Received 10 February 2009 | Accepted 21 June 2009 | Published 17 July 2009
1Department of Ophthalmology, the First Affiliated Hospital of Chongqing Medical University, Chongqing Key Laboratory of Ophthalmology, Chongqing, China; 2Department of Veterinary Medicine and Surgery, Department of Ophthalmology, Mason Eye Institute, University of Missouri, Columbia, MO; 3Department of Molecular Microbiology and Immunology, University of Missouri, Columbia, MO
Correspondence to: Bo Lei, Department of Ophthalmology, The First Affiliated Hospital of Chongqing Medical University, 1 You Yi Road, Yu Zhong District, Chongqing, 400016, China; Phone: 0086 23 68485440; FAX: 0086 23 68485440; email: firstname.lastname@example.org
Purpose: Adeno-associated virus serotype-9 (AAV-9) is a promising gene delivery vector. In this study, we evaluated AAV-9 transduction in the mouse retina.
Methods: Three different AAV vectors were used in our study: AAV-9.RSV.AP, AAV-9.CMV.eGFP, and AAV-9.CMV.∆R4–23/∆C. In these vectors, two different promoters (the cytomegalovirus promoter-CMV promoter and the Rous sarcoma virus-RSV promoter) were used to express three different transgenes including the alkaline phosphatase (AP) gene, the enhanced green fluorescent protein (eGFP) gene, and a therapeutic microdystrophin gene (the ∆R4–23/∆C). Specifically, 1 µl AAV-9 reporter gene vectors (1×109 viral genome particles of AAV-9.RSV.AP or 1×1010 viral genome particles of AAV-9.CMV.eGFP) were administered subretinally to young (2–3-week-old), adult (3-month-old), and old (12-month-old) C57BL/6J mice. To evaluate AAV-9 transduction in a diseased retina, we injected subretinally 1×109 viral genome particles of AAV-9.CMV.∆R4–23/∆C to mdx3cv mice, which we used as a model for Duchenne muscular dystrophy (DMD). Transgene expression was examined by histochemical as well as immunofluorescence staining at three and five weeks after injection. Electroretinograms were recorded five weeks after subretinal AAV-9.RSV.AP injection.
Results: Subretinal injection yielded widespread transduction throughout the retina in all age groups. Robust expression was seen in the retinal pigment epithelium, outer nuclear layer, and in Müller cells. Interestingly a synaptic layer, the outer plexiform layer (OPL), also showed intensive expression. Transduction of the synaptic layer was further confirmed by immunostaining for C-terminal binding protein 2 (CtBP2), a marker for the photoreceptor synaptic ribbon. Dystrophin is normally expressed in the OPL photoreceptor terminals. This expression is lost in DMD patients and mdx3cv mice. Consistent with our findings in normal mice, we observed efficient microdystrophin expression in the OPL after AAV-9.CMV.∆R4–23/∆C infection. At five weeks after subretinal delivery of AAV-9.RSV.AP, no morphology or ERG abnormalities were observed.
Conclusions: We demonstrated that AAV-9 is a potent vector for retinal gene delivery. Furthermore, subretinal AAV-9 administration did not cause appreciable acute retinal damages. In summary, AAV-9-mediated OPL transduction holds promise for treating diseases that primarily affect this layer.
Adeno-associated virus (AAV) is a single-stranded DNA virus. AAV-mediated gene therapy has successfully corrected several degenerative retinal diseases in animal models [1-4]. Recent successes in Leber congenital amaurosis phase I trials have provided the first clinical proof that AAV vector holds great promise in treating retinal diseases [5-8].
Recombinant AAV vectors are generated by replacing the endogenous viral genome with a therapeutic or marker gene expression cassette. The prototype AAV vector is based on AAV serotype-2 (AAV-2). In the last few years, several new AAV serotypes have been developed . Due to the differences in the cellular transduction pathway, these new serotypes have opened additional avenues for tailoring AAV gene therapy to specific clinical applications. Previous studies have evaluated AAV serotypes 1–9 in the retina [3,10-13]. It was found that subretinal injection transduces the outermost retinal structures, such as retinal pigment epithelium (RPE) and the photoreceptors, while intravitreal injection transduces ganglion cells in the innermost layer. Pathology in the retinal synaptic layers such as the outer plexiform layer (OPL) is associated with a wide range of retinal diseases [14-18]. Two recent studies suggest that AAV may transduce the OPL [12,19]. However, these studies did not explicitly explore the specificity of OPL transduction. Considering the importance of the OPL in retinal diseases, a thorough and more focused study would be necessary to establish AAV transduction in the OPL. Such efforts would likely open the door for AAV-mediated OPL disease gene therapy.
