Long-term efficacy of ciliary muscle gene transfer of three sFlt-1 variants in a rat model of laser-induced choroidal neovascularization


Inhibition of vascular endothelial growth factor (VEGF) has become the standard of care for patients presenting with wet age-related macular degeneration. However, monthly intravitreal injections are required for optimal efficacy. We have previously shown that electroporation enabled ciliary muscle gene transfer results in sustained protein secretion into the vitreous for up to 9 months. Here, we evaluated the long-term efficacy of ciliary muscle gene transfer of three soluble VEGF receptor-1 (sFlt-1) variants in a rat model of laser-induced choroidal neovascularization (CNV). All three sFlt-1 variants significantly diminished vascular leakage and neovascularization as measured by fluorescein angiography (FA) and flatmount choroid at 3 weeks. FA and infracyanine angiography demonstrated that inhibition of CNV was maintained for up to 6 months after gene transfer of the two shortest sFlt-1 variants. Throughout, clinical efficacy was correlated with sustained VEGF neutralization in the ocular media. Interestingly, treatment with sFlt-1 induced a 50% downregulation of VEGF messenger RNA levels in the retinal pigment epithelium and the choroid. We demonstrate for the first time that non-viral gene transfer can achieve a long-term reduction of VEGF levels and efficacy in the treatment of CNV.


The eye is an organ of choice for protein therapies because the surface of tissue to be treated is small, requiring low amount of protein and once inside the ocular globe, very little diffusion of the proteins is expected to occur owing to the ocular barriers. Intravitreous injections are now routinely used in clinical practice. Monthly injections of Ranibizumab (Lucentis), an anti-vascular endothelial growth factor (VEGF) Fab, have become the standard of care for patients presenting with wet age-related macular degeneration (AMD).1 More recently, pivotal phase 3 trials with Aflibercept (VEGF-trap, Eylea) have shown that an induction treatment of three monthly injections followed by bimonthly treatment were non-inferior to monthly Ranibizumab at preventing vision loss, inducing vision gains and with respect to safety while allowing for reduced treatment frequency.2, 3 However, anti-VEGF therapy has not shown the ability to fully eradicate the choroidal neovascularization (CNV), so that recurrences are common when intravitreal injections are suspended.4 Anti-VEGFs are also routinely used for the treatment of macular edema secondary to vein occlusion5 and diabetic macular edema.6, 7

Despite major advances in the treatment of these patients, thanks to anti-VEGF, the major unmet need remains a drug delivery solution that would avoid or reduce repeated injections and frequent follow-up visits. From a pharmacokinetic point of view, sustained release of anti-VEGF compounds should require a lower total amount for equal functional and anatomical results while avoiding peak and valley effects. To this effect, attempts to formulate proteins in polymeric particulate or solid sustained release systems are progressing8 and gene therapy is being tested in clinical trials. Viral gene therapy has shown its effectiveness for the transfection of retinal pigment epithelial (RPE) cells in animal models of Leber Congenital Amaurosis as well as in humans presenting the disease.9, 10, 11 Based on these studies, adeno-associated virus (AAV) has been proposed to transfect RPE cells with genes encoding secreted diffusible therapeutic proteins such as antibodies or soluble receptors neutralizing VEGF.12, 13, 14, 15 Viral vectors very efficiently transfect RPE cells, but the long-term persistence of viral particles in the retina and the brain,16, 17 as well as the invasive sub-retinal injection used to target the RPE, continue to raise safety concerns. Alternative solutions are being clinically evaluated, such as anti-VEGF producing encapsulated cell implants (Neurotech Inc, Lincoln, RI, USA), which efficiently release the therapeutic protein for at least 2 years in the vitreous.18 We have developed a non-viral gene therapy strategy based on the fact that ciliary muscle cells can be efficiently transduced by electrotransfer (ET) and serve as a biofactory for the sustained production of therapeutic proteins into the vitreous.19 This non-viral gene therapy technique can be applied using a disposable device and a minimally invasive procedure, being therefore applicable for chronic diseases requiring today frequent intraocular injections such as AMD or diabetic macular edema.8

VEGF’s angiogenic actions are mediated through its high-affinity binding to two endothelium-specific receptor tyrosine kinases, VEGFR1 (or Flt-1) and VEGFR2 (or Flk-1/KDR).20, 21 Flt-1 shows at least a 10-fold higher affinity for VEGF than Flk-1/KDR.21 The secreted extracellular domain of Flt-1 receptor (sFlt-1), formed by alternative splicing, exists naturally and binds to VEGF with an affinity equivalent to that of membrane-bound Flt-1.22, 23, 24 sFlt-1 inhibits VEGF angiogenic action by sequestering VEGF.23, 24 Domain deletion studies have shown that the first and the second domains are involved in PGF (placental growth factor) binding, and the second and third Ig-like domains are involved in VEGF binding, while the fourth domain is needed for receptor dimerization.25, 26

We present evidence that ciliary muscle gene ET leads to sustained intraocular production of three sFlt-1 variants and the neutralization of VEGF in a rat model of CNV. We show that sFlt-1, produced and secreted by the ciliary muscle cells into the vitreous, sequesters enough VEGF to efficiently inhibit CNV for up to 6 months after a single ET, opening a real opportunity for the treatment of VEGF-dependent chronic retinal diseases.


Functionality of the sFlt-1 variants encoding plasmids

The three sFlt-1 variants encoding plasmids were specifically and newly generated for this study (see Materials and Methods for details). Although their sequences have already been checked by DNA sequencing, we first proposed to test whether they encoded for the expected proteins (illustrated in Figure 1b) before assessing their therapeutic efficacy. Production of sFlt-1 variants was evaluated by western blotting on supernatants collected from cells transfected either with the empty plasmid pVAX2, used as a negative control, or with each of the sFlt-1-encoding construct. Detection was carried out with an antibody directed against an epitope found in each form (identified with an asterisk in Figure 1b). Bands were detected only in lanes loaded with the supernatants of cells treated with the sFlt-1 variant-encoding plasmids but not in the pVAX2 control samples, demonstrating the specificity of the detected signal (Figure 2). The three sFlt-1 variants had different molecular weights that were relatively well correlated with their expected number of amino acids (Figure 1b). The full-length sFlt-1 was detected as a 100 kDa protein, which corresponds to the molecular weight described for the glycosylated form of the physiological rat soluble Flt-1.27 The medium-sFlt-1 appears with an intermediate molecular weight of about 55–60 kDa, which is in agreement with a previous study showing that the corresponding amino-acid sequence encoded for a 57 kDa protein.25 The small-sFlt-1 has an apparent molecular weight of 45 kDa. To our knowledge, this form has never been previously described in the literature. These results showed that the three sFlt-1 variants encoded by the newly generated plasmids were produced as secreted forms with expected sizes thus confirming that the plasmids were functional in mammalian cells.

Figure 1

Structure of the three rat sFlt-1 variants compared with rat Flt-1. (a) The transmembrane rat Flt-1, synthesized as a 1336 amino-acid residue (aa) protein, is divided in four regions: a signal peptide (23 aa, in gray), an extracellular region, a transmembrane domain (in black) and a cytoplasmic region (in white). The extracellular region is made of seven immunoglobulin (Ig)-like domains (hatched domains), the first three being involved in ligand (PGF and VEGF) binding and the fourth in receptor dimerization. (b) Expression of the three sFlt-1 variants used in this study. The longer variant (687 aa) corresponds to the full-length soluble Flt-1 (sFlt-1). As an alternatively spliced variant of Flt-1, it contains the first six extracellular Ig-like loops found in the transmembrane form and a unique C-terminal intron-derived extension of 31 aa residues (#).60 Medium-sFlt-1 (434 aa) and small-sFlt-1 (363 aa) are truncated forms of sFlt-1 containing, respectively, the first four Ig-like domains and the first three Ig-like domains of the transmembrane Flt-1. The three sFlt-1 variants share the PGF- and VEGF-binding domains but only full-length sFlt-1 and medium-sFlt-1 comprise the receptor dimerization domain. Note that the antibody used for western blot detection of sFlt-1 variants was raised against the N-terminal aa 23–247 of the human Flt-1. It was known to crossreact with rat Flt-1 and the epitope was found in each rat sFlt-1 variant (*).

