Rescue of photoreceptor function by AAV-mediated gene ... - Nature

Gene Therapy (1997) 4, 683–690  1997 Stockton Press All rights reserved 0969-7128/97 $12.00

Rescue of photoreceptor function by AAV-mediated gene transfer in a mouse model of inherited retinal degeneration C Jomary1, KA Vincent2, J Grist 1, MJ Neal1 and SE Jones1 1

British Retinitis Pigmentosa Society Laboratory, Department of Pharmacology, UMDS, The Rayne Institute, St Thomas’ Hospital, London SE1 7EH, UK; and 2Genzyme Corporation, Framingham, MA, USA

Knowledge of the mutations leading to inherited retinal degenerations provides a foundation for the development of somatic gene therapy in which potentially corrective genes are transferred to the target photoreceptor cells. Towards this end, we have evaluated the efficacy of a recombinant adeno-associated virus (AAV) vector to deliver and express the correct form of the cGMP phosphodiesterase-b (PDE-b) gene in the retinas of rd mice, which suffer rapid retinal degeneration due to recessive mutation in the endogenous gene. A truncated murine opsin

promoter was used to drive expression of the PDE-b cDNA. Following intraocular injection of AAV.PDE-b, increased retinal expression of immunoreactive PDE protein was observed, including within photoreceptor cell bodies. Compared with age-matched controls, treated eyes showed increased numbers of photoreceptors and a twofold increase in sensitivity to light as measured by in vitro electroretinography. These findings provide evidence that rescue of functional photoreceptor neurons can be achieved by somatic gene therapy.

Keywords: adeno-associated virus; photoreceptor; retinal degeneration

Introduction The human photoreceptor cell is a specialized neuron which is the critical site of cellular dysfunction in retinitis pigmentosa (RP), the most prevalent group of retinal genetic diseases. RP is characterised by the development of night blindness due to the progressive loss of the rod photoreceptors, leading to contraction of the visual fields, and reduced or absent electroretinogram. RP affects up to one in 3000 in the populations studied,1,2 and is clinically and genetically heterogeneous, with autosomal dominant and recessive, as well as several X-linked forms. More than 30 chromosomal loci associated with RP and other retinal diseases have been mapped, 3 and mutations in at least 10 genes identified. Consequent on these findings is the possibility of treating certain of these diseases by delivering a functional gene directly to the affected cells. At present, viral vectors are the most efficient gene transfer vehicles for neural retina transduction. Replication-defective adenoviruses can efficiently deliver genes to retinal cells in vivo,4–6 and complementation of the mutant gene via adenovirus-mediated delivery to the retinal degeneration (rd) mouse eye can slow the rate of photoreceptor loss in this model.7 However, adenoviruses are naturally pathogenic in humans, and their use as gene transfer vehicles has been complicated by adverse immunological reactivity and cytotoxicity,8 including following intravitreal injection in dogs.9

Correspondence: SE Jones Received 8 January 1997; accepted 7 March 1997

Gene transfer vectors based on adeno-associated virus (AAV) are emerging as promising vehicles which may overcome some of the limitations of other viral systems.10,11 The AAV genome is a linear single-stranded DNA molecule of 4680 nucleotides, but, for the purposes of gene transfer, the viral genes may be removed leaving only the 145-base inverted terminal repeat sequences (ITRs) flanking the transgene construct. The ITRs are sufficient to ensure packaging of inserted DNA into infectious virions when AAV and adenoviral helper functions are provided in trans. As a potential vector for gene therapy, AAV has the further advantages that human infection has not been associated with any disease, the virus is capable of efficiently infecting both dividing and postmitotic cells including neurons, and is able to persist in cells for long periods. Presently, AAV-based gene transfer protocols are under evaluation preclinically for several genetic diseases, including haemolytic anaemias, and at the clinical level for cystic fibrosis. 11 We and others have demonstrated the ability of AAV to mediate reporter gene transfer to the retina.12–14 We have proceeded to evaluate a recombinant AAV (rAAV) construct in a model of degenerative retinal disease, the rd mouse. In homozygous (rd/rd) animals, degeneration of the rod photoreceptors occurs during the first 3 postnatal weeks of life, due to recessive mutation of the bsubunit of the visual transduction enzyme, cGMP phosphodiesterase (PDE).15 The production of a truncated PDE-b protein causes loss of function of the holoenzyme, resulting in elevated retinal cGMP and leading to rapid photoreceptor degeneration. The relevance of this model to understanding human retinal degenerations has been underscored by the finding of mutations in both the b-

