Scholarly article on topic 'HSV-1 Vector-Delivered FGF2 to the Retina Is Neuroprotective but Does Not Preserve Functional Responses'

HSV-1 Vector-Delivered FGF2 to the Retina Is Neuroprotective but Does Not Preserve Functional Responses Academic research paper on "Biological sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Mol Ther
OECD Field of science
Keywords
{""}

Academic research paper on topic "HSV-1 Vector-Delivered FGF2 to the Retina Is Neuroprotective but Does Not Preserve Functional Responses"

Article

doi:10.1006/mthe.2001.0307, available online at http://www.idealibrary.com on IDEAL

HSV-1 Vector-Delivered FGF2 to the Retina Is Neuroprotective but Does Not Preserve Functional Responses

Brian Spencer,*1 Seema Agarwala,t2 Laura Gentry,* and Curtis R. Brandt*,t 3

*Department of Medical Microbiology & Immunology and rDepartment of Ophthalmology & Visual Sciences, University of Wisconsin, Madison, Wisconsin 53706

Received for publication December 13, 2000; accepted in revised form March 13, 2001

Fibroblast growth factor 2 (bFGF, FGF2) exhibits mitogenic, angiogenic, wound healing, and neuroprotective properties. Infusion of FGF2 in vivo to treat neurodegenerative disorders in animal models results in increased survival of damaged neurons, but these effects are transient. To test the feasibility of HSV vector-delivered FGF2 for neuroprotection, we inserted the FGF2 gene under the control of the HCMV immediate-early promoter into an attenuated avirulent HSV-1 vector. Trans-duction with FGF2/HSV-1 virus promoted survival of PC12 cells, induced differentiation of these cells to the neuronal phenotype in vitro, and protected PC12 neuronal cells from death induced by nerve growth factor withdrawal. The attenuated FGF2/HSV-1 virus was able to deliver and direct expression of the FGF2 gene in the eye. Delivery prior to light exposure in a rat model of retinal degeneration resulted in significant protection against photoreceptor loss. However, functional ERG responses were not detected. Treatment of normal eyes with the vector alone suppressed ERGs, which were only partially restored in eyes receiving the FGF2 vector. Thus, although the FGF2-HSV-1 virus induced preservation of cell and tissue structure, this was not sufficient to protect photoreceptor function.

Key Words: neuroprotection; FGF2; bFGF; HSV vectors; retinal degeneration; electroretinogram; PC12 cells.

Introduction

Fibroblast growth factor 2 (bFGF, FGF2) is a member of a family of 18 growth factors that exhibit mitogenic, angiogenic, wound healing, and neuroprotective properties [reviewed in (1)]. FGF2 also plays an important role in the maintenance and repair of neurons both in vitro and in vivo (2-7). Infusion of FGF2 in vivo in animal models of Parkinson's, Alzheimer's, and motor neuron disease, as well as animal models of axotomy-induced neurodegeneration, results in increased survival of damaged and innervating neurons even when administered as late as 2 days posttrauma (8-14). In both inherited and physically induced animal models of retinal degeneration, injections

1 Currently at The Salk Biological Institute, LOG/V, 10010 North Torrey Pines Road, La Jolla, CA 92037.

2 Currently at University of Chicago, Department of Neurobiology, Pharmacology & Physiology, Ab 210 (MC 0926), 947 East 58th Street, Chicago, IL 60637.

3 To whom correspondence and reprint requests should be addressed at 6630 Medical Sciences Center, 1300 University Avenue, Madison, WI 53706. Fax: (608) 262-0479. E-mail: crbrandt@facstaff.wisc.edu.

of recombinant FGF2 into the vitreal cavity of the eye significantly delay photoreceptor cell death (15-22). These results clearly show that FGF2 is a survival-promoting factor in the central nervous system and the eye. However, the effect of FGF2 appears to wane after approximately 1-3 months, possibly due to degradation or sequestration of the FGF2 in the extracellular matrix. Viral vector-mediated delivery of the gene for FGF2 to affected neurons could possibly provide long-term expression and treatment for many neurodegenerative diseases.

Several rat models of retinal degeneration are available and have been used to test various therapeutic strategies. The Royal College of Surgeons (RCS) rat displays an inherited degeneration characterized by progressive loss of rods beginning 20-60 days postpartum (23-25). The defect results in reduced phagocytosis of shed outer segments and is due to mutations in the c-mer gene (26). Retinal degeneration in the Fischer 344 rat is characterized by a slow progressive loss of photoreceptors that begins at 12 months of age (27-30). Albino rats show significant retinal degeneration following exposure to high-intensity light (31, 32). The photoreceptor death in

this model is thought to involve oxidative damage to photoreceptors (33-35). Phagocytic defects do not appear to be involved (16). More recently, a transgenic rat carrying the P23H mutation found in human retinitis pigmentosa has been developed as a model (36). In the albino bright-light model, onset can be precisely controlled. We, therefore, chose to use the acute bright-light albino rat model for these studies.

The eye is an excellent model system to test neuroprotection due to its accessibility, the ease of delivery, the quantitative methods for measuring protection, and the availability of noninvasive tests of neuronal function. Akimoto et al. (37) tested the ability of FGF2 delivered by an adenovirus (AV) vector to inhibit retinal degeneration in the inherited RCS rat model and reported that photoreceptor degeneration was delayed. However, electroreti-nography (ERG) responses were not examined, thus therapeutic success was not demonstrated. Intravitreal delivery of an empty adenovirus vector caused significant reductions in ERG responses (38), thus AV vectors may not be ideal for ocular, and possibly other, neuronal gene delivery. Similar results have also been reported for FGF2 delivered by adeno-associated virus (AAV) vectors (39).

Herpes simplex virus-based vectors have been studied extensively for delivery to the CNS. Recently, we showed that an HSV vector could efficiently deliver a gene to several cell types in the eye (40) without any apparent toxicity as measured histopathologically even after multiple injections (41). To determine if HSV-based vectors might be more suitable than AV or AAV vectors for ocular gene delivery, we have constructed FGF2-expressing HSV viruses and tested the neuroprotective function of these vectors in the acute bright-light model of retinal degeneration. These results are the first to report in vivo effects of FGF2 using an HSV vector in the eye. The two most significant findings are (i) HSV-delivered FGF2 was neuroprotective, preventing photoreceptor loss, and (ii) HSV vector suppressed functional (ERG) responses. These findings emphasize the importance of testing both structural and functional parameters in studies of neuronal gene delivery.

Materials and Methods

Cell Culture

African green monkey kidney cells (Vero) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% serum (1:1 mixture of fetal bovine and defined, supplemented calf serum), 100 U/ml penicillin, and 100 ^g/ml streptomycin sulfate (42). High-titer stocks of the virus were prepared on Vero cells as previously described (42). Titers were determined by plaque assay on Vero cells. Rat pheochromocytoma cells (PC12, ATCC CRL-1721) were grown in DMEM supplemented with 15% serum (2:1 mixture of horse serum and defined, supplemented calf serum), 100 U/ml penicillin, and 100 ^g/ml streptomycin sulfate (43). Differentiation of PC12 cells was induced by plating cells onto a collagen (Sigma Chemical, St. Louis, MO) layer at a density of 2 X 105 cells per 3.5-cm2 plate (TC60). The cells were then fed DMEM with 1% serum supplemented with nerve growth factor (NGF; 50 ng/ml; Promega, Madison, WI) and incubated for a period of 2 weeks. NGF (50 ng/ml) was readministered every 3 days (43). Differentiation of the PC12 cells was

A 1E0-—-«-:-Ir7.TK3 rrrsi

B KOSHSV-1 Hff ULTO feHP|in.-n)~l»-

Q hrR3 —|p| I llgrtlactosidasc :'->-UL4''

