiquitination. although opsin is prone to self-association and aggregation during SDS-PAGE. The relatively low level of P23H opsin in

K296E opsin — I Merge the untreated insoluble fraction most probably reflects the insolubility of the opsin aggresomes in the DM-insoluble affords maximal stability for folded opsin but does not simply pellets and their subsequent failure to enter the SDS-PAGE gel. produce soluble and aggregated fractions. DM solubilisation These data suggest that a functioning proteasome is required for probably fractionates between folded or nearly folded opsin the degradation of the mutant opsin by ER-associated degradation and opsin folding intermediates, small aggregates and larger (ERAD) and that the proteasome also plays a role in the aggregates. The major opsin component in the DM-soluble retrotranslocation of the mutant protein from the ER. fraction of untreated cells for the wildtype and both mutants coiTesponded to the mature form of the protein, which is Endo H insensitive. The major opsin component in the DM- Glycosylation is required for mutant opsin ERAD but not insoluble fraction of untreated cells had the same electrophoretic quality control mobility as PNGase-F-digested opsin, suggesting that it We examined the effect of tunicamycin, which blocks the was deglycosylated. Deglycosylation is a hallmark of synthesis of precursor oligosaccharides required for N-linked

WT-OPSIN P23H-OPSIN K296E-OPSIN Fig. 7. Inhibition of the proteasome leads to an increase MG132 + 4- 4- + in mutant opsin. Western blot using mAb ID4 of opsin FRACTION S P S P S P S P in DM-soluble (S) and DM-insoluble (P) protein extracts showed that inhibition of the proteasome machinery with MG-132 (5 fiM for 16 hours) (-r) led to an increase in the steady state level of mutant opsin in COS-7 cells. The level of WT protein (WT-opsin) was not affected by proteasome inhibition, whereas the level of mutant proteins (P23H-opsin and K296E-opsin) showed a dramatic increase. The electrophoretic mobilities of different glycoforms of opsin were determined empirically using PNGase F and Endo H: mature (-40 kDa) (arrowhead); and immature forms (>41 kDa); deglycosylated form (-30 kDa) (*) and dimer (-60 kDa) (**). Blot exposures have been adjusted to give equivalent band intensity between WT and mutants, as the WT protein is far more abundant. The positions of the molecular weight markers are indicated on the left in kDa. 2914 Journal of Cell Science 115 (14)

Untreated Untreated Untreated Fig. 8. Inhibition of N-linked glycosylation leads to an increase of mutant opsin ER staining. The colocalisation of WT (WT- opsin) and mutant opsins (P23H-opsin and K296E-opsin) with the ER marker calnexin in untreated cells is shown in the top row (Untreated). Inhibition of protein glycosylation with tunicamycin (0.8 pg/ml for 8 hours) led to an increase in the level of mutant protein retained within the ER. The WT-opsin fCalnexin P23H-opsin -(-Calnexin K296E opsin tCalnexin trafficking of the WT protein (WT-opsin) to the plasma membrane was not affected by -t-Tunlcamycin 4-Tunicamycin ■flunicamycin the treatment with tunicamycin and showed no increase in ER staining. By contrast, the mutant proteins (P23H-opsin and K296E- opsin) accumulated in the ER as shown by colocalisation with the ER-membrane- bound lectin chaperone calnexin. Bar, 10 pm.

formation of the mature diglycoslyated WT-opsin — I Calnexin form of opsin and resulted in the accumulation of non-glycosylated opsin as seen by western blotting (Fig. 9). Although treatment with tunicamycin did not have a major effect on the level of the wild-type protein, the steady state levels of the mutant proteins increased considerably, particularly in the insoluble fraction (data not shown). These observations show that the targeting of mutant opsin for ERAD is dependent on N-linked glycosylation. ER retention of mutant opsin, however, is not affected by tunicamycin and reveals a glycan-independent quality control mechanism that prevents mutant opsin escaping the ER.

9-cis-retinal promotes the targeting of P23H mutant opsin to the plasma membrane Previous studies have suggested that the addition of 11-cis-retinal and 9-cis- glycosylation, on normal and mutant opsin processing and retinal to mutant-opsin-expressing cells can improve the quality control in vitro. As previously described (Kaushal et folding of mutant opsins (Li et al., 1998a). Therefore, we al., 1994), tunicamycin did not affect the transport of the examined the effect of 9-cis-retinal on mutant opsin processing wild-type protein to the plasma membrane and did not lead and aggresome formation in P23H-opsin-expressing cells. The to the retention of the protein within the cell; the small 9-cis-retinal increased the level of mutant opsin as assessed by amount of intra-cellular staining observed corresponded to western blotting and in particular the mature form of the protein trafficking through the Golgi (Fig. 8). The processing protein in the soluble fraction, suggesting efficient transit of the P23H and K296E mutant proteins, in contrast, was through the Golgi apparatus (Fig. 10). This was confirmed by dramatically altered by tunicamycin. Similar to the situation immunocytochemical analysis of opsin localisation (Fig. 10), with MG-132 treatment, tunicamycin led to an increase in the as an increase in plasma membrane staining with P23H could intensity of ER staining in all cells compared with the be observed. However, the incubation of 9-cis-retinal did not untreated cells, as shown by colocalisation with BiP lead to a significant decrease in the formation of aggresomes (data not shown) and calnexin (Fig. 8). The incidence of over the period of the treatment time. The addition of 9-cis- aggresome formation, however, did not appear to change retinal to K296E opsin expressing cells had no effect on the significantly during either 8 or 14 hours of incubation with processing of the mutant opsin (data not shown), as would be tunicamycin. expected as the mutation has removed the site of retinal As expected, treatment with tunicamycin prevented the attachment. Mutant opsin aggresome formation 2915

WT-OPSIN P23H-OPSIN K296E-OPSIN Tunicamycin + + - +

110 90 £

— IP23H + 9-c/s-retinal

Fig. 9. Inhibition of N-linked glycosylation prevents the degradation of mutant opsin. A western blot using mAb 1D4 of opsin in total 9-c/s-retinal protein extracts showed that inhibition of protein glycosylation with tunicamycin (0.8 pg/ml for 16 hours) led to an increase in the steady Fraction state level of mutant protein. The level of total WT opsin was not noticeably altered, whereas the level of mutant (P23H-opsin and K296E-opsin) increased significantly after treatment. The electrophoretic mobilities of different glycoforms of opsin were 110 determined empirically using PNGase F and Endo H: mature (-40 90 kDa) (arrowhead); and immature forms (>41 kDa); deglycosylated form (-30 kDa) (*) and dimer (-60 kDa) (**). Blot exposures have been adjusted to give equivalent band intensity between WT and mutants. Tbe positions of the molecular weight markers are indicated on the left in kDa.

