Introduction

Age-related macular degeneration (AMD) is the most common cause of acquired blindness in the developed countries^1,2. Approximately 90% of patients with AMD have the nonexudative or dry form characterized by atrophy of the retinal pigment epithelium (RPE)³. The exudative or wet form of AMD affects only 10% of those with the disease and is characterized by abnormal blood vessels growing from the choriocapillaris through the disturbed RPE layer, typically resulting in subretinal fluid, subretinal hemorrhage, hard exudates, and/or retinal detachment⁴.

Accumulation of fluorescent lipofuscin granules in the RPE and extracellular deposits called drusen on Bruch's membrane, the basement membrane of the RPE, are frequently observed in diseased eyes⁵. Since the major fluorophore of lipofuscin, N-retinylidene-N-retinylethanolamine (A2E), has potential toxicity to the RPE⁶, the dry form of AMD may be a consequence of age-related lipofuscin accumulation in the RPE. Although soft drusen is widely accepted as a contributor to the etiology of AMD and as a risk factor for developing to the wet form of AMD^7,8, the causal events responsible for AMD are not clearly delineated.

A possible explanation for the pathogenesis of choroidal neovascularization (CNV), a hallmark of the wet form of AMD, is the increased secretion of angiogenic growth factors^9,10, such as vascular endothelial growth factor (VEGF) and fibroblast growth factor 2, and the decreased secretion of the antiangiogenic factor pigment epithelium-derived factor in the RPE^11,12. However, it is difficult to explain the reason angiogenesis occurs only in the choriocapillaris and not in retinal vessels in this disease. Recently endocrine-gland-derived vascular endothelial growth factor (EG-VEGF)¹³, also called prokineticin 1 (hPK1)¹⁴, was identified as a mitogen specific for the endothelium of steroidogenic glands, which often display a fenestrated phenotype. We hypothesized that this growth factor could induce CNV, because the endothelium of the choriocapillaris, but not retinal endothelium, has a fenestrated phenotype¹⁵.

Here we generated transgenic mice expressing hPK1 specifically in the retina. The enlarged and increased choroidal vascular vessels resembling CNV were observed in these eyes without any retinal vascular changes. In addition, enhanced accumulation of A2E in the eyes was detected. This animal model for AMD can be useful for potential future treatments of the disease.

Results

Transgenesis by lentiviral vector, LV-Rho-hPK1

Previously we have shown that lentiviral transgenesis employing the bovine rhodopsin promoter restricts expression of target genes in the retina (photoreceptor cells)¹⁶. Using this technology, we generated transgenic mice expressing hPK1 in the retina by a lentiviral vector, LV-Rho-hPK1 (Fig. 1A). We detected the transgene by PCR and genomic Southern blot analysis (data not shown). Since hPK1 is a secreted protein^13,14, we subjected whole eye extracts to Western blot analysis.

Figure 1.

(A) Diagram of lentiviral vector (LV-Rho-hPK1) used for generation of transgenic mice expressing hPK1 in retina. Rho, bovine rhodopsin gene promoter sequence. The lentiviral vector carries a posttranscriptional regulatory element of the woodchuck hepatitis virus (wPRE)³⁰, a central polypurine tract (cPPT)³¹ of human immunodeficiency virus-1, and self-inactivating deletion mutations (open triangle)³². Arrow indicates BamHI site, a unique site in the vector. RSV, promoter sequence from respiratory syncytial virus; , packing signal. (B) Lysates of eyes derived from each line of transgenic mice were subjected to Western blot analysis. hPK1 expression was confirmed in lines 1 and 5.

Full figure and legend (53K)

hPK1 expression in the eye of a transgenic mouse was detected in lines 1 and 5 (Fig. 1B). We confirmed hPK1 expression in the retina further by immunohistological analysis (see Figs. 5C).

Figure 5.