Figure 2

In vitro production of sFlt-1 variants by transfected culture cells. Four micrograms of pVAX2 control plasmid or plasmids encoding the small-, medium- and full-length rat sFlt-1 (proteins illustrated in Figure 1b) were transfected in ARPE-19 cells using the calcium phosphate method. Each treatment was carried out in duplicate (n=2 wells per experimental group). Western blotting analysis of rat sFlt-1 variants secretion was performed on cell culture supernatants collected 4 days after transfection using an antibody recognizing the epitope illustrated in Figure 1b (see *). No band was detected in culture media collected from control cells treated with the pVAX2 plasmid backbone (lanes 1–2). sFlt-1 was detected with an apparent molecular weight of 45 kDa (lanes 3 and 4), 55–60 kDa (lanes 5 and 6) and 100 kDa (lanes 7 and 8) in supernatants collected from cells treated with the small-, medium- and full-length sFlt-1-encoding plasmid, respectively.

Comparison of the clinical efficacy of sFlt-1 variants on CNV

The primary aims of this study were: (i) to evaluate whether the ciliary muscle ET technology could be useful to manage CNV by producing anti-VEGF proteins intraocularly; (ii) to evaluate and compare the efficacy of truncated forms of sFlt-1 with that of the full-length sFlt-1for which the inhibitory effect on the development on CNV has already been extensively shown in experimental models.28, 29, 30, 31

To address this issue, plasmids encoding the three sFlt-1 variants described previously (full-length, medium- and small-sFlt-1) were electrotransfered into the ciliary muscle 1 week before laser injury, and analyses were carried out when CNV was most prominent in this model, that is, at 2 weeks after laser injury.32

On fluorescein angiography (FA), compared with untreated control eyes (Figures 3a and f) or eyes treated with the pVAX2 plasmid backbone (Figures 3b and g), we observed inhibition of the neovascular response to laser injury in eyes treated with each of the three sFlt-1 variants (Figures 3c–e and Figures 3h–j) (reduced area and intensity of vascular leakage) although laser scars on corresponding fundus appeared similar in all groups (Figures 3k–o). As shown in Figure 3p, clinical grading confirmed that vascular leakage was significantly lower in sFlt-1 (0.36±0.08 arbitrary units (AU)), medium-sFlt-1 (0.21±0.07 AU) and small-sFlt-1 (0.58±0.16 AU) treated eyes than in untreated (1.80±0.19 AU) and pVAX2-treated (2.00±0.19 AU) control eyes. Interestingly, the reduction of leakage observed in sFlt-1 (82%), medium-sFlt-1 (90%) and small-sFlt-1-treated eyes (71%) in comparison with pVAX2-treated ones was not statistically different (Kruskal–Wallis test, P=0.121) demonstrating that the three variants exhibited the same clinical efficacy in the short term (21 days). Another way to express the results is to evaluate the percentage of laser burn without CNV, which means the percentage of laser-induced CNV that does not show any type of leakage (grade 0 on our angiographic scoring). In control groups, 90% of the laser burns were associated with CNV and 10% or less was not associated with CNV. In sFlt-1-treated eyes depending on the molecule size, 64% (sFlt-1), 79% (medium-sFlt-1) and 60% (small-sFlt-1) of burns were not associated with CNV.

Figure 3

Short-term clinical efficacy of sFlt-1 variants evaluated on late FA. (aj) Representative early and late phase fundus angiograms performed at the peak of CNV, 21 days after ciliary muscle ET of pVAX2 (b and g) or of plasmid encoding sFlt-1 variants (sFlt-1 (c and h), medium-sFlt-1 (d and i) or small-sFlt-1(e and j)) (n=8 eyes per experimental group). Comparison was made with untreated control eyes (a and f) (n=4 eyes). (ko) Infrared (IR) photographs taken during fundus angiographies and showing the corresponding laser scars (white arrowheads) in the eye fundus. (p) Grading of vascular leakage performed on fluorescein angiograms showing the mean angiographic score per impact in each experimental group at 21 days (five impacts per eye in each group). Results are expressed in AU (mean+s.e.m.). Statistical analysis: Kruskal–Wallis test: P<0.0001. Mann–Whitney U-test: ###P<0.0001 versus untreated, ***P<0.0001 versus pVAX2.

Moreover, it is worthy to note that ET of the empty plasmid pVAX2 did not diminish nor exacerbate per se the area and intensity of pathological leakage in comparison with untreated controls (Mann–Whitney U-test, P=0.387).

Short-term effect of sFlt-1 variants on CNV growth

At 23 days, on flat-mounted choroids, CNV immunolabeled with a fluorescein isothiocyanate-conjugated tomato lectin extended similarly in untreated control eyes and eyes treated by ET of the empty plasmid pVAX2 (Figures 4a and b) whereas CNV size appeared reduced in eyes treated by each of the three sFlt-1 variant-encoding plasmids (Figures 4c–e). Indeed, CNV area was statistically lower in each sFlt-1 variant-treated group (30 431±2265, 27 912±1416 and 23 866±2036 μm2, respectively, for sFlt-1, medium-sFl-1 and small-sFlt-1) compared with pVAX2-treated eyes (46 457±2536 μm2) (Figure 4f). No statistical difference was shown between the three variants, each inducing a significant reduction of 40% of laser-induced CNV in comparison with the plasmid backbone. Furthermore, measurement analysis highlighted that no decrease or exacerbation of CNV were induced in pVAX2-treated eyes compared with untreated control eyes (50 023±4 428 μm2).

Figure 4

Short-term effects of sFlt-1 variants on CNV development. CNV in RPE cells/choroid flatmounts were labeled with fluorescein isothiocyanate-conjugated lycopersicon esculentum tomato lectin 23 days after ciliary muscle ET. (a–e) Representative images of CNV observed in (a) untreated control eyes (n=4 eyes) and eyes treated 7 days before CNV induction by ciliary muscle ET of (b) pVAX2 or of (c–e) each of the three sFlt-1 variants encoding plasmids (n=8 eyes per group). Scale bar=200 μm. (f) Quantification of neovascular areas by digital images analysis on RPE/choroid flatmount showing the mean (+s.e.m.) of CNV area per impact in each experimental group (five impacts per eye in each group; data expressed in μm2). Statistical analysis: Kruskal–Wallis test: P<0.0001. Mann–Whitney U-test: ###P<0.001 versus untreated, ***P<0.0001 versus pVAX2.

Mid- and long-term efficacy of sFlt-1 variants produced by ciliary muscle ET

Sustained release of the two shortest sFlt-1 variants was evaluated by their therapeutic efficacy in laser-induced CNV performed 6, 12 and 24 weeks after ciliary muscle ET of the therapeutic and control plasmids. Clinical analyses were carried out 14 days thereafter, that is, 2, 3.5 and 6 months after ET.

Therapeutic efficacy was first evaluated by FA as done for the short-term analysis (that is, at 21 days, Figures 3a–e). Treatment with both medium- and small-sFlt-1 variants (medium and small, Figures 5c and e) led to a major reduction in area and intensity of the pathological leakage observed at 2 months compared with the pVAX2-treated eyes (Figure 5a). A similar clinical effect was observed 3.5 months (Figures 5g, i and k) and 6 months (Figures 5m, o and q) after ET of the therapeutic plasmids as compared with the pVAX2 control one. Statistical analyses of clinical scoring highlighted that vascular leakage was reduced by 70% and around 50%, respectively, 2 and 3.5 months after ET compared with eyes treated with the empty plasmid (Figures 5s and t). More interestingly, vascular leakage was still significantly reduced 6 months after ET in both treated groups, with a 50% decrease similar to that observed at 3.5 months (Figure 5u). Note that both variants exhibited the same clinical efficacy at each time point tested (Mann–Whitney U-test, P>0.05).

Figure 5

Mid- and long-term efficacy of ciliary muscle ET of sFlt-1 variants encoding plasmids on vascular leakage and CNV development. CNV was induced at different time points after eyes were pretreated by ciliary muscle ET of the pVAX2 plasmid backbone or of the plasmids encoding medium- and small-sFlt-1 variants. (ar) Representative images of fluorescein angiographies (FA) and infracyanine angiographies (ICG) carried out (af) 2 months (n=4 eyes per group), (gl) 3.5 months (n=6 eyes per group except for pVAX2 group n=4) and (mr) 6 months (n=6 eyes per group except for pVAX2 group n=4) after ET. (su) Grading of vascular leakage performed on fluorescein angiograms showing the mean angiographic score per impact in each experimental group at the three time points (eight impacts per eye in each group). Results are expressed in AU (mean+s.e.m.). (v–x) Quantification of neovascular areas on ICG angiograms, showing the mean (+s.e.m.) of CNV area per impact (in pixel2) in each experimental group at the three time points. Statistical analysis: Kruskal–Wallis test: P<0.0001. Mann–Whitney U-test: ***P<0.0001 versus pVAX2.