Discussion

Mutations in the rod visual pigment rhodopsin are the major Fig. 10. Treatment with 9-cis-retinal leads to an increase in soluble cause of inherited retinal degeneration, and the majority of mutant opsin that reaches the plasma membrane. (A) Confocal these mutations lead to the misfolding of the protein. immunofluorescence showing that overnight treatment of COS-7 Misfolded opsin is degraded by the ubiquitin-proteasome cells expressing P23H-opsin leads to an increase in the amount of system in COS-7 cells and in SH-SY5Y neuroblastoma protein that reaches the plasma membrane. Bar, 10 pm. (B) A cells (R.S.S. and M.E.C., unpublished observations). This western blot using mAh 1D4 of P23H-opsin in total (T), DM-soluble degradation requires the retrotranslocation of the protein from (S) and DM-insoluble (P) protein extracts shows that the addition of 9-cis-retinal (4-) dramatically increases the amount of opsin in the the ER and is dependent on the presence of an N-linked cells and that most of this is in the soluble fraction. The glycosylation Tag’ and a functional proteasome. The electrophoretic mobilities of different glycoforms of opsin are retrotranslocated mutant opsin appears to be degraded less indicated: mature (-40 kDa) (arrowhead); and immature forms (>41 efficiently than mutant CFTR, despite their expression levels kDa); deglycosylated form (-30 kDa) (*) and dimer (-60 kDa) (**). being similar. Mutant rhodopsin, therefore, forms aggresomes The positions of the molecular weight markers are indicated on the spontaneously in COS-7 cells, in contrast to CFTR, where the left in kDa. majority of the mutant protein is rapidly degraded and inhibition of the proteasome is required to stimulate aggresome formation (Johnston et al., 1998). The addition of a GFP tag to their arrangement in the fully folded protein. It is also possible the C-terminus of opsin increases the rate of aggresome that folding of the protein is subjected to particularly tight formation. This would agree with an earlier study of opsin-GFP quality control in order to prevent the incorporation of that showed that although the protein can fold normally and improperly folded protein into the outer segment, as this could produce functional visual pigment, the misfolding rate is lead to an unacceptable level of ‘noise’ through constitutive higher than non-tagged protein (Moritz et al., 2001). activity. Furthermore, in the presence of mutant protein aggresomes, the Glycoslyation plays pivotal roles in protein folding, normal wild-type protein can be recruited to the inclusions. oligomerisation, quality control, sorting and transport The correct folding of rhodopsin appears to be exquisitely (Helenius and Aebi, 2001). Rhodopsin is glycosylated at sensitive to changes in the primary amino acid sequence, as asparagine residues 2 and 15 in the FR, and the glycan groups judged by the large number of single amino acid substitutions are modified in the Golgi on the way to the outer segment in that lead to ADRP (OMIM: http://www.ncbi.nlm.nih.gov/ photoreceptors or the plasma membrane in COS-7 cells. The entrez/dispomim.cgi?id=l80380). At present, the reasons for importance of glycosylation in the folding and the processing this sensitivity are not clear even though the structure of of opsin has been studied extensively but has generated rhodopsin has been solved (Palczewski et al., 2000), but may apparently conflicting data. Inhibition of glycoslyation with relate to the hydrophobicity of the seven transmembrane tunicamycin, which blocks the synthesis of precursor domains and the need for their membrane insertion, coupled to oligosaccharides required for N-linked glycoslyation, has been 2916 Journal of Cell Science 115(14) used to study rhodopsin biogenesis in vivo and in vitro.(Hirsch and Ploegh, 2000). In the case of CFTR, inhibition of Treatment of isolatedXenopus laevis and Rana pipiens retinas the proteasome does not inhibit retrotranslocation but instead with tunicamycin disrupts normal outer segment disc assemblystimulates aggresome formation by increasing the amount of but has no effect on die intracellular transport of rhodopsinundegraded misfolded CFTR in the cytosol (Johnston et al, (Fliesler et al., 1985; Fliesler and Basinger, 1985). Similarly, 1998). By contrast, inhibition of the proteasome appears to tunicamycin does not prevent the transport of rhodopsin toincrease the the amount of mutant opsin that is retained within the plasma membrane in cell culture (Kaushal et al., 1994), and ER, suggesting that the retrotranslocation of mutant opsin is deglycosylated rhodopsin retains its spectral propertiescoupled to the function of the proteasome, as has been (Kaushal et al., 1994; Prasad et al., 1992). By contrast, site- described for several other proteins (Chillaron and Haas, 2000; directed mutagenesis of the asparagine residues required for N- Mancini et al, 2000; Mayer et al, 1998). Another difference linked glycosylation leads to the accumulationDrosophila of between mutant opsin and CFTR is the efficiency of rhodopsin (Rhl) protein within the ER and retinal degenerationdegradation of the misfolded protein. AF508CFTR is degraded (Katanosaka et al., 1998; OTousa, 1992; Webel et al, 2000). very effectively by the proteasome, and it is only on However, such changes in the primary amino acid sequenceproteasome inhibition that the mutant protein aggresomes are may also disrupt protein folding and lead to the protein beingreadily seen in all cells expressing the mutant protein. Mutant retained within the ER because of misfolding, as seen with opsin does not appear to be efficiently degraded by the mutations in the tri-peptide consensus sequence for N-linkedproteasome and forms aggresomes spontaneously in the glycosylation in bovine opsin (Kaushal et al, 1994). Our dataabsence of proteasome inhibition. The incidence of aggresome show that the major effect of inhibiting N-linked glycosylationformation varies between mutations and most probably reflects in COS-7 cells is to prevent the degradation of the mutantthe rate that the proteins achieve a native or quasi-native state protein and lead to its retention in the ER, whereas and their misfolding rate. glycosylation does not appear to be required for the folding andLi et al. demonstrated that the addition of 11-cis-retinal and transport of the normal protein. These results are consistent9-cis-retinal to mutant-opsin-expressing cells could improve with the major role of N-linked glycosylation of rhodopsinthe in folding of T17M mutant opsin but not a mutation in the C- the early secretory pathway being the quality control terminus and of the opsin (Li et a l, 1998a). Incubation of P23H- degradation of misfolded protein. The glycan moieties ofexpressing cells with 9-cis-retinal increased the level of the rhodopsin may be required for the recognition of misfoldedmature form of the protein in the soluble fraction, suggesting opsin by ER-resident lectin chaperones involved in ERAD,efficient transit through the Golgi apparatus, and this correlated such as calnexin and calreticulin (Helenius and Aebi, 2001), or with an increase in plasma membrane staining. However, the the recently discovered ER degradation enhancing a- incubation with 9-cis-retinal overnight did not lead to a mannosidase-like protein, EDEM (Hosokawa et al, 2001). It significant decrease in the formation of aggresomes during the appears, however, that the N-linked glycan moieties oftreatment time. These data suggest that, although the ligand can rhodopsin are not required for the non-glycosylated proteinbe to used to promote mutant opsin folding and prevent its be recognised as misfolded as the mutant protein is retained degradation, once an opsin aggresome has formed treatment within the ER and prevented from exiting to the Golgi.with ligand cannot disaggregate it. Therefore, the quality control of opsin folding is not dependentThe burden of degrading opsin could weigh heavily on a rod on the recognition of the misfolded protein by ER-residentphotoreceptor cell. Normally, the outer segment and the vast lectin chaperones, such as calnexin and calreticulin, but mayquantities of rhodopsin that it contains are degraded by the involve other chaperones such as BiP and Grp94, which haveretinal pigment epithelium (RPE). Given that opsin previously been shown to bind mutant opsins (Anukanthcorresponds and to at least 30% of the protein produced by a rod Khorana, 1994). The major role of rhodopsin glycosylationcell, in even if only a small percentage of this misfolded and the early secretory pathway appears to be the facilitation needed of to be degraded it could represent a significant problem misfolded protein degradation. Later in the secretory pathway,for the ER-resident chaperones that perform the quality control however, N-linked glycosylation may play other pivotal rolesand ERAD, as well as the ubiquitin-proteasome machinery. in post-Golgi protein sorting, targeting to the rod outerThis could compromise the normal biosynthetic activity of segment, disk assembly (Fliesler et al, 1985; Fliesler and secretory pathway in the inner segment and lead to the Basinger, 1985) or even phototransduction (Kaushal et shortening al, of outer segments that has been observed in animal 1994). models with the P23H mutation (Machida et al, 2000). The The misfolded opsin is recognised and bound by the ERAD-rate of mutant opsin misfolding would determine the effect on associated lectins, then the protein is destined for normal protein biosynthesis and thereby outer segment length. retrotranslocation and degradation by the proteasome. BothThe presence of mutant protein also appears to affect the luminal and transmembrane proteins retained in the ER are processing of the normal protein (Colley et al, 1995; Wu et now known to be retrotranslocated into the cytosol in a processal, 1998), and dominant mutations in the rhodopsin gene, that involves ER chaperones and components of the proteinNina-E, ofDrosophila exert a dominant-negative effect on the import machinery (Plemper and Wolf, 1999). Once exposed tobiosynthesis of the normal rhodopsin protein (Colley et al, the cytosolic milieu, retrotranslocated proteins are degraded by1995; Kurada and O’Tousa, 1995). Our data show that mutant the proteasome, in most cases following polyubiquitination.opsin aggresomes can recruit the normal protein. It may be that There is growing evidence that both the ubiquitin-conjugatingthe smaller aggregates of mutant protein that are the precursors machinery and proteasomes may be associated with theof aggresomes and are destined for degradation also recruit the cytosolic face of the ER membrane and that they could bewild-type protein, leading to the degradation of the normal functionally coupled to the process of retrotranslocationprotein. In photoreceptors in vivo, opsin aggresomes may form Mutant opsin aggresome formation 2917 only rarely as the photoreceptor has a protein foldingubiquitin-proteasome system by protein aggregation.Science 292, 1552- machinery that is finely tuned for opsin synthesis, including1555. specialised chaperones (Chappie et al., 2001). Aggregates of Bunt-Milam, A, H., Kalina, R. E. and Pagon, R. A. (1983). Clinical- ultrastructural study of a retinal dystrophy.Invest. Ophthalmol. Vis. Set. 24, electron-dense rhodopsin have been observed in both human458-469. ADRP and animal models of rhodopsin ADRP (Bunt-Milam Chappie, J. P., Grayson, C., Hardcastle, A. J., Saliha, R. S., van der Spuy, et al., 1983; Li et al., 1998b), suggesting their potential J. and Cheetham, M. E. (2001). Unfolding retinal dystrophies:a role for importance to the disease process. The formation of opsinmolecular chaperones?Trends Mol. Med. 7, 414-421. Chappie, J. P., Hardcastle, A. J., Grayson, C., Spackman, L. A., Willison, aggresomes, however, in human and animal models of RP K. R. and Cheetham, M. E. (2000). Mutations in the N-terminus of the X- remains to proven. We propose that aggresome formation couldlinked retinitis pigmentosa protein RP2 interfere with the normal targeting occur as a stochastic event with a frequency that is inverselyof the protein to the plasma membrane.Hum. Mol. Genet. 9, 1919-1926. proportional to the mutant protein’s ability to fold correctly.Chillaron, J. and Haas, I. G. (2000). Dissociation from BiP and This ‘risk’ could in turn be further modified by cellular stress, retrotranslocation of unassembled immunoglobulin light chains are tightly coupled to proteasome activity.Mol. Biol. Cell, 11, 217-226. which would compromise the cellular chaperone machinery Clarke, G., Lumsden, C. J. and Mclnnes, R. R. (2001). Inherited and could explain the apparent ‘one-hit’ nature ofneurodegenerative diseases: the one-hit model of neurodegeneration.Hum. photoreceptor cell death in rhodopsin RP (Clarke et al., 2001). Mol. Genet. 10, 2269-2275. The formation of an aggresome within a photoreceptor couldColley, N. J., Cassill, J. A., Baker, E. K. and Zuker, C. S. (1995). Defective lead rapidly to the death of the photoreceptor, as opposed tointracellular its transport is the molecular basis of rhodopsin- dependent dominant retinal degeneration.Proc. Natl. Acad. Sci. USA 92, 3070-3074. dysfunction. Dryja, T. P., McGee, T, L., Reichel, E., Hahn, L, B., Cowley, G. S., Yandell, The formation of opsin aggresomes may have importantD. W., Sandberg, M. A. and Berson, E. L. (1990). A point mutation of consequences for photoreceptor viability. The presence the of rhodopsin gene in one form of retinitis pigmentosa.Nature 343, 364- protein aggregates has been shown to compromise the ubiquitin-366. proteasome machinery (Bence et al., 2001), and this could be Fliesler, S. J. and Basinger, S. F. (1985). Tunicamycin blocks the incorporation of opsin into retinal rod outer segment membranes.Proc. Natl. detrimental to several cellular pathways. Similarly, the Acad. Sci. USA 82, 1116-1120. disruption of the cytoskeleton could alter intracellular proteinFliesler, S. J., Ray born, M. E. and Hollyfield, J. G. (1985). Membrane and organelle transport. The recruitment of cellular chaperonesmorphogenesis in retinal rod outer segments: inhibition by tunicamycin.J. to the protein aggregates could expose the cells to environmentalCell Biol. 100, 574-587. Garcia-Mata, R., Bebok, Z., Sorscher, E. J. and SztuI, E. S. (1999). stress and may stimulate an unfolded protein response, which Characterization and dynamics of aggresome formation by a cytosolic GFP- could lead to the downstream activation of caspases (Nakagawa chimera. J. Cell Biol. 146, 1239-1254. et al., 2000). We have shown that mutant opsin aggregates can Heath, C. M., Windsor, M. and Wileman, T. (2001). Aggresomes resemble recruit the normal protein, and it is possible that the aggregates sites specialized for virus assembly.J. Cell Biol. 153, 449-455. recruit other proteins. A recent study has shown that opsinHelenius, A. and Aebi, M. (2001). Intracellular functions of N-linked glycans. Science 291, 2364-2369. aggregates appear to be specific, and although they colocaliseHirsch, C. and Ploegh, H. L. (2000). Intracellular targeting of the proteasome. with other protein aggregates, such as CFTR and TCRa Trends Cell Biol. 10, 268-272. subunits, they do not co-aggregate (Rajan et ah, 2001). Hosokawa, N., Wada, I., Hasegawa, K,, Yorihuzi, T., Tremblay, L. O., Nevertheless, the recruitment of specific opsin-binding proteinsHerscovics, A. and Nagata, K. (2001). A novel ER alpha-mannosidase- like protein accelerates ER-associated degradation.EMBO Rep. 2,415-422. [e.g. Tctex (Tai et al., 1999) and other C-termini-binding sorting Humphries, M. M,, Rancourt, D., Farrar, G. J., Kenna, P., Hazel, M., factors (Tam et al., 2000) or phototransduction components] toBush, R. A., Sieving, P. A., Shells, D. M., McNally, N., Creighton, P. et opsin aggresomes could also influence the targeting of theal. (1997). Retinopathy induced in mice by targeted disruption of the normal protein to the outer segment and compromiserhodopsin gene.Nat. Genet. 15, 216-219. photoreceptor viability. Cell-culture-based systems will be Johnston, of J. A., Dalton, M. J., Gurney, M. E. and Kopito, R. R. (2000). Formation of high molecular weight complexes of mutant Cu, Zn- value in determining some of the consequences of opsin superoxide dismutase in a mouse model for familial amyotrophic lateral misfolding and aggresome formation and how these events can sclerosis. Proc. Natl. Acad. Sci. USA 97, 12571-12576. be manipulated. It is vital, however, to determine the fate ofJohnston, J. A,, Ward, C, L, and Kopito, R, R. (1998). Aggresomes: a misfolded opsin in photoreceptors in vivo and determine thecellular response to misfolded proteins.J. Cell Biol. 143, 1883-1898. precise cellular consequences for photoreceptors of rhodopsinKatanosaka, K., Tokunaga, F., Kawamura, S. and Ozaki, K. (1998). N- linked glycosylation ofDrosophila rhodopsin occurs exclusively in the misfolding in order to develop novel therapies for ADRR amino-terminal domain and functions in rhodopsin maturation.FEBS Lett. 424, 149-154. We are grateful to the following for providing reagents: D. Oprian Kaushal, S. and Khorana, H. G. (1994). Structure and function in rhodopsin. for wild-type and K296E rhodopsin in pMT3; R. Molday, L. Molday 7. Point mutations associated with autosomal dominant retinitis pigmentosa. and the National Cell Culture Centre for rhodopsin mAbs 1D4 and Biochemistry 33, 6121-6128. 4D2; Hans-Peter Hauri for anti-ERGIC-53; and R. Kopito for AF508 Kaushal, S., Ridge, K. D. and Khorana, H. G. (1 994). Structure and function CFTR-GFP and c-myc-ubiquitin plasmids (Ward et al., 1995). This in rhodopsin: the role of asparagine-linked glycosylation.Proc. Natl. Acad. work is supported by the EC. Sci. USA 91, 4024-4028. Kaytor, M. D. and Warren, S. T. (1999). Aberrant protein deposition and neurological disease.J. Biol. Chem. 274, 37507-37510. Kopito, R. R. (2000). Aggresomes, inclusion bodies and protein aggregation. References Trends Cell Biol. 10, 524-530. Alves-Rodrigues, A., Gregori, L. and Figueiredo-Pereira, M. E. (1998). Kurada, P. and OTousa, J. E. (1995). Retinal degeneration caused by Ubiquitin, cellular inclusions and their role in neurodegeneration.Trends dominant rhodopsin mutationsDrosophila. in Neuron 14, 571-579. Neurosci. 21, 516-520. Li, T., Sandberg, M. A., Pawlyk, B. S., Rosner, B., Hayes, K. C., Dryja, T, Anukanth, A, and Khorana, H. G. (1994). Stmcture and function in P. and Berson, E. L. (1998a). Effect of vitamin A supplementation on rhodopsin. Requirements of a specific structure for the intradiscal domain.rhodopsin mutants threonine-17 -4 methionine and proline-347 serine in J. Biol. Chem. 269, 19738-19744. transgenic mice and in cell cultures.Proc. N atl Acad. Sci. USA 95, 11933- Bence, N. F., Sampat, R. M. and Kopito, R. R. (2001). Impairment of the 11938. 2918 Journal of Cell Science 115(14)