(A) Cryosections of eyes derived from age-matched wild-type control mice (8 months of age) and transgenic mice (lines 1 and 5). Arrows indicate RPE cell layer. Immunohistochemical analysis for (B) endothelial cell and (C) hPK1. (B) The sections shown in A were stained with FITC-conjugated GSA-I-B4 lectin, which binds to the endothelial cells of nonprimates. Each vessel was FITC-positive (arrowhead), thus ruling out artificial vessels. (C) Each section was reacted with anti-hPK1 antibody and then visualized by RR-conjugated anti-rabbit IgG antibody. hPK1 expression was observed, especially at the interphotoreceptor matrix, in both lines of transgenic mice (C', C").

Full figure and legend (226K)

Morphological analysis of the eyes

Fluorescein angiography

To visualize the vascular bed in retinal and choroidal tissue, we performed fluorescein angiography in each line of transgenic and age-matched wild-type control mice (8 months of age) (Fig. 2). We perfused the mice with fluorescein isothiocyanate (FITC)-conjugated dextran through the left ventricle of the heart and then enucleated the eye. We stripped the retinal tissues from the eyecup and flat-mounted both retina and RPE–choroid–sclera complex. Surprisingly, we observed a massive accumulation of fluorescein in the choroidal vessels of transgenic mice (Figs. 2B' and 2B"), but not in those of control (Fig. 2B), beyond the RPE layer, which seemed to be intact by light microscopy (Figs. 2A). As expected, we detected no obvious disturbance of retinal capillaries in either wild-type control mice (Fig. 2C) or transgenic mice (Figs. 2C' and 2C"). In the deep plexus layer of the retinal capillaries, there was no evidence of retinal angiogenesis (Figs. 2D), thus confirming choroidal vascular-specific morphological changes.

Figure 2.

Fluorescein angiography. Age-matched wild-type control mice (8 months of age) and transgenic mice (lines 1 and 5) were perfused with FITC-conjugated dextran through the left ventricle. Retinal tissues were stripped from the eyecup and then both (A, B) RPE–choroid–sclera complex and (C, D) retina were flat-mounted. (A) Pictures of the RPE–choroid–sclera complex taken under bright field show intact RPE layer in each eye. (B) Choroidal angiography. Accumulation of fluorescein was observed in the choroidal vessels of transgenic mice (B', B"), but not in that of control (B). Retinal angiography. (C) Low- and (D) high-magnification pictures are shown. (C) No obvious fluorescein accumulation was observed in the retinal capillary. (D) In the deep plexus layer of the retinal capillary, no evidence of retinal microvascular changes, such as aneurysm and tuft of neovascular vessels, were detected in either control or transgenics.

Full figure and legend (400K)

Histology

To exclude the possible induction of fenestrations in retinal endothelium by hPK1 transduction, we examined ultrastructural changes in the transgenic mouse retina. The retinal vessel had a lumen lined with a thick continuous retinal endothelium similar to controls (data not shown). We could not observe any fenestrated retinal endothelium (Fig. 3). To demonstrate the fine structure of the CNV-like region, we next made transverse sections of eyes derived from each line of transgenic mouse and age-matched wild-type control mice (8 months of age) (Fig. 4). The number and the diameter of choroidal vessels indicated by arrows were increased in transgenic mice (Fig. 4, middle and bottom) compared to the control (Fig. 4, top). It is worth noting that although these transgenic mice showed choroidal thickening with increased choroidal vascular bed, we detected no evidence of penetration of choroidal vessels through Bruch's membrane, the hallmark of CNV. We also made cryosections of eyes (Fig. 5). The blood vessels were identified by staining with FITC-conjugated Griffonia simplicifolia agglutinin-I-B4 (GSA-I-B4) lectin (Fig. 5B), which binds to the endothelial cells of nonprimates¹⁷. Again, we could observe increased choroidal vascular bed. We examined the expression of hPK1 in the eye further by immunohistochemical analysis using rabbit anti-hPK1 antibody and it was expressed throughout the retina, especially at the interphotoreceptor matrix (IPM), in both lines of transgenic mice (Figs. 5C' and 5C"). Expression of endogenous mouse PK1 was not readily detectable because the antibody was raised against human PK1.

