Technetium-99 Conjugated with Methylene Diphosphonate (⁹⁹Tc-MDP) Inhibits Experimental Choroidal Neovascularization In Vivo and VEGF-Induced Cell Migration and Tube Formation In Vitro

Abstract

Purpose.: To investigate the effects of ⁹⁹Tc-MDP, a decay product of ^99mTc-MDP, on the development of choroidal neovascularization (CNV), together with its underlying mechanisms.

Methods.: C57BL/6J mice were used to induce CNV by laser photocoagulation. ⁹⁹Tc-MDP at the doses of 0.5 × 10⁻¹, 1 × 10⁻¹, and 2 × 10⁻¹ μg/kg or the same volume of PBS was intraperitoneally injected daily after photocoagulation until the end of the experiment. Seven days after laser injury, mice were perfused with fluorescein-labeled dextran, and areas of CNV were measured. Numbers of infiltrating macrophages, protein levels of VEGF, and inflammation-related molecules including intercellular adhesion molecule (ICAM)-1, tumor necrosis factor (TNF)-α, and matrix metalloproteinases (MMPs) in the RPE-choroid complex were detected 3 days after laser photocoagulation. Effects of ⁹⁹Tc-MDP on VEGF-induced endothelial cell migration and tube formation were also studied. Toxicity of ⁹⁹Tc-MDP was evaluated in vivo and in vitro.

Results.: Areas of CNV were significantly suppressed by ⁹⁹Tc-MDP treatment without toxicity to the retina compared with PBS treatment in a dose-dependent manner: ⁹⁹Tc-MDP treatment of 0.5 × 10⁻¹ μg/kg (5698.60 ± 1037.70 μm²), 1 × 10⁻¹ μg/kg (3678.34 ± 1328.18 μm²), and 2 × 10⁻¹ μg/kg (2365.78 ± 923.80 μm²) suppressed the development of CNV by 36.12%, 58.76%, and 73.48%, respectively, compared with that in the PBS treatment group (8920.36 ± 1097.29 μm²; P < 0.001). ⁹⁹Tc-MDP treatment led to significant inhibition of macrophages infiltrating to CNV together with downregulated protein expressions of VEGF, ICAM-1, TNF-α, and MMP-2. ⁹⁹Tc-MDP also showed an inhibitive effect on cell proliferation and VEGF-induced migration and capillary-like tube formation of endothelial cells.

Conclusions.: Anti-inflammatory treatment with ⁹⁹Tc-MDP has therapeutic potential for CNV-related diseases.

Choroidal neovascularization (CNV) is the key pathogenesis in exudative form of age-related macular degeneration (AMD), the leading cause of blindness among the elderly in developed countries. ¹ CNV is immature new blood vessel that penetrates the Bruch's membrane and extends into the subretinal or subretinal pigment epithelium space. It often leads to loss of central vision by exudation and hemorrhage, retinal pigment epithelial (RPE) or retinal detachment, photoreceptor degeneration, and the formation of disciform scars. ²  

The exact mechanisms of CNV formation have not been fully elucidated. Recent studies demonstrate an important role of inflammatory mechanisms in CNV of AMD patients ^3,4 and in experimental CNV models. ^5,6 The inflammatory processes include macrophage infiltration and a variety of inflammatory cytokine networks. This concept of inflammation-mediated neovascularization is further supported by studies that generalized macrophage depletion reduced the size and leakage of laser-induced CNV associated with decreased macrophage infiltration and VEGF protein expression. ^7,8 Inflammation-related molecules, such as intercellular adhesion molecule (ICAM)-1, ^9,10 tumor necrosis factor (TNF)-α, ^11,12 and matrix metalloproteinases (MMPs), ¹³ express at a higher level than normal in CNV patients and laser-induced animal models. Since inflammation-related cells and molecules play an important role in the development of CNV, it is reasonable to hypothesize that the regulation of inflammation may be a promising therapeutic strategy to inhibit CNV. 

Technetium-99 conjugated with methylene diphosphonate (Yunke; ⁹⁹Tc-MDP) is a kind of anti-inflammatory drug patented in China (patent no. ZL94113006.1) that has been demonstrated to be safe and effective in the clinical treatment of rheumatoid arthritis (RA) and ankylosing spondylitis (AS) in China. ^{14 –17} The chemical structure of ⁹⁹Tc-MDP is shown in Figure 1. ⁹⁹Tc-MDP is actually the decay product of ^99mTc-MDP (^99mTc-methylene diphosphonate, a US Food and Drug Administration–approved radioactive agent widely used for bone scintigraphy). Unlike ^99mTc-MDP, ⁹⁹Tc-MDP has little of the intensity of radioactivity demonstrated to be harmless to the human body. ⁹⁹Tc-MDP treatment can lead to decreased serum levels of TNF-α and IL-1β, which are commonly accepted as proinflammatory factors ¹⁸ in the rat model of adjuvant arthritis ¹⁹ and in patients with RA. ¹⁵ Because IL-1 is the chemoattractant for leukocytes, ⁹⁹Tc-MDP is supposed to have the ability to inhibit leukocyte migration. ¹⁵ At the same time ⁹⁹Tc-MDP promotes the scavenging of free radicals and activates superoxide dismutase; therefore, it can prevent the tissue damage caused by free radicals. ¹⁴ Considering the similarity of pathogenesis between AMD and systemic inflammation-related diseases, we used the anti-inflammatory property of ⁹⁹Tc-MDP to investigate its effects on the development of CNV. 

Figure 1.

 

Chemical structure of ⁹⁹Tc-MDP.

