Dual Role of Intravitreous Infliximab in Experimental Choroidal Neovascularization: Effect on the Expression of Sulfated Glycosaminoglycans

Abstract

Purpose.: To determine effects of intravitreous anti-TNF-α (infliximab) in a laser-induced choroidal neovascularization (CNV) model by fluorescein angiogram (FA), immunofluorescence, ELISA, and glycosaminoglycan analyses.

Methods.: CNV induction was performed using argon laser. Rats were divided into eight groups (no-laser no-infliximab; laser; laser with 10, 20, 40, 80, or 320 μg infliximab; and isotype-matched IgG). After 3 weeks, CNV area was measured by FA and von Willebrand factor (vWF) immunofluorescence. VEGF, TGF-β, and syndecan-4 were evaluated by ELISA and immunofluorescence. Glycosaminoglycan expression was determined in retina and choroid of animals metabolically labeled with [³⁵S]-sulfate. Cytotoxicity was investigated using ARPE-19 and endothelial cells.

Results.: FA showed significant reduction in the low-dose infliximab groups (10–40 μg), confirmed by vWF immunofluorescence that showed 49% decrease in the CNV. VEGF and TGF-β decreased expression detected by ELISA and immunofluorescence paralleled these results. Similar data were observed for syndecan-4. The expression of these molecules in the neovascularization area using 320 μg was similar to the no-infliximab laser group or laser with isotype-matched IgG. Heparan sulfate expression in retina and choroid paralleled the observed effects on angiogenesis. Increased expression of chondroitin sulfate in retina and dermatan sulfate in choroid reflects the effects of injury and fibrosis using high doses of anti-TNF-α. Infliximab showed no cytotoxic effect in ARPE-19 cells, whereas high doses led to 20% decrease in endothelial cell viability.

Conclusions.: Intravitreal infliximab shows dual effect on the development of laser-induced CNV. It reduces angiogenesis and glycosaminoglycan expression at low doses, whereas opposite effects are observed at high doses.

Choroidal neovascularization is an intricate process controlled by myriad angiogenic agents such as growth factors, cytokines, and extracellular matrix (ECM) components including glycosaminoglycans. ^1,2  

Several growth factors have been implicated in pathologic vessel formation in ocular diseases, such as age-related macular degeneration, including fibroblast growth factor (FGF), platelet-derived endothelial growth factor (PEDF), and vascular endothelial growth factor (VEGF), among others. ³  

The role of tumor necrosis factor-alpha (TNF-α) in angiogenesis has been controversial. It inhibits endothelial cell proliferation or microvascular sprouts in vitro. Conversely, sustained release of TNF-α in the cornea or injection of recombinant TNF-α into the vitreous cavity of rabbits causes cellular infiltration and neovascularization (NV) in the cornea, ^4,5 possibly by induced expression of other proangiogenic proteins such as interleukin-8 (IL-8), VEGF, FGF-2, ⁶ and TGF-β. ⁷ TNF-α also induces the expression of VEGF receptor 2 and neuropilin-1 in cultured endothelial cells. ⁸ Furthermore, TNF-α induces capillary-tubule formation in HMEC-1 (human microvascular endothelial cells). ⁹  

In part, this dual effect may be related to the fact that TNF-α can trigger opposite signaling pathways involved in angiogenesis, evoking apoptosis ^10,11 as well as cell survival and proliferation, depending on the conditions. ^12–14 In addition, TNF-α recruits inflammatory cells, which can stimulate or inhibit NV, leading to the breakdown of the blood-retinal barrier. ^15,16 Anti-TNF-α suppresses experimental autoimmune uveoretinitis, suggesting another therapeutic use for this antibody. ¹⁷ Such variable findings indicate that this cytokine can exert different effects depending on the tissue and pathologic conditions, though the exact role in ocular disease is yet to be clarified. ¹⁸  

Infliximab is a chimeric human immunoglobulin IgG1 with a mouse Fv variable fragment of high TNF-α affinity and neutralizing capacity. ¹⁷ It inhibits functional TNF-α activity in a variety of in vitro bioassays using human fibroblasts, endothelial cells, neutrophils, lymphocytes, and epithelial cells. ¹⁹ In vivo, endovenous infliximab has been indicated in the treatment of rheumatologic, gastrointestinal, and dermatologic diseases, and recent studies have described its efficacy in the treatment of chronic ocular inflammation. ²⁰  