AAV-mediated photoreceptor rescue C Jomary et al

684

and the a-subunit of PDE in families with recessively inherited RP.16,17 In addition, transgenic rd/rd mice in which expression of wild-type, bovine PDE-b was targeted to the photoreceptors using the opsin promoter, showed rescue of these cells from degeneration.18 This suggested that an approach using somatic delivery of the corrective gene might also provide a therapeutic effect. We have constructed a rAAV to deliver the human PDEb cDNA, under the control of a segment of the murine opsin promoter, to the retinas of degenerative rd mice and control strains, and used a series of evaluative measures to establish the potential for this approach to mediate rescue of the degenerative photoreceptors, particularly in terms of their functional response to light.

Results Expression of the transgene The plasmid pTOP8 (Figure 1), containing the full-length human PDE-b cDNA driven by 1.4 kb of the murine opsin promoter, was used to generate the AAV.PDE-b for intravitreal injection. Expression of the PDE-b transgene was initially evaluated following treatment of adult nondegenerative (rd/+) animals. PDE protein expression was detected by immunocytochemistry on frozen tissue sections using a rabbit polyclonal anti-PDE antibody (a gift of Dr Y-K Ho, University of Illinois, Chicago) (Figure 2). This antibody does not distinguish between the mouse and human proteins, and reacts strongly with the a- and b-subunits and weakly with the g-subunit of the enzyme.19 Endogenous PDE protein was detected at the outer segments of the photoreceptors in untreated rd/+ retina (Figure 2a). Following AAV.PDE-b injection, immunoreactivity was similarly observed at the photoreceptor outer segments but with much greater intensity of staining (Figure 2b). Most interestingly, in the vicinity of the injection site, intense immunostaining extending from the photoreceptor nuclei to the outer segments was observed. Slight immunoreactivity was detected at the plexiform layers and nerve fibre layer in both injected and non-injected animals. Control sections lacking primary antibody displayed none of the above reactivity (data not shown). In the homozygous rd mouse, the effect of the mutant PDE-b gene results in only one row of photoreceptor cell nuclei remaining at 21 days after birth (Figure 3a) compared with eight to 10 rows in the heterozygote.15 In non-

Figure 1 Representation of part of the plasmid clone pTOP8 used to generate recombinant AAV.PDE-b. 1.4 kb of the murine opsin promoter was cloned upstream of the full-length (2.6 kb) human PDE-b cDNA between the inverted terminal repeat sequences (ITRs) and SV40 polyadenylation signal of the vector pTR. Positions of the oligonucleotide primers for reverse transcription (APDETG), and for PCR (MUSOPS4 and HPDE6) are shown as filled arrows; open arrow indicates the origin of transcription.

injected eyes, slight PDE immunoreactivity was detected in the outer and inner plexiform layers (Figure 3a). Similar results were obtained in the PBSS-injected eyes (data not shown). In contrast, up to three rows of photoreceptor nuclei were observed in retinas of AAV.PDE-binjected animals, and increased PDE immunostaining was observed in the outer and inner plexiform layers, and at the nerve fibre layer (Figure 3b). In parallel studies (unpublished), AAV-mediated delivery of the lacZ reporter gene was examined in rd mice, using similar viral titres and injection and analytical protocols. We found no evidence of rescue of photoreceptors in these treated eyes by either histological or electrophysiological assessments. We examined the expression of the PDE-b transgene at the mRNA level using the reverse transcription polymerase chain reaction (RT-PCR) technique. Following amplification the PCR products were separated by agarose gel electrophoresis in the presence of ethidium bromide. The gel was then blotted and hybridized with a human PDE-b cDNA probe to confirm that the positive reactions corresponded to specific amplification from the transgene. As shown in Figure 4, expression was detectable in retinas from AAV.PDE-b-injected eyes, in both rd/rd and C57BL/6 mice, at up to 6 weeks after injection, the latest time-point studied. Non-injected and PBSSinjected control eyes were negative for transgene expression. To assess the possibility of AAV-mediated gene transfer to extraocular tissues, the analysis was also performed on total RNA from the brain, liver, heart and lung of virus-injected animals, also at 6 weeks after injection. No expression of the transgene was detectable in any of these tissues (data not shown). That negative results were not due to degradation of RNA was determined by separate RT-PCR experiments in which the detection of transferrin receptor mRNA was confirmed (not shown).

