Targeting of Interferon to Choroidal Neovascularization by Use of Dextran and Metal Coordination

Abstract

purpose. Bioactive proteins such as interferon (IFN) have been reported to be combined with water-soluble polymers, such as dextran, through metal coordination, without need for complicated procedures. In the current study, the targeting and inhibitory effects of IFN combined with dextran on experimental choroidal neovascularization (CNV) were studied in vivo.

methods. Interferon (IFN)β was conjugated to dextran, which has metal-chelating, diethylenetriaminepentaacetic acid (DTPA) residues. Based on metal coordination, conjugation of IFNβ with DTPA-dextran resulted from simply mixing both substances in an aqueous solution containing Zn²⁺. The effects of IFNβ on the proliferation of human umbilical vein endothelial cells (HUVECs) and bovine retinal pigment epithelial cells (BRPECs) were evaluated. To evaluate the activity loss of IFNβ by conjugation, the effect of the conjugate on HUVECs was compared with that of free IFNβ. Experimental CNV was induced by subretinal injection of gelatin microspheres containing basic fibroblast growth factor in rabbits. The rabbits with CNV were intravenously treated twice weekly with 7.5 million international units (MIU)/kg per day free IFNβ (for 4 weeks), with IFNβ-DTPA-dextran conjugate containing 7.5 (for 2 weeks) or 0.75 (for 4 weeks) MIU/kg per day IFNβ, or with saline. The effects of these substances were evaluated by fluorescein angiography and histology. To observe the accumulation of conjugate, the doses of IFNβ in CNV tissues were measured by enzyme-linked immunosorbent assay.

results. IFNβ inhibited the growth of HUVECs and enhanced the proliferation of BRPECs. The conjugate seemed to preserve approximately 44% of IFNβ activity. Although both doses of IFNβ-DTPA-dextran inhibited progression of CNV in rabbits, longer term administration of a lower dose of IFNβ-DTPA-dextran had a sustained inhibitory effect on progression of CNV (P < 0.05). Histologic studies revealed the inhibitory effect of IFNβ-DTPA-dextran on progression of CNV. This conjugate prolonged the plasma half-life of IFNβ and enabled IFNβ to accumulate in the CNV in rabbits.

conclusions. In this study, human IFNβ was successfully used to target CNV, an enhanced antiangiogenic effect was achieved by combining it with dextran, based on metal coordination. This targeted delivery of IFNβ may have potential as a treatment modality for CNV.

Age-related macular degeneration (AMD) is a major cause of blindness in patients more than 50 years of age. Severe visual loss is often caused by angiogenesis originating from the choriocapillaris. ¹ Laser photocoagulation, vitreous surgery, radiation, and photodynamic therapy have been used to treat AMD. ^{2 3 4 5 6 7 8 9 10 11 12 13 14 15 16} However, none of the current forms of therapy is effective against blindness from choroidal neovascularization (CNV). ^{2 3 4 5 6 7 8 9 10 11 12 13 14 15 16} These treatments may accelerate angiogenesis or injure normal retinal tissue ¹⁷ and therefore cannot be used to treat early-stage AMD in patients with good visual acuity. It is important to develop potent antiangiogenic drugs to arrest AMD in the incipient stage. 

Several antiangiogenic drugs, such as interferon (IFN)-α, thalidomide, and TNP-470, have been investigated. ^{18 19 20 21 22} Although in vivo studies using animal models of angiogenesis have demonstrated the therapeutic feasibility of these drugs, recently reported studies have shown that IFNα and thalidomide do not always sufficiently inhibit the development of CNV membranes, ^{23 24 25} primarily because the drugs generally do not have organ-specific affinity and their in vivo half-lives are too short. Moreover, systemic administration of these substances may result in serious side effects when sufficient drug concentrations reach the tissues with neovascularization, because angiogenesis occurs not only at the pathologic site but also during wound healing and tissue development. ^{20 24} This history of attempted development of pharmaceutical therapy indicates that future novel drugs also may not be efficacious. The development of drug delivery systems to facilitate therapeutic efficacy and to minimize side effects is desired as eagerly as the identification of a new potent antiangiogenic drug. 