Glycosaminoglycans are negatively charged polysaccharides composed of repeating disaccharide units. Glycosaminoglycans are present in almost all cell types, where they can be found in the ECM, associated with the plasma membrane, or in cytoplasmic granules. ²¹  

Several studies have shown that there is a specificity directing the interactions of glycosaminoglycans and target proteins regarding both the fine structure of the polysaccharide chain and precise protein motifs. Thus, they can interact with a diverse range of proteins leading to various biologic activities. ² Among the sulfated glycosaminoglycans, heparan sulfates have been shown to be involved in the modulation of the neovascularization that takes place in different physiological and pathologic conditions. ^22,23 This modulation occurs through the interaction of glycosaminoglycans with angiogenic growth factors or with negative regulators of angiogenesis, such as VEGF, FGF, placenta growth factor, transforming growth factor-β (TGF-β), interferon-γ (IFN-γ), TNF-α, and thrombospondin, suggesting that the study of the biochemical bases of protein-glycosaminoglycan interaction may help to design synthetic analogs endowed with angiostatic properties. 

Based on these findings, this experimental study was designed to determine the effects of intravitreal infliximab in angiogenesis modulation and glycosaminoglycans expression in CNV lesions. 

Materials and Methods

Animals

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Care Ethics Committee of the Federal University of São Paulo (number 1388/06). This study was carried out on female heterozygote pigmented Zucker rats weighing between 180 and 220 g. The animals were kept under 12-hour light/12-hour dark room conditions and had free access to food and water. 

Induction of Choroidal Neovascularization

Animals were anesthetized with intramuscular injection of a mixture of 56 mg/kg ketamine and 6 mg/kg xylazine. Afterward, the pupil was dilated by topical application of a combination of 1% tropicamide and 2.5% phenylephrine hydrochloride (Allergan, Guarulhos, Brazil). Photocoagulation was performed using argon laser equipment connected to a slit lamp delivery system (SL 115; Carl Zeiss Meditec, Jena, Germany). A handheld coverslip was used as a contact lens. Four lesions, located at the 3, 6, 9, and 12 o'clock meridians centered on the optic nerve head and located approximately 2 disc diameters from the optic nerve head were created using a power of 200 mW, spot size of 100 μm, and duration of 100 ms. The aim was to create a rupture in Bruch's membrane that is indicated by a bubble at the time of lasering. 

Intravitreous Infliximab Injection

Immediately after the laser procedure, the animals were injected with infliximab (Centocor Inc., Horsham, PA) using a microsyringe (Hamilton Co., Reno, NV). They were assigned to the following experimental groups according to the dose: 10 μg, 20 μg, 40 μg, 80 μg, or 320 μg in 5 μL balanced salt solution (BSS; Alcon, São Paulo, Brazil); only BSS (no-infliximab); or isotype-matched IgG1 control (BD PharMingen, San Diego, CA). The injection was visualized with the slit lamp to confirm proper placement. Animals with traumatic lens injury, vitreous hemorrhage, or retinal hemorrhage were excluded from the study. 

Glycosaminoglycan Analyses

Immediately after laser exposure and intravitreous injection of infliximab, four animals of the experimental groups (BSS 10 μg, 80 μg, and 320 μg and a control corresponding to no-laser no-infliximab) were injected intraperitoneally with [³⁵S]-inorganic sulfate (IPEN, São Paulo, Brazil; 5 μCi/g). After 48 hours, the eyes were enucleated, and the retina and choroid tissue removed. The tissues were ground with 10 vol acetone, dried, and weighed. The sulfated glycosaminoglycans were extracted according to a procedure previously described ²¹ with some modifications, as follows. The dry tissue was incubated at 56°C for 24 hours with a proteolytic enzyme isolated from Bacillus subtilis 4 mg/mL in 0.05 M Tris-HCl buffer, pH 8.0, containing 1 M NaCl (Maxatase; Biocon Laboratories, São Paulo, Brazil). After this incubation, trichloroacetic acid was added up to 10% final concentration, and the precipitate formed was removed by centrifugation. Three volumes of methanol were slowly added to the supernatant. After standing overnight at −20°C, the precipitate containing the free glycosaminoglycan chains was collected by centrifugation. 