In vitro electroretinography To determine the therapeutic effect of transgene expression on retinal function, we used an in vitro electroretinography assay to measure the responsiveness of the retina to light. Retinas from untreated control C57BL/6 mice (n = 4), and from rd/rd non-injected (n = 23), PBSSinjected (n = 4) and AAV.PDE-b-injected (n = 23) animals were compared. Since individual variation in fast PIII response amplitudes has been reported,20 the data were normalized to facilitate comparison. The PIII amplitude– response curves (V-log I function) for the tested retinas are illustrated in Figure 5. The fast PIII responses from control C57BL/6 retinas reached saturation (Vmax) at a light intensity 1.2 log units less than in the rd/rd control and injected mice. Fast PIII response threshold in rd/rd animals was 1.2 log units higher than in control C57BL/6. The reduced sensitivity of the non-injected rd/rd retinas to light compared with the non-injected C57BL/6 is demonstrated by a shift of the V-log I function of the rd/rd retinas to the right, with a 1.99 log unit decrease in the s value (light filter density at 50% Vmax) compared with the C57BL/6. No significant light sensitivity difference was observed between PBSS-injected and non-injected rd/rd animals. In contrast, for retinas from AAV.PDE-binjected rd/rd mice, there was a 0.343 log unit increased light sensitivity, significantly different from that of control non-injected or PBSS-injected mice (P , 0.005). This


685

Figure 2 Immunostaining for PDE in retinal sections of adult rd/+ mice. Frozen retinal sections 10 mm from non-injected (a) and AAV.PDE-b-injected (b) animals were processed for immunocytochemistry using anti-PDE antibody. In treated mice, immunoreactivity is observed in the outer segment of photoreceptor cells (black arrow) and at photoreceptor cell nuclei (white arrow). In (b), outer segment damage was due to cryo-sectioning. Original magnification × 1640.

Figure 3 Immunostaining for PDE in 10 mm frozen retinal sections of 21-day-old rd/rd mice. Sections from non-injected (a) and AAV.PDE-b-injected (b) animals. Note increased thickness of photoreceptor outer nuclear layer (arrows) in (b). Original magnification × 1720.


686

Figure 4 Detection of transgene expression by RT-PCR of total RNA from individual retinas at 6 weeks following intraocular injection. RNA samples were incubated with (+) or without (−) reverse transcriptase and subjected to amplification using specific primers. PCR products were blotted and hybridized with a 32P-labelled human PDE-b cDNA probe and autoradiographed. Retinal RNA was derived from: lane 1, PBSS-injected eye; lane 2, non-injected eye; lanes 3–5, AAV.PDE-b-injected eyes from three rd/rd animals; lane 6, one AAV.PDE-b-injected rd/+ eye. Lane 7 was the PCR negative control (water). Samples run in lanes 2 and 3 were derived from the same treated animal. The positive control (C) was obtained by direct amplification from pTOP8 plasmid.

Figure 5 Light sensitivity of isolated retinas measured by in vitro electroretinography, expressed as normalized fast PIII amplitude–response curves (V-log I function) of 18-day-old control C57BL/6 (n = 4: –r–), rd/rd non-injected (n = 23: –p–), rd/rd PBSS-injected (n = 4: –l–) and rd/rd AAV.PDE-b-injected (n = 23: –g–) animals. Error bars: ± s.e.m.

difference corresponds to an approximately two-fold increase in sensitivity of the retinas from AAV.PDE-btreated versus control rd/rd eyes.