D FCpF2.;HSV-1 -nn~Hi'l FGF2 i;DNA ITMT»-

■ •'l fc_ "*■

E -I iBGHpAl

F _I i ? f I f

FIG. 1. Schematic representation of the viruses used in this study. Line A shows the structure of the HSV-1 genome; Ul, unique long segment; Us, unique short segment; Rl, terminal repeat long segment; Rs, terminal repeat short segment. Line B shows an enlargement of the region encoding the UL39 and UL40 genes, which encode the large and small subunits of ribonucleotide reductase, respectively. P denotes the promoters of the UL39 and UL40 genes. Line C shows the structure of the hrR3 virus in which the ^-galactosidase gene has been inserted under the control of the UL40 promoter. Line D shows the structure of the FGF2 expression cassette recombined into the hrR3 virus to replace the ^-galactosidase gene. P represents the UL39 promoter; C, the HCMV major immediate-early promoter; T, the splice-polyadenylation signal. Line E shows the expanded 3' end of the FGF2 cassette with the locations of the PCR primers used to detect expression of the vFGF2 message and vector genome; 1 and 2 denote the forward and reverse primers, respectively. The location of relevant restriction sites used to confirm integration of the FGF2 cassette are depicted (line F) as BamH\ (B), EcoRI (R), and EcoRV (V).

confirmed by microscopic examination for the presence of neurite extensions. Differentiated PC12 cells were transduced in DMEM with 1% serum. After 1 h exposure to the virus, the supernatant was aspirated and the cells were refed with DMEM with 1% serum. Cultures transduced in the presence of acycloguanosine (ACG; Sigma Chemical) were pretreated (70 jjuM) 24 h prior to transduction and again at the time of transduction (44, 45).

Construction of a FGF2-Expressing HSV Vector

The FGF2 cDNA plasmid, pBS-bFGF (Dr. Judy Abraham, California Biotechnology, Inc., Mountain View, CA), was digested with EcoRI and the FGF2 cDNA fragment was cloned into the pcDNA3 expression vector (Invitrogen, Carlsbad, CA), generating an expression cassette with the FGF2 cDNA flanked by the HCMV major immediate-early promoter at the 5' end and the bovine growth hormone splicing and polyadenylation site at the 3' end. The FGF2 expression cassette was inserted into the plasmid p-MAK (Dr. Sandra Weller, University of Connecticut, Farmington, CT), generating plasmid pMAK-FGF2. The pMAK-FGF2 plasmid was linearized by digestion with ScaI and then cotransfected into Vero cells with purified hrR3 viral DNA. Homologous recombination between pMAK-FGF2 and RR sequences in the hrR3 virus resulted in insertion of the FGF2 expression cassette and deletion of the lacZ sequence in hrR3 (Fig. 1D). A total of 14 colorless plaques were picked and screened by Southern blotting with a FGF2-specific probe isolated from the pBS-bFGF plasmid. Two of these plaque isolates contained the FGF2 gene and were further plaque-purified a total of three times prior to further evaluation These viruses were designated 2526/FGF2 and 5042/FGF2. Insertion of FGF2 into the viral genome was confirmed by Southern blotting with a probe from the FGF2 gene following digestion with several enzymes (data not shown).

DNA Isolation and PCR

Dissected eye cups were resuspended in cell lysis buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 25 mM EDTA, 0.5% SDS, 125 ^g/ml proteinase K, 1.5% 2-mercaptoethanol) and incubated at 55°C for 24 h. The DNA was then purified by phenol/chloroform extraction and ethanol precipitation. PCR was performed on 250 ng of total DNA with the HSV-1 ICP27B primers (46) using PCR SuperMix (Life Technologies, Rockville, MD) for 37

cycles and the product was visualized after electrophoresis in a 1.8% agarose gel and staining with ethidium bromide. Mouse p-actin primers from the p-actin Control Amplimer Set (Clontech Laboratories, Palo Alto, CA) and whole genomic DNA of the HSV-1 strain OD4 (42) were used as controls for the PCR in addition to template-minus and primer-minus controls.

Southern Blotting

Southern blots were performed essentially as described previously (47). DNA was separated by electrophoresis through agarose, transferred to a nylon membrane (Micron Separations, Westborough, MA), and then probed with a digoxigenin (dig)-dNTP-labeled probe prepared according to the manufacturer's directions (Dig DNA Labeling Kit; Boehringer Mannheim, Indianapolis, IN). The labeled probe was detected with an anti-dig alkaline phosphatase-conjugated antibody (Boehringer Mannheim) and the BCIP/NBT substrate (Sigma Chemical) or the CSPD chemilumines-cence reagent (Boehringer Mannheim), followed by exposure to X-ray film.

RNA Isolation and RT-PCR

Cells or dissected eye cups were treated with RNAzol (Biotecx Laboratories, Houston, TX) according to the manufacturer's directions to extract total RNA. One microgram of total RNA was added to RT-PCR buffer (50 mM KCl, 10 mM Tris-HCl, pH 8, 1% Triton X-100, 1.25 mM MgCl2, 5% DMSO), 2.5 mM dNTPs, 5 mM DTT, 10 U RNasin, 5 U AMV-RT, and 1.25 U TaqPol and RT-PCR was carried out using 37 cycles of amplification. Specific amplification of the virus-expressed FGF2 (vFGF2) RNA was carried out with the forward primer complementary to the FGF2 cDNA (GGG TTG TGT CTA TCA AAG GAG TGT GTG CAA) and the reverse primer complementary to the human growth hormone polyadenylation sequence (GCA TTT AGG TGA CAC TAT AGA ATA GGG CCC) (see Fig. 1E). A positive control was performed for each RT-PCR using an identical amount of RNA with the human GAPDH primers (forward, ACA GCC TCA AGA TCA TCA GC, and reverse, ATG AGT CCT TCC ACG ATA CC) for the Vero cells, rat GAPDH primers (forward, GAA CAT CAT CCC TGC ATC, and reverse, TGC TTC ACC ACC TTC TTG) for the PC12 cells, and mouse p-actin primers (Clontech Laboratories) for the mouse tissues. RT-PCR products were electrophoresed in a 1.8% agarose gel and stained with ethidium bromide for visualization.

Protein Isolation and Analysis

Total protein was isolated by scraping the cells and transferring the suspension to a 15-ml polypropylene centrifuge tube. The cells were pelleted by centrifugation at 850g for 15 min at 4°C and then resuspended in protein loading buffer (50 mM Tris-HCl, 10% glycerol, 1% SDS, 1% 2-mer-captoethanol, 0.001% bromphenol blue). The samples were sonicated for 10 s, boiled for 10 min, and centrifuged for 5 min at 14,000g to remove insoluble cellular debris. The samples were then electrophoresed in a 12% denaturing polyacrylamide gel and transferred to nitrocellulose. Serial dilutions of purified recombinant human FGF2 (rhFGF2; Cat. No. G5071; Promega) were also electrophoresed in the absence of cellular extracts as standards for quantification. Western blotting was carried out as described previously (48, 49) using a rabbit anti-FGF2 primary antibody (Cat. No. SC-79; Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:100 and an anti-rabbit-conjugated alkaline phosphatase secondary antibody diluted 1:2000 (Sigma Chemical). The blots were developed using the BCIP/NBT substrate (Promega). Following development of the blot, FGF2 bands were quantified by densitometric scanning.

Cell Survival Assays

PC12 survival and differentiation assay. Confluent T-25 flasks of undiffer-entiated PC12 cells were resuspended and washed twice with PBS to remove serum. The cells were then resuspended in DMEM with 1% serum and ACG (70 pM), transduced with either 2526/FGF2 or 5042/FGF2 or infected with hrR3 virus at a m.o.i. of 5 or treated with rhFGF2 (15 ng/ml) or NGF (50 ng/ml) and plated onto a collagen-coated 96-well plate at a density of 10,000 cells per well. The cells were treated daily with ACG (70

pM). On day 7 posttreatment, the cells were photographed and then assayed for cell survival with the CellTiter 96 Aqueous One Solution cell proliferation assay, which is an MTT-based assay and thus measures both viable and proliferating cells (Promega).