The newly formed vasculature in tumor tissue has high substance permeability compared with that of normal tissues and incomplete lymphatic systems. These features enable macromolecules to accumulate and remain in the perivascular regions of solid tumors longer than in normal tissue, which is referred to as the enhanced permeability and retention (EPR) effect. ^{26 27 28 29} It has been demonstrated that conjugation of drugs with water-soluble polymers such as poly(vinyl alcohol), poly(ethylene glycol), and dextran prolongs the circulating life of these drugs and increases the accumulation of drugs in tumor mass, according to the pharmacokinetics of macromolecules. ^{30 31 32} This passive drug targeting is supposed not only to facilitate the treatment effect, but also to attenuate adverse effects, because substances combined with large molecules are nearly unable to access nontargeted tissues across capillaries with normal permeability, and the dose necessary for systemic administration decreases in association with the enhanced therapeutic effect. In ocular angiogenesis, choroidal neovascular membranes (CNVMs) have anatomic characteristics similar to tumor tissue, because the retinal tissue surrounding the CNVMs has prelymphatic systems, ³³ and CNVMs are stained by fluorescein in late-phase fluorescein angiography. Based on these concepts, we demonstrated that passive targeting of the antiangiogenic agent TNP-470 to experimental CNV through chemical conjugation with a water-soluble polymer significantly enhances its therapeutic potential for treating AMD. ³⁴  

However, in bioactive proteins, chemical conjugation causes inevitable loss of protein activity, because it involves complicated multistep procedures. ^{35 36} It is therefore of utmost importance to develop a new and simple method for conjugating proteins to carriers without using chemical coupling. We developed a new method of preparation of the drug–polymer conjugate based on metal coordination, which has been conventionally used in metal-chelating affinity chromatography to separate proteins and peptides. ^{37 38} We have already applied this conjugation method to tumor necrosis factor (TNF)-α and IFNα to demonstrate enhanced antitumor and antiviral effects. ^{39 40}  

In the present study, we attempted the targeted delivery of IFNβ, which is expected to exhibit stronger antiangiogenic effects than IFNα, to treat AMD. We investigated the in vitro effect of IFNβ on the proliferation of vascular endothelial cells and retinal pigment epithelial (RPE) cells and the inhibitory effect of IFNβ on the progression of experimental CNV in rabbits through simple mixing with dextran, involving diethylenetriaminepentaacetic acid (DTPA) residues. We also evaluated the retention effect of conjugate with IFNβ in CNV lesions. 

Materials and Methods

Chemicals

IFNβ and basic fibroblast growth factor (bFGF) were kindly supplied by Toray (Kanagawa, Japan) and Kaken Pharmaceutical (Osaka, Japan), respectively. Dextran with an average molecular weight of 200,000 was purchased from Nacalai Tesque (Kyoto, Japan); dimethylaminopyridine and zinc chloride (ZnCl₂) from Wako Pure Chemicals Industries (Osaka, Japan); and DTPA anhydride from Dojindo Laboratories (Kumamoto, Japan). All other chemicals were reagent-grade products obtained commercially. 

Preparation of DTPA-Introduced Dextran

For metal chelation, DTPA residues were introduced to a hydroxyl group of dextran, as described previously. ³⁹ Briefly, 662 mg DTPA and 22.6 mg dimethylaminopyridine were added to 500 mL dehydrated dimethyl sulfoxide containing 500 mg dextran. The solution mixture was agitated at 40°C for 16 hours, followed by dialysis against distilled water for 2 days and freeze drying to obtain dextran with DTPA residues introduced (DTPA-dextran). DTPA introduction was quantitated using a conductivity meter (model DS-12; Horiba, Kyoto, Japan). DTPA residues were estimated to be introduced into 15% of the repeated glucose residues of dextran. 