To study the extent of neovascular membrane in vivo independently of their diffusion activity, infracyanine was injected together with fluorescein so that infracyanine angiography (ICG) angiographies could be performed at the same time. Careful observations of ICG images highlighted that the neovascular membranes filled quite entirely the lesion scar in pVAX2-treated eyes (Figures 5b, h and n) whereas they were clearly smaller than the scars in medium- (Figures 5d, j and p) and small-sFlt-1- (Figures 5f, l and r) treated eyes at the three time points. Measurements performed on ICG images highlighted that at 2 months CNV area was significantly reduced by 40% in each treated group compared with controls (Figure 5v), without any statistical difference between the two sFlt-1 variants. This effect was maintained at later time points, with decreases of CNV estimated around 30% both at 3 months (Figure 5w) and 6 months (Figure 5x) after ET of each therapeutic plasmid. The efficacy of both variants was not statistically different at each time point.

Sustained decrease of intraocular VEGF levels over months

As VEGF is known to have a key role in the development of CNV, but also is a target of sFlt-1, the biological effects of sFlt-1 variants produced by the transfected ciliary muscle were followed up over months by measuring VEGF concentration in ocular media at each time point using enzyme-linked immunosorbent assay (ELISA) (Figure 6). At 23 days, intraocular VEGF was detected at significantly lower levels in the three groups treated with sFlt-1 variants compared with untreated control eyes (315±25 pg ml−1) and pVAX2-treated control eyes (341±36 pg ml−1), whereas no difference could be detectable between both control groups (Figure 6a). At this time point, levels of VEGF were similarly decreased by 55–60% with the full-length and medium-sFlt-1 (139±11 and 157±14 pg ml−1, respectively) while the small-sFlt-1 induced a significantly more important decrease of about 77% (80±12 pg ml−1).

Figure 6

Quantification of intraocular levels of VEGF. Levels of VEGF (in pg ml−1, mean+s.e.m.) in ocular media were measured by ELISA at the peak of CNV, CNV being induced at different time points after ciliary muscle ET of the plasmids encoding sFlt-1 variants (sFlt-1, medium-sFlt-1 or small-sFlt-1). (a) At 23 days, intraocular VEGF levels in the three sFlt-1 variant-treated eyes (n=8 eyes per group) were compared with that of untreated control eyes (n=4) and eyes treated with pVAX2 plasmid backbone (n=8). Statistical analysis: Kruskal–Wallis test: P<0.0001. Mann–Whitney U-test: ##P<0.01 versus untreated, ***P<0.001 versus pVAX2. Analyses at (b) 2 months (n=4 eyes per group), (c) 3.5 months and (d) 6 months (n=6 eyes per group except for the pVAX2 group n=4) after ET were performed only in small- and medium-sFlt-1 variant-treated eyes and eyes treated with pVAX2. Statistical analysis: Kruskal–Wallis test: P<0.05. Mann–Whitney U-test: *P<0.05 and **P<0.01 versus pVAX2.

At 2 months (Figure 6b), a significant reduction of VEGF levels by almost 40% was still measured both in the medium- and small-sFlt-1-treated eyes. Interestingly, the drop of VEGF levels measured in medium-sFlt-1-treated eyes remains stable between 2 and 6 months (38%, 36% and 38% at 2, 3.5 and 6 months, respectively) showing that the biological effect was maintained over months (Figures 6b–d). The ability of small-sFlt-1 to reduce intraocular VEGF levels was relatively close to that of the medium variant at 2 and 3.5 months (Figures 6b and c), whereas some differences could be detectable between the two at 23 days and 6 months (Figures 6a and d).

Transcriptional regulation of pro-angiogenic mediators in RPE/choroid complexes

As RPE and choroid are directly involved in the development on CNV, these tissues were collected at 2 months to quantify the transcriptional expression of the two main known pro-angiogenic mediators (VEGF and PGF) as well as their receptors (VEGFR1 and VEGFR2) using quantitative reverse transciptase PCR (RT-PCR). As shown on Figure 7a, treatment with medium-sFlt-1 (0.53±0.06) and small-sFlt-1 (0.48±0.04) resulted in a 50% reduction in VEGF messenger RNA (mRNA) levels in RPE/choroid complexes compared with control eyes treated with the empty plasmid pVAX2 (1.00±0.03). No statistical difference in the expression of PGF, VEGFR1 and VEGFR2 could be measured between control and treated eyes at the mRNA level (Figures 7b–d; Kruskal–Wallis test, P>0.05).

Figure 7

mRNA expression of pro-angiogenic mediators in retinal pigment epithelium/choroid complexes. Levels of VEGF (a), PGF (b), VEGFR1 (c) and VEGFR2 (d) mRNA in retinal pigment epithelium/choroid complexes were measured by quantitative real-time PCR 2 months after ciliary muscle ET. Results, normalized with the mRNA levels of the actine housekeeping gene, were expressed in arbitrary units (mean+s.e.m.). Comparison was made between eyes treated with medium- and small-sFlt-1-encoding plasmids, and eyes treated with the corresponding empty plasmid pVAX2 (n=4 eyes per group). Kruskal–Wallis test: P<0.05. Mann–Whitney U-test: *P<0.05 versus pVAX2.

Characterization of the chemokines/cytokines profiles in ocular media

To evaluate the effect of the treatment on the production of pro- and anti-inflammatory mediators, chemokines and cytokines were measured in ocular media 23 days after treatment using multiplex assay.

Among the 17 molecules tested, nine of them (MCP-1, MIP-1α, interferon-γ, interleukin (IL)-4, IL-5, IL-6, IL-12, IL-13 and IL-17) were below detectable levels both in control and in treated eyes. IL-1β was detected at levels of about 40 pg ml−1 (range: 15–80 pg ml−1) in sFlt-1 variant-treated eyes whereas it was undetectable in untreated and pVAX2-treated eyes, showing that an increase in IL-1β levels seemed to be induced by sFlt-1 overproduction. Among mediators that were detectable in all groups, no statistical difference was seen for GRO-KC, RANTES and tumour necrosis factor (TNF)-α. IL-10 levels were increased significantly in the full-length sFlt-1-treated eyes. IL-2 was detected at levels that were around two-fold higher in eyes treated with plasmids encoding sFlt-1 variants compared with those treated with the corresponding empty plasmid or those left untreated (P<0.05). The most significant effect was seen on VEGF for which a reduction of about 60% could be measured in eyes belonging to the sFlt-1, medium-sFlt-1 and small-sFlt-1 compared with pVAX2 control eyes (P<0.05), confirming the results obtained at the same time point using a rat VEGF-specific ELISA assay (see Figure 6a).


During the last decade, neutralization of VEGF or VEGF and PGF using intravitreous injections of neutralizing proteins has taken a major place in the treatment of exudative AMD, diabetic macular edema and macular edema secondary to vein occlusion.33 Indeed, VEGF has been identified as a central mediator in the development of retinal neovascularization and macular edema during diabetic retinopathy.34 Its exact role in choroidal angiogenesis during AMD remains less understood as enhanced expression of VEGF in photoreceptors or RPE cells does not induce CNV.35, 36, 37 However, the role of VEGF in CNV has been supported by the clinical efficacy of anti-VEGF drugs on leakage and subsequent retinal edema.38 Increased levels of PGF have been measured in the vitreous of diabetic patients and shown to contribute to the pathogenesis of diabetic retinopathy.39 Its role in CNV is suggested by the fact that PGF is produced by hypoxic RPE cells, but there is no clinical evidence that PGF is directly involved in CNV.40, 41 To date, no clear differential efficiency or safety parameters have emerged from clinical studies to define which isoforms of VEGF or PGF should be neutralized. However, no study has been designed specifically to answer this question and an optimal strategy would include both VEGF and PGF neutralization.

The soluble form of Flt-1 (sFlt-1, 100 kDa), which consists of the extracellular domains of Flt-1, efficiently traps each of its ligands with different affinities (the affinity for VEGF-A is 5-fold to >100-fold higher than for PGF).23 In our study, we have cloned two truncated variants of rat sFlt-1: the medium length variant has a molecular weight of about 55–60 kDa and was already described,25 the small-sFlt-1 has an apparent molecular weight of 45 kDa and has not been previously described. Both variants contain the first three Ig-like domains of the transmembrane Flt-1 that are involved in PGF and VEGF binding. The differential VEGF/PGF trapping efficiency of each variant has not been evaluated in vitro but we have compared their clinical efficacy in a rat model of CNV and showed no significant difference in the prevention of CNV formation and VEGF neutralization between the two variants as compared with the full-length sFlt-1. The biological and clinical efficacy of the small variant suggests that the first three Ig-like domains of sFlt-1 are sufficient to inhibit CNV formation. Additional studies are needed to better characterize the potency of each variant and to optimize treatment efficacy.