To our knowledge, no data have been presented concerning the effects of ⁹⁹Tc-MDP on CNV development. Here we report the first evidence of the in vivo and in vitro inhibitive effects of ⁹⁹Tc-MDP on the development of CNV together with the underlying cellular and molecular anti-inflammatory mechanisms. 

Materials and Methods

Preparation of ⁹⁹Tc-MDP

Yunke (Chengdu Yunke Pharmaceutical, Sichuan, China) contains two preparations. Preparation A is a 5-mL vial of colorless transparent solution containing 0.05 μg ⁹⁹Tc. Preparation B is a lyophilized white powder that contains 5 mg methylene diphosphonic acid (MDP) and 0.5 mg stannous chloride. ⁹⁹Tc-MDP should be prepared before use by injecting preparation A into preparation B. According to the instruction manual, both preparations should be stored at 2°C to 8°C in a dark environment. 

Animals

Male C57BL/6J mice aged 6 to 8 weeks (SPF level) were purchased from the Animal Experiment Center of Sun Yat-sen University. All animals used in this study were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Induction of CNV and Intraperitoneal Injection of ⁹⁹Tc-MDP

Mice were anesthetized by intramuscular injection of 100 mg/kg body weight (BW) ketamine hydrochloride, and pupils were dilated with 0.5% tropicamide-phenylephrine ophthalmic solution (Mydrin; Santen Pharmaceutical, Osaka, Japan). Four laser burns (50-μm spot size, 0.05-second duration, 250-mW power) were placed at a distance of 2 to 3 PD around the optic nerve with a frequency-doubled Nd:YAG laser (Novus Spectra; Lumenis Inc, Santa Clara, CA) using the slit lamp delivery system (Haag-Streit, Mason, OH) and a coverslip as a contact lens, as previously described. ²⁰ Production of a cavitation bubble at the laser point is thought to be the sign of disruption of Bruch's membrane, which leads to the formation of CNV. ²⁰ Therefore, only burns with bubbles were included in this study. The treatment groups were intraperitoneally injected daily until the end of experiment with ⁹⁹Tc-MDP at the dose of 0.5 × 10⁻¹ μg/kg BW, 1 × 10⁻¹ μg/kg BW, or 2 × 10⁻¹ μg/kg BW immediately after laser photocoagulation and the control group was intraperitoneally injected daily with the same volume of PBS. 

Preparation and Quantitative Analysis of Choroidal Flatmounts

One week after laser photocoagulation, mice were anesthetized and perfused with 1 mL PBS containing 50 mg/mL fluorescein-labeled dextran (2 × 10⁶ average MW; Sigma, St. Louis, MO) through the left ventricle, as previously described. ²¹ The eyes were enucleated and fixed in 4% paraformaldehyde for 1 hour at room temperature. The anterior segment and neurosensory retina were carefully removed; after that, four to six radial relaxing incisions were made from the edge to the equator to allow the remaining RPE-choroid-sclera complex to be flattened on a glass slide with the RPE facing up. Then a coverglass was placed on the choroidal flatmount with mounting medium (Vectashield; Vector Laboratories, Burlingame, CA). CNVs on the flatmounts were examined by fluorescence microscopy (Axioplan 2 imaging; Zeiss, Thornwood, NY), and the CNV images were digitized. Areas of CNV that were associated with hyperfluorescence (n = 13 in each group) were then measured (Image-Pro Plus 6.0 software; Media Cybernetics, Silver Spring, MD). 

Immunofluorescence for Infiltrating Macrophages

Whole mounts of RPE-choroid-sclera complexes were obtained 3 days after photocoagulation (11–12 mice were used in each group). The complexes were incubated at 4°C overnight with a goat polyclonal antibody against mouse PECAM-1 (CD31; Santa Cruz Biotechnology, Santa Cruz, CA) and a rat polyclonal antibody against F4/80 (Serotec, Oxford, UK) followed by incubation with Avidin-Alexa 488–tagged and Avidin-Alexa 555–tagged secondary antibodies (Alexa 488 and Alexa 555; Invitrogen, Carlsbad, CA) for 2 hours at room temperature, as described previously. ²² After that, PECAM-1–stained CNV areas and F4/80 positive infiltrating macrophages were scanned with a scanning laser confocal microscope (LSM710; Zeiss) in 5-μm steps, and numbers of macrophages per laser lesion were calculated by image processing and analysis software (Imaris 7.0; Bitplane, Zurich, Switzerland). 

ELISA for VEGF and Inflammation-Related Molecules

Three days after laser photocoagulation, RPE-choroid complexes were harvested from the RPE-choroid-sclera flatmount under a dissecting microscope, using a slit knife with its edge completely dull. After that, two RPE-choroid complexes from the right and left eyes of the same mouse were placed together into 500 μL lysis buffer supplemented with protease inhibitors (Sigma), as previously described, ²³ and then were carefully homogenized with glass tissue homogenizer. Then the lysate was cleared of debris by centrifugation at 12,000 rpm for 10 minutes at 4°C. The levels of VEGF (n = 7 or 8 mice in each group), ICAM-1 (n = 12), and TNF-α (n = 12) in the supernatant were determined with the mouse VEGF, ICAM-1, and TNF-α ELISA kits (R&D Systems Inc., Minneapolis, MN) and normalized to total protein (determined by DC protein assay kit; Bio-Rad Laboratories) according to the instruction manual. Specially, the VEGF ELISA kit recognizes all the isoforms of VEGF. 