Figure 3.

Electron microscopy. Retinal blood vessels of a transgenic mouse (8 months of age, line 1) were examined. The retinal vessel in the deep plexus layer had a lumen (^*) lined with thick continuous retinal endothelium (En) similar to controls (data not shown). Although the retinal endothelium was thinner at the sites indicated by arrows, no evidence of induced fenestration was observed. Photoreceptor cells had dark nuclei at the outer nuclear layer (ONL) and bipolar cells had relatively light nuclei at the inner nuclear layer (INL).

Full figure and legend (172K)

Figure 4.

Transverse sections of eyes derived from age-matched wild-type control mice (8 months of age) and transgenic mice (lines 1 and 5). The number and the diameter of choroidal vessels (arrows) were increased in transgenic mice compared to controls. Note that the choroidal vessels are uneven in diameter and in vascular wall thickness. OS, outer segment of photoreceptor cell; RPE, retinal pigment epithelium.

Full figure and legend (246K)

Analysis of lipofuscin fluorophores in the ocular tissues

Although we did not observe apparent morphological changes in the RPE layer of the eyes (Figs. 2, 4, and 5), we further examined biochemical alterations in the RPE cells of these transgenics as well as age-matched wild-type control mice (8 months of age). Representative chromatographic tracings obtained from posterior eyecup, including retina/RPE, phospholipid extracts of control mice (Fig. 6A) and transgenic mice (Fig. 6B) are shown. While there was no significant difference in abundance of the major phospholipid classes, a major lipofuscin fluorophore (A2E) and its precursor (N-retinylidene-N-retinylphosphatidylethanolamine, A2PE) were significantly elevated in the transgenic mouse eyes (Fig. 6B). When we normalized the data to the amount of phosphatidylcholine, a single phospholipid class, we obtained a similar result (data not shown). We confirmed the identification of A2E and A2PE by spectral analysis of the eluted peaks (Fig. 6B, inset). Comparison of the internal standard peak heights suggests that the observed difference was not due to differences in sample extraction or loading amounts. Quantitative analyses revealed a 2.4-fold increase in A2E (Fig. 6C) and a 3.4-fold increase in A2PE (Fig. 6D) in transgenic mice relative to control mice. The absolute levels of A2E found in control mice in the present study are consistent with previous reports of A2E in wild-type mice of this age and reflect the natural accumulation of this age pigment^18,19.

Figure 6.

Analysis of lipofuscin fluorophores (A2E and its precursor A2PE) in the eyes of age-matched wild-type control mice (8 months of age) and transgenic mice (line 1). Representative chromatographic tracings obtained from posterior eyecup, including retina/RPE, phospholipid extracts of (A) control mice and (B) transgenic mice are shown. The arrows indicate the peaks derived from A2E, A2PE, and internal standard. Identification of A2E and A2PE was confirmed by spectral analysis of the eluted peaks (B, inset). Quantitative analyses revealed (C) a 2.4-fold increase in A2E and (D) a 3.4-fold increase in A2PE in transgenic mice relative to control mice.

Full figure and legend (191K)

Discussion

Macular degeneration is the physical disturbance of the center of the retina called the macula. The macula is the part of the retina that is capable of our most acute and detailed vision. AMD is the leading cause of legal blindness in people over age 55. (Legal blindness means that a person can see only 20/200 or worse with eyeglasses.) Even with a loss of central vision, however, peripheral vision may remain clear. The root causes of macular degeneration are still unknown but abnormal blood vessels growing in the choroid are likely to be the main culprit. hPK1 was originally cloned by Li et al. as an endogenous regulator of gastrointestinal motility¹⁴. Later, LeCouter et al. cloned a novel growth factor called EG-VEGF, which was identical to hPK1¹³. It was reported that hPK1 induced proliferation and migration in capillary fenestrated endothelial cells. Furthermore expression of hPK1 is hypoxia inducible due to the presence of a hypoxia inducible factor-1 binding site in its promoter region¹³. We focused on this growth factor because its site of action is the fenestrated endothelium but not the continuous endothelium. CNV, a hallmark of the exudative form of AMD, is characterized by abnormal blood vessels growing from the choriocapillaris⁴ whose endothelium is fenestrated¹⁵. The other reason is its hypoxia inducibility, since hypoxia is one of the factors implicated in the development of AMD²⁰.