Histological analyses To assess whether the enhanced light sensitivity of the treated rd/rd retinas correlated with histomorphological characteristics, retinas used for in vitro electroretinography were subsequently examined by light and electron microscopy (Figures 6 and 7). Incubation of isolated retinas in the Krebs-buffer medium was somewhat detrimental to morphological preservation (Figure 6a, b), compared with retinas fixed immediately after killing (Figure 6c, d). However, no morphological differences were observed in PBSS-injected animals (Figure 6d), in comparison with control, non-injected animals (Figure 6c). Notably, in AAV.PDE-b-injected rd/rd eyes, the number of rows of photoreceptor nuclei (3–4, Figure 6b) was increased approximately two-fold compared with noninjected eyes (1–2, Figure 6a). No significant difference was observed at the inner nuclear layer level. Photoreceptor segments were frequently detectable by light microscopy in AAV.PDE-b-treated animals (Figure 6b), compared with only isolated instances in non-injected

animals (Figure 6a). The electron microscopy examination confirmed this difference. In virus-treated retinas, the inner photoreceptor segments showed characteristic accumulation of mitochondria and Golgi cisternae (Figure 7). In addition, compact piles of disc structures were frequently observed in outer photoreceptor segments of the treated animals (Figure 7 inset). Cone nuclei surrounded by large amounts of cytoplasm with numerous large mitochondria were also preponderant. Photoreceptor outer segment material was rarely seen in analysis of comparable regions of retinas from control rd/rd mice.

Discussion We report here the first demonstration that an AAV vector can transfer and express a therapeutic gene in the murine retina. Significant physiological photoreceptor response can be rescued following injection of a human PDE-b AAV construct in the rd/rd mouse model of inherited retinal degeneration, and this rescue is correlated with increased survival of photoreceptors. During the course of our study, a similar approach using adenovirus-mediated transfer in the rd mouse was reported.7 However, the disadvantages of adenoviruses have been indicated, and further, while slowing of degeneration occurred, the function of the residual photoreceptors was not investigated. Additionally, the expression of the PDE-b was driven by the cytomegalovirus (CMV) immediate–early promoter, which is not specific for photoreceptor cells. We have designed a rAAV construct in which expression of human PDE-b is driven by a 1.4 kb promoter segment of the photoreceptor-specific murine opsin gene, in order to attempt to counteract the damaging effect of the mutated endogenous protein in the rd mouse photoreceptors by somatic transfer and expression of the PDE-b subunit. Localization of PDE has previously been reported to be only in the photoreceptor outer segments of non-degenerative mouse retina.21 In our investigation, intraocular delivery of the AAV.PDE-b construct in the nondegenerative rd/+ heterozygote results in a higher level of immunodetectable PDE at the photoreceptor outer segments, as well as at the photoreceptor nuclei level, in comparison with controls. These results suggest that the rAAV construct is capable of transducing murine photoreceptors. Elevated PDE protein expression is also observed in AAV.PDE-b-injected rd/rd animals, though to a lesser degree, presumably due to still-extensive outer segment loss. However, the different localization, in the plexiform layers and nerve fibre layer, may be due to the loss of structural and biochemical polarity of the photoreceptor cells undergoing degeneration in this model, as previously postulated.22 Misrouting of the visual transduction proteins rhodopsin, transducin and PDE has also been reported in transgenic mice carrying a mutant rhodopsin gene.21 At the transcriptional level, in situ hybridization for transgene mRNA will assist in verifying constraint of expression to the surviving photoreceptors, since promoter specificity may be influenced by somatic versus germline delivery. It is nevertheless likely that the increased PDE immunoreactivity in rAAV-injected eyes results from translation of transgene mRNA, and transcripts remain detectable for at least 6 weeks after administration in both heterozygous and homozygous mice. Taken together, these