Li, Z. Y,, Wong, E, Chang, J. H,, Possln, D. E., Hao, Y., Fetters, R. M. and in intracellular protein aggregation and inclusion body formation.Proc. Milam, A. H. (1998b). Rhodopsin transgenic pigs as a model for human Natl. Acad. Sci. USA 98, 13060-13065. retinitis pigmentosa.Invest. Ophthalmol. Vis. Sci. 39, 808-819. Roof, D. J., Adamian, M, and Hayes, A.(1994). Rhodopsin accumulation at Machida, S., Kendo, M., Jamison, J. A., Khan, N. W., Kononen, L. T,, abnormal sites in retinas of mice with a human P23H rhodopsin transgene. Sugawara, T., Bush, R. A. and Sieving, P. A. (2000). P23H rhodopsin Invest. Ophthalmol. Vis. Sci. 35,4049-4062. transgenic rat; correlation of retinal function with histopathology.Invest. Sherman, M. Y. and Goldberg, A. L. (2001). Cellular defenses against Ophthalmol. Vis. Sci. 41, 3200-3209. unfolded proteins: a cell biologist thinks about neurodegenerative diseases. Mancini, R., Fagioli, C., Fra, A. M,, Maggioni, C. and Sitia, R. (2000). Neuron 29, 15-32. Degradation of unassembled soluble Ig subunits by cytosolic proteasomes:Sung, C. H., Schneider, B. G,, Agarwal, N., Papermaster, D. S. and evidence that retrotranslocation and degradation are coupled events.FASEB Nathans, J. (1991). Functional heterogeneity of mutant rhodopsins J. 14, 769-778. responsible for autosomal dominant retinitis pigmentosa.Proc. Natl. Acad. Mayer, T. U., Braun, T. and Jentsch, S. (1998). Role of the proteasome in Sci. USA 88, 8840-8844. membrane extraction of a short-lived ER- transmembrane protein.EMBO J. Sung, C. H,, Davenport, C. M. and Nathans,J. ( 1993). Rhodopsin mutations 17, 3251-3257. responsible for autosomal dominant retinitis pigmentosa. Clustering of Moritz, 0. L., Tam, B. M., Papermaster, D. S. and Nakayama, I.(2001). functional classes along the polypeptide chain.J. Biol. Chem. 268, 26645- A functional rhodopsin-green fluorescent protein fusion protein localizes26649. correctly in transgenicXenopus laevis retinal rods and is expressed in a time- Sung, C. H., Makino, C., Baylor, D. and Nathans, J.(1994). A rhodopsin dependent pattern.J. Biol. Chem. 276, 28242-28251. gene mutation responsible for autosomal dominant retinitis pigmentosa Nakagawa, T., Zhu, H., Morishima, N., Li, E,, Xu, J., Yankner, B. A. and results in a protein that is defective in localization to the photoreceptor outer Yuan, J, (2000). Caspase-12 mediates endoplasmic-reticulum-specific segment. J. Neurosci. 14, 5818-5833. apoptosis and cytotoxicity by amyloid-beta,Nature 403, 98-103. Tai, A. W., Chuang, J, Z., Bode, C., Wolf rum, U. and Sung, C. H, (1999). O’Tousa, J. E. (1992). Requirement of N-linked glycosylation site in Rhodopsin’s carboxy-terminal cytoplasmic tail acts as a membrane receptor Drosophila rhodopsin.Vis. Neurosci. 8, 385-390. for cytoplasmic dynein by binding to the dynein light chain Tctex-1.Cell Olsson, J, E., Gordon, J. W., Pawlyk, B, S., Roof, D,, Hayes, A., Molday, 97, 877-887. R. S., Mukai, S., Cowley, G. S., Berson, E. L. and Dryja, T, P, (1992). Tam, B. M., Moritz, 0. L., Hurd, L. B. and Papermaster, D. S. (2000). Transgenic mice with a rhodopsin mutation (Pro23His): a mouse model ofIdentification of an outer segment targeting signal in the COOH terminus autosomal dominant retinitis pigmentosa.Neuron 9, 815-830. of rhodopsin using transgenicXenopus laevis. J. Cell Biol. 151, 1369- Palczewski, K., Kumasaka, T., Hori, T., Behnke, C. A., Motoshima, H., 1380. Fox, B. A., Le TVong, L, Teller, D. C., Okada, T., Stenkamp, R. E. et al. Waelter, S., Boeddrich, A., Lurz, R,, Scherzinger, E,, Lueder, G., Lehrach, (2000). Crystal structure of rhodopsin: A G protein-coupled receptor.H, and Wanker, E, E, (2001). Accumulation of mutant huntingtin Science 289, 739-745. fragments in aggresome-like inclusion bodies as a result of insufficient Plemper, R. K. and Wolf, D. H. (1999). Retrograde protein translocation: protein degradation.Mol. Biol. Cell 12, 1393-1407. ERADication of secretory proteins in health and disease.Trends Biochem. Ward, C. L., Omura, S. and Kopito, R. R. (1995). Degradation of CFTR by Sci. 24, 266-270. the ubiquitin-proteasome pathway.Cell 83, 121-127. Portera Cailliau, C., Sung, C. H., Nathans, J. and Adler, R. (1994). Webel, R., Menon, L, O’Tousa, J. E. and Colley, N. J. (2000). Role of Apoptotic photoreceptor cell death in mouse models of retinitis pigmentosa.asparagine-linked oligosaccharides in rhodopsin maturation and association Proc. Natl. Acad. Sci. USA 91, 974-978. with its molecular chaperone, NinaA.J. Biol. Chem. 275, 24752-24759. Prasad, A. V., Plantner, J. J. and Kean,E. L. (1992). Effect of enzymatic Wu, T. H., Ting, T. D., Okajima, T. L, Pepperberg, D. R., Ho, Y. K., Ripps, deglycosylation on the regenerability of bovine rhodopsin.Exp. Eye Res. 54, H. and Naash, M. I. (1998). Opsin localization and rhodopsin 913-920. photochemistry in a transgenic mouse model of retinitis pigmentosa. Raj an, R. S., filing, M. E., Bence, N. F. and Kopito, R. R. (2001). Specificity Neuroscience 87, 709-717. 414 Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001 Unfolding retinal dystrophies: a role for molecular chaperones?