AAV serotype-9 (AAV-9) was discovered a few years ago from human tissue [9,20]. Due to its unique serological property, it was classified as clade F, a clade distinctive from all known AAVs . AAV-9 has been shown to efficiently transduce several tissues including the heart, liver, lung, kidney, pancreas, and skeletal muscle [21-27]. Furthermore, it was reported recently that AAV-9 is capable of bypassing the blood-brain barrier and efficiently targeting cells of the central nervous system . This unique feature may enable the development of gene therapies for a range of neurodegenerative diseases. Two studies evaluated AAV-9 transduction in the retina following subretinal administration [12,13]. Both studies demonstrated efficient transduction of RPE and Müller cells [12,13]. Interestingly, one group showed photoreceptor transduction . However, the other group did not detect photoreceptor transduction . The reasons for these differences are not clear but may relate to animal age, the promoter and the reporter gene used, and the time frame of observation. To further characterize AAV-9 transduction in the retina, we performed a comprehensive study in young (2–3-week-old), adult (3-month-old), and old (12-month-old) mice using either an Rous sarcoma virus promoter (RSV) driving alkaline phosphatase reporter gene vector (AAV-9.RSV.alkaline phosphate [AP]) or a Cytomegalovirus promoter (CMV) driving enhanced green fluorescent protein (AAV-9.CMV.enhanced green fluorescent protein [eGFP]) reporter gene vector. To further extend the study, we also evaluated subretinal delivery of a therapeutic microdystrophin vector (AAV-9.CMV.∆R4–23/∆C) in adult mdx3cv mice, a model for Duchenne muscular dystrophy (DMD). We observed widespread transduction throughout the retina following subretinal injection. Interestingly, we observed remarkable transduction in the synaptic OPL irrespective of the vector used. Colocalization experiments with a marker of the photoreceptor presynaptic ribbon further confirmed the OPL transduction. To our knowledge, this is the first demonstration of efficient retinal synaptic layer transduction by an AAV vector. Besides tissue tropism, we also evaluated acute toxicity of subretinal AAV-9 injection. Our results suggest that subretinal AAV-9 delivery was not associated with deleterious effect on retinal morphology and electroretinogram (ERG) function.
All experiments were conducted in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. The experimental protocols were reviewed and approved by institutional animal care and use committee at the University of Missouri. C57BL/6J and mdx3cv mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME). Experimental mice were obtained from local breeding colonies. All mice were housed in specific pathogen-free (SPF) animal care facilities and kept under a 12 h:12 h light-dark cycle (25 lx) with free access to food and water.
AAV-9 stocks were generated using an adenoviral free triplasmid transfection protocol described previously [22,23]. Briefly, 70% to 80% confluent 293 cells were cotransfected with a cis-plasmid, a pRep2/Cap9 helper plasmid (a gift from Dr. James M. Wilson, University of Pennsylvania, Philadelphia, PA) and an adenoviral helper plasmid (pHelper; Stratagene, La Jolla, CA). Crude viral lysate was harvested 72 h later and purified through two rounds of CsCl isopycnic ultracentrifugation. Purified virus was dialyzed through three exchanges of HEPES buffered saline at 4 °C. Viral titer determination and quality control were performed as described before [22,23]. Briefly, viral DNA was extracted and slot blot hybridization was performed using P32-labeled radioactive probe. Vector genome was determined according to the plasmid copy number standards. The absence of significant wild type AAV contamination was confirmed by immunofluorescence staining with an antibody that recognizes AAV Rep protein.