One of the shortcomings of the currently available treatments is the need for repeated injections (monthly or bimonthly). Thus, one major unmet immediate need is to reduce the frequency of intravitreous injections of anti-VEGF drugs, to avoid peak and valley effects with potential recurrence of leakage and associated vision loss, and reduce the need for frequent patient follow-up. Several viral and non-viral gene transfer strategies are in preclinical or clinical evaluation to inhibit CNV. However, few of them were able to achieve a long-term efficacy and some of them are associated with safety concerns. Up to now, the dendrimer peptide-mediated delivery of an anti-VEGF oligonucleotide was the only non-viral gene therapy that inhibited laser-induced CNV for up to 4 months in rats.42 Regarding viral gene therapy, long-term suppression of CNV (up to 16 months) was achieved with intraocular injection of AAV-sFlt-1 in mice and monkeys.30 Recently, intravitreal AAV2-sFLT01 was used in mouse and monkey laser-induced CNV models.43 In the mouse model, 84% of the laser burns were associated with CNV in AAV2 control vector-treated eyes. In AAV2-sFLT01-treated eyes, 39% of burns were associated with CNV, which is about a 54% reduction. A phase I dose-ranging trial of intravitreous injection of AAV2-sFLT01 is ongoing in patients with advanced neovascular AMD (registration no. NCT01024998; http://clinicaltrials.gov/). In the present study, we have shown, using non-viral gene therapy, that two truncated variants of sFlt-1 efficiently inhibited CNV for up to 6 months after a single electroporation. The treatment with these variants not only led to a major decrease of CNV associated with laser burn (more than 70% reduction with medium length variant) as observed by FA but also in a reduction of the CNV growth as shown by RPE/choroid flatmounts and ICG angiography.

An important unanswered question for clinical application is to define how much inhibition of neovessel growth and/or leakage is required to maintain a healthy retina, without inducing choroidal atrophy. Interestingly, in human eyes treated by repeated anti-VEGF injections during 2 years, the size of CNV did not change as compared with baseline.44 Despite the fact that the CNV areas did not decrease, patients significantly gained vision, and the retina anatomy was maintained because leakage was significantly reduced by anti-VEGF treatments. Furthermore, other clinical trials have confirmed that reduction of leakage is the main aim of the treatment in order to avoid photoreceptor loss and subsequent vision alteration, and that indirect leakage signs, observed on OCT, have become the common standard method to decide for re-treatment. Therefore, there is no evidence that suppression of CNV would be superior in term of retinal function as compared with leakage suppression and subsequent sub- and/or intraretinal fluid reduction. On the other hand, excess of VEGF suppression can lead to RPE/choroid atrophy.45, 46 In this laser-induced CNV model, we could not evaluate the integrity of photoreceptors and other retinal integrity owing to the fact that the argon laser itself induces retinal damages; however, in previous experiments and other animal models, we have shown that ciliary muscle electroporation is safe for the retina.47

We have developed ciliary muscle ET as a biofactory platform for the sustained intraocular production of therapeutic proteins. The concept is to transduce ciliary muscle cells using minimally invasive microinjection with an electroporation device. Once transduced, the smooth muscle cells secrete the therapeutic proteins into the ocular media, and owing to their location, secretion is preferentially achieved into the vitreous, as previously demonstrated.19 Long-term expression of TNF-α soluble receptors up to 9 months was previously shown48 and this ability of sustained production is confirmed in our present experiment for up to 6 months.

Not only was clinical efficacy maintained up to 6 months after electroporation of the two variants, but it was also correlated with VEGF neutralization in the ocular media. This reduction of VEGF levels in ocular media can be explained by two main mechanisms: the reduction of free VEGF owing to the formation of sFlt-1–VEGF complexes and/or the decrease of VEGF production supported by the reduction of VEGF mRNA expression detected using RT-PCR. Our results, showing a 6-month therapeutic effect, would change the treatment regimen and be a significant improvement compared with the currently available anti-VEGF therapies.

The mechanisms of action of anti-VEGF agents are not fully understood and few studies have fully evaluated cytokines and growth factor expression changes in the neuroretina and RPE cells after short- or long-term anti-VEGF administration. In a recent paper, the effect of acute bevacizumab treatment on cytokine expression was measured in the retina of rats with central retinal vein occlusion. Anti-VEGF not only controlled VEGF-A expression but also IL-1β upregulation, suggesting that beside neutralization of VEGF, bevacizumab act on retinal gene expression.49 In our experiments, 2 months after sFlt-1 transfection, VEGF mRNA expression was reduced in RPE/choroid complexes of rats with CNV but no modification was observed on VEGF receptor expression. Interestingly, only full-length sFlt-1 induced IL-10 upregulation in the ocular media while TNF-α did not change. In human placenta, it was shown that exogenous sFlt-1 also induced IL-10 production but associated with increased TNF-α.50 Sustained production of sFlt-1 in the ocular media also induced the expression of other cytokines, such as IL-2 and IL-1β, suggesting that besides neutralizing VEGF, sFlt-1 may influence the Th1/Th2 balance. IL-2 is a known anti-tumor and anti-angiogenic cytokine, used in anti-cancer therapy.51, 52 It is expressed in activated brain microglia53 and more importantly it has been shown that IL-2 levels are decreased in aqueous humor of patients with recurrent CNV.54 One can hypothesize that macrophages/microglia activation state is modulated by VEGF as Flt-1 is expressed in macrophage and brain microglia.1, 55 IL-1β was also found increased in sFlt-1-treated eyes. We can suspect that IL-1β intervenes as a counter regulation to VEGF as it has been shown that a crosstalk between Il-1β and VEGF intervenes in the control of the blood–retinal barrier.56 Interestingly, the full picture of cytokine modification induced by anti-VEGF treatment has not been studied so far and could be more complex than a simple reduction of VEGF as, in the retina, VEGF controls the activation of immunoreactive cells, such as microglia.

In conclusion, this is the first study showing that non-viral gene therapy after single treatment can be efficient in preventing CNV for up to 6 months. Ciliary muscle ET offers many advantages: it is minimally invasive, can be realized using a disposable device at low cost, it is safe and treatment can be repeated as needed. Ongoing non-human primate studies are evaluating devices for human use.

Materials and methods

Plasmid DNA constructs

pVAX2, a plasmid backbone carrying the immediate early cytomegalovirus promoter (CMV IE),57 was used as a negative control in all experiments and used as an expression plasmid to subclone the complementary DNA sequences of three different variants of the rat soluble VEGF receptor type I (sFlt-1) (small-sFlt-1, medium-sFlt-1 and sFlt-1, respectively) corresponding to different forms of the extracellular part of this receptor (Figure 1b). The cloning process of the three constructs is described in detail as follows.

Total mRNAs were extracted from rat brain using TriZol reagent (Invitrogen, Cergy Pontoise, France) according to the manufacturer’s instructions. Ten microgram of total mRNAs were subjected to RT with a 20 base pair (bp) polyT DNA primer and the Superscript II enzyme (Invitrogen) in a 20 μl reaction volume. Then, 2 μl of this reaction were subjected to a PCR amplification reaction using the rat sFlt-1-specific primers, sense: 5′-IndexTermCGGGATCCGCCACCATGGTCAGCTGCTGGGACAC-3′ and antisense 5′-IndexTermCGGAATTCTTAGGGGAAGGCCTTCACTTTC-3′ derived from the Genbank sequence N° AF157595, with the Accuzyme polymerase enzyme (Bioline, Abcys, Paris, France). The resulting 1114 bp PCR product was digested with BamHI. This fragment was then ligated into a BamHI–EcoRV-digested pVAX2 fragment to give the pVAX2–small-sFlt-1 construct encoding the first three immunoglobulin domains of the VEGFR1 receptor (protein named ‘small-sFlt-1’) as described by Wiesmann et al.58

The second plasmid, encoding the ‘medium-sFlt-1’ variant, was constructed as follows: 2 μl of the RT reaction described above were subjected to a PCR amplification reaction using the rat sFlt-1-specific primers, sense: 5′-IndexTermGAAAGTGAAGGCCTTCCCCT-3′ and antisense: 5′-IndexTermCGGAATTCTTAAGACTTTTCGTAGATCTGAGGT-3′. One microlitre of the resulting 242 bp PCR product was ligated to 1 μl of the previous 1114 bp PCR product described above for 2 h using T4 DNA ligase (Takara Shuzo, Kyoto, Japan). Then, 1 μl of the ligation mixture was subjected to PCR amplification using the rat sFlt-1-specific primers, sense: 5′-IndexTermCGGGATCCGCCACCATGGTCAGCTGCTGGGACAC-3′ and antisense: 5′-IndexTermCGGAATTCTTAAGACTTTTCGTAGATCTGAGGT-3′. The resulting 1327 bp product was then digested with BamHI and ligated into a BamHI–EcoRV-digested pVAX2 fragment. The resulting plasmid was digested with XmnI and re-ligated to remove the redundant XmnI site to yield the pVAX2–medium-sFlt-1 construct encoding the first four immunoglobulin domains of the VEGFR1 receptor.