J. Paul Chappie, Gelena Grayson, Alison J. Hardcastle, Richard S. Saliba, Jacqueline van der Spuy and Michael E. Cheetham

Inherited retinal dystrophy is a major cause of blindness worldwide. Recent almost entirely restricted to the specialized rod molecular studies have suggested that protein folding and molecular photo-sensing organelle, the outer segment. In chaperones might play a major role in the pathogenesis of these degenerations. contrast, mutant rhodopsin is localized within the cell Incorrect protein folding could be a common consequence of causative body of photoreceptors in animal models'* ^ In mutations in retinal degeneration disease genes, particularly mutations in the transfected cells, rhodopsin with mutations in the visual pigment rhodopsin. Furthermore, several retinal degeneration disease intradiscal, transmembrane and cytoplasmic domains genes have recently been identified as putative facilitators of correct protein fails to translocate to the plasma membrane, and folding, molecular chaperones, on the basis of sequence homology. We also accumulates in the endoplasmic reticulum (ER) and consider whether manipulation of chaperone levels or chaperone function Golgi (Fig. 1). These mutant proteins trapped within might offer potential novel therapies for retinal degeneration. the cell cannot form the visual pigment with 11 -Cis-retinal and are found in complex with the Diseases affecting the function of the retina leading to molecular chaperones BiP and Grp94 (members of the visual impairment and blindness display remarkable Hsp70 and Hsp90 families, respectively), supporting genetic and clinical heterogeneity'. Indeed, Online the notion that they are incorrectly folded® (Fig. 2b). Mendelian Inheritance in Man (OMIM, It appears, therefore, that these mutant proteins fail www3.ncbi.nlm.nih.g0v/0mim/) and the online to translocate because of protein misfolding. retinal information network RetNet The failure of rhodopsin to translocate to the outer (www.sph.uth.tmc.edu/RetNet/) list over 120 distinct segment, perse, does not appear to be enough to cause disorders that include some form of retinal RP, as heterozygous rhodopsin knockout mice display degeneration, which can be inherited as autosomal little photoreceptor cell death^. Rather, it would appear dominant, autosomal recessive, X-linked, that misfolded rhodopsin acquires a gain of function’ mitochondrial, digenic, polygenic and syndromal that leads to cell death. The nature of this gain of diseases affecting over 1 in 2000 people (Fig. i. Box 1). function is unclear at present, but could be related to a Here, we will review recent developments in the saturation of the normal protein processing, transport molecular and cellular understanding of these and degradation machineiy in such a highly-specialized diseases to examine the potential role of protein cell. Defining the cellular consequences of incorrect folding and molecular chaperones, and whether rhodopsin folding and targeting might help bring to light molecular chaperones might represent novel targets the pathogenetic mechanisms that link rhodopsin for therapeutic intervention. misfolding with photoreceptor apoptosis. Given the importance of correct rhodopsin folding for Protein misfolding and retinal disease photoreceptor survival, it is perhaps not surprising that Correct protein folding is an essential biological specialized chaperones of opsin biogenesis have been process. To ensure this happens in the intracellular identified, such as the Drosophila protein NinaA (Ref. 8). milieu many proteins interact with molecular

Alison J. Hardcastle chaperones (Box 2). The importance of correct protein Specialized opsin chaperones Dept of Molecular folding and the potential involvement of molecular NinaA is a photoreceptor-specific integral membrane Genetics chaperones in retinal degeneration are emphasized by glycoprotein with a central cyclophilin (CyP) homology J.Paul Chappie the most common and studied form of inherited retinal domain (Box 2) that extends into the ER lumen®. The Celene Grayson degeneration, retinitis pigmentosa (RP) caused by major visual pigment in Drosophila, rhodopsin 1 Richard S. Saliba mutations in rhodopsin - the molecular sensor for light. (Rhl), is expressed in photoreceptor cells R1 -6 and is Jacqueline van der Spuy Michael E. Cheetham* synthesized in the ER, and then transported through Dept of Pathology, Rhodopsin and misfolding in RP the secretory pathway to the rhabdomere (a light Institute of Mutations in the gene encoding rhodopsin were the first transducing organelle similar to the mammalian rod Ophthalmology, identified in 1990^, and over 100 mutations have now outer segment). NinaA and Rh 1 colocalize in the ER University College London, 11-43 Bath been described that account for approximately 15% of and transport vesicles within photoreceptor cells Street, London, all inherited human retinal degenerations (see OMIM). (Fig. 2e). Mutations in NinaA lead to the accumulation UK EC1V9EL. Many studies have suggested that the vast majority of of immature, misfolded Rhl in the ER that is *e-mail michaei.cheetham® these mutations cause disease through the misfolding prevented from reaching the rhabdomere. The ucl.ac.uk or mislocalisation of the protein^. Wild-type rhodopsin is misfolded Rh 1 is degraded, resulting in reduced levels

http://tm m .trends.com 1471-4914/01/$-s e e fro n t m atter C 2001 Elsevier Science Ltd. A ll rights reserved. PH: 81471-4914(01)02103-7 Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001 415