687

Figure 6 Light micrographs of 1 mm retinal sections of 18-day-old rd/rd mice. (a, b) following in vitro electroretinography; (c, d) not processed for electroretinography. Sections from non-injected animals (a, c), PBSS-injected (d) and AAV.PDE-b-injected (b) mice are shown. The solid arrows highlight the number of rows of photoreceptor nuclei. The open arrows in (b) indicate photoreceptor segments. Original magnification × 1770.

results confirm that the AAV.PDE-b vector is able to transfer and express the transgene in vivo in the murine retina. Compared with non-injected and vehicle-injected animals, AAV.PDE-b-injected rd/rd mouse retinas have a greater survival of photoreceptor cells, which correlates with preservation of retinal function as assessed by electroretinography. This was not due to a nonspecific influence of injection since neither virus suspension buffer nor (in previous studies) similar doses of AAV-lacZ, were found to have any detectable rescue effect. The increased amount of outer segment material in AAV.PDE-b-treated animals implies that there are more photoreceptor Na+ channels, capable as a result of higher photocurrents and increased Vmax for a light stimulus of given intensity. The decreased semi-saturation luminance further suggests an increase in retinal quantal catch, and greater light sensitivity indicates a higher level of rhodopsin in photoreceptors of rAAV-injected rd/rd mice. A 1.0 log unit loss of sensitivity has been shown to correspond to a loss of approximately 50% of rhodopsin.23 Therefore, the extent of rescue could be estimated to be about 17% of the normal (C57BL/6) photoreceptor function in rAAV-injected rd/rd animals. This corresponds to an approximately twofold increase in light sensitivity of the retina, in agree-

ment with the approximate doubling of the number of photoreceptor nuclear layers across much of the treated retinas. Our present study provides an evaluation of the functional rescue at a single dose and early post-injection time-point, but clearly dose–response and time-course experiments will be needed to determine the maximum and long-term functional recovery in this mouse model. Wild-type AAV is capable of integrating into the host genome following infection,10,11 primarily in a sitespecific manner at chromosome 19 in human cells. The possibility of recombinant AAV integration in vivo remains to be evaluated, although in vivo persistence in part as an episome has recently been reported.24 In our study, all treated animals remained clinically healthy throughout the experiments, suggesting that AAV.PDEb was not acutely toxic at the dose used. It will nevertheless be important to assess the need for and possibility of repeated delivery of the vector and to determine any potential humoral immune response. Indeed, a more thorough evaluation of safety, toxicity and host response aspects will be required, as has been done for in vivo delivery of the cystic fibrosis transmembrane regulator (CFTR) gene.24 In particular, the possibility that the recombinant virus may be systemically spread should be assessed by analysing the vector DNA distribution and


688

cassette was then cloned into the AAV vector pTR (source: Dr N Muzyczka, University of Florida, Gainesville, USA). The structure and orientation of the final construct (pTOP8, Figure 1) were confirmed by restriction digest analysis.

Preparation of the recombinant AAV Recombinant AAV virions were produced using standard protocols.28 Briefly, the insert and ITRs of pTOP8 were packaged into AAV particles by complementation with a system expressing AAV rep and cap genes in adenovirus (Ad5ts149)-infected 293 cells (ECACC, Salisbury, UK). Recombinant virus was then harvested, purified by double caesium chloride density gradient centrifugation, dialysed, and the viral titre determined by infectious centre assay.29 No helper adenovirus was detectable in purified rAAV by immunocytochemistry using anti-adenoviral antibody (Chemicon International, Harrow, UK) on AAV-infected 293 cells.