Finally, the full-length ‘sFlt-1’-encoding plasmid was constructed as follows: 2 μl of the RT reaction described above was subjected to a PCR amplification reaction using the rat sFlt-1-specific primers, sense: 5′-IndexTermACCTCAGATCTACGAAAAGTC-3′ and antisense: 5′-IndexTermTTAATGTTTGACATTACTTTGTG-3′. The resulting 784 bp product was digested with BglII and ligated into the previously described pVAX2–medium-sFlt-1 (containing the four IgG-like domains) digested with BglII and PmeI to yield the final construct pVAX2–sFlt-1.

All plasmids were checked by DNA sequencing. They were then amplified in Escherichia coli bacteria, prepared endotoxin-free (NucleoBond EF Kit; Macherey-Nagel, Hoerd, France) and diluted in endotoxin-free water containing 77 mM of NaCl as previously described.19 Plasmid DNA concentration was determined by spectroscopy measurement (absorbance at 260 nm).

Western blot analysis of sFlt-1 variants in vitro secretion

A spontaneously immortalized human RPE cell line displaying many differentiated properties typical of RPE cells in vivo (ARPE-19 (ATCC CRL- 2302)) was cultured in Dulbecco’s modified Eagle’s medium:F12 supplemented with 10% fetal calf serum and antibiotics (100 U ml−1penicillin; 100 mg ml−1 streptomycin; Invitrogen). The cultures were kept at 37 °C in a 5% CO2 humidified atmosphere. Cells (passage 27) were seeded in 12-well plates (1 ml per well) at a density of 60 000 cells per cm2 and grown for 24 h. Medium was removed and cells were then transfected with 4 μg of plasmid (pVAX2, pVAX2–small-sFlt-1, pVAX2–medium-sFlt-1or pVAX2–sFlt-1) using the calcium phosphate transfection method. After 24 h, cells were washed with phosphate-buffered saline (PBS) and medium was refreshed (400 μl per well). At 96 h, cell culture supernatants were collected, clarified by centrifugation and stored at −20 °C until being analyzed by western blotting as described below.

After the addition of LDS sample buffer (NuPAGE; Invitrogen) and heating for 5 min at 100 °C, equal volumes of culture supernatants were electrophoresed in a NuPAGE 3–8% Bis-Tris gel using MOPS SDS Running Buffer (Invitrogen). Proteins were then ET onto nitrocellulose membranes (Schleicher and Schuell BioScience, Dassel, Germany) using NuPAGE Transfer Buffer (Invitrogen). Membranes were blocked with 5% skimmed milk in PBS/0.5%Tween-20/0.5%Triton-X100 (PBS–TT) and incubated for 2 h at room temperature with a rabbit polyclonal anti-Flt-1 antibody (sc-9029; Santa Cruz Biotechnology; Tebu-Bio, Le Perray en Yvelines, France) diluted 1:1000 in PBS–TT containing 5% skimmed milk. Horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G (H+L) secondary antibody (Vector Laboratories, Clinisciences, Montrouge, France), diluted 1:15 000 in PBS–TT containing 1% skimmed milk, was used as a secondary antibody. Bands were visualized with the Amersham ECL Plus Western Blotting Detection System (GE Healthcare Europe, Orsay, France) used according to the manufacturer’s instructions.


Six- to eight-week-old female-pigmented Brown Norway rats weighing 220–250 g (Harlan Laboratories B.V., The Netherlands) were used and handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. For all in vivo experiments, rats were anesthetized by intramuscular injection of ketamine (75 mg kg−1) (Virbac, Magny-en-Vexin, France) and largactil (0.5 mg kg−1) (Sanofi-aventis, Paris, France). Unless stated otherwise, animals were killed with a lethal intraperitoneal dose of pentobarbital (80 mg kg−1) (Sanofi-aventis) at the end of the experiments.

In vivo ET to rat ciliary muscle

Thirty micrograms of plasmid (pVAX2, pVAX2–small-sFlt-1, pVAX2–medium-sFlt-1 or pVAX2–sFlt-1), in a total volume of 10 μl, were injected into the ciliary mucle using a 30-gauge disposable needle transsclerally posterior to the limbus as already described.19 For electric pulse delivery, a specially designed iridium/platinum electrode was introduced into the preformed intrasscleral tunnel and a semi-annular iridium/platinum anode electrode was placed on the limbus, facing the active cathode. ET was performed immediately after injection by applying eight electrical pulses (15 V, 10 ms in duration, 10 ms interval) generated by a 830 BTX electropulsator (Genetronics, San Diego, CA, USA) as previously described in the adult rat eye.19 All eyes were treated at the temporal side of the eyeball on day 0 and laser photocoagulation was performed at different time points thereafter.

Experimentally induced CNV

CNV was induced by laser photocoagulation at different time points after ciliary muscle ET (see ‘Experimental design’ section for details). Pupils were dilated by instillation of Mydriaticum (Tropicamide, Thea Laboratory, Clermont-Ferrand, France). Five or eight burns of 532-nm diode laser photocoagulation (50 μm spot size, 100 ms duration, 175 mW power; Argon Laser 532 nm, Viridis, Quantel Medical, Clermont-Ferrand, France) were performed in both eyes of each animal. The laser spots were applied around the optic nerve using a slit lamp delivery system (BQ-9000; Haag-Streit AG, Koeniz, Switzerland) and a hand-held cover slide as a contact lens. The reactive blebs observed at the retinal surface after laser delivery were considered to be evidence of the appropriate focusing and as an indicator of the effective rupture of Bruch’s membrane.

Experimental design

Three sets of experiments were carried out on a total of 58 Brown Norway rats (that is, 116 eyes).

The two first independent experiments, evaluating the short-term effects of the treatment, comprised 72 eyes randomized into five experimental groups. The first group received no injection and no ET (untreated eyes, n=10). The second group, used as an additional control, received an injection of the control empty plasmid pVAX2 in the ciliary muscle followed by ET (n=14). In the three last groups, plasmid DNA encoding for each sFlt-1 variant (small-sFlt-1, medium-sFlt-1 and sFlt-1; n=16 eyes per group) were injected and electrotransfered into the ciliary muscle. CNV was induced by laser burns 7 days after treatment and FA was carried out 14 days thereafter, that is, 21 days after ciliary muscle ET. Eyes from the first set of experiments were used for CNV immunolabeling and VEGF quantification in ocular media (aqueous and vitreous humors) whereas those from the second one were used for multiplex assay on ocular media (aqueous and vitreous humors).

In the third set of experiments, 44 eyes were used to evaluate mid- and long-term effects of the treatment. They were treated on the same day by ET of pVAX2 backbone (n=12), pVAX2–medium-sFlt-1 (n=16) and pVAX2–small-sFlt-1 (n=16). CNV was induced by laser photocoagulation 45 (n=4 eyes per group), 90 and 165 days (n=4 eyes for the pVAX2 group; n=6 eyes for each sFlt-1 variant-treated group) after treatment. Fluorescein/infracyanine angiographies were performed 14 days later, that is, at 2, 3.5 and 6 months. Ocular media were collected at each time point for VEGF quantification by ELISA, and RPE/choroid/sclera were collected only at 2 months for RT-PCR analyses.