To date, there is no generally accepted experimental animal model for CNV. The existing models, such as laser-induced CNV^21,22 and angiogenic agent-induced CNV^23,24, are commonly used to determine the therapeutic effect of CNV treatments; however, reproducibility of those models is limited because induction of CNV was accompanied by a nonspecific, local inflammatory reaction^21,22,23,24. Okamoto et al. reported a transgenic mouse overexpressing VEGF in photoreceptors using the rhodopsin promoter that we employed in this study. These mice developed retinal neovascularization but failed to develop CNV²⁵. Recently Schwesinger et al. demonstrated a transgenic mouse exhibiting CNV by overexpressing VEGF in RPE using the RPE65 promoter²⁶. They hypothesized that RPE may serve as a barrier that blocks VEGF diffusion from the photoreceptors to the choroids. Therefore, they tried to overexpress VEGF in the RPE, expecting it to induce CNV efficiently, and thus obtained a mouse showing CNV.

We could observe a CNV-like region (Figs. 2, 4, and 5) in lines of transgenic mice expressing hPK1 throughout the retina (Figs. 5C' and 5C"). We do not know the mechanism by which hPK1 reaches and affects choroidal vessels. Since hPK1 is secreted and detected particularly in the IPM (Figs. 5C' and 5C"), it is possible that secreted hPK1 could affect choroidal vessels directly through the blood–retinal barrier at the level of the RPE. Alternatively hPK1 could also have an effect indirectly through the systemic circulation. Interestingly, we demonstrated the accumulation in these transgenic mouse eyes of A2E and its precursor, A2PE (Figs. 6C and 6D), which are reported to have several potential cytotoxic effects on RPE cells⁶. This finding would suggest the existence of barrier breakdown allowing hPK1 to affect choroidal vessels directly. Through either pathway, hPK1 was detected in both the retina and the choroid (Figs. 5C' and 5C"). Morphologically, only the choroidal vasculature appeared to be different between wild-type and transgenic mice. hPK1 seems to be specific to choroidal endothelium (fenestrated endothelium), a premise on which this work was initiated, since no evidence of retinal angiogenesis (Fig. 2) or induction of fenestration of retinal endothelium (Fig. 3) was detected.

In the early occult stage of human AMD, abnormal choroidal vessel penetrates Bruch's membrane and the RPE layer where certain damage exists. Then, in the symptomatic stage of the disease, the choroidal vessel grows and forms a neovascular membrane underneath the neural retina. Both events, disturbance of the RPE layer and CNV, are necessary for the development of the disease. Although we found increased A2E accumulation, which has potent toxic effects on the RPE layer, in the transgenic mouse eyes, it would not be strong enough to induce visible RPE damage. Prokineticin 1 could affect the step of choroidal neovascular membrane formation based on the evidence of increased choroidal vascular thickening in the transgenic mice. Unlike in human AMD, we were not able to observe the invasion of a CNV-like region into the subretinal space in transgenic mice because of a lack of visible RPE damage and presumably the much shorter life span of mice as well. Using a laser with minute power, which never causes CNV by itself, on the RPE layer or continuous lightning to stress the RPE and/or photoreceptor cells would be potential methods to accelerate subretinal CNV invasion. Furthermore, it will be also interesting to determine the effects of hPK1 in human AMD by examining the expression pattern of hPK1 in excised choroidal neovascular membranes and/or vitreous derived from patients. Here we propose a novel animal model for AMD, which facilitates the screening of putative therapies for the wet form of the disease.