IFNβ Conjugation to DTPA-Dextran under Zn²⁺ Coordination

IFNβ was conjugated to DTPA-dextran (IFNβ-DTPA-dextran) in ZnCl₂ aqueous solution. Briefly, the dose of IFNβ required for one intravenous injection in a rabbit was diluted with distilled water containing DTPA-dextran to 1.35 mL final volume. To this resultant solution, 0.15 mL ZnCl₂ aqueous solution was added (ratio of IFNβ, DTPA-dextran, and ZnCl₂ = 1 million international units [MIU]:1.5 mg:0.126 mg). The mixture was gently agitated at 25°C for 1 hour to allow IFNβ to conjugate to the DTPA-dextran under Zn²⁺ coordination (Fig. 1) . Before intravenous administration, 1.5 mL of 1.8 N normal saline was added to this solution. For in vitro use, this solution was diluted with culture medium by 0.1% or less. 

Cell Culture

Human umbilical vein endothelial cells (HUVECs) were purchased from Kurabo (Okayama, Japan). HUVECs were grown in monolayer cultures in medium (HuMedia EG-2; Kurabo) containing 1% fetal bovine serum. Bovine RPE cells (BRPECs) were obtained from bovine eyes, as described previously. ³⁴ The cells of passages 3 to 5 were used for the experiments. HUVECs and BRPECs were maintained in 10-cm cell culture dishes. 

Cell Viability

To assess the original effect of IFNβ on cell viability, a tetrazolium-based colorimetric assay (XTT assay; Boehringer Mannheim, Tokyo, Japan) method was used according to the instructions of the manufacturer. ⁴¹ HUVECs and BRPECs were plated into 96-well cell culture plates (Iwaki Glass, Tokyo, Japan), and IFNβ (0.1 IU/mL to 1.0 MIU/mL) was added to the cultures the next day. After 5 days, 50 μL of XTT solution was added to the culture. After an additional 4-hour incubation, the absorbance at 450 nm was determined by spectrophotometry (model DU-64; Beckman Instruments, Tokyo, Japan). Each experiment was performed in quadruplicate and repeated three times. 

To evaluate the loss of activity of IFNβ by conjugation with dextran, HUVECs were incubated for 2 days with free IFNβ (0.1 MIU/mL), IFNβ-DTPA-dextran with the same concentration of IFNβ, the mixture of IFNβ and DTPA-dextran without ZnCl₂, or DTPA-dextran and ZnCl₂ without IFNβ. As described previously, cell viability was measured using the XTT assay. Each experiment was performed in quadruplicate and repeated four times. 

Animals and Anesthesia

Eighty-eight pigmented rabbits, weighing 1.9 to 2.8 kg, were used in this study. The animals were treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Only the right eye of each rabbit was used. The rabbits were anesthetized with intramuscular ketamine (5 mg/kg) and xylazine (2 mg/kg). Topical 1% tropicamide and 2.5% phenylephrine hydrochloride were instilled for mydriasis during surgery and fluorescein angiography and to observe the fundus. The rabbits were killed with an overdose of intravenous pentobarbital sodium. 

CNV Induction in Rabbits

Experimental CNV in rabbits was induced as previously reported. ⁴² In this model, CNV is induced by bFGF, a direct angiogenesis factor, to exclude inflammation as the predominant angiogenic stimulus. 

In 45 (70.3%) of the 64 eyes that received the bFGF-impregnated microspheres, CNV lesions developed with mild or moderate fluorescein leakage 3 weeks after induction. Because the fluorescein leakage in eyes persists for 4 to 6 weeks, ⁴² we used these eyes in the following in vivo studies. 

Efficacy of IFNβ-DTPA-Dextran on Experimental CNV

To determine the antiangiogenic efficacy of IFNβ-DTPA-dextran, rabbits were treated with saline (n = 9, 4 weeks), free IFNβ (single dose, 7.5 MIU/kg; n = 5, 4 weeks), or IFNβ-DTPA-dextran with 0.75 (n = 8, 4 weeks) and 7.5 (n = 10, 2 weeks) MIU/kg of IFNβ. Each substance, beginning 3 weeks after implantation of the gelatin microspheres, was administered intravenously twice weekly. Fluorescein angiography was repeated weekly. To evaluate the treatments’ effects, all angiographic images taken 7 to 9 minutes after injection to document the degree of late fluorescein leakage were converted to digital images, and the area of fluorescein leakage was quantified in a masked fashion by computer (Image 1.52; NIH Image is provided in the public domain by the National Institutes of Health, Bethesda, MD and is available at http://rsb.info.nih.gov/nih-image/). To prevent variability during image processing, the density of the choroidal background on the images from the same eye was harmonized without changing the contrast. Each measurement was scored as follows: 0, no or faint fluorescein leakage (less than choroidal background in density; 1, mild leakage (as low as choroidal background in density); 2, focal intense leakage (20% or less than the area with microspheres); 3, intense leakage of moderate size (20%–50%); and 4, intense leakage of large size (50% or more). 