Survival ofPC12 neuronal cells. PC12 cells were grown and differentiated by addition of exogenous NGF as described above. After differentiation was confirmed by microscopic examination at 10 days, the cells were washed three times with PBS to remove remaining NGF and then triplicate cultures were transduced with the 2526/FGF2, 5042/FGF2, or hrR3 virus at a m.o.i. of 1 or were mock transduced in the presence of ACG (70 pM) in DMEM with 1% serum and 2 pg/ml anti-NGF antibody (Cat. No. SC-549; Santa Cruz Biotechnology) to block residual NGF in the PC12 cell cultures. Cells were treated with ACG daily for 7 days and cell viability was determined with the CellTiter 96 Aqueous One Solution cell proliferation assay.

In Vivo Analysis of Neuroprotection

Intravitreal injection. Albino adult rats (300 g) (Harlan Sprague-Dawley, Indianapolis, IN) were anesthetized by an intramuscular injection of xy-lazine (9 mg/kg) and ketamine-HCl (90 mg/kg). Five microliters of the 2526/FGF2 [1 X 107 plaque-forming units (pfu)] or hrR3 virus (1 X 107 pfu) in PBS, rhFGF2 (Promega; 1000 ng), or PBS alone was then injected intra-vitreally. Mice (3-6 months, B6F1) were anesthetized with Halothane (3-5%) by inhalation and injected with virus as described above (41).

Retinal degeneration. Sprague-Dawley albino rats (300 g) were housed for 7 days in a low-level (400 lux) cyclical (12 h on/12 h off) light environment with food and water ad libitum. The rats then received an intravitreal injection of virus, purified human rhFGF2, or PBS as described above and were returned to the low-level, cyclical light environment for 48 h. Retinal degeneration was induced with continuous (24 h) bright light (1200 lux) produced by two 40-W fluorescent bulbs suspended approximately 1 meter above the floor of the cage (16, 19, 50). After 7 days of constant bright light, rats were returned to low-level cyclical light for 7 or 28 days prior to electroretinography (see below). Rats were sacrificed, and the eyes were removed and fixed in 10% formalin in PBS. The samples were embedded in paraffin and sectioned (10 pm) at five random positions along the pupil/optic nerve axis and stained with nuclear fast red. All animal procedures conformed to NIH guidelines for responsible care and use of animals.

Retinal degeneration was quantified as described previously (16, 18). Briefly, the thickness of the outer nuclear layer (ONL) plus outer segments (OS) of the photoreceptors was measured at three points each in the superior and inferior hemispheres of the eye with one each at 20 pm from the ora serata and optic nerve and the third at the midpoint between these. At least six animals were analyzed in each group and three sections were measured per animal. The results were averaged and plotted as the mean thickness of the ONL + OS as a function of distance from the optic nerve.

Electroretinography. ERG was measured in the rats 7 or 28 days after exposure to bright light by a modification of the protocol by Goto et al. (51). Rats were anesthetized as described above, and pupils were dilated with 2.5% phenylephrine HCl. ERGs were recorded using a specialized DTL fiber electrode (LKC Technologies, Inc., Gaithersburg, MD). Reference and ground needle electrodes were placed at the base of the neck and tail, respectively. Signals were amplified by an LKC Utas 2000 (LKC Technologies) ERG/VEP system at gains of 20 and 50 pp/div and filtered between 1 and 1000 Hz. Animals were tested in a LKC ganzfeld under dark-adapted conditions. Flashes were provided by a Grass PS22 xenon discharge lamp and associated power supply with a duration of approximately 800 ps. Flash intensity was controlled by a series of stops and filters built into the ganzfeld unit. Animals were dark-adapted in a light-proof room for 1 h. Under red light illumination (Wratten filter 445), anesthesia and dilating drops were administered. A dark-adapted luminance-response function was obtained. The ERGs were then recorded to flashes of three luminance levels ranging from —3 to 0 log cd-s/m2 in order of increasing luminance. In order to maintain adaptation level, a 30-s period was inserted between flashes for the lowest luminance flashes. The highest intensities had a minimum interstimulus interval of 1 min. The response to four flashes at each intensity level was averaged by the LKC Utas 2000 unit and stored on disk for further analysis.

FIG. 2. vFGF2 protein expression in Vero and PC12 neuronal cells. Immunoblots showing vFGF2 expression following transduction with the 2526/FGF2 or the 5042/FGF2 or infection with hrR3 virus in Vero or differentiated PC12 neuronal cells. Cells were transduced or infected at a m.o.i. of 5 in the presence or absence of ACG (70 ,uM) for 25 h. Total cell protein from 107 Vero cells or 105 PC12 cells was extracted and electrophoresed on a 12% polyacrylamide gel. vFGF2 migrates at approximately 18 kDa on a 12% polyacrylamide gel.

Statistical Analysis

All samples were compared using Student's t test, assuming equal variances.

Results

of an eclipse period). Both of the FGF2/HSV-1 viruses grew as well as the hrR3 parental vector in serum-fed (Fig. 3A) and serum-starved Vero cells (Fig. 3B) with no significant difference between the samples (P < 0.05, n = 3 for each time point). Differentiated PC12 neuronal cells did not

The FGF2 HSV-1 Viruses Express vFGF2

To confirm expression of the viral FGF2 protein in cells transduced with the FGF2-expressing HSV-1 viruses, Vero cells or differentiated PC12 neuronal cells were transduced with the 2526/FGF2, 5042/FGF2, or hrR3 virus at a m.o.i. of 5 and then 25 h later vFGF2 protein production was examined by immunoblotting (Fig. 2). The 18-kDa vFGF2 protein was specifically detected in the 2526/FGF2-and 5042/FGF2-transduced cells. There did not appear to be an appreciable difference of vFGF2 protein expression between the two FGF2 viruses in Vero cells. To determine if vFGF2 protein was synthesized in the absence of viral replication, transductions were carried out as described above in the presence of the anti-HSV-1 nucleoside analog acycloguanosine ACG (Fig. 2). The vFGF2 was expressed in Vero cells transduced with 2526/FGF2 and 5042/FGF2 in the presence of ACG. Similarly, differentiated PC12 cells transduced with 2526/FGF2 or 5042/FGF2 viruses expressed FGF2 in the presence of ACG. Examination of vFGF2 RNA by RT-PCR with the vFGF2-specific primers under similar conditions confirmed that ACG treatment did not affect the expression of the vFGF2 gene (data not shown).

Expression of vFGF2 Does Not Affect Virus Replication in Vitro

One-step growth curves were performed to determine if expression of the vFGF2 gene altered replication of the vector virus. Vero cells or differentiated PC12 neuronal cellswere infected atam.o.i. of 1, and at0, 3, 6, 12, 18, 24, and 48 h postinfection samples were collected and titered on Vero cells. Growth curves for both serum-fed and serum-starved Vero cells were carried out to determine if the expression of vFGF2 would have an effect under either condition. The PC12 cells were not rinsed after adsorption to remove virus, since differentiated PC12 cells are only loosely attached even when differentiated (hence the lack

FIG. 3. One-step growth curves of the FGF2 viruses in Vero cells. One-step growth curves of the 2526/FGF2, 5042/FGF2, and hrR3 viruses were performed in serum-fed (5% serum) Vero cells (A), serum-starved Vero cells (B), or differentiated PC12 cells (C). Cells were transduced with 2526/FGF2 (■) or 5042/FGF2 (•), or infected with hrR3 (♦), at a m.o.i. of 1, and at various times, samples were taken and the virus was titered on Vero cells. The PC12 cells were not rinsed after the adsorption period, hence the lack of an eclipse phase. There was no significant difference in titers (Student's t test, P < 0.05, n = 3) for each time point between any of the viruses.