Box 1. Retinal dystrophies

Retinitis pigmentosa (RP) affect primarily cone photoreceptors (cone and RP (OMIM 180380 for rhodopsin RP and 312600 for cone-rod dystrophies). RP2 RP) is the most common cause of inherited blindness with over 25 genetic loci identified Leber congenital amaurosis (LCA) (see RetNet). Clinically, the disorder is characterized LCA (OMIM 604393 for AIPL1 LCA) accounts for by night-blindness, loss of peripheral vision, followed approximately 10% of inherited retinal disease and is by loss of central vision. Examination typically shows the most common cause of congenital visual pigmentation in a 'bone-spicule' pattern in the retina, impairment in infants and children. Five distinct genetic pallor of the optic nerve and narrowing of retinal loci have been reported. This severe form of inherited blood vessels. There is much variability in the retinal dystrophy presents within a few months of birth severity of symptoms and signs between different when a non-seeing infant has a normal appearing families as well as between different individuals of fundus but a flat electroretinogram. Later, the retina can the same family, and although total blindness can appear to resemble that seen in RR and a subset of the occur, this is rare. Electroretinography, which disease can present with an area of deficient retinal involves the detection and characterization of the tissue at the macula (macular coloboma). microvoltage potentials that occur on light stimulation of the retina, reveal a reduced response from rod Bardet- BiedI syndrome (BBS) photoreceptors and later cone photoreceptors, and BBS [OMIM 605231 for McKusick-Kaufman syndrome confirm that the primary pathology involves the (MKKS)-BBS] has an ocular component, which is maintenance of rod photoreceptor function. Fig. I characterized by pigmentary retinopathy similar to that shows a fundus photograph of normal retina seen in RP as a result of rod-cone disease. Associated alongside a patient fundus with RP. Genetically, symptoms of BBS include obesity, polydactyly, there is considerable overlap between RP and other hypogonadism, learning difficulties and renal failure. retinal dystrophies such as those that affect the Six loci have been described to date, with MKKS macula (macular dystrophies) and those that representing the first cloned gene for this disorder. ■T"

Fig. I. A comparison of normal and RP fundus. Retinal dystropfiies are classified by (i) ophthalmoscopic findings (fundus examination), (ii) psychophysics and electrophysiology, (iii) age of onset, and (iv) genetics or family history.

of Rhl found in NinaA mutant flies‘°. Furthermore, A mammalian orthologue of NinaA has not yet NinaA forms a stable and highly specific complex with been identified. Ferreira e ta /*' identified a putative Rhl and the amount of functional NinaA directly retina-specific cyclophilin from a bovine retina cDNA reflects the production of mature rhodopsin®. libraiy with two splice variants encoding isoform Mutations in the CyP homology domain are thought to types I and II, which were predominately expressed in compromise the peptidyl-prolyl cis-tra/75 isomerase cone photoreceptors. The type I I isoform is identical to (PPIase) activity of NinaA, preventing proline RanBP2 (Ran-binding protein 2), a large protein that isomerization within Rhl that might be required for was identified as a component of the nuclear its proper folding®. Collectively, these observations transport machinery*^. RanBP2 might also play a role suggest that processing and transport of Rh 1 are in opsin processing. Two adjacent domains in dependent on NinaA, and that NinaA acts as a specific RanBP2, the Ran-binding domain 4 (RBD4) and the molecular chaperone for Rhl. cyclophilin domain, form a complex with the bovine http://tmm.trends.com 416 Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001

Box 2. What are molecular chaperones?

By the simplest definition, molecular chaperones are The different families of chaperone proteins recognise various facilitators of protein conformational change that, in intermediates of non-native polypeptides and interact through themselves, provide no information as to the final protein different modes of binding. Hsp70 proteins bind short regions of structure. Several more precise definitions exist but these peptides with a certain position and pattern of hydrophobic seldom cover the breadth of molecular chaperone action. residues in a substrate-binding pocket^°. By contrast, chaperonins Molecular chaperones are a group of functionally related but can facilitate folding by enclosing non-native polypeptides in the otherwise diverse protein families. They bind to and stabilize central cavity of a double ring structure formed from identical or conformers of other proteins and, through cycles of regulated closely related rotationally symmetrical subunits^^ ^°. Another binding and release, are able to facilitate the correct fate of example of chaperone activity is the peptidyl-prolyl cis-trans their client. Via this mechanism, molecular chaperones, in isomerase (PPIase) activity found in the cyclophilins (e.g. NinaA) conjunction with their cochaperones, play an essential role in that overcomes a rate-limiting step in protein folding, the correct many cellular processes (Fig. 2a-f). They have a principal role orientation o f proline residues®®. in protein folding, where they are involved in thede novo For many chaperones, cycles of client protein binding and synthesis of polypeptides (Fig. 2a), transport across release are coupled to conformational changes in the chaperone m em branes (Fig. 2b and e) and the refolding o f proteins protein, which are dependent on the hydrolysis and exchange of denatured by adverse environmental conditions (Fig. 2f). ATP that is regulated by cochaperones'”’. Cochaperones function Chaperones also function in oligomeric assembly and syner gistically with the major chaperones in protein folding and disassembly of protein complexes, controlled switching often have independent chaperone activity, but their major role between active and inactive conformations of client proteins, might be to provide these folding machines with specificity. For intracellular transport and proteolytic degradation example, the Hsp70 protein machinery achieves its multiple (Fig. 2c,d and cellular functions because of various cochaperone proteins, such More than 20 different families of proteins have been shown to as the DnaJ family, which stim ulate Hsp70 ATP hydrolysis have chaperone activity. Principal chaperone families include the through their conserved J domain®®. Several other cochaperones Hsp90, Hsp70, chaperonins (e.g. Hsp60), DnaJ (e.g. Hsp40) and (e.g. Hip, Hop, Chip and some im m unophilins including AIPL1) small heat-shock proteins (e.g. Hsp27, a-crystallin). The utilise a degenerate 34-amino-acid repeat motif, the best-studied and mechanistically understood chaperone tetratricopeptide repeat (TPR), in tandem arrays to prom ote machines are Hsp70 and the chaperonins. These chaperones chaperone-cochaperone interactions®’ and modify chaperone recognize and bind to unfolded or partially folded polypeptides function. In addition, there is functional cooperation between the by binding to exposed hydrophobic regions preventing them individual chaperone machines. For example, Hsp90 and Hsc70 from aggregating and maintaining them in a folding competent cooperate in the assembly of steroid receptor and transcription state until release. This is particularly im portant as the nascent factor complexes (Fig. 2c), and the Hsp70 chaperone machine polypeptide emerges from the ribosome, where efficient folding might pass nascent chains onto chaperonins to complete folding of the newly synthesized chain is achieved by a transient (Fig. 2a). Through these com plem entary but distinct roles in interaction with the chaperone that prevents aggregation due to protein folding, molecular chaperones can facilitate changes in unwanted interactions with hydrophobic regions of other protein conformation from initial folding through function and proteins or w ithin the extending polypeptide (Fig. 2a). ultimately to degradation.

long- and medium-wavelength (red/green) sensitive misfolding. Point mutations in other essential opsin‘3. Similar to NinaA, the cyclophilin domain photoreceptor proteins might destabilize the protein might function to induce a d5->tra/7s isomerization of structure and lead to misfolding. It is also possible one or more peptidyl-prolyl bonds in the that retinal degeneration can result from a long-wavelength (red) opsin*'’. dysfunction in the specialized chaperone machinery A specific molecular chaperone for mammalian of the retina, and in the following sections we will rhodopsin biogenesis and transport has yet to be examine some recently identified retinal disease identified, but other, less specialized, molecular genes as putative chaperones. chaperones might participate in rhodopsin biogenesis. For example, BiP and Grp94 might Putative chaperones as causes of retinal degeneration interact transiently with wild-type opsin as part of RPZandRP normal protein folding and quality control in the ER. Mutations in RP2 account for 15-20% of X-linked In addition, the small heat shock protein family retinitis pigmentosa (XLRP)'®. RP2 was identified as member a-crystallin has been found to be enriched in a putative chaperone on the basis of its similarity post-golgi membranes in the inner segment of frog rod (30.4% identity over 151 amino acids) to cofactor C, a cells along with newly synthesized rhodopsin'®, component of the tubulin folding pathway'^*®. The suggesting that a-crystallin might associate with significance of this similarity is supported by rhodopsin and be involved in its processing (Fig. 2e). pathogenic amino-acid substitutions in RP2 at Rhodopsin is not the only example of a retinal conserved residues'® '® ®", suggesting that RP2 has disease protein that can cause disease through functional homology with cofactor C.