Histologic Studies

The eyes treated with saline (n = 2) and IFNβ-DTPA-dextran (n = 2) were studied histologically. These rabbits were killed 7 weeks after implantation of the microspheres. The eyes were enucleated and immediately placed in 2.5% glutaraldehyde and 2% formaldehyde in 0.1 M phosphate-buffered saline (PBS; pH 7.4) for 15 minutes. The cornea, lens, and vitreous were carefully removed and placed in the fixative for an additional 24 hours at 4°C. The area in which the hydrogel had been implanted was dissected from each eye under the dissecting microscope. Each specimen was embedded in paraffin. Sections 2- to 3-μm thick were prepared and stained with hematoxylin-eosin for light microscopy. 

Accumulation of IFNβ-DTPA-Dextran in CNV Lesions

To investigate the plasma half-life, blood samples were obtained and centrifuged 1, 2, and 3 days after intravenous administration of 7.5 MIU/kg IFNβ (n = 12) or IFNβ-DTPA-dextran, with the same amount of IFNβ (n = 12). The plasma samples were diluted with PBS containing 1 M NH₄Cl and analyzed by enzyme-linked immunosorbent assay (ELISA). The dose of IFNβ-DTPA-dextran in the CNV lesions was also measured by ELISA. Other rabbits with CNV were treated with 7.5 MIU/kg IFNβ (n= 4) or IFNβ-DTPA-dextran with the same amount of IFNβ (n = 5). Twenty-four hours later, these rabbits were killed with an overdose of intravenous pentobarbital sodium, and the eyes were immediately enucleated. The chorioretinal tissue (5 mm in diameter including CNV lesions) was punched out after freezing on acetone ice and homogenized with PBS containing 1 M NH₄Cl and 0.1% 3-[(3-cholamidopropyl) dimethylammonio]-1-propane sulfonate. The tissue homogenates were centrifuged at 3000 rpm, and the supernatant was collected. The remaining tissue was freeze dried and weighed. The concentration of IFNβ in this supernatant was then measured by ELISA. 

Statistical Analysis

All data are mean ± SEM. Student’s t-test was used to analyze the in vitro data and the recovery of IFNβ from CNV lesions. The Wilcoxon signed rank test was used to compare the results of fluorescein angiography at different time points in each group and the Mann-Whitney test to compare the results at each time point with the control. Differences were considered statistically significant at P < 0.05. 

Results

In Vitro Effect of IFNβ

RPE cells may play a partial role in vascular nonperfusion and the reconstruction of the outer BRB, as well as in the progression of CNVMs. ^{43 44} Proliferation of RPE cells may promote the regression of CNVMs by covering the neovascular vessels. We investigated the effect of IFNβ on the proliferation of HUVECs and BRPECs (Fig. 2) . At concentrations of more than 10 IU/mL, IFNβ inhibited the proliferation of HUVECs. Cytostatic inhibition was observed, with complete inhibition at concentrations higher than 0.1 MIU/mL. However, IFNβ enhanced the proliferation of BRPECs in a dose-dependent manner. If these properties of IFNβ were reflected in the in vivo relation between choroidal endothelial cells and RPE cells, they would be advantageous in the treatment of AMD. 

At the concentration of 0.1 MIU/mL, whereas IFNβ reduced the cell viability of HUVECs to 68% of the control with cytostatic inhibition, the conjugate retained approximately 44% of this activity of IFNβ. The solution of DTPA-dextran and ZnCl₂ without IFNβ exhibited no inhibitory effect on HUVECs (Fig. 3) . 