FIG. 4. Survival of undifferentiated PC12 cells transduced with 2526/FGF2 or 5042/FGF2. Undifferentiated PC12 cells were transduced with 2526/FGF2 or 5042/FGF2 or infected with hrR3 at a m.o.i. of 5 in the absence of serum or exogenous growth factors. Control PC12 cells were treated with rhFGF2 (15 ng/ml) or NGF (50 ng/ml) or were mock transduced in the absence of serum (n.t.). Seven days posttreatment cell viability was assayed. All samples were compared with the Student t test and (*) denotes samples that were significantly different (P < 0.05) from the samples with no treatment or with hrR3 infection.

appear to support the replication of 2526/FGF2, 5042/ FGF2, or hrR3 such that at 48 h postinfection virus titers were at the limit (<100 pfu/ml) of our detection assay (Fig. 3C).

FGF2 HSV-1 Viruses Promote PC12 Cell Survival and

Differentiation in Vitro

Undifferentiated PC12 cells were transduced at a m.o.i. of 5 with 2526/FGF2, 5042/FGF2, or hrR3 in the presence of ACG. In addition, replicate cultures of undifferentiated PC12 cells were treated with ACG and rhFGF2 (15 ng/ml), rNGF (50 ng/ml), or PBS as controls (Fig. 4). Seven days posttreatment the 2526/FGF2-transduced PC12 cells showed an approximately sixfold greater cell survival compared to the untreated controls and the 5042/FGF2-transduced cells showed a fivefold greater cell survival. Transduction of the PC12 cells by either FGF2 virus resulted in significant cell survival (P < 0.05, n = 3 for each time point), compared to the untreated control cells or hrR3-infected cells. A significant difference (P < 0.05, n = 3 for each time point) was noted between the two FGF2 viruses such that the 2526/FGF2 virus promoted greater PC12 cell survival than the 5042/FGF2 virus. Treatment with the hrR3 virus alone promoted some cell survival over the control PBS-treated culture; but this was significantly less than the cultures that had been treated with either FGF2/HSV-1 virus (P < 0.05). Two FGF2/ HSV-1 viruses also induced differentiation of the PC12 cells to a neuronal morphology in the absence of NGF (Fig. 5). The 5042/FGF2 virus appeared to induce greater neurite extension in the PC12 cells compared to 2526/ FGF2-treated cells.

To test the ability of 2526/FGF2 and 5042/FGF2 to promote the survival of neuronal cells during conditions of serum withdrawal, fully differentiated PC12 neuronal cells were rinsed to remove serum and exogenous growth factors and then transduced with 2526/FGF2, 5042/FGF2, or hrR3 at a m.o.i. of 1 or left untreated. Seven days later cells that had been treated with either 2526/FGF2 or 5042/FGF2 showed significantly greater cell survival than replicate cultures that had been left untreated or infected with the hrR3 control (P < 0.05, Fig. 6). The 2526/FGF2 virus-transduced cells had 17-fold greater cell survival compared to the untreated cells and 5042/FGF2 virus-transduced PC12 neuronal cells had 10-fold greater cell survival compared to the untreated control cells. The difference between the 2526/FGF2 and the 5042/FGF2 virus-transduced PC12 neuronal cell survival was significant (P < 0.05).

The 2526/FGF2 Virus Expresses the vFGF2 Gene in Vivo

In order to reduce the number of animals required for in vivo testing, we chose to perform in vivo studies with only the 2526/FGF2 virus. Three mice were transduced by in-travitreal injection with 1 X 107 pfu of virus. Three days posttransduction, RT-PCR was performed on total RNA extracted from the retina using the vFGF2-specific primers (Fig. 1E). The vFGF2 message was detected in the retina of three of three animals treated with the 2526/FGF2 virus, but not in hrR3- or PBS-injected eyes (data not shown). When we tested for vFGF2 RNA at 7 days posttreatment, no mRNA was detected. Analysis of retinal DNA by PCR at 28 days postinjection revealed the presence of vector DNA in three of three samples, indicating that loss of vector did not explain the loss of FGF2 expression.

In Vivo Neuroprotection

To test the neuroprotective ability of 2526/FGF2 in vivo, we used a well-established rat model wherein continuous exposure to high-intensity light for 1 week results in acute loss of photoreceptors (16, 19, 50). The 2526/FGF2 or hrR3 virus, rhFGF2, or PBS was delivered via intravitreal injection into the right eye and 48 h later, rats were subjected to constant bright light (1200 lux). Seven days later, the rats were sacrificed and the eyes were removed for sectioning and staining (Fig. 7). The thickness of the ONL plus photoreceptor OS was measured as described (Figs. 7 and 8). Animals that had no treatment displayed marked reduction in the photoreceptor cells with only one or two rows of cell nuclei remaining (Figs. 7A and 8). In addition, there was a shortening of the OS such that measurements of the ONL + OS across the retina averaged only 12 ^m (n = 6). Intravitreal delivery of PBS (Figs. 7B and 8) or the parental HSV-1 vector, hrR3 (data not shown), resulted in limited cell survival of the photore-ceptor layer similar to previously published reports (15, 16, 52). This was most evident near the site of injection where four or five rows of photoreceptor nuclei were

FIG. 5. Differentiation of PC12 cells transduced with 2526/FGF2 or 5042/FGF2. PC12 cells were transduced with 2526/FGF2 or 5042/FGF2 or infected with hrR3 at a m.o.i. of 5 and treated daily for 7 days with ACG (70 jM). Control cells were treated with NGF (50 ng/ml) or rhFGF2 (15 ng/ml) or were left untreated. Arrows indicate neurite extensions from cells treated with 2526/FGF2, 5042/FGF2, NGF, and FGF2 (phase contrast, original magnification 100X).

no treatment

observed. On average, only three or four rows of photo-receptor nuclei were preserved across the retina of these animals, which was not significantly different from untreated animals (P > 0.05).

Animals that received the 2526/FGF2 virus displayed enhanced photoreceptor cell survival (Figs. 7C and 8) with an average of eight or nine rows of photoreceptor nuclei remaining. Measurements of the ONL + OS in 2526/FGF2-treated rats showed normal-length OS near the optic nerve and slightly longer OS (1.5 X) near the ora serata. Delivery of the 2526/FGF2 virus resulted in an average ONL + OS measurement of 38 ^m, which was nearly twice that in rats treated with PBS or vector alone (P < 0.01) and was not significantly different from control animals exposed to normal light. Control animals that had not been exposed to bright light had ONL + OS measurements of 35 ^m near the optic nerve (Figs. 7D and 8). Measurements of the ONL + OS in normal eyes were significantly greater at points near the optic nerve and at the midpoint of the superior and inferior hemisphere, compared to animals that had received either PBS

FIG. 6. Protection of neurons from cell death following NGF and serum withdrawal. Fully differentiated PC12 neuronal cells were transduced in the absence of exogenous growth factors with either 2526/FGF2 or 5042/FGF2 or infected with hrR3 at a m.o.i. of 1 in the presence of ACG (70 jM). Seven days posttreatment, neuronal cells were assayed for cell viability. All samples were compared with the Student t test and (*) denotes samples that are significantly different (P < 0.05) from the mock-transduced or hrR3-infected PC12 cells.

A B C D E

FIG. 7. Photoreceptor cell survival in albino rats subjected to constant bright light for 7 days. Animals received an intravitreal injection of PBS (B), 2526/FGF2 (C), or rhFGF2 (E) or were left uninjected as controls for exposure to bright light (A). Following 7 days of exposure to constant bright light and a 7-day recovery period in normal light, animals were sacrificed and eyes removed, sectioned (10 fim), and stained with nuclear fast red. A normal animal not exposed to bright light is shown in (D). Brackets denote the length of the outer nuclear layer + outer segments. The black bar represents 20 fim.

or hrR3 or no treatment (P < 0.01). Intravitreal delivery of rFGF2 (Figs. 7E and 8) promoted significant photoreceptor cell survival as previously reported (15). In the animals that received rFGF2, an average of eight rows of photoreceptor nuclei was observed in the ONL. In addition, these animals displayed significantly longer OS than PBS-injected animals, with measurements of ONL + OS of approximately 30 ^m near the optic nerve.