Figure 7 Electron micrographs of 18-day-old rd/rd mouse retina following treatment with AAV.PDE-b. Indicated are C, cone cell nucleus; is, photoreceptor inner segment; os, photoreceptor outer segment; R, rod photoreceptor nucleus. Open arrow, photoreceptor cilium; solid arrow, outer limiting membrane. Original magnification × 7470. Inset: detail of an outer and an inner photoreceptor segment at higher magnification (× 13 500).

potential pathological effects. While our initial studies did not detect any transgene expression in extraocular tissues, there may, nevertheless, be extraocular infection by rAAV without expression, and, given the possibility of insertional mutagenesis following integration, this aspect must be more thoroughly examined. Evaluation of the inflammatory cytokines IL-6 or IL-8 will also give an estimation of the cellular immunological response, and seropositivity for AAV will be established before and after rAAV injections. One further aspect to be considered is that mobilization of the CFTR rAAV has been described in vitro.24 However, Afione and collaborators25 have shown that vector rescue rarely occurs following in vivo wild-type AAV2 and adenovirus coinfection. Nevertheless, it will be essential to determine potential rescue and spread to other organs in animals intraocularly injected with AAV.PDE-b. These initial issues should be carefully examined in appropriate models to constitute the basis of preclinical studies towards eventual clinical trials.

Materials and methods Assembly of the pTOP8 plasmid The human PDE-b cDNA full-length coding sequence was amplified by ligation PCR from two cDNA clones26 (gift of Dr M Hayden, University of British Columbia, Vancouver, Canada), using sequence-specific primers and 28 cycles of PCR amplification. The resulting 2.6 kb product was cloned into the vector pCRII (Invitrogen BV, Leek, The Netherlands) and then transferred into plasmid pRho4–4 containing 4.4 kb of the murine opsin promoter27 (gift of Dr J Chen, California Institute of Technology, Pasadena, USA). A truncated promoter–cDNA

Virus delivery All animal procedures were carried out in accordance with the requirements of a Home Office Licence. Eightday-old C57BL/6, rd/+, and rd/rd mice, anaesthetized with Hypnorm (Janssen Pharmaceutical, Oxford, UK) and diazepam (Phoenix Pharmacia, UK), were injected intravitreally with 2 ml of AAV.PDE-b at a titre of approximately 1 × 107 infectious units/ml, using a fine microcapillary needle directed posteriorly through the limbus. Virus resuspension buffer (phosphate-buffered saline containing sucrose at 5% w/v; PBSS) was used for control injections. One eye only was injected per animal, the other serving as an untreated control. Immunocytochemistry Enucleated eyes were fixed in fresh 4% paraformaldehyde in 0.1 m phosphate buffer overnight, embedded and frozen.30 Cryostat sections (10 mm) were processed for immunocytochemistry using the classical peroxidaseanti-peroxidase technique.31 Primary antibody (1:500 dilution) used was a polyclonal anti-PDE (gift from Dr Y-K Ho, Chicago, USA). Reverse transcription PCR Total RNA was isolated from dissected, frozen retinas and portions of extraocular tissues using an RNeasy kit (Qiagen, Crawley, UK), and treated with DNaseI (Pharmacia Biotech, St Albans, UK) to degrade genomic DNA. Samples were then reverse transcribed in 20 ml reactions using 3.8 units of AMV reverse transcriptase (Promega, Southampton, UK) and specific antisense primer APDETG (5’ CTGGAATTCGGCTTGTCCCCACGGGACCAGT 3′), for 90 min at 42°C, followed by heat inactivation of the enzyme. Controls for genomic DNA contamination were identically processed but in the absence of enzyme. Primers for amplification of the transgene were MUSOPS4: 5′ AGCTCGCCAAGCAGCCTTGGT 3′ and HPDE6: 5′ CCTTGAAGACCACGCGCTCCA 3′. Amplifications were performed in 30 ml reactions using 0.75 units Taq polymerase (GibcoBRL, Paisley, UK) in a thermal cycler (Hybaid, Teddington, UK) with initial denaturation of 94°C for 3 min, followed by cycling parameters of (94°C 45 s, 56°C 45 s, 72°C 30 s) × 40, and final step of 72°C for 5 min. Products were analysed by agarose gel electrophoresis, blotted on to


nylon membrane (Amersham, Little Chalfont, UK), hybridized with a fragment of human PDE-b cDNA labelled with 32P dCTP (ICN Biomedicals, Thame, UK) using a Redi-Prime kit (Amersham), and autoradiographed at −70°C. For controls to confirm RNA integrity, reverse transcription was initiated using random hexanucleotide primers (Pharmacia Biotech) and amplification performed using primers for the ubiquitously expressed transferrin receptor: MMTR.1, 5′ TGAGGCTGGATCTCAAA AAG 3′ and MMTR.2, 5′ CCTTTAATAACTCCAAAGAT 3′, to yield a product of 720 bp. Conditions were as above but with an annealing temperature of 54°C, and products analysed by ethidium bromide-stained agarose gel electrophoresis.