http://tmm.trends.com Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001 417

subsequently designated the aryl hydrocarbon receptor-interacting protein-like (AIPL) 1. AIP facilitates the signalling of a transcription factor, aryl hydrocarbon receptor (AhR) (Fig. 2 c) 2s 29 in the absence of ligands, inactivated AhR exists in a cytosolic multiprotein complex, which includes AIP and the molecular chaperone Hsp90. In the presence of ligands, AhR translocates to the nucleus where it dissociates from Hsp90 and AIP, and subsequently heterodimerizes with a structurally related protein termed the aryl hydrocarbon receptor nuclear translocator (Arnt). Importantly, both the transactivation activity and Wild-type rhodopsin P23H rhodopsin subcellular compartmentalization of AhR are modulated by AIP (Refs 30-32). AIP is able to protect fig . 1. Comparison of the Molecular chaperones, in particular the cytosolic the inactivated AhR from ubiquitination, thereby cellular localization of chaperonin containing TCPl (CCT), play an essential stabilizing the cytosolic AhR from degradation and wild-type and mutant role in the biogenesis of several components of the enhancing the levels of total and functional AhR in the rhodopsin in transfected COS-7 cells. Wild-type cytoskeleton, including actin (microfilaments) and cytosop2-34. Furthermore, an AlP-dependent delay in protein translocates to the tubulin (microtubules)^'. The production of native a-P the nuclear accumulation of the AhR is observed in the plasma membrane tubulin heterodimers before their assembly into presence of ligand. AhR-mediated transactivation is (arrowhead), whereas microtubules depends on the action of several cofactors thus significantly enhanced in the presence of AIP protein with a P23H mutation accumulates (A-E), including cofactor C, following the release of near through an increased availability of AhR ligand-binding within the ER/Golgi (star) native folded subunits from CCT (Fig. 2d) These sites. AIP can also help prevent AhR aggregation by and fails to reach the cofactors might also play a role in sequestering tubulin thermal dénaturation. These data, in addition to plasma membrane. or modifying microtubule function and dynamics, for sequence similarities between AIP and the example, by acting as a GTPase activating protein immunophilin FKBP52 (see below), support a function (GAP) for the tubulin heterodim er^2-24 However, the for AIP as a molecular chaperone and suggests that precise role of cofactor C in tubulin and microtubule AIPL 1 also acts as a chaperone or cochaperone. dynamics remains unclear. FKBP52 is a high-molecular-weight member of RP2 is ubiquitously expressed (i.e. it is not retina the family of immunophilins. Analogous to the AhR specific) and is post-translationally modified at its signalling system, a direct interaction of the N-terminus by the addition of two acyl moieties that high-molecular-weight immunophilins with the target the protein to the plasma membrane^^. This molecular chaperone Hsp90 mediates their acyl-mediated membrane targeting is disrupted by a association with cytosolic steroid hormone receptor pathogenic mutation in RP2 (AS6), suggesting that the complexes. These immunophilins are thought to membrane localization is essential for protein function assist the targeted translocation of the associated in the retina (Fig. 3)^^ Therefore, it seems unlikely activated steroid receptor to the nucleus. that RP2 functions exclusively in tubulin folding. It is Furthermore, these immunophilins might function possible that RP2 does still interact with tubulin as molecular chaperones in their own right, as they and/or microtubules and might provide a link between can both inhibit the aggregation of thermally membranes and the cytoskeleton, potentially as part denatured substrates and maintain them in a of the cellular protein traffic machinery or a signalling folding-competent conformation^^. cascade. This hypothesis is supported by the Although there are no obvious structural identification of ADP ribosylation factor (ARF)-like similarities, AhR and members of the steroid receptor proteins and src as interacting partners of RP2 superfamily exhibit similarities in their signalling (Ref. 26), a function that appears to be conserved with mechanism. Both the AIP and immunophilin cofactors C and D (Refs 23,24). As more functional containing complexes represent the penultimate stage information on RP2 and cofactor C is ascertained, of a procedure that involves the maturation of the their putative chaperone function can be evaluated. receptor to a high-affinity ligand-binding conformation. Two other proteins homologous to known chaperones This procedure is tightly regulated by the molecular have more recently been identified as causes of chaperones Hsp90 and Hsp70 and their associated non-syndromic and syndromic retinal degenerations. cochaperones, which collectively comprise a cytosolic heterocomplex chaperoning machine^®. Similarly, a AIPL1 and Leber congenital amaurosis retina-specific system of interacting chaperones and Mutations in the gene encoding a novel cochaperones might exist and fulfil an essential photoreceptor-pineal-specific protein, AIPL 1, cause chaperone function in the retina. Within this system, approximately 10% of recessive Leber congenital jAIPLI could fulfil a molecular chaperone function in amauroses (LCA)^L This protein shares 49% amino retinal protein folding, or assist receptor nuclear acid identity with the human aryl hydrocarbon translocation in a manner analogous to that of AIP and receptor-interacting protein (AIP), and was FKBP52 in their respective signalling systems.

http://tmm.trencJs.com 418 Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001

Type II chaperonin mediated folding

ADP A-rp Unfolded Folded actin actin o o Photoreceptor specific substrate?

(b) Post golgi Co-translational Post-translational transport? translocation translocation 0,0

\ o .

ER to golgi O' Recycle? transport? / ( T o \ b

(c)

sHsp oligomer sHsp O Stress O (e.g. heat UVA light)

Unfolding

ADP ATP Nucleus J Q 0 Cochaperones of Target gene Hsp70 required

Q Hsc70 Q BiP/Grp78 Q Hsp70 0 Coat protein [^T ra n slo co n (Sec63p DnaJ protein component) Qp> ER membrane chaperone (e.g. calnexin) ^ Hsp40 (specialized DnaJ proteins in photoreceptors?) MKKS □ ER lumen chaperone (e.g. calreticulin) □ Immunophilin (AIPL1 in photoreceptor specific complex?) Folded glycoprotein (e.g. ROS proteins, rhodopsin?) ( 2 ) a-Tubulin Q Hsp 90 Cochaperones O a-crystallin O Unfolded protein O p-Tubulin

Signalling molecule O Folded protein 0 Cofactor Intermediate filament Receptor Nina A mammalian orthologue for “ ^RiRibosome — Actin filament Rhodopsin folding/transport ? h TRENDS in Molecular Medicine

MKKS and Bardet-Biedi Strongly suggests that MKKS is a member of the family Mutations in A/A7CShave recently been identified as of type II chaperonins (Fig. 4)^^. Chaperonins can been the cause of the developmental diseases, classified into two families: type I (or GroEL subclass) McKusick-Kaufman syndrome (MKKS)^^, and a rare chaperonins found in eubacteria and in organelles of form of the more severe disease Bardet-Biedl syndrome eubacterial descent, and; type II (or TCP-1 subclass) (BBS6), which also has a characteristic retinal chaperonins found in archaea and the eukaryotic dystrophy (Box 1)^. Amino acid sequence homology cytosoP®. Chaperonins are large homo- or

http://tmm.trends.com Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001 419

Fig. 2. Schematic of chaperone-facilitated events in a photoreceptor cell, including putative specialized Two of the mutations found in MKKS patients (Y37C pathways in the outer segment (OS), the cilium (C), mitochondria (M), the golgi (G), the endoplasmic and H84Y) are predicted to lie in the highly conserved reticulum (ER), the nucleus (N) and the synapse (S). (a) Folding in the cytosol: nascent chains emerging from the ribosome are bound by the Hsp70 chaperone machine, in particular the constitutive Hsp70, equatorial domain of the ring structure. The structure of Hsc70, and members of the DnaJ family including Hsp40. The nascent polypeptides can then be passed this domain of the MKKS protein can be efficiently onto the cytosolic chaperonin CCT, or putatively MKKS, to complete folding, (b) Folding in the ER: modelled on the thermosome structure (Fig. 4). The proteins can be cotranslationally inserted into the ER with the help of the translocon (including Sec63) Y37C mutation is in a region of the protein involved in or post-translationally with the assistance of Hsc70. Once within the ER, the unfolded proteins are bound by the ER resident Hsp70, BiR and the glycan-binding chaperones calnexin and calreticulin that making intra-ring contacts in the thermosome, whereas promote and monitor correct folding, (c) Signalling: nuclear receptors, such as the glucocorticoid the H84Y mutation lies in a region that is responsible receptor, form a complex with HspSO and Hsc70 and several other cochaperoncs (Hop, DnaJ proteins). for ATP hydrolysis in type I I chaperonins. It seems This complex matures', through the exchange of several components (e.g. Hop and Hsc70 exchange for immunophilins and p23), to a complex that facilitates the binding of ligand. Upon ligand binding the likely, therefore, that the correct assembly of the complex changes conformation again and the activated receptor is transported to the nucleus where toroidal structure and the hydrolysis of ATP are the chaperones are released, (d) Cytoskeleton: actin and tubulin require CCT for correct folding. essential for the function of MKKS. The putative Near-native a and p tubulin subunits are released from CCT and bound by cofactors that facilitate identity of MKKS as a type II chaperonin strongly formation of the heterodimer and stimulate GTP hydrolysis. Small heat shock proteins (sHsps) also interact with components of the cytoskeleton. (e) Vesicular transport: immature, unfolded rhodopsin is suggests a function in the cellular protein folding bound by NinaA in the ER. The NinaA remains associated with the rhodopsin through the secretory machinery in the retina and other tissues (Fig. 2a). pathway and is recycled to the ER upon release of the mature folded opsin, a crystallin associates with Chaperonins are thought to assist the folding of post-golgi rhodopsin vesicles, (f) Stress response: proteotoxic stress (e.g. heat) leads to the partial unfolding of proteins. The unfolded proteins are bound by the stress-induced Hsp70 (Hsp70) or sHsp their ‘client proteins’ in the central cavity of their oligomers. sHsps can hold the protein until Hsp70 is ready to refold' them with the assistance of other toroidal structure (Fig. 2a and d. Fig. 4 and Box 2). If chaperones and cochaperones (e.g. Hsp90 and Hsp40). MKKS is indeed a component of a novel chaperonin it will be important to identify its client proteins, as they hetero-oligomeric complexes that promote the folding of could also represent BBS causative genes. It has been proteins to native states in an ATP-dependent cycle of estimated that ~ 15% of newly synthesised eukaryotic binding and release^^ '*'^ (Fig. 2a and d). The subunits proteins interact with CCT (Fig. 2a)^^, although other are arranged in two rings stacked together back-to-back studies suggest that CCT predominantly has the such that they form a toroidal structure, with a central cytoskeletal proteins actin and tubulin as client cavity in each ring where it is believed protein folding proteins (Fig. 2d)'*'*. Therefore, it is difficult to predict can occur in isolation from the rest of the cell. whether MKKS will facilitate the folding of a wide In type II chaperonins, the rings consist of eight or range of proteins or only a few specialised clients. nine subunits and, although some thermosomes Recently, two other BBS genes (BBS2and BBS4) have (chaperonins from high temperature living archaea) are been identified'*^ '*®. Neither protein has homology to homo-oligomeric, the majority are hetero-oligomeric MKKS, suggesting that they are not other subunits of consisting of at least two different homologous subunits a hetero-olgomeric complex, however, BBS4 contains a (Fig. 4). The eukaryotic cytosolic chaperonin (CCT or putative tetratricopeptide repeat (TPR) domain TriC) has eight unique subunits per ring, with each (Box 2). BBS4, therefore, could also be a chaperone or subunit occupying a well-defined position'**. Currently, it cochaperone involved in the same folding pathway, is not known whether the MKKS protein is a subunit in a whereas BBS2 could represent a putative client. homo- or hetero-oligomeric ring. The thermosomes from The diseases MKKS and BBS both show variable Thermoplasma acidophilum, whose fold best resembles penetrance and expressivity®^. The underlying the MKKS protein (Fig. 4)3^, are hexadecamers composed genetic fault might result in compromised cellular of alternating a and P subunits'*^. If the MKKS protein chaperone machinery, and this in turn could is one subunit of a hetero-oligomeric type II compromise the stress handling capacity (Fig. 2f ). chaperonin, then other as yet unidentified subunits The variation in disease penetrance and expression could represent candidate genes for other BBS loci. could, therefore, reflect differences in stress during embryonic development that expose the inability of the developing embryo to compensate for protein damaging insults, such as elevated temperatures caused by maternal illness.