Gelatin Zymography for Activity of MMP-2 and MMP-9

Gelatin zymography was used to detect the protein of MMP-2 and MMP-9 expressed in RPE-choroid complexes 3 days after laser injury. Zymography was performed as described by Munaut et al. ²⁴ with some modifications. Briefly, 30 μL homogenate supernatant of the RPE-choroid complex (contained 30 μg total protein, determined by DC protein assay kit; Bio-Rad Laboratories) was added to each lane and electrophoresed in 10% sodium dodecyl sulfate (SDS)-polyacrylamide with 1 mg/mL gelatin (Sigma). The gels were soaked in wash buffer (Triton; Sigma), then stained overnight with Coomassie brilliant blue and destained with methanol and acetic acid. Zones of clearing in the gel indicated MMP activity, and the gray intensity value of the proteolytic zone of the gels compared with those of the PBS control group were calculated (ImageQuant software, version 5.2; Molecular Dynamics, Sunnyvale, CA). 

Real-Time PCR for mRNA Expression of MMP-2

Total RNA was isolated (Trizol; Invitrogen) from the RPE-choroid complexes 3 days after photocoagulation. RPE-choroid complexes from both eyes of a mouse were pooled for analysis as the sample for this mouse (five mice in each group). Yield and purity were determined photometrically. Complementary DNA synthesis was performed with a first-strand DNA synthesis kit (QuantiScript; Qiagen, Valencia, CA) and then was quantified by real-time PCR analysis (ABI7500; Applied Biosystems, Carlsbad, CA). Sequences of the PCR primers were as follows: (MMP-2) forward primer, 5′-GCCCCCATGAAGCCTTGT-3′; reverse primer, 5′-TTGTAGGAGGTGCCCTGGAA-3′; probe, 5′-FAM-TGGCAATGCAGATGGACAGCCC-TAMRA-3′; (β-actin) forward primer, 5′-GCGCGGCTACAGCTTCA-3′; reverse primer, 5′-TCTCCTTAATGTCACGCACGAT-3′; probe, 5′-FAM-CACCACGGCCGAGCGGGA-TAMRA-3′. The PCR mix was denatured at 93°C for 3 minutes, followed by 40 cycles of melting at 93°C for 15 seconds, annealing at 55°C for 32 seconds, and elongation at 72°C for 32 seconds. The data were normalized relative to β-actin mRNA levels. 

ERG for Toxicity of ⁹⁹Tc-MDP on Retina In Vivo

ERG recordings were performed both at baseline (before the first injection of ⁹⁹Tc-MDP) and 7 days after intraperitoneal injection daily with ⁹⁹Tc-MDP at the doses of 0.5 × 10⁻¹ μg/kg BW (n = 5), 1 × 10⁻¹ μg/kg BW (n = 5), 2 × 10⁻¹ μg/kg BW (n = 5), or the same volume of PBS (n = 5). The mice were kept in the dark environment for 12 hours and were prepared under dim red illumination, and then flash ERG recordings were performed in the right eye (RETIport; Roland Consult, Brandenburg, Germany), as previously described. ²⁵ Briefly, mice were anesthetized and were placed on a heating pad with a constant temperature of 39°C throughout the experiment. Pupils were dilated with 0.5% tropicamide-phenylephrine ophthalmic solution (Mydrin; Santen Pharmaceutical, Osaka, Japan), and corneas were anesthetized with a drop of 0.5% proparacaine hydrochloride (Alcon, Puurs, Belgium). Custom gold wire electrodes were placed on the corneas. Reference and ground electrodes were placed subcutaneously between the eyes and inserted into the tails, respectively. Eyes were stimulated with 3 cd · s/m² intensity levels of white light by a light-emitting diode photostimulator (Ganzfeld Q450; Roland Consult) at 10-μs duration and 9-second interval for each stimulus. Averages of 20 responses were calculated, and the a-wave and b-wave amplitudes and b-wave implicit times were recorded; a-wave and b-wave amplitudes and b-wave implicit times of the flash ERGs were compared between the pre- and post-⁹⁹Tc-MDP treatment groups. 

In Vitro Assays

Proliferation and Cytotoxicity of Cells.

Effects of ⁹⁹Tc-MDP on the proliferation of RF/6A cells (rhesus macaque choroid-retinal endothelial cell; purchased from the Institute of Biochemistry and Cell Biology, Shanghai, China) were estimated by bromodeoxyuridine (BrdU) assay. Cells were seeded into 96-well plates with a density of 1 × 10⁴ cells/well. Blank control (wells without cells) and background (wells with cells but without ⁹⁹Tc-MDP treatment) were used. One day after plating (cells were still nonconfluent and dividing at this time), the medium was removed and the cells were incubated with serum-free medium with or without ⁹⁹Tc-MDP (n = 6) (0.50 × 10⁻¹ μg/mL, 0.25 × 10⁻¹ μg/mL, or 0.125 × 10⁻¹ μg/mL) for 12 hours. BrdU label (20 μL; Sigma) was added to each well (except background wells) together with the ⁹⁹Tc-MDP treatment. Then cells were fixed and permeabilized and the DNA was denatured, followed by incubation with anti-BrdU antibody (Sigma) for 1 hour. The unbound antibody was washed away, and horseradish peroxidase-conjugated goat anti-mouse antibody (Boster, Wuhan, China) was added to each well and incubated for 30 minute. Finally, chromogenic substrate tetra-methylbenzidine (R&D Systems Inc.) was added, and the absorbance of the color was measured with a microplate reader at 450-nm wavelength. 

MTT assay was used to determine the cytotoxicity of ⁹⁹Tc-MDP on nondividing RF/6A cells. Cells at a density of 5 × 10³ per well were plated in 96-well plates and grown until confluence. After confluence, cells were incubated with ⁹⁹Tc-MDP solutions from 0.05, 0.025, and 0.0125 μg/mL or PBS (n = 8) for 12 hours. ⁹⁹Tc-MDP solutions were then drilled out, and 180 μL medium with 20 μL MTT (5 mg/mL) was added to each well; this was followed by incubation for 4 hours at 37°C. After that, all media were carefully removed. Then 150 μL dimethyl sulfoxide was added to each well to dissolve the formazan crystals. Ten minutes later, the plate was read at a wavelength of 490 nm. Cell viability was compared on the basis of the OD₄₉₀ value. The experiment was repeated three times. 