Suppressive Effect of IFNβ-DTPA-Dextran on Experimental CNV

A CNV rabbit model was prepared by local injection of gelatin microspheres containing bFGF into the subretinal space. We demonstrated that controlled-release gelatin microspheres enable bFGF to induce neovascularization from the choriocapillaris and that the pathologic course is similar to that of human AMD. ⁴² In this model, CNV was observed 3 weeks after the injection and developed for approximately 4 weeks. We designed the administration regimen to treat CNV in the incipient stage. The rabbits with CNV received intravenous injections of IFNβ-DTPA-dextran containing 0.75 or 7.5 MIU/kg IFNβ for 4 or 2 weeks, respectively, or free IFNβ (7.5 MIU/kg) for 4 weeks. All substances were administered intravenously twice weekly. Fluorescein angiography revealed that IFNβ-DTPA-dextran significantly inhibited progression of CNV, whereas free IFNβ had no significant effect (Figs. 4 5) . When the conjugate containing a higher IFNβ dose (7.5 MIU/kg) was intravenously injected twice weekly for 2 weeks, a significant inhibitory effect on progression of CNV was observed 5 weeks after the induction of CNV, but CNV recurred with the end of treatment (Fig. 5C) . In contrast, 4-week administration even with the lower IFNβ dose (0.75 MIU/kg) exhibited a prolonged inhibitory effect (P < 0.01; Fig. 5D ). These findings demonstrate that the conjugate of IFNβ and DTPA-dextran based on metal coordination successfully inhibited CNV in rabbits compared with free IFNβ, in a manner dependent on the injection frequency and dose. 

Histologic Observation

After injection of the gelatin microspheres, fibrovascular membranes, which are mainly composed of RPE-like cells, an extracellular matrix, and new vessels, were observed beneath the retina. Numerous capillaries had newly formed from the choriocapillaris. After treatment with the IFNβ-DTPA-dextran conjugate, the subretinal membrane consisted mainly of a large number of RPE-like cells with pseudoacinar structures, and the newly formed capillaries were hardly visible in the membrane (Fig. 6) . 

Bodily Distribution of IFNβ-DTPA-Dextran and Retention Effect in CNV Lesions

To investigate the retention’s effect, an experimental CNV model in rabbits was used. After the CNV lesions were examined by fluorescein angiography 3 weeks after the subretinal injection, rabbits were killed 1 day after intravenous injection of free IFNβ or IFNβ-DTPA-dextran. The eyes were enucleated and the chorioretinal tissues were dissected and homogenized. The concentration of IFNβ in the supernatant of this homogenate was measured by ELISA. More IFNβ was detectable in rabbits injected with IFNβ-DTPA-dextran than in those injected with free IFNβ (Fig. 7) . The plasma half-life of IFNβ in the terminal-elimination phase increased to 16.9 hours with the DTPA-dextran conjugation with metal coordination, whereas that of free IFNβ was 4.5 hours. Therefore, the presence of IFNβ-DTPA-dextran was prolonged in the CNV lesions because of the prolonged plasma half-life and the EPR effect, compared with free IFNβ. 

Discussion

CNV is a major cause of visual loss, and a new therapy for CNV in AMD has been eagerly sought. Recently, antiangiogenic drugs, such as IFNα and thalidomide, have been used to suppress the progression of CNVMs. However, systemic IFNα is not effective in patients with exudative AMD. ²⁴ Thalidomide also did not prevent CNV’s recurrence in patients with punctuate inner choroidopathy, and difficulties in its continuous use have been reported, including such side effects as drowsiness, constipation, and peripheral neuropathy. ^{23 25}  