The cis-plasmids (pcisRSV.AP, pcisCMV.eGFP, and pcisCMV.∆R4–23/∆C) have been reported previously [24,25]. pcisRSV.AP expresses the heat-resistant human placental AP gene under the transcriptional control of the RSV promoter and the SV40 polyadenylation signal (pA). pcisCMV.eGFP expresses the eGFP gene under the transcriptional control of the CMV promoter and the SV40 pA. pcisCMV.∆R4–23/∆C encodes a highly abbreviated ∆R4–23/∆C human microdystrophin gene under the transcriptional control of the CMV and the SV40 pA . Specifically, a large portion of the dystrophin rod domain (from spectrin-like repeat 4 to 23) and the entire C-terminal domain are deleted. The recombinant AAV vectors are called AAV-9.RSV.AP, AAV-9.CMV.eGFP, and AAV-9.CMV.∆R4–23/∆C.
AAV-9.RSV.AP and AAV-9.CMV.eGFP injection were delivered to young (2- to 3-week-old), adult (3-month-old), and old (12-month-old) C57BL/6J mice. AAV-9.CMV.∆R4–23/∆C injection was delivered to 3-month-old adult mdx3cv mice. HEPES-buffered saline was used as vehicle control in a subset of animals.
Mice were given 1% atropine eyedrops 3 h before they were anesthetized with an intraperitoneal injection of a mixture of 75 mg/kg ketamine and 13.6 mg/kg xylazine. Following general anesthesia, 2.5% phenylephrine hydrochloride eyedrops were applied. One drop of 1% proparacaine hydrochloride was administered as local anesthesia, followed by 2.5% hydroxypropyl methylcellulose. Subretinal injection was performed according to a previous publication with modifications . Briefly, an aperture within the pupil area was made through the superior cornea with a 30-gauge needle. A 33-gauge blunt needle mounted on a 10-μl syringe was introduced through the corneal opening, avoiding the lens and penetrating the neuroretina to reach the posterior subretinal space. The NanoFil™ sub-microliter injection system (WPI, Sarasota, FL) was used to inject 1 µl of AAV vector in 30 s. The injection was considered successful when retinal blebs occupied more than half of the retina. Evaluation was performed only in mice that were successfully injected. Following subretinal injection, 1% atropine eyedrops and antibiotic ophthalmic ointment were administered daily for three days. Seven days after injection, the mice were anesthetized with ketamine and xylazine as described previously, and their eyes were examined under microscope. Eyes that exhibited any sign of surgical complications, including anterior or posterior synechia, cataract, vitreous and retinal hemorrhage, and unresolved retinal detachment, were excluded from the study. Such signs were observed in 20% to 30% of the eyes.
For AP and dystrophin expression, mouse eyes were enucleated and snap frozen in liquid nitrogen–cooled isopentane in optimal cutting temperature (OCT) compound (Sakura Finetek Inc., Torrance, CA). Next, 10 µm cryosections were cut along the optic nerve head and stained with hematoxylin and eosin (H&E) for routine retinal morphology. After endogenous heat labile AP activity was heat inactivated, AAV-9 mediated AP expression was examined by histochemical staining using a previously described protocol [22-24]. Microdystrophin expression was evaluated by immunofluorescence staining according to a previously published protocol . Briefly, cryosections were sequentially blocked with Papain-digested rabbit anti-mouse immunoglobulin and rabbit serum. Then sections were stained with the primary anti-dystrophin antibody and signal was detected with appropriate secondary antibodies. The present study employed 1:30 Dys-2 antibody (Novocastra, Newcastle, UK), which recognizes dystrophin C-terminal domain and 1:20 Dys-3 antibody (Novocastra), which recognizes the hinge 1 region in human dystrophin. Also used was 1:100 Alexa Fluor 594-conjugated goat anti-mouse IgG (H+L) F(ab’)2 fragment antibody as the secondary antibody in immunofluorescence staining (Invitrogen, Carlsbad, CA). For eGFP and CtBP2 detection, the eyeballs were first fixed in 4% paraformaldehyde and phosphate-buffer saline (PBS,, 0.01 M phosphate buffer, 0.0027 M potassium chloride and 0.137 M sodium chloride, pH 7.4) then cryoprotected in 30% sucrose/PBS overnight. Next, 10 µm cryosections were stained with 1:200 goat anti-CtBP2 (C-terminal binding protein 2, Santa Cruz, Biotechnology, Santa Cruz, CA) followed by 1:500 Texas Red conjugated donkey anti-goat IgG secondary antibody (Jackson ImmunoResearch, West Grove, PA). Nuclei were revealed with 1:500 4’,6-diamidino-2-phenylindole, dihydrocholoride (DAPI, Molecular Probes, Eugene, OR). The specificity of immunostaining was confirmed by staining cryosections in the absence of the primary antibody. Photomicrographs were taken with a digital camera using a Leica fluorescence microscope (DMR, Deerfield, IL).