Materials and methods

Preparation of viral plasmid and viral vector production

Poly(A)⁺ RNAs were isolated from 293 T cells using a QuickPrep Micro mRNA purification kit (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Human PK1 transcript was amplified by RT-PCR using the primer pair hPK1-F and hPK1-R. Primer sequences were as follows: hPK1-F, 5'-GACTGCAGATGAGAGGTGCCACGCGAGTCTCAATCATGC-3'; hPK1-R, 5'-GAAAGCTTCTAAAAATTGATGTTCTTCAAGTCCATGGAGC-3'. Amplicons were subcloned into the pBluescript II SK(+) plasmid (Stratagene, La Jolla, CA, USA), resulting in pBS-SK-hPK1. The lentivirus vector LV-Rho-hPK1 was constructed by replacing the EGFP fragment in LV-Rho-EGFP¹⁶ with the XbaI–SalI fragments of pBS-SK-hPK1. All constructs were confirmed by direct sequencing. Vesicular stomatitis virus G (VSVG) envelope protein-pseudotyped lentiviral vector was generated as described²⁷. In brief, 293 T cells were transfected with the vector, packing plasmids, and a plasmid coding for the VSVG envelope protein by the calcium phosphate method. Virus was harvested over the following 2 days and concentrated by ultracentrifugation (68,000g). The titer of lentiviral vector was determined by measuring the amount of HIV p24 gag antigen by ELISA (NEN Life Science Products, Boston, MA, USA). The multiplicity of infection of lentiviral vector was estimated using the following equation: 1 ng of p24 = 10⁵ infectious units.

Generation of transgenic mice

All experiments were conducted in accordance with the standards of the Salk Institute and the Association for Research in Vision and Ophthalmology statement for the use of animals in ophthalmic and vision research. We generated transgenic mice using lentiviral vector, a novel method developed recently¹⁶. In brief, B6D2F1 females were superovulated by intraperitoneal injection of pregnant mare's serum gonadotropin (5 units) followed by human chorionic gonadotropin (5 units) 48 h later and then mated with B6D2F1 males. Two-cell-stage embryos were collected from the oviducts of the copulated female 36 h after injection of human chorionic gonadotropin. To remove the zona pellucidae, the embryos were placed in acidic Tyrode's solution²⁸ for 30 s to 1 min. After a dissociation of the zona pellucidae was confirmed, embryos were washed three times with kSOM²⁹ and placed in a 5-l drop of kSOM containing lentiviral vector, LV-Rho-hPK1. Blastocysts developed from infected embryos were transferred into 2.5-day pseudopregnant females.

PCR, Southern blot analysis, and Western blot analysis

We performed 40 cycles of PCR with Rho-Tg-F (5'-GGTGCCAGATAAAACTCACAGCTGG-3') and Rho-Tg-R (5'-GCTTAGTGCCAGGGCTTTATCCAGG-3') primers to check the existence of the transgene(s) in the mouse genome. For Southern blot analysis, genomic DNA (20 g) was digested with BamHI, separated by electrophoresis in a 0.8% agarose gel, and blotted onto a nylon membrane before hybridization with the ³²P-random-prime-labeled 598-bp SalI–EcoRI fragment of LV-Rho-hPK1. Lysates of the eyes were subjected to SDS–PAGE under reducing conditions. Proteins transferred to the nitrocellulose membrane were probed by anti-hPK1 antibody and anti-actin antibody (Sigma, St. Louis, MO, USA). The immunoreactive bands were visualized with an ECL detection system (Amersham Pharmacia Biotech). Anti-hPK1 rabbit polyclonal antibody was raised against synthetic peptides corresponding to the hPK1 (60–77; N'-HPGSHKVPFFRKRKHHT-C').