Functional ERG Responses

ERGs were measured to determine if the structural preservation of the photoreceptors observed following treatment with rhFGF2 or 2526/FGF2 virus correlated with preservation of retinal function. Animals exposed

FIG. 8. Measurements of photoreceptor ONL + OS thickness in sections from albino rats following 7 days of exposure to constant bright light or from control rats that had been exposed to low-level cyclical light only. Albino rats received an intravitreal injection of PBS, rhFGF2, hrR3 virus, or 2526/FGF2 virus and were exposed to constant bright light for 7 days. Animals were then sacrificed, and the eyes were removed, fixed, sectioned (10 fim), and stained with nuclear fast red. Measurements of ONL + OS were made at eight points across the retina, including four each in the inferior and superior hemispheres, and plotted as a function of distance from the optic nerve.

to constant bright light for 7 days were returned to low-level cyclical light for 7 or 28 days to allow for recovery prior to ERG measurements. Controls included animals that received no treatment or PBS and were exposed to bright light and animals that had not been exposed to bright light (normal). Animals that received either the 2526/FGF2 virus or rhFGF2 prior to exposure to bright light displayed similar ERG patterns (Fig. 9A). The ERG responses from these animals were significantly greater than ERGs from animals that received PBS. However, b-wave and a-wave amplitudes from animals that received either the 2526/FGF2 virus or rhFGF2 were significantly depressed compared to those of animals that had not been exposed to bright light (P < 0.05) (Fig. 9B). Allowing the retina to recover for 28 days did not alter the ERG patterns in any of the groups (data not shown).

To determine whether the 2526/FGF2 or hrR3 virus could alter retinal function in normal rat eyes, ERGs were measured from rats that had received the 2526/ FGF2 or hrR3 virus via intravitreal injection without subsequent exposure to bright light. ERGs were measured 16 days following virus delivery, similar to animals that had been exposed to bright light with a 7-day recovery period. Delivery of the hrR3 virus resulted in nearly complete abrogation of the b-wave and a-wave responses in the ERG compared to the normal rat (P < 0.05) (Figs. 10A and 10B). In animals that received the 2526/FGF2 virus, ERG responses (Figs. 10A and 10B) were partially restored compared to hrR3-treated eyes (P < 0.05). The a-wave response in these animals was significantly greater compared to that of animals that received the hrR3 vector virus alone (P < 0.05) and was similar in magnitude to the a-waves in normal eyes. However, the b-wave responses in rats treated with 2526/FGF2 were significantly reduced compared to those of normal eyes. Histopathologic examination of sections of the eyes from these animals revealed no abnormalities (Fig. 7).

vitro and in vivo to specific growth factors or neurotro-phins can prevent or delay neuronal cell death. For example, FGF2 is neuroprotective for several types of neuronal cells in vitro and in vivo [reviewed in (1)]. The neuropro-tective effects of these proteins, however, are transient, probably due to turnover of the factors. Thus, therapeutic use of these proteins would require frequent reapplication (e.g., injection), which is clinically undesirable or difficult to achieve. Viral vector-mediated delivery of the genes encoding FGF2 with the resulting longer-term expression

FIG. 9. (A) Representative ERGs from albino rats subjected to constant bright light. Albino rats received an intravitreal injection of 2526/FGF2, rhFGF2, or PBS and were subjected to constant bright light for 7 days. Following exposure to bright light, animals were returned to cyclical low-level light for 7 days and ERGs were measured. Animals housed in normal light were used as controls (normal). The baseline for ERG measurements is identified (vertical line) and the peak amplitudes of the a-wave (1) and the b-wave (2) are identified. (B) Mean peak ERG a-wave and b-wave amplitudes from animals subjected to bright light. Albino rats received an intravitreal injection of 2526/FGF2, rh-FGF2, or PBS and were subjected to constant bright light for 7 days. Following exposure to bright light, animals were returned to cyclical low level for 7 days and ERGs were measured. A-wave amplitudes were measured as the difference between the baseline and the peak of the a-wave. B-wave amplitudes were measured as the difference between the baseline and the peak of the b-wave. Samples denoted with (*) are significantly different from the those of the PBS-injected animals, and samples denoted with (+) are significantly different from those of the normal animals (P < 0.05).

Discussion

Neuronal degenerative diseases are significant causes of morbidity and mortality. Although there are several known neuronal degenerative diseases with diverse causes, one feature common to many is the induction of neuronal apoptosis by various insults (53, 54). Previous work has shown that exposure of damaged neurons in

FIG. 10. (A) Representative ERGs from albino rats subjected to cyclical light. Albino rats received an intravitreal injection of either 2526/FGF2 or hrR3 and were housed in a cyclical low-level-light environment for 16 days at which time ERGs were measured. Animals that did not receive an injection and were housed in cyclical light were used as controls (normal). The baseline for ERG measurements is identified (0) and the peak amplitudes of the a-wave (a) and the b-wave (b) are identified. (B) Mean peak ERG a-wave and b-wave amplitudes from animals subjected to cyclical light. Albino rats received an intrav-itreal injection of either 2526/FGF2 or hrR3 and were housed in a cyclical low-level-light environment for 16 days at which time ERGs were measured. A-wave amplitudes were measured as the difference between the baseline and the peak of the a-wave. B-wave amplitudes were measured as the difference between the baseline and the peak of the b-wave. Samples denoted with (*) are significantly different from those of the hrR3-injected animals, and samples denoted with (+) are significantly different from those of the normal animals (P < 0.05).

of the neurotrophic factor may be a clinically superior alternative. As a first step in determining the suitability of HSV-based vectors for delivery of the neuroprotective factor FGF2, we tested the ability of HSV-delivered FGF2 to protect neuronal cells in culture and for expression and protective effects in vivo. The two most significant findings of our study are that (i) HSV-mediated delivery of FGF2 was neuroprotective and prevented photoreceptor death and (ii) vector delivery suppressed functional ERG responses in the absence of any histological evidence of retinal damage.

The results clearly emphasize the point that assaying anatomical protection of neuronal cells following gene delivery is not sufficient to claim a therapeutic effect and that functional studies are essential. In addition, several other novel findings, including: (i) HSV-mediated delivery and expression of vFGF2 in differentiated PC12 cells prevent NGF withdrawal-induced cell death, (ii) delivery and expression of the vFGF2 gene in PC12 cells induced differentiation to the neuronal phenotype and protected the cells from death, and (iii) expression from the CMV promoter lasted less than 7 days in vivo even though the vector genome persisted for at least 28 days, suggesting promoter shutoff is occurring.

The PC12 cell line is a useful in vitro model for studying neuronal cell survival due to its dependence on exogenous growth factors for survival under low-serum conditions and because it will differentiate to a neuronal phe-notype in the presence of NGF or FGF2 (43, 55, 56). PC12 cells transduced with 2526/FGF2 and 5042/FGF2 displayed significantly greater survival than control untreated or hrR3-infected cells in the absence of exog-enously added growth factor. The FGF2 viruses were also able to induce differentiation as evidenced by neurite extensions similar to that seen in rNGF- or rhFGF2-treated cells. These studies were performed with a m.o.i. of 5, which would result in less than 0.1 ng of vFGF2 production per 1 X 104 PC12 cells. This is equivalent to a concentration of 1 ng/ml of vFGF2 in the transduced PC12 cells. The minimum effective concentration of FGF2 previously reported to induce differentiation of PC12 cells is 10 ng/ml (56). Thus in our studies, presumably low concentrations of FGF2 were neuroprotective, which suggests that lower concentrations of virally delivered FGF2 may yield clinically significant results. The 2526/FGF2 virus-transduced cells exhibited significantly greater cell survival compared to the 5042/FGF2 virus, even though the two viruses produced similar quantities of the vFGF2 protein in Vero or PC12 neuronal cells and replicated to virtually identical titers in Vero cells (Fig. 3). The reason for the difference between the two viruses is not readily apparent and will require further study.