Data are presented as mean±standard deviation (mean±SD). Statistical significance was examined with ANOVA followed by the Tukey post hoc test. A p value of less than 0.05 was considered as significant.
We first examined subretinal AAV-9 AAV.RSV.AP injection in normal C57BL/6J mice. Figure 1 shows the representative retinal cross-sections from young (n=6 eyes, 3 weeks post-injection), adult (n=12 eyes, 5 weeks post-injection) and old (n=5 eyes, 5 weeks post-injection) mice. AAV-9 resulted in widespread (peripheral to central to peripheral) and throughout (RPE to retinal ganglion cells [RGC]) AP expression in all age groups (Figure 1A, data not shown for young and old mice). Nevertheless, the highest expression was observed at the injection site (Figure 1A, arrow). AP expression was observed in the RPE, outer nuclear layer (ONL), inner nuclear layer (INL), OPL, inner plexiform layer (IPL), RGC layer and Müller cells but not the outer and inner segments of the photoreceptor (Figure 1C,E-H). Remarkably, two retinal synaptic layers (OPL and IPL) were highly transduced (Figure 1C,E-H). AP expression was not detected in the mock-infected control eyes (5 eyes each for young and adult mice; Figure 1B,D).
To exclude the potential bias from the transgene as well as the promoter, we evaluated AAV-9.CMV.eGFP transduction (4 eyes in 2-week-old and 4 eyes in 3-month-old mice). Four weeks after subretinal injection, we observed intense eGFP expression in young and adult groups in the RPE, ONL, INL, OPL, IPL, RGC layers, and Müller cells (Figure 2B,C). Interestingly, we also observed robust expression in the photoreceptors (Figure 2B-E). To further confirm OPL transduction, we stained AAV-9.CMV.eGFP-infected eyes with CtBP2, a marker for the photoreceptor synaptic ribbon (Figure 2 E,G,H). The colocalization of eGFP and CtBP2 strongly suggest that the OPL layer was transduced (Figure 2F,H).
OPL defects are associated with several retinal diseases. To determine whether AAV-9 can be used to deliver a therapeutic gene to the OPL, we performed subretinal injection in mdx3cv mice using AAV-9.CMV.∆R4–23/∆C vector. Mdx3cv mice are models for DMD, a lethal childhood genetic disease caused by mutations in the dystrophin gene [33,34]. Besides muscle disease, DMD patients also suffer from pathology in other systems such the central nerve system. A 260 kDa dystrophin isoform (Dp260) is normally expressed in the OPL. In the eyes of DMD patients and mdx3cv mice, Dp260 expression is lost [35-37]. The absence of Dp260 has been associated with the abnormal ERG seen in DMD patients.
The 3.8 kb ∆R4–23/∆C microgene encodes a highly truncated dystrophin. This microgene has been extensively studied as a candidate gene for DMD gene therapy . At 5 weeks after subretinal injection (n=5 mice; AAV-9.CMV.∆R4–23/∆C to one eye and HEPES buffer to the contralateral eye), we evaluated dystrophin expression by immunofluorescence staining. Two epitope-specific antibodies were used in the study. The Dys-2 antibody recognizes endogenous Dp260 in the wild type retina (Figure 3G) , while the Dys-3 antibody only reacts with microdystrophin. Consistent with our findings in reporter AAV vector-infected normal eyes, we observed efficient OPL transduction in the eyes of AAV-9.CMV.∆R4–23/∆C infected mdx3cv mice (Figure 3B,C). No dystrophin was detected in HEPES buffer-injected eyes (Figure 3E,F).