In vivo FA and ICG

FA and ICG were performed on anesthetized rats 14 days after CNV induction at the time points stated below. Pupils were dilated with Mydriaticum (Tropicamide, Thea Laboratory) eye drops. Both analyses were conducted using a SPECTRALIS High-resolution, spectral domain digital imaging system (Heidelberg Retina Angiograph II, Heidelberg Engineering, Inc., Dossenheim, Germany), which uses a confocal scanning laser ophthalmoscope. To adapt the system for the small size of the rat eye, an additional customized lens was added to the system.

FA was performed at 21 days (first and second sets of experiments) but also at 2, 3.5 and 6 months (third set of experiments). Fluorescein (0.1 ml of 10% fluorescein in saline, SERB Laboratories, Paris, France) was administered intravenously. Early- and late-phase angiograms were recorded, respectively, 1–3 and 6–8 min after fluorochrome injection. For each laser-induced lesion, fluorescein leakage was graded qualitatively by evaluating the increase in size/intensity of dye leakage between the early and late phase. Angiographic scores were established by two masked observers according to the following scale ranging from 0 to 3: grade 0 indicated no leakage; grade 1 indicated a slight hyperfluorescence with no leakage nor increase in size and intensity; grade 2 indicated a moderate leakage (hyperfluorescence increasing in intensity but not in size); and grade 3 indicated a severe leakage (hyperfluorescence increasing in size and intensity).

ICG was performed only at 2, 3.5 and 6 months (third set of experiments) to visualize CNV in vivo. At these time points, infracyanine (0.2 ml of infracyanine 25 mg per 10 ml, SERB Laboratories) was injected intravenously together with fluorescein. Two-dimensional images of laser scars were captured with a digital video camera coupled to a computer system. All digital images were taken under the same conditions, resolved at 1024 × 768 pixels and converted to Bitmap Image Files (.bmp). CNV areas (in pixels2) were measured in a masked fashion by outlining the margins of the hyperfluorescence area for each laser-induced scar using the National Institutes of Health (NIH) ImageJ software (Bethesda, MD, USA). To reduce measurement bias, the mean value obtained from three different images was calculated for each impact analyzed.

Choroidal neovessels immunolabeling and RPE/choroid flatmounts preparation

At 23 days, intravenous injection of fluorescein isothiocyanate-conjugated lycopersicon esculentum tomato lectin (Sigma-Aldrich, St-Quentin Fallavier, France) was performed on anesthetized rats (first set of experiments). Five minutes later, rats were killed by carbon dioxide inhalation and freshly enucleated eyes were fixed in 4% paraformaldehyde for 15 min at room temperature for flatmount preparation. They were carefully dissected by transverse sectioning at 1 mm from the limbus. Anterior segments were discarded and the neuroretinas were carefully separated from the remaining RPE/choroid/sclera complexes. After eight relaxing incisions had been performed to allow the flattening of the specimens, RPE/choroid/sclera complexes were mounted in Fluoromount-G (Southern Biotechnology, Birmingham, AL, USA).

Flatmounts were examined with a confocal microscope (Zeiss LSM 710, Le Pecq, France) using a Cy3 filter (excitation 550 nm; emission 570 nm). Images of the laser scars were captured with a digital video camera coupled to a computer system. Photographs were taken using the same exposure times and contrast settings. CNV areas (in μm2) were determined by fluorescein isothiocyanate–lectin fluorescence and measured by outlining the margins of the labeled area on flatmounts images using the National Institutes of Health (NIH) ImageJ software.

Quantitative real-time PCR

At 2 months (rats from the third set of experiments), RPE/choroid/sclera complexes were carefully dissected from enucleated eyes, as described above, snap-frozen in liquid nitrogen and stored at 80 °C until use. Total RNA was isolated from tissues using RNeasy Plus Mini Kit (Qiagen, Courtaboeuf, France) according to the manufacturer’s instructions. After treatment with DNase I (Qiagen), first-strand complementary DNA was synthesized from total mRNA using random primers (Invitrogen) and SuperScript II reverse transcriptase (Invitrogen) as recommanded. Transcript levels of VEGF, PGF, VEGFR1 and VEGFR2 were analyzed by quantitative real-time PCR performed in 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with fluorescent DNA-binding dye (SYBR Green; Invitrogen) detection. The sequences of forward and reverse rat-specific primers (Sigma-Aldrich, Lyon, France) were: 5′-IndexTermACGAAAGCGCAAGAAATCCC-3′ and 5′-IndexTermTTAACTCAAGCTGCCTCGCC-3′ for VEGF; 5′-IndexTermCATGGACTTTGACCACTGC-3′ and 5′-IndexTermCAAGAGAATCTGGCTTGGC-3′for PGF; 5′-IndexTermCGACACTCTTTTGGCTCCTTCTAAC-3′ and 5′-IndexTermTGACAGGTAGTCCGTCTTTACTTCG-3′ for VEGFR1; and 5′-IndexTermTCTCGTACGGACCGTTAAGC-3′ and 5′-IndexTermCTCATCCAAGGGCAGTTCAT-3′ for VEGFR2. The β-actin gene (forward primer: 5′-IndexTermAAGTCCCTCACCCTCCCAAAAG-3′; reverse primer: 5′-IndexTermAAGCAATGCTGTCACCTTCCC-3′) was used as the reference gene. According to Pfaffl technique,59 serial dilutions were used for calibration and the slope was used to calculate gene amplification efficiency. Relative quantification of gene expression was calculated using Pfaffl formula as a ratio of the treatment group to the pVAX2 control group, followed by the ratio of the target gene to the reference gene.

VEGF, chemokines and cytokines quantification in ocular media

Eyes were removed, rinsed in PBS and dried. Once opened under an operating microscope, aqueous humor and vitreous were collected and pooled for each eye. Intraocular media were immediately centrifuged and the cell-free fractions were stored at −20 °C until use.

To determine VEGF concentration, ocular media were collected at 23 days (first set of experiments) but also at 2, 3.5 and 6 months (third set of experiments). Intraocular rat VEGF levels were quantified using a commercially available ELISA kit for rat VEGF (Duoset; R&D Systems, Lille, France) according to the manufacturer’s instructions (detection threshold around 50 pg ml−1).

For chemokine/cytokine analysis, ocular media were collected at 23 days (second set of experiments: n=8 eyes per group except for the untreated and the pVAX2 groups n=6), diluted to obtain a final volume of 25 μl and subjected to multiplex bead analysis. For each sample, 17 analytes were quantified simultaneously using the rat cytokine/chemokine-17plex kit (Milliplex Map Kit; Millipore, Saint-Quentin-en-Yvelines, France) according to the manufacturer’s instructions, as follows: chemokines MCP-1/CCL2, MIP-1α/CCL3, RANTES/CCL5 and GRO/KC; interleukins IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12, IL-13, IL-17 and IL-18; other cytokines interferon-γ, TNF-α and growth factor VEGF. The assay was performed in a 96-well filter plate, and standard curves for each cytokine were generated with a rat cytokine standard provided in the kit. All incubation steps were performed under medium orbital agitation in the dark to protect the beads from light. Data acquisition and analysis was performed with the manager software version 4.1 (Bio-plex, Bio-Rad, Hercules, CA, USA) with four or five logistic parameters for standard curves. Detection thresholds for all the analytes were estimated to be 1 to 10 pg ml−1.

Statistical analysis

For each experiment, the number of eyes or impacts analyzed per experimental condition was written in the legend of the figures. For numerical data, results were expressed as means±s.e.m. Statistical analyses were carried out using GraphPad Prism5 for Windows (GraphPad Software Inc., San Diego, CA, USA). Nonparametric Mann–Whitney, and Kruskal–Wallis tests were used to compare continuous data when appropriate. P-values of 0.05 or less were considered statistically significant.


  1. 1

    Dhoot DS, Kaiser PK . Ranibizumab for age-related macular degeneration. Expert Opin Biol Ther 2012; 12: 371–381.

    CAS  Article  Google Scholar 

  2. 2

    Browning DJ, Kaiser PK, Rosenfeld PJ, Stewart MW . Aflibercept for age-related macular degeneration: a game-changer or quiet addition? Am J Ophthalmol 2012; 154: 222–226.

    CAS  Article  Google Scholar 

  3. 3

    Ohr M, Kaiser PK . Intravitreal aflibercept injection for neovascular (wet) age-related macular degeneration. Expert Opin Pharmacother 2012; 13: 585–591.

    CAS  Article  Google Scholar 

  4. 4

    Muether PS, Hermann MM, Viebahn U, Kirchhof B, Fauser S . Vascular endothelial growth factor in patients with exudative age-related macular degeneration treated with ranibizumab. Ophthalmology 2012; 119: 2082–2086.