In vitro electroretinography In vitro electroretinography was performed using the method previously described20 on rd/rd mouse retinas (17–18 days old) non-injected (n = 23), and injected either with PBSS (n = 4), or AAV.PDE-b (n = 23). Retinas of C57BL/6 mice (n = 4) of matching age were normal controls. Transretinal fast PIII photoreceptor responses were recorded from freshly dissected retinas of dark-adapted mice. ERG recordings were then analysed using previously defined methods.20 Statistical analysis was performed using Student’s t test. Retinal morphology After ERG recording, the retinas were stored in 4% paraformaldehyde solution until fixation in 3.5% glutaraldehyde, followed by processing for light and electron microscopy.32

Acknowledgements We gratefully acknowledge the support of this work by grants from the British Retinitis Pigmentosa Society, the Guide Dogs for the Blind Association, and the Iris Fund for Prevention of Blindness. We thank Dr Sam C Wadsworth and Dr Alan E Smith of Genzyme Corporation for their collaborative contributions. For the provision of clones, we thank Drs N Muzyczka, J Chen and MR Hayden, and for anti-PDE antibody, Dr Y-K Ho. For assistance with microscopy, we are grateful to Ms A Patmore and Dr S Pathak, and for technical help we thank Mr D Thomas. Dr AA Hussain provided us with the equipment to perform in vitro electroretinography. Lastly, we are indebted to Prof J Marshall for advice during the course of this study.

References 1 Heckenlively JR. The diagnosis and classification of retinitis pigmentosa. In: Heckenlively JR (ed). Retinitis Pigmentosa. Lippincott: Philadelphia, 1988, pp 6–24. 2 Kaplan J et al. Clinical and genetic heterogeneity in retinitis pigmentosa. Hum Genet 1990; 85: 635–642. 3 Papermaster DS. Necessary but insufficient. Nature Med 1995; 1: 874–875. 4 Bennett J et al. Adenovirus vector-mediated in vivo gene transfer into adult murine retina. Invest Ophthalmol Vis Sci 1994; 35: 2535–2542. 5 Li T et al. In-vivo gene-transfer to normal and rd mouse retinas mediated by adenoviral vectors. Invest Ophthalmol Vis Sci 1994; 35: 2543–2549. 6 Jomary C et al. Adenovirus-mediated gene transfer to murine retinal cells in vitro and in vivo. FEBS Lett 1994; 347: 117–122.