These treatment failures may occur partly because there is no organ-specific affinity and because of the short plasma half-lives of low-molecular-weight drugs. A new treatment for AMD may be developed without overcoming these limitations. To the best of our knowledge, the application of drug-targeting technology to treat ocular angiogenesis has been reported only by our group. ³⁴ Our previous studies have demonstrated that various water-soluble polymers with a molecular weight of approximately 200,000 accumulates in tumor tissue in significantly higher amounts and for longer periods than larger or smaller polymers. ^{28 29} Based on these results and the hypothesis that CNV may have surroundings similar to that of neovascular vessels in tumors, we have demonstrated that the chemical conjugation of TNP-470 with poly(vinyl alcohol) with a molecular weight of 220,000 accumulates in the CNV lesion and inhibits progression. ³⁴ Therefore, we used dextran, with an average molecular weight of 200,000, in the present study. Dextran has been used as a plasma expander to prevent blood platelet aggregation. ⁴⁵ Recently, another study showed that dextran with molecular weights of 70,000 and higher tends to remain in the perivascular area longer than several weeks, ⁴⁶ which makes it possible to enhance the targetability of antiangiogenic drugs to new vessels. 

In this study, we demonstrated that a simple mixing procedure enabled bioactive protein to conjugate to macromolecules, based on metal coordination. In our previous studies, the coexistence of a polysaccharide with DTPA residues and metal ions, such as Zn²⁺ and Cu²⁺, shifted the peak of IFN or TNF shown in gel filtration chromatography to a shorter retention time, which suggested that these bioactive proteins could be conjugated to DTPA-polysaccharide, based on metal coordination. ^{39 40} This metal ion-coordinated protein-DTPA-polysaccharide conjugate was structurally stable in PBS. In metal-chelating affinity chromatography, IFNβ that was bound to the gel through Zn²⁺ coordination was released slightly by application of serum (data not shown). Therefore, it is possible that the conjugate may release the protein gradually by displacing or chelating agents in the plasma and target tissue. However, bodily distribution and in vivo studies have revealed a prolonged plasma half-life and targeting of bioactive proteins and their biological activities only by mixing proteins and DTPA-polysaccharide in an aqueous solution containing metal ions. Therefore, it is likely that the metal coordination bond enables proteins to conjugate to a polysaccharide strongly enough to carry them to the target tissue without dissociation in the body. 

The present study demonstrated that the conjugation of IFNβ with dextran could prolong the plasma half-life of IFNβ, enhance the accumulation of IFNβ in CNV lesions, and inhibit progression of CNV in rabbits. When IFN was chemically conjugated by the cyanuric chloride method to another polysaccharide (pullulan) with a high inherent affinity for the liver, IFN activity decreased by 9% or less. ^{35 36} If hexamethylendiamine was introduced between pullulan and IFN molecules as a spacer group to increase the distance between these large molecules, IFN activity was recovered to 18.8%, ³⁶ which suggests that this loss of IFN activity results not only from complicated procedures but also from steric hindrance by conjugated polymer. Thus, conjugation may reduce the biological activity of IFNβ because of steric hindrance by dextran. In vitro, IFNβ-DTPA-dextran had approximately 56% of the apparent activity loss of IFNβ through conjugation. As described previously, when the liberation of IFNβ from the conjugate in culture medium is taken into consideration, the activity loss by conjugation may be greater. However, even if this activity loss is considered, the conjugate was still highly effective in suppressing the progression of CNV. The in vivo efficacy of the conjugate in suppression of CNV is ascribed to the prolonged plasma half-life and enhanced targetability of IFNβ to CNV. IFNβ that is retained longer and possibly slowly liberated in the CNV lesion may easily interact with its receptor on the cell surface. Because the efflux of free IFNβ from the target tissue is relatively faster than that of the conjugate with a higher molecular weight, the conjugate may continue to release IFNβ slowly, different from its activity in vitro. Another possibility is that the internalization and metabolization of IFNβ after binding to the receptor may be suppressed because of its binding to dextran. The IFNβ molecule may be released from the receptor without internalization and again bind to another receptor. Enhanced concentrations and reuse effects in CNV lesions probably enabled IFNβ to be remarkably effective in vivo. 