Treatment of neuronal degeneration in the adult brain will require delivery of FGF2 to nondividing neurons to prevent death. Thus, an in vitro model that utilizes rapidly dividing PC12 cells may not be ideal for testing potential therapies, and a model that relies on terminally differentiated, nondividing neuronal-like cells would be more appropriate. Differentiation of PC12 cells to the neuronal

phenotype by addition of NGF followed by growth factor withdrawal provides such a model. Previous studies have reported that concentrations of exogenously added FGF2 as low as 1 ng/ml can protect dissociated and organotypic cultures of cortical, hippocampal, and dopaminergic neurons from apoptosis (2-7). Similarly, investigators have found that concentrations of exogenously added FGF2 as low as 1 ng/ml prior to injury can promote the survival of neurons in vivo (8, 9). In our studies, endogenously expressed FGF2 from the FGF2/HSV-1 viruses appears to protect differentiated PC12 neuronal cells at lower concentrations than exogenously added growth factor, suggesting that low titers of FGF2-expressing virus may be able to protect neurons from apoptosis in the adult mammalian brain. We also found that the hrR3 virus alone appeared to have a protective effect in cell cultures. The hrR3 virus expresses several HSV antiapoptotic genes, which would explain this effect (57-59). It is also possible that hrR3 infection induces other neuroprotective factors. Further studies are needed to explain these effects.

The 18-kDa secreted form of FGF2 lacks a conventional secretory signal, and it is not clear how FGF2 is released from cells. Addition of a traditional secretory signal pep-tide to the FGF2 results in tumorigenesis (60, 61); therefore, this modification poses significant potential risks if used to improve the release of FGF2 from cells for gene therapy. We were unable to detect FGF2 released into the culture medium from the cells transduced with 2526/ FGF2 or 5042/FGF2 even using heparin affinity columns; thus, it is not clear if the protective effect is occurring via autocrine or paracrine mechanisms. We also attempted to inhibit neuroprotection and differentiation activity by including antibodies specific for FGF2 in the culture medium. Three different antibodies from two manufacturers (Cat. No. SC79 and SC79-G, Santa Cruz Biotechnology, and Cat. No. F5537, Sigma Chemical) were unable to inhibit the effects of the FGF2/HSV viruses. The anti-FGF2 antibodies are described by the manufacturers as neutralizing antibodies in the NIH3T3 cell survival assay; however, this neutralizing effect was not observed in our PC12 cell assay. Thus, we were unable to determine if the effects of vFGF2 were due to virus-directed endogenous synthesis or to vFGF2 release from cells. A unique secretory pathway may exist for the 18-kDa FGF2 (62, 63). Since un-transduced PC12 cells do not normally express FGF2, the secretion pathway may not exist in these cells and attempts to detect vFGF2 in the culture medium would be futile. We have previously shown that the parental vector, hrR3, is capable of infecting glial cells, astrocytes, and neurons in the rat CNS and retinal pigment epithelium (RPE) cells in the rodent and primate retina (40, 64), which are the primary sources of endogenous FGF2 in the adult animal (65, 66). Because these cells normally release FGF2, delivery of the FGF2/HSV virus to these cells may allow for release of the vFGF2 protein by the normal pathway.

The previous studies have shown that FGF2 delivered by either AV or AAV vectors can block or delay photore-ceptor degeneration as assessed histologically (37, 39). We

have now shown that the same neuroprotection effect can be achieved with HSV vectors. In the study by Aki-moto et al. (37), functional ERG responses were not measured, thus a complete therapeutic effect was not demonstrated. Sakamoto et al. (38) showed that intravitreal delivery of an AV vector caused a significant reduction in ERG responses. A similar effect was also reported with AAV vectors (39), and we now report that HSV vectors also reduce ERGs; thus, three major viral gene delivery vectors share this characteristic. The AAV vectors were delivered by subretinal injection and express FGF2 in photoreceptors, thus ERG suppression by AAV may be through direct effects on the photoreceptors. With AV and HSV vectors, the effects are likely indirect, since pho-toreceptors do not express the transgene following intra-vitreal delivery (37, 40). Future studies will be needed to determine the mechanism of ERG suppression and whether different vectors use different mechanisms.

There are several potential explanations for the vector-mediated ERG suppression. The HSV vector transduces retinal ganglion cells (RGC) and cells in the RPE (40). For RGC, the transduction itself could disrupt neuronal signaling; however, components of the ERG generated by photoreceptors (67) were also suppressed even though the vector does not transduce these cells, suggesting that the suppression is indirect. The RPE cells are responsible for providing several functions critical for photoreceptor physiology (68). Since HSV vectors transduce the RPE, it is possible that alterations in RPE cell function could indirectly affect photoreceptors. Viral vectors have the potential to alter cell function via signaling events generated from binding to cells (69) and this issue has received little, if any emphasis to date. Vector delivery could also result in synthesis or release of proinflammatory cytokines that could affect retinal function. Adenovirus vectors induce a transient inflammatory response which could include chemokine or cytokine release. Although we have been unable to detect an inflammatory response following administration of the hrR3 vector in rodent eyes [(40, 41), this study], this does not preclude the induction of cyto-kines or chemokines in the absence of inflammatory cell infiltration due to the immunosuppressive environment of the eye [reviewed in (70)]. It is also possible that limited viral gene expression could cause toxicity to the RPE, which in turn could affect ERGs. We previously showed that RR-null vectors did not replicate when injected in-travitreally in rodent eyes (41). So suppression is unlikely due to replicating virus.

It is unlikely that the suppression of ERG responses in eyes treated with 2526/FGF2 and exposed to bright light was due to the expression of FGF2. We found that ERG responses were suppressed almost completely in eyes from animals kept under constant light but treated with hrR3, which expresses p-galactosidase (71, 72). However, in the eyes of rats kept in normal light and given 2526/ FGF2, the ERG responses were restored to 40% of normal values. These results suggest that expression of FGF2 counteracts the ERG suppression caused by the vector in normal eyes. Conflicting reports about the effect of FGF2

on ERGs exist. Masuda et al. showed that injection of rFGF2 preserved function of the retina after constant light exposure, restoring ERG values to 60% of normal (52). In contrast, Gargini et al. (73) showed that FGF2 reduced the ERG peak b-wave amplitude 25-fold. The reason for the discrepancy is not clear. The fact that Masuda et al. and we (this study) showed partial restoration of ERG responses after rFGF2 injection or viral FGF2 delivery suggests that FGF2 is not causing the suppression of ERG responses.

Many viral vectors are currently being examined for the introduction of exogenous genes to the retina, CNS, or PNS; however, few studies have been undertaken to assess the impact on synaptic transmission in the transduced tissues. Thus, previous reports claiming therapeutic efficacy as measured only by structural preservation of neurons in the retina or CNS need to be reconsidered. Our results clearly show that functional studies are critical in assessing therapeutic efficacy; thus, future examination of viral vectors constructed for gene delivery to the nervous system will have to be assessed for their impact on synaptic transmission and other cell functions.

Acknowledgments

The authors thank Janice Lokken for preparing the tissue sections and Inna Larsen for administrative assistance. Drs. Donna Peters, Robert Nickells, and James VerHoeve kindly provided critical comments on the manuscript. This work was supported by a grant from the Retina Research Foundation, Houston, Texas, to C.R.B., a grant to the Department of Ophthalmology and Visual Sciences from Research to Prevent Blindness, Inc., and Grant EY07736 from the NEI to C.R.B. C.R.B. is the Retina Research Foundation Alice McPherson Professor.

References

1 Bikfalvi, A., Klein, S., Pintucci, G., and Rifken, D. B. (1997). Biological roles of fibroblast growth factor-2. Endocr. Rev. 18: 26-45.