    Article  Google Scholar 

  5. 5

    Campochiaro PA . Anti-vascular endothelial growth factor treatment for retinal vein occlusions. Ophthalmologica 2012; 227 (Suppl 1): 30–35.

    CAS  Article  Google Scholar 

  6. 6

    Ho AC, Scott IU, Kim SJ, Brown GC, Brown MM, Ip MS et al. Anti-vascular endothelial growth factor pharmacotherapy for diabetic macular edema: a report by the american academy of ophthalmology. Ophthalmology 2012; 119: 2179–2188.

    Article  Google Scholar 

  7. 7

    Zechmeister-Koss I, Huic M . Vascular endothelial growth factor inhibitors (anti-VEGF) in the management of diabetic macular oedema: a systematic review. Br J Ophthalmol 2012; 96: 167–178.

    Article  Google Scholar 

  8. 8

    El Sanharawi M, Kowalczuk L, Touchard E, Omri S, de Kozak Y, Behar-Cohen F . Protein delivery for retinal diseases: from basic considerations to clinical applications. Prog Retin Eye Res 2010; 29: 443–465.

    CAS  Article  Google Scholar 

  9. 9

    Cideciyan AV . Leber congenital amaurosis due to RPE65 mutations and its treatment with gene therapy. Prog Retin Eye Res 2010; 29: 398–427.

    CAS  Article  Google Scholar 

  10. 10

    Hufnagel RB, Ahmed ZM, Correa ZM, Sisk RA . Gene therapy for Leber congenital amaurosis: advances and future directions. Graefes Arch Clin Exp Ophthalmol 2012; 250: 1117–1128.

    CAS  Article  Google Scholar 

  11. 11

    Smith AJ, Bainbridge JW, Ali RR . Gene supplementation therapy for recessive forms of inherited retinal dystrophies. Gene Ther 2012; 19: 154–161.

    CAS  Article  Google Scholar 

  12. 12

    Bainbridge JW, Mistry A, De Alwis M, Paleolog E, Baker A, Thrasher AJ et al. Inhibition of retinal neovascularisation by gene transfer of soluble VEGF receptor sFlt-1. Gene Ther 2002; 9: 320–326.

    CAS  Article  Google Scholar 

  13. 13

    Ideno J, Mizukami H, Kakehashi A, Saito Y, Okada T, Urabe M et al. Prevention of diabetic retinopathy by intraocular soluble flt-1 gene transfer in a spontaneously diabetic rat model. Int J Mol Med 2007; 19: 75–79.

    CAS  PubMed  PubMed Central  Google Scholar 

  14. 14

    Mao Y, Kiss S, Boyer JL, Hackett NR, Qiu J, Carbone A et al. Persistent suppression of ocular neovascularization with intravitreal administration of AAVrh.10 coding for bevacizumab. Hum Gene Ther 2011; 22: 1525–1535.

    CAS  Article  Google Scholar 

  15. 15

    Pechan P, Rubin H, Lukason M, Ardinger J, DuFresne E, Hauswirth WW et al. Novel anti-VEGF chimeric molecules delivered by AAV vectors for inhibition of retinal neovascularization. Gene Ther 2009; 16: 10–16.

    CAS  Article  Google Scholar 

  16. 16

    Provost N, Le Meur G, Weber M, Mendes-Madeira A, Podevin G, Cherel Y et al. Biodistribution of rAAV vectors following intraocular administration: evidence for the presence and persistence of vector DNA in the optic nerve and in the brain. Mol Ther 2005; 11: 275–283.

    CAS  Article  Google Scholar 

  17. 17

    Stieger K, Schroeder J, Provost N, Mendes-Madeira A, Belbellaa B, Le Meur G et al. Detection of intact rAAV particles up to 6 years after successful gene transfer in the retina of dogs and primates. Mol Ther 2009; 17: 516–523.

    CAS  Article  Google Scholar 

  18. 18

    Tao W . Application of encapsulated cell technology for retinal degenerative diseases. Expert Opin Biol Ther 2006; 6: 717–726.

    CAS  Article  Google Scholar 

  19. 19

    Touchard E, Kowalczuk L, Bloquel C, Naud MC, Bigey P, Behar-Cohen F . The ciliary smooth muscle electrotransfer: basic principles and potential for sustained intraocular production of therapeutic proteins. J Gene Med 2010; 12: 904–919.

    CAS  Article  Google Scholar 

  20. 20

    de Vries C, Escobedo JA, Ueno H, Houck K, Ferrara N, Williams LT . The fms-like tyrosine kinase, a receptor for vascular endothelial growth factor. Science 1992; 255: 989–991.

    CAS  Article  Google Scholar 

  21. 21

    Mustonen T, Alitalo K . Endothelial receptor tyrosine kinases involved in angiogenesis. J Cell Biol 1995; 129: 895–898.

    CAS  Article  Google Scholar 

  22. 22

    He Y, Smith SK, Day KA, Clark DE, Licence DR, Charnock-Jones DS . Alternative splicing of vascular endothelial growth factor (VEGF)-R1 (FLT-1) pre-mRNA is important for the regulation of VEGF activity. Mol Endocrinol 1999; 13: 537–545.

    CAS  Article  Google Scholar 

  23. 23

    Kendall RL, Thomas KA . Inhibition of vascular endothelial cell growth factor activity by an endogenously encoded soluble receptor. Proc Natl Acad Sci USA 1993; 90: 10705–10709.

    CAS  Article  Google Scholar 

  24. 24

    Kendall RL, Wang G, Thomas KA . Identification of a natural soluble form of the vascular endothelial growth factor receptor, FLT-1, and its heterodimerization with KDR. Biochem Biophys Res Commun 1996; 226: 324–328.

    CAS  Article  Google Scholar 

  25. 25

    Barleon B, Totzke F, Herzog C, Blanke S, Kremmer E, Siemeister G et al. Mapping of the sites for ligand binding and receptor dimerization at the extracellular domain of the vascular endothelial growth factor receptor FLT-1. J Biol Chem 1997; 272: 10382–10388.

    CAS  Article  Google Scholar 

  26. 26

    Cunningham SA, Stephan CC, Arrate MP, Ayer KG, Brock TA . Identification of the extracellular domains of Flt-1 that mediate ligand interactions. Biochem Biophys Res Commun 1997; 231: 596–599.

    CAS  Article  Google Scholar 

  27. 27

    Basile DP, Fredrich K, Chelladurai B, Leonard EC, Parrish AR . Renal ischemia reperfusion inhibits VEGF expression and induces ADAMTS-1, a novel VEGF inhibitor. Am J Physiol Renal Physiol 2008; 294: F928–F936.

    CAS  Article  Google Scholar 

  28. 28

    Gehlbach P, Demetriades AM, Yamamoto S, Deering T, Xiao WH, Duh EJ et al. Periocular gene transfer of sFlt-1 suppresses ocular neovascularization and vascular endothelial growth factor-induced breakdown of the blood-retinal barrier. Hum Gene Ther 2003; 14: 129–141.

    CAS  Article  Google Scholar 

  29. 29

    Igarashi T, Miyake K, Masuda I, Takahashi H, Shimada T . Adeno-associated vector (type 8)-mediated expression of soluble Flt-1 efficiently inhibits neovascularization in a murine choroidal neovascularization model. Hum Gene Ther 2010; 21: 631–637.

    CAS  Article  Google Scholar 

  30. 30

    Lai CM, Shen WY, Brankov M, Lai YK, Barnett NL, Lee SY et al. Long-term evaluation of AAV-mediated sFlt-1 gene therapy for ocular neovascularization in mice and monkeys. Mol Ther 2005; 12: 659–668.

    CAS  Article  Google Scholar 

  31. 31

    Lai YK, Shen WY, Brankov M, Lai CM, Constable IJ, Rakoczy PE . Potential long-term inhibition of ocular neovascularisation by recombinant adeno-associated virus-mediated secretion gene therapy. Gene Ther 2002; 9: 804–813.

    CAS  Article  Google Scholar 

  32. 32

    Edelman JL, Castro MR . Quantitative image analysis of laser-induced choroidal neovascularization in rat. Exp Eye Res 2007; 71: 523–533.

    Article  Google Scholar 

  33. 33

    Kaiser PK . Strategies for inhibiting vascular endothelial growth factor. Retina 2009; 29: S15–S17.

    Article  Google Scholar 

  34. 34

    Antonetti DA, Klein R, Gardner TW . Diabetic retinopathy. N Engl J Med 2012; 366: 1227–1239.