Cell Migration Assay.

RF/6A cell migration assay was performed using 50 μg/mL fibronectin (Biological Industries Ltd., Kibbutz Beit Haemek, Israel)–coated, 24-well permeable cell inserts (Boyden chamber, Transwell; Corning Costar, Corning, NY) with 8-μm pore size, as described previously with some modifications. ²⁶ Briefly, chemotaxis was induced by the addition of 800 μL Dulbecco's modified Eagle medium (DMEM) with 0.1% bovine serum albumin (BSA; Sigma), with or without 10 ng/mL VEGF (R&D Systems Inc.) to the lower compartment. Serum-free DMEM (200 μL) with 0.1% BSA of 1 × 10⁵ RF/6A cells per well were loaded onto the upper compartment with or without ⁹⁹Tc-MDP (0.50 × 10⁻¹ μg/mL, 0.25 × 10⁻¹ μg/mL, or 0.125 × 10⁻¹ μg/mL). The transwell units were then incubated at 37°C (95% air/5% CO₂) for 6 hours. The filters were fixed and stained with hematoxylin; nonmigrated cells (on top of the filters) were wiped with a cotton bud, and then the remaining filters with the migrated cells were mounted on a glass slide. Four randomly chosen fields were photographed per filter (× 20 objective), and the numbers of migrated cells were counted. The experiment was repeated three times. 

Tube Formation Assay.

For capillary-like tube formation assay, 50 μL basement membrane extract (BME; R&D Systems Inc.) was used to coat each well of a 24-well plate, followed by polymerization at 37°C for 30 minutes. RF/6A cells were cultured in DMEM with 1% fetal bovine serum (FBS) overnight. Trypsin-harvested cells (3.3 × 10⁵) suspended in 300 μL fresh 1% FBS assay medium in the absence or presence of 10 ng/mL VEGF (R&D Systems Inc.) with or without ⁹⁹Tc-MDP (0.50 × 10⁻¹ μg/mL, 0.25 × 10⁻¹ μg/mL, or 0.125 × 10⁻¹ μg/mL) were seeded onto BME-coated plates and incubated at 37°C for 6 hours. To evaluate the lengths of the capillary-like vessels on the BME, four photographs in different randomly selected fields around the center of each well were taken with a phase-contrast microscope (×10 objective). The lengths of the tube were measured (Image-Pro Plus 6.0 software; Media Cybernetics). The experiment was also repeated three times. 

Statistical Analysis

Data were expressed as mean ± SD and were tested for significance using one-way ANOVA. A confidence level of P < 0.05 was considered statistically significant. All analyses were performed statistical software (SAS 8.1; SAS Institute, Inc., Cary, NC). 

Results

Suppression of CNV in Mice Receiving ⁹⁹Tc-MDP

Seven days after laser photocoagulation, CNVs on RPE-choroid-sclera flat mounts examined by fluorescence microscopy showed that ⁹⁹Tc-MDP significantly suppressed the development of CNV in a dose-dependent manner (Fig. 2). ⁹⁹Tc-MDP treatment of 0.5 × 10⁻¹ μg/kg (5698.60 ± 1037.70 μm², P < 0.001; n = 13), 1 × 10⁻¹ μg/kg (3678.34 ± 1328.18 μm², P < 0.001; n = 13), and 2 × 10⁻¹ μg/kg (2365.78 ± 923.80 μm², P < 0.001; n = 13) could respectively suppress the CNV areas by 36.12%, 58.76%, and 73.48% compared with PBS treatment group (8920.36 ± 1097.29 μm²; n = 13). 

Figure 2.

 

Suppression of CNV in mice receiving ⁹⁹Tc-MDP. Seven days after laser photocoagulation, mice were perfused with 1 mL of 50 mg/mL fluorescein-labeled dextran through the left ventricle, and CNVs on RPE-choroid-sclera complex flat mounts were examined by fluorescence microscopy. (A–D) Representative images for the groups. (A) PBS control group. (B) ⁹⁹Tc-MDP treatment of 0.5 × 10⁻¹ μg/kg BW. (C) ⁹⁹Tc-MDP treatment of 1.0 × 10⁻¹ μg/kg BW. (D) ⁹⁹Tc-MDP treatment of 2.0 × 10⁻¹ μg/kg BW. (E) Quantitative assessment of CNV areas. Note that the development of CNV is significantly suppressed by the treatment with ⁹⁹Tc-MDP in a dose-dependent manner. ***P < 0.001 compared with the PBS control group; n = 13. Scale bars, 50 μm (A–D).

Inhibitive Effects of ⁹⁹Tc-MDP on Macrophage Infiltration

We quantitated the infiltration of macrophages in CNV with an immunofluorescence method. The numbers of macrophages per laser lesion were calculated to determine whether ⁹⁹Tc-MDP had an inhibitive effect on the macrophage infiltration. Immunofluorescence intensity for F4/80 tended to be lower in the ⁹⁹Tc-MDP–treated mice than in the PBS-treated mice (Fig. 3A), and ⁹⁹Tc-MDP had the ability to decrease the numbers of infiltrating macrophages in a dose-dependent manner (n = 11–12) (Fig. 3B). 

Figure 3.