Fluorescein angiography

Each mouse was anesthetized by an intraperitoneal injection of saline solution containing ketamine (70 mg/kg body wt) and xylazine (20 mg/kg body wt) before 1 ml of PBS containing 50 mg of FITC-conjugated dextran (Sigma) was perfused through the left ventricle of heart. The eyes were removed and placed in 4% paraformaldehyde for 1 h. Then retinal tissue was isolated from the eyecup that consists of the RPE–choroid–scleral complex. Both segments were further incubated in 4% paraformaldehyde for 3 h before they were flat-mounted on a glass slide with glycerol gelatin (Sigma) for viewing by fluorescence microscope.

Histological and immunohistochemical analysis

For electron microscopy, retinas were fixed with 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 24 h at 4°C. The retinal segments were postfixed with 1% osmium tetraoxide/potassium ferrocyanide in 0.1 M cacodylate buffer (pH 7.4), stained with 2% uranyl acetate, dehydrated through a series of graded glycol methacrylate, and embedded in standard hard Epon. Ultrathin sections were cut and examined with a transmission electron microscope. For light microscopy, eyes were fixed with 4% paraformaldehyde, embedded in paraffin, and sectioned. Routine hematoxylin and eosin staining was performed. For immunohistochemistry, eyes were fixed with 4% paraformaldehyde, frozen in OCT compound (Sakura Finetek USA, Torrance, CA, USA), and sectioned on a cryostat. Each section was stained with FITC-conjugated GSA-I-B4 (Vector Laboratories, Burlingame, CA, USA). For the detection of hPK1 expression, each section was incubated with rabbit anti-hPK1 antibody. The immunoreaction was visualized by rhodamine red (RR)-conjugated anti-rabbit IgG antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). Each section was observed by fluorescence microscope and the images were recorded with the Axio Vision imaging system (Zeiss).

Analysis of A2E and A2E precursor

Eye samples were incubated on ice in 1 ml of PBS (pH 7.2) and then hemisected to separate the anterior and posterior segments. The anterior portion containing cornea and lens was discarded and the posterior portion was homogenized in 1 ml of PBS (pH 7.2) using a dual glass–glass homogenizer. One milliliter of chloroform/methanol (2/1, v/v) was added and the samples were rehomogenized. Samples were transferred to a borosilicate tube and the lipid-soluble components were extracted into 4 ml of chloroform with vigorous mixing for 10 min. The samples were washed with 3 ml of PBS (pH 7.2) and then centrifuged at 3000g for 10 min to partition organic and aqueous phases. The chloroform (lower) phase was decanted and the aqueous phase was reextracted with another 4 ml of chloroform. Following centrifugation, the chloroform phases were combined and taken to dryness under argon gas. Sample residues were resuspended in hexane and analyzed by HPLC as described below. Chromatographic separations were achieved on an Agilent Zorbax Rx-Sil column (5 m, 4.6 250 mm) using an Agilent 1100 series liquid chromatograph equipped with fluorescence and diode array detectors. The mobile phase (hexane/2-propanol/ethanol/25 mM KH₂PO₄ (pH 7.0)/acetic acid; 485/376/100/50/0.275, v/v) was delivered at 1 ml/min. Sample peak identification was made by comparison to retention time and absorbance spectra of authentic standards. Sample peak quantitation was performed by comparing peak areas to calibration curves constructed with authentic standards.