    CAS  Article  Google Scholar 

  35. 35

    Ohno-Matsui K, Hirose A, Yamamoto S, Saikia J, Okamoto N, Gehlbach P et al. Inducible expression of vascular endothelial growth factor in adult mice causes severe proliferative retinopathy and retinal detachment. Am J Pathol 2002; 160: 711–719.

    CAS  Article  Google Scholar 

  36. 36

    Okamoto N, Tobe T, Hackett SF, Ozaki H, Vinores MA, LaRochelle W et al. Transgenic mce with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am J Pathol 1997; 151: 281–291.

    CAS  PubMed  PubMed Central  Google Scholar 

  37. 37

    Yamada E, Tobe T, Yamada H, Okamoto N, Zack DJ, Werb Z et al. TIMP-1 promotes VEGF-induced neovascularization in the retina. Histol Histopathol 2001; 16: 87–97.

    CAS  PubMed  Google Scholar 

  38. 38

    Kovach JL, Schwartz SG, Flynn HW Jr., Scott IU . Anti-VEGF treatment strategies for wet AMD. J Ophthalmol 2012; 2012: 786870.

    Article  Google Scholar 

  39. 39

    Kowalczuk L, Touchard E, Omri S, Jonet L, Klein C, Valamanes F et al. Placental growth factor contributes to micro-vascular abnormalization and blood-retinal barrier breakdown in diabetic retinopathy. PLoS One 2011; 6: e17462.

    CAS  Article  Google Scholar 

  40. 40

    Miyamoto N, de Kozak Y, Jeanny JC, Glotin A, Mascarelli F, Massin P et al. Placental growth factor-1 and epithelial haemato-retinal barrier breakdown: potential implication in the pathogenesis of diabetic retinopathy. Diabetologia 2007; 50: 461–470.

    CAS  Article  Google Scholar 

  41. 41

    Miyamoto N, de Kozak Y, Normand N, Courtois Y, Jeanny JC, Benezra D et al. PlGF-1 and VEGFR-1 pathway regulation of the external epithelial hemato-ocular barrier. A model for retinal edema. Ophthalmic Res 2008; 40: 203–207.

    CAS  Article  Google Scholar 

  42. 42

    Marano RJ, Toth I, Wimmer N, Brankov M, Rakoczy PE . Dendrimer delivery of an anti-VEGF oligonucleotide into the eye: a long-term study into inhibition of laser-induced CNV, distribution, uptake and toxicity. Gene Ther 2005; 12: 1544–1550.

    CAS  Article  Google Scholar 

  43. 43

    Lukason M, DuFresne E, Rubin H, Pechan P, Li Q, Kim I et al. Inhibition of choroidal neovascularization in a nonhuman primate model by intravitreal administration of an AAV2 vector expressing a novel anti-VEGF molecule. Mol Ther 2011; 19: 260–265.

    CAS  Article  Google Scholar 

  44. 44

    Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser PK, Chung CY et alMARINA Study Group. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006; 355: 1419–1431.

    CAS  Article  Google Scholar 

  45. 45

    Saint-Geniez M, Kurihara T, Sekiyama E, Maldonado AE, D’Amore PA . An essential role for RPE-derived soluble VEGF in the maintenance of the choriocapillaris. Proc Natl Acad Sci USA 2009; 106: 18751–18756.

    CAS  Article  Google Scholar 

  46. 46

    Lois N, McBain V, Abdelkader E, Scott NW, Kumari R . Retinal pigment epithelial atrophy in patients with exudative age-related macular degeneration undergoing anti-vascular endothelial growth factor therapy. Retina 2013; 33: 13–22.

    CAS  Article  Google Scholar 

  47. 47

    Touchard E, Heiduschka P, Berdugo M, Kowalczuk L, Bigey P, Chahory S et al. Non-viral gene therapy for GDNF production in RCS rat: the crucial role of the plasmid dose. Gene Ther 2012; 19: 886–898.

    CAS  Article  Google Scholar 

  48. 48

    Touchard E, Bloquel C, Bigey P, Kowalczuk L, Jonet L, Thillaye-Goldenberg B et al. Effects of ciliary muscle plasmid electrotransfer of TNF-alpha soluble receptor variants in experimental uveitis. Gene Ther 2009; 16: 862–873.

    CAS  Article  Google Scholar 

  49. 49

    Drechsler F, Koferl P, Hollborn M, Wiedemann P, Bringmann A, Kohen L et al. Effect of intravitreal anti-vascular endothelial growth factor treatment on the retinal gene expression in acute experimental central retinal vein occlusion. Ophthalmic Res 2012; 47: 157–162.

    CAS  Article  Google Scholar 

  50. 50

    Xu B, Thornton C, Tooher J, Hennessy A . Exogenous soluble VEGF receptor-1 (sFlt-1) regulates Th1/Th2 cytokine production from normal placental explants via intracellular calcium. Hypertens Pregnancy 2009; 28: 448–456.

    Article  Google Scholar 

  51. 51

    Rocha FG, Calvo FB, Chaves KC, Peron JP, Marques RF, de Borba TR et al. Endostatin- and interleukin-2-expressing retroviral bicistronic vector for gene therapy of metastatic renal cell carcinoma. J Gene Med 2011; 13: 148–157.

    CAS  Article  Google Scholar 

  52. 52

    Alì G, Boldrini L, Lucchi M, Picchi A, Dell’Omodarme M, Prati MC et al. Treatment with interleukin-2 in malignant pleural mesothelioma: immunological and angiogenetic assessment and prognostic impact. Br J Cancer 2009; 101: 1869–1875.

    Article  Google Scholar 

  53. 53

    Girard S, Larouche A, Kadhim H, Rola-Pleszczynski M, Gobeil F, Sébire G . Lipopolysaccharide and hypoxia/ischemia induced IL-2 expression by microglia in neonatal brain. Neuroreport 2008; 19: 997–1002.

    CAS  Article  Google Scholar 

  54. 54

    Roh MI, Kim HS, Song JH, Lim JB, Koh HJ, Kwon OW . Concentration of cytokines in the aqueous humor of patients with naive, recurrent and regressed CNV associated with amd after bevacizumab treatment. Retina 2009; 29: 523–529.

    Article  Google Scholar 

  55. 55

    Ryu JK, Cho T, Choi HB, Wang YT, McLarnon JG . Microglial VEGF receptor response is an integral chemotactic component in Alzheimer’s disease pathology. J Neurosci 2009; 29: 3–13.

    CAS  Article  Google Scholar 

  56. 56

    Vinores SA, Xiao WH, Zimmerman R, Whitcup SM, Wawrousek EF . Upregulation of vascular endothelial growth factor (VEGF) in the retinas of transgenic mice overexpressing interleukin-1beta (IL-1beta) in the lens and mice undergoing retinal degeneration. Histol Histopathol 2003; 18: 797–810.

    CAS  PubMed  Google Scholar 

  57. 57

    Khoury M, Bigey P, Louis-Plence P, Noel D, Rhinn H, Scherman D et al. A comparative study on intra-articular versus systemic gene electrotransfer in experimental arthritis. J Gene Med 2006; 8: 1027–1036.

    CAS  Article  Google Scholar 

  58. 58

    Wiesmann C, Fuh G, Christinger HW, Eigenbrot C, Wells JA, de Vos AM . Crystal structure at 1.7A resolution of VEGF in complex with domain 2 of the Flt-1 receptor. Cell 1997; 91: 695–704.

    CAS  Article  Google Scholar 

  59. 59

    Pfaffl MW . A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29: e45.

    CAS  Article  Google Scholar 

  60. 60

    Shibuya M, Yamaguchi S, Yamane A, Ikeda T, Tojo A, Matsushime H et al. Nucleotide sequence and expression of a novel human receptor-type tyrosine kinase gene (flt) closely related to the fms family. Oncogene 1990; 5: 519–524.

    CAS  Google Scholar 

Download references


This work was funded in part by Fondation pour la Recherche Médicale (FRM).

Author information



Corresponding author

Correspondence to M El Sanharawi.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

El Sanharawi, M., Touchard, E., Benard, R. et al. Long-term efficacy of ciliary muscle gene transfer of three sFlt-1 variants in a rat model of laser-induced choroidal neovascularization. Gene Ther 20, 1093–1103 (2013). https://doi.org/10.1038/gt.2013.36

Download citation


  • gene therapy
  • anti-VEGF
  • choroidal neovascularization

Further reading