 

Inhibitive effects of ⁹⁹Tc-MDP on macrophage infiltration together with downregulated protein levels of VEGF and inflammatory molecules. Three days after photocoagulation, the infiltration of macrophages in CNV was quantitated by immunofluorescence, and the protein levels of VEGF, ICAM-1, and TNF-α in the RPE-choroid complex were measured by ELISA. (A) F4/80-positive macrophages (top, green, white arrows) and PECAM-1–stained CNV areas (bottom, red) showed that immunofluorescence intensity for F4/80 tended to be lower in the ⁹⁹Tc-MDP–treated mice than in the PBS-treated mice. Scale bars, 50 μm. (B) ⁹⁹Tc-MDP inhibited the infiltration of macrophages into the CNV compared with the PBS-treated group by calculating the numbers of macrophages per laser lesion (*P < 0.05, **P < 0.01; n = 11–12). (C–E) The protein levels of VEGF (*P < 0.05, **P < 0.01; n = 7–8), ICAM-1 (*P < 0.05, **P < 0.01; n = 12), and TNF-α (**P < 0.01; n = 12) were significantly suppressed by the treatment of ⁹⁹Tc-MDP compared with the PBS control group. For VEGF, there was no statistical significance between the 1 × 10⁻¹ μg/kg BW group and the 2 × 10⁻¹ μg/kg BW group (P > 0.05) or between the 1 × 10⁻¹ μg/kg BW group and the 0.5 × 10⁻¹ μg/kg BW group (P > 0.05). For ICAM-1, there was no statistical significance between the 1 × 10⁻¹ μg/kg BW group and the 2 × 10⁻¹ μg/kg BW group (P > 0.05). For TNF-α, there was no statistical significance between any two of the ⁹⁹Tc-MDP–treated groups (P > 0.05).

Inhibition of VEGF and Inflammatory Molecules by the Treatment of ⁹⁹Tc-MDP

Macrophages invaded the site of laser injury within 1 day, and peak macrophage numbers paralleled the maximal amount of VEGF protein detected in the RPE-choroid complexes 3 days after laser photocoagulation. ⁷ Therefore, to determine whether the protein level of VEGF and inflammatory molecules had the parallel relationship with the numbers of infiltrating macrophages, we collected RPE-choroid complexes 3 days after laser photocoagulation and detected protein levels of VEGF, ICAM-1, and TNF-α in the RPE-choroid complex by using ELISA. The results showed that ⁹⁹Tc-MDP treatment significantly suppressed protein levels of VEGF (n = 7–8), ICAM-1 (n = 12), and TNF-α (n = 12) compared with the PBS treatment group (P < 0.05 for all) (Figs. 3C-E). However, for VEGF, there was no statistical significance between the 1 × 10⁻¹ μg/kg BW group and the 2 × 10⁻¹ μg/kg BW group (P > 0.05) or between the 1 × 10⁻¹ μg/kg BW group and the 0.5 × 10⁻¹ μg/kg BW group (P > 0.05). For ICAM-1, there was no statistical significance between the 1 × 10⁻¹ μg/kg BW group and the 2 × 10⁻¹ μg/kg BW group (P > 0.05). For TNF-α, although there was statistical significance between the ⁹⁹Tc-MDP–treated groups and the PBS-treated group (P < 0.01), there was no statistical significance between any two of the ⁹⁹Tc-MDP–treated groups (P > 0.05). 

Inhibition of MMP-2 Activity by Zymography

Gelatin zymography is a sensitive technique that detects activated gelatin-degrading metalloproteinases such as MMP-2 and MMP-9. ²⁷ Cultured supernatant of the human HT1080 cell was used as a molecular weight marker. The gray intensity value of the proteolytic zone of the gels in the ⁹⁹Tc-MDP–treated group was much lower than in the PBS control group for active MMP-2 (P < 0.01; n = 9) but not for pro-MMP-9 (P > 0.05; n = 9), as is shown in Figure 4. 

Figure 4.

 

Zymographic analysis of active MMP-2 and pro-MMP-9 in RPE-choroid complexes of laser-induced neovascularization. (A) Representative image of zymography for MMPS. Cultured supernatant of human HT1080 cell was used as a molecular weight marker for the identification of active MMP-2 and pro-MMP-9. (B) Relative gray intensity = (treatment group gray intensity value − background gray intensity value)/(marker gray intensity value − background gray intensity value). The relative gray intensity value of the proteolytic zone of the gels in the ⁹⁹Tc-MDP–treated group was significantly lower than that in the PBS control group for the active MMP-2 (**P < 0.01; n = 9) but not for pro-MMP-9 (P > 0.05; n = 9).

Inhibition of mRNA Expression of MMP-2 by Real-Time PCR

Real-Time PCR was used to determine MMP-2 mRNA expression in RPE-choroid complexes 3 days after photocoagulation. MMP-2 mRNA expression was increased after photocoagulation (Fig. 5), and ⁹⁹Tc-MDP inhibited MMP-2 mRNA expression in a dose-dependent manner compared with the PBS-treated group (P < 0.001; n = 5). 

Figure 5.

 

Real-time PCR analysis for the MMP-2 mRNA expression in RPE-choroid complexes 3 days after laser photocoagulation. Right and left RPE-choroid complexes from a mouse were pooled for analysis as the sample for this mouse. Data were normalized relative to β-actin mRNA levels. The y-axis represents the ratio of the mRNA level in ⁹⁹Tc-MDP–treated group to the PBS-treated group. MMP-2 mRNA expression was increased after photocoagulation (PBS-treated group compared with basal group [normal mice without laser photocoagulation]; ***P < 0.001; n = 5). ⁹⁹Tc-MDP inhibited MMP-2 mRNA expression in a dose-dependent manner (⁹⁹Tc-MDP–treated group compared with the PBS-treated group; ***P < 0.001; n = 5).