2 Hou, J. G., and Mytilineou, C. (1996). Secretion of GDNF by glial cells does not account for the neurotrophic effect of bFGF on dopamine neurons in vitro. Brain Res. 724: 145-148.

3 Lowenstein, D. H., and Arsenault, L. (1996). The effects of growth factors on the survival and differentiation of cultured dentate gyrus neurons. J. Neurosci. 16: 1759-1769.

4 Morrison, R. S., Sharna, A., DeVellis, J., and Bradshaw, R. A. (1986). Basic fibroblast growth factor supports the survival of cerebral cortical neurons in primary cultures. Proc. Natl. Acad. Sci. USA 83: 7537-7541.

5 Nakagami, Y., Saito, H., and Matsuki, N. (1997). Basic fibroblast growth factor and brain-derived neurotrophic factor promote survival and neuronal circuit formation in orga-notypic hippocampal culture. Jpn. J. Pharmacol. 75: 319-326.

6 Prehn, J. H. M. (1996). Marked diversity in the action of growth factors on N-methyl-d-aspartate-induced neuronal degeneration. Eur. J. Pharmacol. 306: 81-88.

7 Walicke, P., Cowan, W. M., Ueno, N., Baird, A., and Guillemin, R. (1986). Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc. Natl. Acad. Sci. USA 83: 3012-3016.

8 Sievers, J., Hausmann, B., Unsicker, K., and Berry, M. (1987). Fibroblast growth factors promote the survival of adult retinal ganglion cells after transection of the optic nerve. Neurosci. Lett. 76: 157-162.

9 Agarwala, S., and Kalil, R. E. (1998). Long-term protection ofaxotomized neurons in the dorsal lateral geniculate nucleus in the rat following a single administration of basic fibroblast growth factor or ciliary neurotrophic factor. J. Comp. Neurol. 392: 264-272.

10 Anderson, K. J., Dam, D., Lee, S., and Cotman, C. W. (1988). Basic fibroblast growth factor prevents death of lesioned cholinergic neurons in vivo. Nature 232: 360-361.

11 Otto, D., Unsicker, K., and Grothe, C. (1987). Pharmacological effects of nerve growth factor and fibroblast growth factor applied to the transectioned sciated nerve on neuron death in adult rat dorsal root ganglia. Neurosci. Lett. 83: 156-160.

12 Wen, T. C., etal. (1995). Protective effect of basic fibroblast growth factor-heparin and neurotoxic effect of platelet factor 4 on ischemic neuronal loss and learning disability in gerbils. Neuroscience 65: 513-521.

13 Emmett, C. J., Aswani, S. P., Stewart, G. R., Fairchild, D., and Johnson, R. M. (1995). Dose-response comparison of recombinant human nerve growth factor and recombinant human basic fibroblast growth factor in the fimbria fornix model of acute cholinergic degeneration. Brain Res. 673: 199-207.

14 Ikeda, K., et al. (1995). Neuroprotective effect of basic fibroblast growth factor on wobbler mouse motor neuron disease. Neurol. Res. 17: 445-448.

15 Faktorovich, E. G., Steinberg, R. H., Yasumura, D., Matthes, M. T., and LaVail, M. M.

(1990). Photoreceptor degeneration in inherited retinal dystrophy delayed by basic fibroblast growth factor. Nature 347: 83-86.

16 Faktorovich, E. G., Steinberg, R. H., Yasumura, D., Matthes, M. T., and LaVail, M. M. (1992). Basic fibroblast growth factor and local injury protect photoreceptors from light damage in the rat. J. Neurosci. 12: 3554-3567.

17 Lin, N., Fan, W., Sheedlo, H. J., and Turner, J. E. (1997). Basic fibroblast growth factor treatment delays age-related photoreceptor degeneration in Fischer 344 rats. Exp. Eye Res. 64: 239-248.

18 LaVail, M. M., Gorrin, G. M., Repaci, M. A., and Yasumura, D. (1987). Light induced retinal degeneration in albino mice and rats: Strain and species differences. In Degenerative Retinal Disorders: Clinical and Laboratory Investigations (J. G. Hollyfield, R. E. Anderson, and M. M. LaVail, Eds.), Column 247, pp. 439-454. Liss, New York.

19 LaVail, M. M., et al. (1991). Basic fibroblast growth factor protects photoreceptors from light-induced degeneration in albino rats. Ann. N. Y. Acad. Sci. 638: 341-347.

20 LaVail, M. M., et al. (1992). Multiple growth factors, cytokines, and neurotrophins rescue photoreceptors from damaging effects of constant light. Proc. Natl. Acad. Sci. USA 89: 11249-11253.

21 Smith, S. B., Titelman, R., and Hamasaki, D. I. (1996). Effects of basic fibroblast growth factor on the retinal degeneration of the mivit/mivit (vitiligo) mouse: A morphologic and electrophysiological study. Exp. Eye Res. 63: 565-577.

22 Uteza, Y., et al. (1999). Intravitreous transplantation of encapsulated fibroblasts secreting the human fibroblast growth factor 2 delays photoreceptor cell degeneration in Royal College of Surgeons rats. Proc. Natl. Acad. Sci. USA 96: 3126-3131.

23 Lucas, D. R., Attfield, M., and Davey, J. B. (1955). Retinal dystrophy in the rat. J. Pathol. Bacteriol. 70: 469-474.

24 Dowling, J. E., and Sidman, R. L. (1962). Inherited retinal dystrophy in the rat. J. Cell Biol. 14: 73-106.

25 Bourne, M. C., Campbell, D. A., and Tansley, K. (1938). Hereditary degeneration of the rat retina. Br. J. Ophthalmol. 22: 613-623.

26 Nandrot, E., et al. (2000). Homozygous deletion in the coding sequence of the c-mer gene in RCS rats unravels general mechanisms of physiological cell adhesion and apoptosis. Neurobiol. Dis. 7: 586-599.

27 Lee, E. W., Render, J. A., Garner, C. D., Brady, A. N., and Li, L. C. (1990). Unilateral degeneration of retina and optic nerve in Fischer-344 rats. Vet. Pathol. 27: 439-444.

28 Shinowara, N. L., London, E. D., and Rapoport, S. I. (1982). Changes in retinal morphology and glucose utilization in aging albino rats. Exp. Eye Res. 34: 517-530.

29 Lai, Y.-L., Jacoby, R. O., and Jonas, A. M. (1978). Age-related and light-associated retinal changes in Fischer rats. Invest. Ophthalmol. Visual Sci. 17: 634-638.

30 DiLoreto, D., Jr., et al. (1994). The influences of age, retinal topography, and gender on retinal degeneration in the Fischer 344 rat. Brain Res. 647: 181-191.

31 Organisciak, D. T., Darrow, R. M., Noell, W. K., and Blanks, J. C. (1995). Hyperthermia accelerates retinal light damage in rats. Invest. Ophthalmol. Visual Sci. 36: 997-1008.

32 deRaad, S., Szczesny, P. J., Munz, K., and Reme, C. E. (1996). Light damage in the rat retina: Glial fibrillary acidic protein accumulates in Muller cells in correlation with photore-ceptor damage. Ophthalm. Res. 28: 99-107.

33 Noel, W. K. (1980). Possible mechanisms of photoreceptor damage by light in mammalian eyes. Vision Res. 20: 1163-1171.

34 Anderson, R. E., Rapp, L. M., and Wiegand, R. D. (1984). Lipid peroxidation and retinal degeneration. Curr. Eye Res. 3: 223-227.

35 Handelman, G. J., and Dratz, E. A. (1986). The role of antioxidants in the retinal and retinal pigment epithelium and the nature of prooxidant-induced damage. Adv. Free Radical Biol. Med. 2: 1-89.

36 Steinberg, R. H. (1996). Transgenic rat models of inherited retinal degeneration caused by mutant opsin genes. Invest. Ophthalmol. Visual Sci. 37: S698.