Next, we studied whether subretinal AAV-9 injection causes acute retinal damage. At 5 weeks after subretinal injection (AAV-9.RSV.AP or HEPES) in adult and old C57BL/6J mice (n=5 for each age group), we examined retinal histology and recorded dark-adapted and light-adapted ERGs. Compared with untreated eyes, neither HEPES buffer injection nor AAV-9.RSV.AP injection resulted in appreciable morphology alterations (Figure 4). In each age group, the thresholds and amplitudes of the dark-adapted ERG a-wave, dark-adapted b-wave and light-adapted b-wave of the AAV-9 or HEPES injected eyes were comparable to those of the untreated eyes (p>0.05, Figure 4, Figure 5, and Figure 6). On the ascending limb of dark-adapted b-wave, the amplitude of the oscillatory potentials (OPs) in the AAV-9 vector injected eyes appeared smaller than the untreated eyes (Figure 4). However, similar reduction was also observed in HEPES-injected eyes.
AAV-mediated gene therapy holds great promise for rescuing vision loss in retinal degenerative diseases. Remarkable progress has been made in animal models and human patients [1,2,4,6-8]. However, most studies have focused on diseases that primarily affect the photoreceptor, RPE, and RGC layer.
There are two synaptic layers in the retina including the OPL in the outer retina and the IPL in the inner retina. The OPL and IPL are indispensable for mediating visual signal transmission. While no human diseases have been associated with the IPL, several retinal diseases originate in as well as primarily affect the OPL. Examples of these diseases include congenital stationary night blindness, retinoschisis, DMD retinopathy, and melanoma-associated retinopathy [14-18]. Gene therapy for these diseases is largely dependent on a vector that can efficiently deliver the therapeutic genes to the OPL.
Several different AAV serotypes have been evaluated for retinal gene delivery. In this study we examined AAV-9 transduction in young, adult, and old mice. Consistent with recent publications from other investigators, we observed robust AAV-9 transduction in the outer retina (RPE and ONL), inner retina (Müller cells), and RGC layers [12,13]. Surprisingly, we also observed quite efficient and widespread transduction in the synaptic layers (Figure 1, Figure 2, and Figure 3). A colocalization study suggested that transduction occurred in the photoreceptor terminals, but not in the bipolar cell and horizontal cell dendrites in the OPL (Figure 2).
To explore therapeutic gene delivery to the OPL we delivered AAV-9.CMV.∆R4–23/∆C to adult mdx3cv mice. We focused on transgene expression rather than functional correction in this study. Consistent with our findings in normal mice, we observed robust microdystrophin expression in the OPL. Taken together, our results suggest that AAV-9-mediated OPL transduction represents a promising approach to treat retinal diseases that are related to OPL defects.
It is important to understand whether subretinal administration of a new viral vector causes any side effects. We found that subretinal delivery of AAV-9 did not alter retinal structure. In addition, rod and cone photoreceptor and bipolar cell functions were not affected by subretinal AAV-9 injection (Figure 4, Figure 5, and Figure 6). Interestingly the OPs, highly sensitive functional indicators of the inner retina , were reduced following AAV-9 injection. Since similar reductions were also observed in HEPES-buffer-injected eyes. We suspect that the OP reduction may relate to the injection procedure rather than AAV-9 vector.
Taken together, our data suggest that AAV-9 is a potent vector for retina gene delivery, and subretinal administration does not cause acute damage. Efficient transduction of the photoreceptor terminals opens the door to develop AAV-9 gene therapy for degenerative retinal diseases that affect this synaptic structure.
This work was supported in part by University of Missouri Research Board, an unrestricted grant from Research to Prevent Blindness, Inc. (New York, NY) to the Department of Ophthalmology, University of Missouri-Columbia, National Institutes of Health AR-49419 (D.D.), and Muscular Dystrophy Association (D.D.). We thank Dr. Jijing Pang for sharing subretinal injection technique, Dr. Lixing W. Reneker for help with histology studies, and Mrs. Chun Long for technical assistance. We also thank Dr. James M. Wilson for providing the AAV-9 packaging plasmid pRep2/Cap9.