ERG for Toxicity of ⁹⁹Tc-MDP on Retina In Vivo

Representative ERG responses were shown in Figure 6. There was no statistically significant difference in a- and b-wave amplitudes or in b-wave implicit times of post-⁹⁹Tc-MDP treatment compared with those of pre-⁹⁹Tc-MDP treatment in each group (P > 0.05; n = 5) (Table 1). ERG data suggested that ⁹⁹Tc-MDP had no influence on the function of retinal outer and inner layers. 

Figure 6.

 

Flash ERGs of C57BL/6J mice pretreated and posttreated with ⁹⁹Tc-MDP. ERG recordings were performed both at baseline (before the first injection of 99Tc-MDP) and 7 days after daily intraperitoneal injection with ⁹⁹Tc-MDP at the dose of 0.5 × 10⁻¹ μg/kg BW (n = 5), 1 × 10⁻¹ μg/kg BW (n = 5), 2 × 10⁻¹ μg/kg BW (n = 5), or the same volume of PBS (n = 5). The male C57BL/6J mice aged 6 to 8 weeks were kept in the dark environment for 12 hours, and then flash ERG recordings were performed. There was no statistically significant difference in a- and b-wave amplitudes or in b-wave implicit times between pretreatment and posttreatment with ⁹⁹Tc-MDP in each group (P > 0.05; n = 5).

Table 1.

 

Amplitudes and Implicit Times of Pre-⁹⁹Tc-MDP Treatment and Post-⁹⁹Tc-MDP Treatment

Groups	a-Wave Amplitudes (μv, n = 5)			b-Wave Amplitudes (μv, n = 5)			b-Wave Implicit Times (ms, n = 5)
	Pretreatment	Posttreatment	P *	Pretreatment	Posttreatment	P *	Pretreatment	Posttreatment	P *
PBS-treated group	107.4 ± 27.1	98.4 ± 33.6	0.370	405.0 ± 74.3	348.0 ± 54.7	0.235	38.8 ± 0.95	39.5 ± 1.4	0.253
0.5 × 10−1 μg/kg BW	92.0 ± 16.3	101.0 ± 21.2	0.541	391.8 ± 102.4	374.0 ± 124.7	0.855	38.2 ± 1.1	35.8 ± 1.2	1.0
1 × 10−1 μg/kg BW	114.4 ± 28.9	108.8 ± 33.1	0.837	355.4 ± 75.6	398.0 ± 100.3	0.342	38.7 ± 0.97	38.0 ± 1.1	0.600
2 × 10−1 μg/kg BW	101.8 ± 28.0	111.4 ± 26.9	0.674	352.2 ± 128.9	374.6 ± 112.4	0.660	38.3 ± 0.96	39.6 ± 1.5	0.324

 Data are expressed as the mean ± SD

*   t-test statistic.

Proliferation and Cytotoxicity of Cells In Vitro

In the BrdU assay we used dividing (not confluent) cells, which represented the pathologic condition of active CNV to some extent, to determine the antiproliferative effect of ⁹⁹Tc-MDP. The results showed that ⁹⁹Tc-MDP treatment had a significantly inhibitive effect on cell proliferation in a dose-dependent manner (Fig. 7A) (P < 0.001; n = 6). However, ⁹⁹Tc-MDP had no influence on completely confluent (nondividing) cells in the MTT assay (Fig. 7B) (P > 0.05; n = 8). The data suggested that ⁹⁹Tc-MDP was cytotoxic only to dividing cells. 

Figure 7.

 

Antiproliferative effects of ⁹⁹Tc-MDP on dividing RF/6A cells by BrdU assay and cytotoxicity of ⁹⁹Tc-MDP on nondividing RF/6A cell viability by MTT assay. (A) BrdU assay data revealed that ⁹⁹Tc-MDP treatment had a significantly inhibitive effect on cell proliferation in a dose-dependent manner. (***P < 0.001 compared with the PBS control group; n = 6). (B) Percentage data compared with PBS control group were expressed as mean ± SD in the MTT assay. ⁹⁹Tc-MDP had no influence on completely confluent cells (nondividing cells) compared with the PBS-treated mice (P > 0.05; n = 8).

In Vitro Inhibition of VEGF-Induced Migration and Tube Formation of RF/6A Cells

In the cell migration assay, as is shown in Figures 8A and 8B, stimulation by 10 ng/mL VEGF resulted in an increase of approximately 100% migration activity compared with the basal group (without VEGF in the culture medium). ⁹⁹Tc-MDP (0.0125–0.05 μg/mL) significantly inhibited cell migration chemoattracted by 10 ng/mL VEGF in a concentration-dependent manner (P < 0.001; n = 9) (Figs. 8C-F). At the highest concentration of ⁹⁹Tc-MDP (0.05 μg/mL), the number of migrated cells was suppressed by 76.78% compared with the PBS control group. 

Figure 8.

 

⁹⁹Tc-MDP inhibited VEGF-induced cell migration in vitro. We used 8-μm pore sized transwell inserts in the cell migration study. The photographs were taken (×20 objective) 6 hours after transwell inserts incubated in a 37°C incubator. (A) Basal group without VEGF or ⁹⁹Tc-MDP. (B–E) All received 10 ng/mL VEGF as the chemoattractants. (C–E) Received increasing doses of ⁹⁹Tc-MDP (0.0125, 0.025, and 0.05 μg/mL in (C), (D), and (E), respectively. and (B) Received PBS as the control group. (F) ANOVA revealed that each of the ⁹⁹Tc-MDP treatment groups had statistically significant differences compared with the PBS treatment group (***P < 0.001; n = 9), and the numbers of the migrated cells were significantly inhibited by ⁹⁹Tc-MDP in a concentration-dependent manner. Scale bars, 50 μm (A–E).