The therapeutic potential of manipulating molecular chaperones in retinal degeneration The upregulation of many molecular chaperones as part of the cellular response to stress is one of their characteristic features, and led to their initial identification as heat shock proteins (Fig. 2f ). This is not only a reactive but also an adaptive response and can protect the cell against future stress, a phenomenon Wild-type RP2 S6A RP2 known as thermotolerance. Even before the identification of the first rhodopsin mutations, it had Fig. 3. Targeting of the N-terminus of RP2-GFP in CHO cells. Wild-type protein localizes to the plasma membrane (arrowhead), whereas RP2, which harbours the SerBdeletion mutation, is localized to the been demonstrated that thermotolerance could be cytoplasm and nucleus, as a result of the disruption of N-terminal acylation of the protein^. exploited to protect the retina from damage. Barbe etalT’

http://tmm.trends.com 420 Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001

Outstanding questions MKKS equatorial domain model a-subunit of thermosome - - ' X . How does the misfolding of rhodopsin lead to photoreceptor cell death by apoptosis? Does the retina contain specialized molecular chaperone machines and what are their client proteins? Can molecular chaperone manipulation be used as a general treatm ent for m isfolded proteins in retinal degeneration; in particular, can the folding o f rhodopsin be manipulated to improve photoreceptor survival (by molecular chaperones or ligand, such as vitamin A)? Side view Top view Can the anti-apoptotic effects of molecular chaperones be used to prom ote photoreceptor viability?

have two modes of action, preventing protein aggregate formation and protection from environmental stress. The balance and potential synergy between these two nriodes of action are perhaps best exemplified by the effect that molecular chaperones have on neurodegenerations mediated by polyglutamine expansions®*. In mammals, only SCA7 polyglutamine expansions cause retinal degeneration, in addition to central nervous system or peripheral neuropathy®^, but in Drosophila, polyglutamine expansions result primarily in retinal degeneration. This model organism Thermosome homo-oligomer has been used in several elegant studies to show the Fig. 4. The McKusick-Kaufman syndrome-causing protein (MKKS) shares structural homology with protective effects of chaperones. Overexpression of type II chaperonins. A three-dimensional ribbon model of the equatorial domain of the MKKS protein Hsp70 in transgenic flies resulted in a reduction of (red), which includes residues mutated in MKKS and BBS, was based on the crystal structure of the polyglutamine induced retinal degeneration®®. a-subunit of the thermosome from Thermoplasma acidophilum (blue) and modelled using SwissModel (http://www.expasy.ch/swiss-mod/SWISS-MODEL.html). Note the importance of this Furthermore, complementary studies using P-element domain for inter-subunit and interring contacts. These subunits are then arranged into a insertional mutagenesis in Drosqp/irVa identified two hexadecamer. The arrangement of the a-subunits in the thermosome homo-oligomer in its open cochaperones, Hsp40/Wdjl and TPR2/Hcp, as conformation’ is shown in side and top view (reproduced, with permission, from Ref. 43). Client proteins are bound in the central cavity. suppressors of polyglutamine mediated degeneration in the eye®**. In these studies, the incidence of the hallmark showed that a prior mild heat shock could protect the intranuclear inclusions, which are thought to be a retina from light-induced damage. Although the response to misfolded protein, was not reduced. Many mechanisms of this protection are not known, the other studies, however, have shown that the overexpression of individual or combinations of formation of these inclusions can be modulated by molecular chaperones can mimic many of the protective chaperone overexpression®* and it is unclear which properties engendered by heat shock. Next, we will chaperone mechanisms are important for protection discuss the potential of manipulating the chaperone from polyglutamine mediated retinal degeneration. machinery for treatment of retinal degeneration. There are now several pharmacological agents that The fate of misfolded protein in retinal can manipulate chaperone expression and/or function, degenerations is not known, and the accumulation of so that gene transfer mediated overexpression might misfolded protein could be a slow process - RP not be necessary to develop therapeutic strategies sometimes does not develop until midlife. Granular involving chaperones®®. In particular, bimoclomol can aggregates, immunoreactive for rhodopsin, form in potentiate the endogenous stress response and increase human**® and animal models of RP (Ref. 49), a chaperone levels®®, and has been shown to protect phenomenon that begs parallels with other photoreceptor cells in an experimental model of diabetic neurodegenerations in which intracellular protein retinopathy®^. The potential of chaperone-modulating Acknowledgements aggregates form and have been linked to cell death. The compounds in treating acquired and inherited retinal We wish to thank Andrew Webster, mechanism by which misfolded proteins lead to degenerations needs to be further explored, but the Bart Leroy and Alan Bird apoptosis is not known and could differ between developing basic science suggests that manipulating for fundus photographs proteins. Photoreceptors in rhodopsin RP, however, are protein folding through molecular chaperones should be and The Wellcome Trust, BBSRC and Fight for Sight more sensitive to stress such as light damage®**. widely applicable in the treatment of retinal for funding. Manipulating molecular chaperones, therefore, could degenerations.

http://tmm.tfends.com Review TRENDS in Molecular Medicine Vol.7 No.9 September 2001 421