37 Akimoto, M., et al. (1999). Adenovirally expressed basic fibroblast growth factor rescues photoreceptor cells in RCS rat. Invest. Ophthalmol. Visual Sci. 40: 273-279.

38 Sakamoto, T., et al. (1998). Retinal functional change caused by adenoviral vector-mediated transfection of lacZ gene. Hum. Gene Ther. 9: 789-799.

39 Lau, D., et al. (2000). Retinal degeneration is slowed in transgenic rats by AAV-mediated delivery of FGF-2. Invest. Ophthalmol. Visual Sci. 41: 3622-3633.

40 Spencer, B., Agarwala, S., Miskulin, M., Smith, M., and Brandt, C. R. (2000). Herpes simplex virus-mediated gene delivery to the rodent visual system. Invest. Ophthalmol. Visual Sci. 41: 1392-1401.

41 Brandt, C. R., et al. (1997). The herpes simplex virus type 1 ribonucleotide reductase is required for acute retinal disease. Arch. Virol. 142: 883-896.

42 Grau, D. R., Visalli, R. J., and Brandt, C. R. (1989). Herpes simplex virus stromal keratitis is nottiter-dependentand does not correlate with neurovirulence. Invest. Ophthalmol. Visual Sci. 30: 2474-2480.

43 Greene, L. A., and Tischler, A. S. (1976). Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc. Natl. Acad. Sci. USA 73: 2424-2428.

44 Wilcox, C. L., and Johnson, E. M. (1987). Nerve growth factor deprivation results in the reactivation of latent herpes simplex virus in vitro. J. Virol. 61: 2311-2315.

45 Wilcox, C. L., and Johnson, E. M. (1988). Characterization of nerve growth factor-dependent herpes simplex virus latency in neurons in vitro. J. Virol. 62: 393-399.

46 Tal-Singer, R., et al. (1997). Gene expression during reactivation of herpes simplex virus type 1 from latency in the peripheral nervous system is different from that during lytic infection of tissue cultures. J. Virol. 71: 5268-5276.

47 Brandt, C. R., and Grau, D. R. (1990). Mixed infection with herpes simplex virus type 1 generates recombinants with increased ocular and neurovirulence. Invest. Ophthalmol. Visual Sci. 31: 2214-2223.

48 Sramek, S. J., Wallow, I. H., Tewksbury, D. A., Brandt, C. R., and Poulsen, G. L. (1992). An ocular renin-angiotensin system. Immunohistochemistry of angiotensinogen. Invest. Ophthalmol. Visual Sci. 33: 1627-1632.

49 Visalli, R. J., and Brandt, C. R. (1993). The HSV-1 UL45 18 kDa gene product is a true late protein and a component of the virion. Virus Res. 29: 167-178.

50 Noel, W. K., Walker, V. S., Kang, B. S., and Berman, S. (1966). Retinal damage by light in rats. Invest. Ophthalmol. Visual Sci. 5: 450-473.

51 Goto, Y., Peachey, N. S., Ripps, H., and Naash, M. I. (1995). Functional abnormalities in transgenic mice expressing a mutant rhodopsin gene. Invest. Ophthalmol. Visual Sci. 36: 62-71.

52 Masuda, K., Watanabe, I., Unoki, K., Ohba, N., and Muramatsu, T. (1995). Functional rescue of photoreceptors from the damaging effects of constant light by survival-promoting factors in the rat. Invest. Ophthalmol. Visual Sci. 36: 2142-2150.

53 Mizuno, Y., Mochizuki, H., Sugita, Y., and Goto, K. (1998). Apoptosis in neurodegenerative disorders. Intern. Med. 37: 192-193.

54 Roy, M., and Sapolsky, R. (1999). Neuronal apoptosis in acute necrotic insults: Why is this subject such a mess? Trends Neurosci. 22: 419-422.

55 Togari, A., Baker, D., Dickens, G., and Guroff, G. (1983). The neurite-promoting effect of fibroblast growth factor. Biochem. Biophys. Res. Commun. 114: 1189-1193.

56 Togari, A., Dickens, G., Kuzuya, H., and Guroff, G. (1985). The effect of fibroblast growth factor on PC12 cells. J. Neurosci. 5: 307-316.

57 Leopardi, R., and Roizman, B. (1996). The herpes simplex virus major regulatory protein ICP4 blocks apoptosis induced by the virus or by hyperthermia. Proc. Natl. Acad. Sci. USA 93: 9583-9587.

58 Leopardi, R., VanSant, C., and Roizman, B. (1997). The herpes simplex virus 1 protein kinase US3 is required for protection from apoptosis induced by the virus. Proc. Natl. Acad. Sci. USA 94: 7891-7896.

59 Aubert, M., and Blaho, J. (1999). The herpes simplex virus type 1 regulatory protein ICP27 is required for the prevention of apoptosis in infected human cells. J. Virol. 73: 2803-2813.

60 Blam, S. B., et al. (1988). Addition of growth hormone secretion signal to basic fibroblast growth factor results in cell transformation and secretion of aberrant forms of the protein. Oncogene 3: 129-136.

61 Rogej, A., Weinberg, R. A., Fanning, P., and Klagsbrun, M. (1988). Basic fibroblast growth factor fused to a signal peptide transforms cells. Nature 331: 173-175.

62 Florkiewicz, R. Z., Majack, R. A., Buechler, R. D., and Florkiewicz, E. (1995). Quantitative export of FGF-2 occurs through an alternative, energy-dependent, non-ER/Golgi pathway. J. Cell. Physiol. 162: 388-399.

63 Mignatti, P., Morimoto, T., and Rifkin, D. B. (1992). Basic fibroblast growth factor, a protein devoid of secretory sequence signal, is released by cells via a pathway independent of the endoplasmic reticulum-Golgi complex. J. Cell. Physiol. 151: 81-93.

64 Liu, X., et al. (1999). Herpes simplex virus mediated gene delivery to primate ocular tissues. Exp. Eye Res. 69: 385-395.

65 Pettman, B., Labourdette, G., Weibel, M., and Sensenbrenner, M. (1986). The brain fibroblast growth factor (FGF) is localized to neurons. Neurosci. Lett. 68: 175-179.

66 Ferrara, N., Ousley, F., and Gospodarawicz, D. (1988). Bovine brain astrocytes express basic fibroblast growth factor, a neurotropic and angiogenic mitogen. Brain Res. 462: 223-232.

67 Granit, R. (1933). The components of the retinal action potential in mammals and their relation to the discharge in the optic nerve. J. Physiol. 77: 14-39.

68 Van Soest, S., Westerveld, A., De Jong, P. T. V. M., Bleeker-Wagemakers, E. M., and Bergen, A. A. B. (1999). Retinitis pigmentosa: Defined form a molecular point of view. Surv. Ophthalmol. 43: 321-334.

69 Greber, U. F., Willetts, M., Webster, P., and Heleniu, S. A. (1993). Stepwise dismantling of adenovirus 2 during entry into cells. Cell75: 477-486.

70 Niederkorn, J. Y. (1990). Immune privilege and immune regulation in the eye. Adv. Immunol. 48: 191-227.

71 Goldstein, D. J., and Weller, S. K. (1988). Factor(s) present in herpes simplex virus type 1-infected cells can compensate for the loss of the large subunit of the viral ribonucleotide reductase: Characterization of an ICP6 deletion mutant. Virology 166: 41-51.

72 Goldstein, D. J., and Weller, S. K. (1988). Herpes simplex virus type-1 ribonucleotide reductase activity is dispensable for virus growth and DNA synthesis: Isolation and characterization of an ICP6 lacZ insertion mutant. J. Virol. 62: 196-205.

73 Gargini, C., et al. (1999). The impact of basic fibroblast growth on photoreceptor function and morphology. Invest. Ophthalmol. Visual Sci. 40: 2088-2099.