In the VEGF-induced capillary-like vessel formation assay, VEGF significantly enhanced tube formation (Figs. 9A, 9B). At the concentration of 0.025 μg/mL ⁹⁹Tc-MDP, the length of the formed capillary-like vessel was suppressed to the basal group level (without VEGF in the culture medium) (Figs. 9A, 9D, 9E). Figures 9C to 9E showed a decreasing tendency of annular structures and numbers of branch points of capillary-like vessels, especially in the 0.05-μg/mL ⁹⁹Tc-MDP–treated group compared with the PBS control group. Vessel lengths were suppressed by the treatment of ⁹⁹Tc-MDP in a concentration-dependent fashion (Fig. 9F) (P < 0.001; n = 12). 

Figure 9.

 

⁹⁹Tc-MDP inhibited VEGF-induced tube formation in vitro. Photographs were taken 6 hours after the RF/6A cells were seeded onto the BME with a phase-contrast microscope (×10 objective). (A) Basal group without VEGF or ⁹⁹Tc-MDP. (B–E) All received 10 ng/mL VEGF as the chemoattractants. (C–E) Received increasing doses of ⁹⁹Tc-MDP (0.0125, 0.025, and 0.05 μg/mL, respectively. (B) Received PBS as the control group. (C–E) Showed a decreasing tendency of annular structures and the numbers of branch points of capillary-like vessels. (F) Showed the lengths of the vessels were significantly inhibited by ⁹⁹Tc-MDP in a concentration-dependent manner (***P < 0.001; n = 12). Scale bars, 200 μm (A–E).

These results demonstrated that ⁹⁹Tc-MDP had the ability to block VEGF-induced angiogenesis in vitro. 

Discussion

In this present study, our data for the first time revealed several important findings concerning the antiangiogenic effects of ⁹⁹Tc-MDP on the development of CNV with its anti-inflammatory property. First, ⁹⁹Tc-MDP significantly suppressed the development of CNV in murine models in vivo (Fig. 2) and had an antiproliferative effect on the RF/6A cell in vitro in a dose-dependent manner (Fig. 7A). Second, ⁹⁹Tc-MDP inhibited VEGF-induced RF/6A cell migration and tube formation in vitro (Figs. 8, 9), both in a dose-dependent manner. Third, ⁹⁹Tc-MDP treatment decreased the numbers of macrophages infiltrating the CNV (Figs. 3A, 3B), together with downregulated protein expression of VEGF (Fig. 3C) and inflammatory molecules including ICAM-1 and TNF-α (Figs. 3D, 3E). In addition, ⁹⁹Tc-MDP treatment inhibited the mRNA expression and gelatin-degrading activity of MMP-2 (Figs. 4, 5). Importantly, electroretinography and cytotoxicity study showed that any dose of ⁹⁹Tc-MDP used in our study was not toxic to the mouse retina (Fig. 6) or to nondividing (confluent) cells (Fig. 7B). 

Angiogenesis is the formation of new blood vessels from preexisting vasculature. ²⁸ As we know, the well-defined steps of angiogenesis include extracellular matrix (ECM) degradation, endothelial cell migration, cell proliferation, tube formation, and vessel wall remodeling. ²⁹ Migration and tube formation of endothelial cells have been thought to be important steps in angiogenesis. ³⁰ Our in vitro experimental data had shown that ⁹⁹Tc-MDP inhibited not only cell proliferation but also VEGF-induced cell migration and tube formation in a dose-dependent manner (Figs. 7A, 8, 9), which could be an important cellular mechanism for the inhibitive effects of ⁹⁹Tc-MDP on the CNV in vivo murine model. 

As discussed, ECM degradation is the initial step for endothelial cell migration. ECM degradation, which facilitates endothelial cell migration further into the subretinal space during angiogenesis, is thought to be a very important step in angiogenesis, ³¹ and this process is potentiated by the secretion of matrix metalloproteinases (MMPs). ³² Among MMPs, MMP-2 and MMP-9 are particularly noted by researchers because they preferentially degrade type IV collagenase (the main component of the basement membrane). ³³ Mice with MMP-2 and MMP-9 gene deficiency have a decreased incidence and severity of laser-induced CNV. ^34,35 Our data showed that MMP-2 mRNA expression was upregulated in the PBS control group (with laser photocoagulation) compared with the non-laser–treated mouse group (basal group), and ⁹⁹Tc-MDP had the property of inhibiting MMP-2 mRNA expression in a dose-dependent manner (Fig. 5). Results of gelatin-degrading activity of MMP-2 paralleled the real-time PCR results (Figs. 4, 5), and these results suggested that decreased gelatin-degrading activity of MMP-2 might be related to the reduced mRNA expression of MMP-2. Because MMP-2 can be directly secreted by macrophages ³⁶ and MMP-2 mRNA expression can be upregulated by VEGF, ³⁷ based on our data we propose that the inhibitive effects of ⁹⁹Tc-MDP on the mRNA expression of MMP-2, which directly leads to the decreased gelatin-degrading activity of MMP-2, are due to the suppression of infiltrating macrophages and the inhibited protein expression of VEGF. 