References

1. Leibowitz, H. M., et al. (1980). The Framingham Eye Study monograph: an ophthalmological and epidemiological study of cataract, glaucoma, diabetic retinopathy, macular degeneration, and visual acuity in a general population of 2631 adults, 1973–1975. Surv. Ophthalmol. 24(Suppl.): 335–610. | Article | PubMed | ChemPort |
2. Klein, R., Klein, B. E. and Linton, K. L. (1992). Prevalence of age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology. 99: 933–943. | PubMed | ISI | ChemPort |
3. Kahn, H. A., et al. (1977). The Framingham Eye Study. II. Association of ophthalmic pathology with single variables previously measured in the Framingham Heart Study. Am. J. Epidemiol. 106: 33–41. | PubMed | ChemPort |
4. Macular Photocoagulation Study Group (1991). Subfoveal neovascular lesions in age-related macular degeneration: guidelines for evaluation and treatment in the macular photocoagulation study. Macular Photocoagulation Study Group. Arch. Ophthalmol. 109: 1242–1257.
5. Marmorstein, A. D., Marmorstein, L. Y., Sakaguchi, H. and Hollyfield, J. G. (2002). Spectral profiling of autofluorescence associated with lipofuscin, Bruch's membrane, and sub-RPE deposits in normal and AMD eyes. Invest. Ophthalmol. Visual Sci. 43: 2435–2441.
6. Sparrow, J. R., Parish, C. A., Hashimoto, M. and Nakanishi, K. (1999). A2E, a lipofuscin fluorophore, in human retinal pigmented epithelial cells in culture. Invest. Ophthalmol. Visual Sci. 40: 2988–2995. | ChemPort |
7. Abdelsalam, A., Del Priore, L. and Zarbin, M. A. (1999). Drusen in age-related macular degeneration: pathogenesis, natural course, and laser photocoagulation-induced regression. Surv. Ophthalmol. 44: 1–28. | Article | PubMed | ChemPort |
8. Midena, E., Deglii Angeli, C., Blarzino, M., Valenti, M. and Segato, T. (1997). Macular function impairment in eyes with early age-related macular degeneration. Invest. Ophthalmol. Visual Sci. 38: 469–477. | ChemPort |
9. Amin, R., Puklin, J. E. and Frank, R. N. (1994). Growth factor localization in choroidal neovascular membranes of age-related macular degeneration. Invest. Ophthalmol. Visual Sci. 35: 3178–3188. | ChemPort |
10. Frank, R. N., Amin, R., Eliott, D., Puklin, J. E. and Abrams, G. W. (1996). Basic fibroblast growth factor and vascular endothelial growth factor are present in epiretinal and choroidal neovascular membranes. Am. J. Ophthalmol. 122: 393–403. | PubMed | ISI | ChemPort |
11. Ogata, N., et al. (2002). Expression of pigment epithelium-derived factor in normal adult rat eye and experimental choroidal neovascularization. Invest. Ophthalmol. Visual Sci. 43: 1168–1175.
12. Renno, R. Z., Youssri, A. I., Michaud, N., Gragoudas, E. S. and Miller, J. W. (2002). Expression of pigment epithelium-derived factor in experimental choroidal neovascularization. Invest. Ophthalmol. Visual Sci. 43: 1574–1580.
13. LeCouter, J., et al. (2001). Identification of an angiogenic mitogen selective for endocrine gland endothelium. Nature. 412: 877–884. | Article | PubMed | ISI | ChemPort |
14. Li, M., Bullock, C. M., Knauer, D. J., Ehlert, F. J. and Zhou, Q. Y. (2001). Identification of two prokineticin cDNAs: recombinant proteins potently contract gastrointestinal smooth muscle. Mol. Pharmacol. 59: 692–698. | PubMed | ISI | ChemPort |
15. Spitznas, M. and Reale, E. (1975). Fracture faces of fenestrations and junctions of endothelial cells in human choroidal vessels. Invest. Ophthalmol. 14: 98–107. | PubMed | ChemPort |
16. Ikawa, M., Tanaka, N., Kao, W.W.-Y. and Verma, I. M. (2003). Generation of transgenic mice using lentiviral vectors: a novel preclinical assessment of lentiviral vectors for gene therapy. Mol. Ther. 8: 666–673. | Article | PubMed | ChemPort |
17. Laitinen, L. (1987). Griffonia simplicifolia lectins bind specifically to endothelial cells and some epithelial cells in mouse tissues. Histochem. J. 19: 225–234. | Article | PubMed | ChemPort |