7 Bennett J et al. Photoreceptor cell rescue in retinal degeneration (rd) mice by in vivo gene therapy. Nature Med 1996; 2: 649–654. 8 Zabner J et al. Safety and efficacy of repetitive adenovirusmediated transfer of CFTR cDNA to airway epithelia of primates and cotton rats. Nat Genet 1994; 6: 75–83. 9 Maguire AM et al. Adenovirus-mediated gene-transfer to canine photoreceptors – effects of inflammation. Invest Ophthalmol Vis Sci 1995; 36: S777 (Abstr. 3591). 10 Xiao X, deVlaminck W, Monahan J. Adeno-associated virus (AAV) vectors for gene transfer. Adv Drug Delivery Rev 1993; 12: 210–215. 11 Flotte TR, Carter BJ. Adeno-associated virus vectors for gene therapy. Gene Therapy 1995; 2: 357–362. 12 Jomary C et al. Adenoassociated virus vector-mediated gene transfer to retinal cells in vitro and in vivo. Invest Ophthalmol Vis Sci 1995; 36: S772 (Abstr. 3569). 13 Hauswirth WW, Zolotukhin S, Muzyczka N, Flannery JG. Adeno-associated virus delivery of an opsin promoter driven reporter gene to the mouse and rabbit retina. Invest Ophthalmol Vis Sci 1995; 36: S845 (Abstr. 3884). 14 Ali RR et al. Gene transfer into the mouse retina mediated by an adeno-associated viral vector. Hum Mol Genet 1996; 5: 591–594. 15 Farber DB, Flannery JG, Bowes-Rickman C. The rd mouse story: seventy years of research on an animal model of inherited retinal degeneration. In: Osborne NN, Chader GJ (eds). Progress in Retinal and Eye Research. Pergamon Press: Oxford, 1994, vol. 13, pp 31–64. 16 McLaughlin ME, Sandberg MA, Berson EL, Dryja TP. Recessive mutations in the gene encoding the beta subunit of rod phosphodiesterase in patients with retinitis pigmentosa. Nat Genet 1993; 4: 130–134. 17 Huang SH et al. Autosomal recessive retinitis pigmentosa caused by mutations in the a subunit of rod cGMP phosphodiesterase. Nat Genet 1995; 11: 468–471. 18 Lem J et al. Retinal degeneration is rescued in transgenic rd mice by expressing the cGMP phosphodiesterase beta-subunit. Proc Natl Acad Sci USA 1992; 89: 4422–4426. 19 Hingorani VN, Tobias DT, Henderson JT, Ho Y-K. Chemical cross-linking of bovine retinal transducin and cGMP phosphodiesterase. J Biol Chem 1988; 263: 6916–6926. 20 Leon A, Hussain AA, Curtis R. Autosomal dominant rod-cone dysplasia in the Rdy cat: 2. Electrophysiological findings. Exp Eye Res 1991; 53: 489–502. 21 Roof DJ, Adamian M, Hayes A. Rhodopsin accumulation at abnormal sites in retinas of mice with a human P23H rhodopsin transgene. Invest Ophthalmol Vis Sci 1994; 35: 4049–4062. 22 Cantera R et al. Postnatal development of photoreceptor-specific proteins in mice with hereditary retinal degeneration. Exp Biol 1990; 48: 305–312. 23 Ernst W, Kemp CM. The effects of rhodopsin decomposition on PIII responses of isolated rat retinae. Vision Res 1972; 12: 1937–1946. 24 Carter BJ, Flotte TR. Development of adeno-associated virus vectors for gene therapy of cystic fibrosis. In: Berns KI, Giraud C (eds). Adeno-associated Virus (AAV) Vectors in Gene Therapy. Springer-Verlag: Berlin, 1996, pp 119–144. 25 Afione S et al. AAV-CFTR vector rescue in Rhesus monkeys. Pediatr Pulmonol 1994; 10 (Suppl.): 220. 26 Collins C et al. The human b-subunit of rod photoreceptor cGMP phosphodiesterase: complete retinal cDNA sequence and evidence for expression in brain. Genomics 1992; 13: 698–704. 27 Woodford BJ, Chen J, Simon MI. Expression of rhodopsin promoter transgene product in both rods and cones. Exp Eye Res 1994; 58: 631–635. 28 Kaplitt MG et al. Long term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994; 8: 148–154. 29 Einerhand MPW et al. Regulated high-level human b-globin gene expression in erythroid cells following recombinant adenoassociated virus-mediated gene transfer. Gene Therapy 1995; 2: 336–343.

689


690

20 Jomary C et al. Spatio-temporal pattern of ocular clusterin mRNA in the rd mouse. Mol Brain Res 1995; 29: 172–176. 31 Robinson G, Ellis IO, MacLennan KA. Immunocytochemistry. In: Bancroft JD, Stevens A (eds). Theory and Practice of Histological Techniques, 3rd edn. Churchill-Livingstone: Edinburgh, 1990, pp 413–436.

32 Marshall J, Hamilton AM, Bird AC. Histopathology of ruby and argon laser lesions in monkey and human retina: a comparative study. Br J Ophthalmol 1972; 59: 610–630.