IFNβ promoted the progression of RPE cells, whereas it inhibited the proliferation of HUVECs. Although further in vitro studies should be undertaken in human RPE cells and choroidal endothelial cells, these effects of IFNβ may be more helpful in causing the regression of CNV, because the reconstruction of the outer BRB must have the proliferation of RPE cells and because RPE cells may have some effects on apoptosis in developed CNV. ^{43 44} Oral administration of 200 mg zinc sulfate daily for 24 months increased the mean serum zinc level by 37% but did not produce a significant change in red blood cell counts and hemoglobin levels or other side effects. ⁴⁷ Rebizak et al. ⁴⁸ investigated the feasibility of DTPA-introduced dextran to prolong the plasma half-life of DTPA and gadolinium, which has been used for magnetic resonance imaging. Thus, the content of Zn²⁺ and dextran with DTPA residues is expected to be low enough that no adverse effects occur. Although further animal studies should be performed to investigate the toxicology, immunology, dose, and frequency of use of conjugate, this novel passive targeting system for bioactive proteins may eliminate the problems encountered in treating patients with this serious ocular disease. 

 Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2000.

 Supported part by a 1999 Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.

 Submitted for publication July 11, 2001; revised October 25, 2001; accepted November 14, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Conjugation of IFNβ and DTPA-dextran through Zn²⁺ coordination.

Figure 2.

 

The effects of IFNβ on the proliferation of HUVECs (▪) and BRPECs (•). Arrow: average initial number of HUVECs. Data are means ± SEM. IFNβ inhibited proliferation of HUVECs, whereas it enhanced that of BRPECs. *P < 0.05 compared with the control.

Figure 3.

 

Retained activity of IFNβ in IFNβ-DTPA-dextran. Compared with the group with free IFNβ, IFNβ-DTPA-dextran preserved approximately 44% of IFNβ activity in HUVECs. DTPA-dextran caused no toxicity in HUVECs. Data are mean ± SEM. *P < 0.05 compared with the control; †P < 0.05 compared with the group with IFNβ-DTPA-dextran.

Figure 4.

 

Fluorescein angiograms from rabbits treated with (A, D) saline, (B, E) free IFNβ (7.5 MIU/kg, 4 weeks), and (C, F) IFNβ-DTPA-dextran conjugate (0.75 MIU/kg, 4 weeks). Late-phase angiograms 3 weeks after implantation of bFGF-impregnated gelatin microspheres (before treatment) showed actively leaking fluorescein with the grades 2 (A), 1 (B), and 1 (C). Late-phase angiograms 7 weeks after implantation suggest that IFNβ-DTPA-dextran suppressed progression of CNV, with the lesion remaining at grade 1 fluorescein leakage (F), compared with saline (D) and free-IFNβ (E), with lesions progressing to grade 4 leakage.

Figure 5.

 

Progression of CNV in (A) the control group and the groups treated with (B) free IFNβ (7.5 MIU/kg, 4 weeks) or (C) IFNβ-DTPA-dextran involving 7.5 or (D) 0.75 MIU/kg IFNβ (after 2 and 4 weeks, respectively). Each arrow indicates a time point of intravenous administration. A higher dose of IFNβ-DTPA-dextran (C) inhibited progression of CNV, but fluorescein leakage resumed after 2 weeks of treatment, whereas longer treatment, even with a lower dose of the conjugate (D), sustained the inhibitory effect for 4 weeks.* P < 0.05 compared with data at different time points. The significant changes indicate progression of CNV in all cases.† P < 0.05, §P < 0.01, compared with the control at each time point.

Figure 6.

 

Light micrographs of (A) a control eye and (B) an eye treated with IFNβ-DTPA-dextran 7 weeks after implantation of the microspheres. (A) The control eye had a massive subretinal neovascular membrane in the subretinal space. The membrane consisted of RPE-like cells, an extracellular matrix, and vascular formation (arrow). (B) The subretinal membrane consisted mainly of a large number of RPE-like cells with pseudoacinar structures. Few new vessels were observed in the membrane. Hematoxylin-eosin; scale bar, 100 μm.

Figure 7.

 

Targeting of IFNβ to CNV lesions. IFNβ (7.5 MIU/kg), with or without DTPA-dextran and ZnCl₂, was administered intravenously to rabbits with induced CNV. Recovery of IFNβ from CNV lesions was enhanced by the conjugation of IFNβ with DTPA-dextran. Data are means ± SEM. *P = 0.0099, Student’s t-test.