Increasing evidence has supported the notion that inflammatory mechanisms play a role in the pathogenesis of CNV. ^3,4,38,39 As one of the most important inflammatory cells, macrophages play a critical role in the development of CNV. ^4,40,41 By secreting growth factors and inflammatory cytokines, macrophages can influence every phase of the angiogenic process, such as alterations of the local extracellular matrix, migration, and proliferation of endothelial cells. ⁴² Blood-derived macrophages infiltrating the CNV are thought to be a rich source of VEGF. ⁴³ Generalized macrophage depletion reduced the size and leakage of laser-induced CNV associated with decreased macrophage infiltration and VEGF protein expression. ^7,8 The process of angiogenesis is controlled by a variety of cells and molecular mediators. Macrophages can promote angiogenesis, and this promotive effect is thought to be related to its production of VEGF and a variety of inflammatory cytokines, such as ICAM-1, TNF-α, and MMPs. ^{9 –12,44} Importantly, ICAM-1 is expressed on RPE and the vascular endothelial cell surface and is an important component of cell-to-cell interactions during inflammatory responses, mediating leukocyte (including macrophages) adhesion. ⁴⁵ The expression of ICAM-1 in cultured human RPE cells is upregulated by the stimulation of IL-1β and TNF-α. ⁴⁶ TNF-α is a pleiotropic cytokine that mediates inflammatory, proliferative, and cytotoxic effects in a variety of cell types. ⁴⁷ TNF-α contributes to the development of choroidal neovascularization through triggering VEGF production by RPE cells. ⁴ In addition, TNF-α enhances the expression of other angiogenic regulators such as MCP-1 and interleukins. Based on these facts, we can see there is a closed but amplifying circuit among VEGF, inflammatory cells (especially for macrophages), and inflammatory cytokines. ^4,48 The initial objective of this study was to use the anti-inflammatory property of ⁹⁹Tc-MDP to interrupt this inflammatory cascade circuit to inhibit CNV formation. Our cellular and molecular data showed that ⁹⁹Tc-MDP can inhibit macrophages infiltrating to the CNV and the downregulated expression of VEGF, ICAM-1, and TNF-α. Based these data, we put forward a hypothesis for the mechanisms of inhibiting the effect of ⁹⁹Tc-MDP on CNV as follows. First, ⁹⁹Tc-MDP decreases the number of macrophages infiltrating the CNV. Macrophages are the main sources of VEGF, which is a critical angiogenic factor in CNV formation; therefore, the production of VEGF protein will be reduced. Second, reductions in the release of TNF-α due to the decreased numbers of macrophages can weaken ICAM-1 expression in endothelial cells and RPE cells, which will mediate fewer macrophages adhering to the CNV area. As a result, less VEGF protein will be produced, and CNV development will be suppressed. In addition, the in vitro experiments showed that ⁹⁹Tc-MDP could affect the downstream signaling transduction that stimulates the proliferation, migration, and tube formation of endothelial cells after VEGF binding to its receptors. Therefore, ⁹⁹Tc-MDP not only reduces VEGF expression in vivo (Fig. 3C), it also blocks the function of already existing VEGF in vitro (Figs. 8, 9). This multiple targeting sites of ⁹⁹Tc-MDP on VEGF may partially explain the inhibitive effects of ⁹⁹Tc-MDP on the development of CNV, though the exact mechanisms have not been disclosed completely. 

The ⁹⁹Tc-MDP used in our study is different from the agent ^99mTc-MDP (a radioactive imaging agent widely used for bone scintigraphy). Actually, ⁹⁹Tc-MDP is the decay product of ^99mTc-MDP. ⁹⁹Tc-MDP has little radioactivity, and this small amount of radioactivity has been demonstrated to be harmless to the human body in prolonged systemic administration in the clinical treatment of systemic inflammatory diseases such as RA and AS in China. ^{14 –17} Because of its powerful anti-inflammatory effect and its safety, we chose ⁹⁹Tc-MDP but not ^99mTc-MDP for our experiment. Significantly, electroretinography and cytotoxicity showed that any dose of ⁹⁹Tc-MDP used in our study was not toxic to the mouse retina (Fig. 6) or to nondividing (confluent) cells (Fig. 7B). Furthermore, in our in vivo study, we found no body weight changes (data not shown) or mouse deaths caused by the injection of ⁹⁹Tc-MDP. Based on these facts, we believe that ⁹⁹Tc-MDP is a safe and effective anti-inflammatory agent. 

In summary, our experimental data showed that ⁹⁹Tc-MDP can affect multiple inflammatory steps during CNV development, including MMP-2–mediated extracellular matrix degradation, recruitment of macrophages mediated by ICAM-1, and downregulated expressions of VEGF and inflammatory cytokines derived from macrophages. However, though ⁹⁹Tc-MDP reduces CNV formation significantly, the incidence rate of CNV between the ⁹⁹Tc-MDP treatment groups and the PBS treatment group are of no statistical significance (data not shown), suggesting that mechanisms other than inflammatory ones are involved in the angiogenesis of laser-induced CNV. In addition, as was shown in Figures 7A and 7B, ⁹⁹Tc-MDP had a specific inhibiting effect only on proliferating cells but not on nonproliferating cells, suggesting that ⁹⁹Tc-MDP might help to inhibit those active CNV lesions. However, pathologic mechanisms of CNV murine models are different from those of human beings. Therefore, the effects of ⁹⁹Tc-MDP on human beings still must be confirmed in clinical studies. 

All these findings support the idea that inflammatory mechanisms play an important role in the pathogenesis of CNV and suggest that ⁹⁹Tc-MDP has therapeutic potential to inhibit CNV. Whereas there is still little knowledge about the effects of ⁹⁹Tc-MDP on CNV, further studies concerning long-term inhibitive effects and more detailed mechanisms of CNV inhibition are needed. 

Footnotes

 Supported by Grant 2008B08070306 from the Technology Program of Guangdong Province.

Footnotes

 Disclosure: K. Lai, None; L. Xu, None; C. Ji, None; K. Wu, None; Z. Tian, None; C. Huang, None; X. Zhong, None; H. Ye, None