References 22 Tian, G. e tal (1999) Tubulin folding cofactors as 41 Liou, A.K. and Willison, K.R. (1997) Elucidation of 1 Sullivan, L.S. and Daiger, S.P. (1996) Inherited GTPasc-activating proteins. GTP hydrolysis and the subunit orientation in CCT (chaperonin retinal degeneration; exceptional genetic and the assembly of the a/P-tubulin heterodimer. containing TCP 1) from the subunit composition of clinical heterogeneity.Mol. Med. TodayZ, 380-386 J. B iol Chem. 274,24054-24058 CCT micro-complexes.EMBOJ. 16,4311-4316 2 Dryja,T.P.etal. (1990) A point mutation of the 23 Radcliffe, PA. e tal (2000) A conserved small 42 Nitsch, M. etal. (1997) The thermosome: rhodopsin gene in one form of retinitis GTP-binding protein AIp41 is essential for the alternatinga and ^-subunits within the pigmentosa.NatureZAZ, 364-366 cofactor-dependent biogenesis of microtubules in chaperonin of the archaeonThermoplasma 3 Berson, E.L. (1996) Retinitis pigmentosa: fission yeast.FEBS Lett. 468,84-88 acidophilum. J. M ol B iol267,142-149 unfolding its mystery.Proc. Natl. Acad. Sci. U. S. A. 24 Bhamidipati, A.etal. (2000) ADP ribosylation 43 Thulasiraman, V.etal. (1999) In wVonewly 93,4526-4528 factor-like protein 2 (Arl2) regulates the translated polypeptides are sequestered in a 4 Roof, D.J.etal. (1994) Rhodopsin accumulation at interaction of tubulin folding cofactor D with protected folding environment.EMBOJ. 18.85-95 abnormal sites in retinas of mice with a human native tubulin.J. Cell Biol 149,1087-1096 44 Llorca, 0. eta! (2000) Eukaryotic chaperonin P23H rhodopsin transgene.Invest. Ophthalmol. 25 Chappie, J.Petal. (2000) Mutations in the CCT stabilizes actin and tubulin folding Vis. Sci. 35,4049-4062 N-terminus of the X-linked retinitis pigmentosa intermediates in open quasi-native 5 Colley, N.J.etal. (1995) Defective intracellular protein RP2 interfere with the normal targeting of conformations.EMBOJ. 19,5971-5979 transport is the molecular basis of rhodopsin the protein to the plasma membrane.Hum. Mol. 45 Nishimura, D.Y.eta l(2001) Positional cloning of dependent dominant retinal degeneration.Proc. Genet. 9,1919-1926 a novel gene on chromosome 16q causing N atl Acad. Sci. U. S. A. 92,3070-3074 26 Breuer, D.K.eta! (2001) Identification of factors Bardet-Biedl syndrome (BBS2).Hum. Mol. 6 Anukanth, A. and Khorana, H.G. (1994) Structure that interact with the RP2 protein involved in Genet. 10,865-874 and function in rhodopsin. Requirements of a X-linked retinitis pigmentosa.Invest. 46 Mykytyn, K.etal. (2001) Identification of the gene specific structure for the intradiscal domain, Ophthalmol. Vis. Sci 42,3509 that, when mutated, causes the human obesity J. Biol Chem. 269,19738-19744 27 Sohocki, M.M.etal. (2000) Mutations in anew syndrome BBS4.Nat. Genet. 28,188-191 7 Humphries, M.M.etal (1997) Retinopathy photoreceptor-pineal gene on 17p cause Leber 47 Barbe, M.F. etal. (1988) Hyperthermia protects induced in mice by targeted disruption of the congenital amaurosis.Nat. Genet. 24,79-83 against light damage in the rat retina.Science rhodopsin gene.Nat. Genet. 15,216-219 28 Ma, Q. and Whitlock, J.P. (1997) A novel 241,1817-1820 8 Baker, E.K. et ai. (1994) The cyclophilin homolog cytoplasmic protein that interacts with the Ah 48 Bunt-Milam,A.H. eta7. (1983) NinaA functions as a chaperone, forming a stable receptor, contains tetratricopeptide repeat motifs, CUnical-uItrastructural study of a retinal complexin wVowith its protein target rhodopsin. and augments the transcriptional response to dystrophy.Invest. Ophthalmol Vis. Sci 24,458-469 EMBOJ. 13,4886-4895 2,3,7,8-tetrachlorodibenzo-p-dioxin.J. Biol. 49 Li, Z.Y. eta l(1998) Rhodopsin transgenic pigs as 9 Stamnes, M.A. etal. (1991) The cyclophilin Chem. 272,8878-8884 a model for human retinitis pigmentosa.Invest. homolog ninaA is a tissue-specific integral 29 Carver, L.A. and Bradfield, C.A. (1997) Ophthalmol Vis. Sci 39, 808-819 membrane protein required for the proper Ligand-dependent interaction of the aiyl 50 Wang, M. etal. (1997) Expression of a mutant synthesis of a subsetDrosophila of rhodopsins. hydrocarbon receptor with a novel immunophilin opsin gene increases the susceptibility of the Ce//65,219-227 homologin vivo. J. Biol. Chem. 272,11452-11456 retina to light damage.Vis. Neurosci. 14,55-62 10 Colley, N.J. etal. (1991) The cyclophilin homolog 30 Meyer, B.K. etal. (1998) Hepatitis B virus 51 Orr, H.T and Zoghbi, H.Y. (2000) Reversing ninaA is required in the secretory pathway.Cell X-associated protein 2 is a subunit of the neurodegeneration: a promise unfolds.Cell 101,1-4 67,255-263 unliganded aryl hydrocarbon receptor core 52 Yvert, C.etal. (2000) Expanded polyglutamines 11 Ferreira, P.A. etal. (1995) Retina-specifically complex and exhibits transcriptional enhancer induce neurodegeneration and trans-neuronal expressed novel subtypes of bovine cyclophilin. activity.Mo! Cell Biol. 18,978-988 alterations in cerebellum and retina of SCA7 J. B iol Chem. 270,23179-23188 31 Petrulis, J R. etal. (2000) Subcellular localization transgenic mice. Hum. Mol. Genet. 9,2491-2506 12 Yokoyama, N.etal (1995) A giant nucleopore of the aryl hydrocarbon receptor is modulated by53 Warrick, J.M. etal. (1999) Suppression of protein that binds Ran/TC4.Nature376. 184-188 the immunophilin homolog hepatitis B virus polyglutamine-mediated neurodegeneration in 13 Ferreira, P.A. etal. (1996) Cyclophilin-related X-associated protein 2.J. Bio! Chem. Drosophila by the molecular chaperone HSP70. protein RanBP2 acts as chaperone for red/green 275, 37448-37453 Nat. Genet. 23,425-428 opsin.Nature383, 637-640 32 Kazlauskas, A.e tal (2000) The 54 Kazemi-Esfaijani, P. and Benzer, S. (2000) 14 Ferreira, PA. etal. (1997) Interconversion of red immunophilin-like protein XAP2 regulates Genetic suppression of polyglutamine toxicity in opsin isoforms by the cyclophilin-related ubiquitination and subcellular localization of the Drosophila. Science287,1837-1840 chaperone protein Ran-binding protein 2. dioxin receptor.J. Biol. Chem. 275,41317-41324 55 Morimoto, R.I. and Santoro, M.C. (1998) Proc. N atl Acad. Sci U. S. A. 94,1556-1561 33 Meyer, B.K. and Perdew, G.H. (1999) Stress-inducible responses and heat shock 15 Deretic, D. e tal (1994) a A- and a B-crystallin in Characterization of the AhR-hsp90-XAP2 core proteins: new pharmacologic targets for the retina. Association with the post-Golgi complex and the role of the immunophilin-related cytoprotection.Nat. Blotechnol. 16,833-838 compartment of frog retinal photoreceptors. protein XAP2 in AhR stabilization.Biochemistry 56 Vigh, L.eta l(1997) Bimoclomol: a nontoxic, ,/. B iol Chem. 269,16853-16861 38,8907-8917 hydroxylamine derivative with stress protein- 16 Schwahn, U.etal. (1998) Positional cloning of the 34 Meyer, B.K.eta! (2000) Aiyl hydrocarbon (Ah) inducing activity and cytoprotective effects.Nat. gene for X-linked retinitis pigmentosa Nat.2. receptor levels are selectively modulated by Med 3,1150-1154 Genet. 19,327-332 hsp90-associated immunophilin homolog XAP2.57 Biro, K.etal. (1998) Bimoclomol improves early 17 Tian, G. etal. (1996) Pathway leading to correctly Cell Stress Chaperones 243-254 electrophysiological signs of retinopathy in folded p-tubulin. Ce//86,287-296 35 Freeman, B.C.etal. (1996) Molecular chaperone diabetic rats. Neuroreport9,2029-2033 18 Tian, G. etal. (1997) Tubulin subunits exist in an machines: chaperone activities of the cyclophilin58 Agashe, V.R. and Hard, F.U. (2000) Roles of activated conformational state generated and Cyp-40 and the steroid aporeceptor-associated molecular chaperones in cytoplasmic protein maintained by protein cofactors.J. Cell Biol. protein p23.ScienceZIA, 1718-1720 folding. Semin. Cell Dev. Biol. 11,15-25 138, 821-832 36 Pratt, W.B. and Toft, D.O. (1997) Steroid receptor 59 Cothel, S.F. and Marahiel, M.A. (1999) 19 Hardcastle, A.J. etal. (1999) Mutations in the RP2 interactions with heat shock protein and Peptidyl-prolyl cis-trans isomerases, a gene cause disease in 10% of families with familial immunophilin chaperones.Endocr. Rev. 18,306-360 superfamily of ubiquitous folding catalysts.Cell X-linked retinitis pigmentosa assessed in this 37 Stone, D.L.etal. (2000) Mutation of a gene M ol Life Sci 55, 423-436 study.Am. J. Hum. Genet. 64,1210-1215 encoding a putative chaperonin causes McKusick-60 Cheetham, M.E. and Caplan, A.J. (1998) 20 Sharon, D.etal. (2000) X-linked retinitis Kaufman syndrome.Nat. Genet. 25,79-82 Structure, function and evolution of DnaJ: pigmentosa: mutation spectrum of the RPGR and 38 Slavotinek, A.M.etal. (2000) Mutations in MKKS conservation and adaptation of chaperone RP2 genes and correlation with visual function. cause Bardet-Biedl syndrome.Nat. Genet. 26,15-16 function.Cell Stress. Chaperones. 3,28-36 Invest. Ophthalmol Vis. Sci 41,2712-2721 39 Braig, K. (1998) Chaperonins.Curr. Opln. Struct. 61 Blatch, C.L. and Lassie, M. (1999) The 21 Liang, P. and MacRae, TH. (1997) Molecular B iol 8,159-165 tetratricopeptide repeat: a structural motif chaperones and the cytoskeleton.J. Cell Sci 40 Bukau, B. and Horwich.A.L. (1998) The Hsp70 mediating protein-protein interactions.Bioessay. 110 (Pt 13), 1431-1440 and Hsp60 chaperone machines.Cell9Z, 351-366 21,932-939

http://tmm.trends.com