18. Mata, N. L., Weng, J. and Travis, G. H. (2000). Biosynthesis of a major lipofuscin fluorophore in mice and humans with ABCR-mediated retinal and macular degeneration. Proc. Natl. Acad. Sci. USA. 97: 7154–7159. | Article | PubMed | ChemPort |
19. Kim, S. R., et al. (2004). Rpe65 Leu450Met variant is associated with reduced levels of the retinal pigment epithelium lipofuscin fluorophores A2E and iso-A2E. Proc. Natl. Acad. Sci. USA. 101: 11668–11672. | Article | PubMed | ChemPort |
20. Mousa, S. A., Lorelli, W. and Campochiaro, C. A. (1999). Role of hypoxia and extracellular matrix–integrin binding in the modulation of angiogenic growth factors secretion by retinal pigmented epithelial cells. J. Cell. Biochem. 74: 135–143. | Article | PubMed | ISI | ChemPort |
21. Pollack, A., Korte, G. E., Weitzner, A. L. and Henkind, P. (1986). Ultrastructure of Bruch's membrane after krypton laser photocoagulation. I. Breakdown of Bruch's membrane. Arch. Ophthalmol. 104: 1372–1376. | PubMed | ChemPort |
22. Pollack, A., Korte, G. E., Heriot, W. J. and Henkind, P. (1986). Ultrastructure of Bruch's membrane after krypton laser photocoagulation. II. Repair of Bruch's membrane and the role of macrophages. Arch. Ophthalmol. 104: 1377–1382. | PubMed | ChemPort |
23. Spilsbury, K., Garrett, K. L., Shen, W. Y., Constable, I. J. and Rakoczy, P. E. (2000). Overexpression of vascular endothelial growth factor (VEGF) in the retinal pigment epithelium leads to the development of choroidal neovascularization. Am. J. Pathol. 157: 135–144. | PubMed | ISI | ChemPort |
24. Soubrane, G., et al. (1994). Basic fibroblast growth factor experimentally induced choroidal angiogenesis in the minipig. Curr. Eye Res. 13: 183–195. | PubMed | ChemPort |
25. Okamoto, N., et al. (1997). Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am. J. Pathol. 151: 281–291. | PubMed | ISI | ChemPort |
26. Schwesinger, C., et al. (2001). Intrachoroidal neovascularization in transgenic mice overexpressing vascular endothelial growth factor in the retinal pigment epithelium. Am. J. Pathol. 158: 1161–1172. | PubMed | ISI | ChemPort |
27. Dull, T., et al. (1998). A third-generation lentivirus vector with a conditional packaging system. J. Virol. 72: 8463–8471. | PubMed | ISI | ChemPort |
28. Naito, K., Toyoda, Y. and Yanagimachi, R. (1992). Production of normal mice from oocytes fertilized and developed without zonae pellucidae. Hum. Reprod. 7: 281–285. | PubMed | ChemPort |
29. Ho, Y., Wigglesworth, K., Eppig, J. J. and Schultz, R. M. (1995). Preimplantation development of mouse embryos in KSOM: augmentation by amino acids and analysis of gene expression. Mol. Reprod. Dev. 41: 232–238. | Article | PubMed | ISI | ChemPort |
30. Ramezani, A., Hawley, T. S. and Hawley, R. G. (2000). Lentiviral vectors for enhanced gene expression in human hematopoietic cells. Mol. Ther. 2: 458–469. | Article | PubMed | ISI | ChemPort |
31. Zennou, V., et al. (2000). HIV-1 genome nuclear import is mediated by a central DNA flap. Cell. 101: 173–185. | Article | PubMed | ISI | ChemPort |
32. Miyoshi, H., Blomer, U., Takahashi, M., Gage, F. and Verma, I. M. (1998). Development of a self-inactivating lentivirus vector. J. Virol. 72: 8150–8157. | PubMed | ISI | ChemPort |

Acknowledgements

The authors thank Anand Swaroop for significant discussion and Richard Jacobs for technical help in electron microscopy. This work was supported by the Japan Eye Bank Association (N.T.). I.M.V. is an American Cancer Society Professor of Molecular Biology. He is supported in part by grants from the NIH, the Larry L. Hillblom Foundation, Inc., the Lebensfeld Foundation, the Wayne and Gladys Valley Foundation, and the H. N. and Frances C. Berger Foundation.