The sulfated glycosaminoglycans were separated, identified, and quantified by a combination of agarose gel electrophoresis in 0.05 M 1,3-diaminopropane acetate buffer, pH 9.0, ²⁴ and enzymatic degradation with specific enzymes as previously described. ^25–27 Nonradioactive sulfated glycosaminoglycan quantification was performed by densitometry at 560 nm of the agarose gel after electrophoresis and toluidine blue staining. The extinction coefficients of the sulfated glycosaminoglycans were calculated using chondroitin sulfate, dermatan sulfate, and heparan sulfate standards. For the quantification of [³⁵S]-labeled sulfated glycosaminoglycans, the radioactive bands were visualized by exposure to storage phosphor screen (Cyclone; Packard Instruments Co. Inc., Meriden, CT) for 24 hours, and the images were acquired with a storage phosphor system (Cyclone; Packard). The radioactive bands were excised from the agarose gels and counted in 5 mL liquid scintillation cocktail (Ultima Gold; Packard) in a liquid scintillation analyzer (TRI-CARB 2100 TR; Packard). The specific activity represents the turnover of the sulfated glycosaminoglycans; it reports the amount of [³⁵S]-sulfate incorporated into glycosaminoglycans divided by the amount of glycosaminoglycan present in the tissue expressed in micrograms. 

Fluorescein Angiography

Eight animals of all experimental groups were submitted to laser exposure and infliximab administration, as described. Three weeks after CNV laser induction, the animals were anesthetized and the pupils were dilated as described. Fluorescein angiography (FA) was performed with a retina angiograph (HRA-2; Heidelberg Engineering, Dossenheim, Germany). ²⁸  

Rats were held manually in front of the retina angiograph (HRA-2; Heidelberg Engineering) in an upright position, and FA was then performed. Fluorescein injections were administered intraperitoneally (0.2 mL of 2% fluorescein; Ophthalmos, São Paulo, Brazil), and the timer was started as soon as the fluorescein bolus was injected. All early images were captured immediately after fluorescein injection, and all late-phase images were obtained 12 minutes after injections. Two masked observers with experience in FA evaluation, not involved in the experiments, assessed the FA obtained 21 days after laser induction of CNV with angiographic standards. For quantitative analysis of fluorescein leakage, the image of the entire area of leakage, determined as area of hyperfluorescence, was measured in the late phase in square pixels (pixel²) using retina angiograph (HRA-2; Heidelberg Engineering) software. Data were expressed as mean ± SEM. 

Immunofluorescence Analysis

Eight animals of all experimental groups were submitted to laser shots and infliximab treatment, as described. After 3 weeks, clinical examination was performed and the eyes were enucleated. The eyecups were fixed with 2% paraformaldehyde for 30 minutes, washed in 0.1 M glycine in PBS, and incubated with anti-von Willebrand factor (vWF) 1:50 (sc-8068; Santa Cruz Biotechnology, Santa Cruz, CA) in PBS containing 0.1% saponin at room temperature for 2 hours. The eyecups were then incubated with anti-goat IgG conjugated with Alexa Fluor 594 (Molecular Probes, Carlsbad, CA). Double staining was performed using anti-VEGF (1:50; Santa Cruz), anti-TGF-β (1:50; Santa Cruz), or mouse monoclonal anti-syndecan 4 IgM (culture supernatant; 4E12A8 clone). The monoclonal antibody was raised in our laboratory against a synthetic peptide containing the first 29 N-terminal amino acids of syndecan-4. The antibodies were incubated in PBS containing 0.1% saponin at room temperature for 2 hours. Afterward, anti-rabbit IgG or anti-mouse IgM conjugated with Alexa Fluor 488 (Molecular Probes) in PBS were incubated for 30 minutes. The eyecups were flatmounted in mounting medium (Fluoromont G; Electron Microscopy Sciences, Hatfield, PA), and the slides were analyzed by confocal microscopy (LSM 510 META; Carl Zeiss). The area (μm²) of lesions was measured using LSM image browser (Carl Zeiss) software by two different analyzers. 

ELISA

The CNV laser induction was performed as described. Six animals from each group (laser, laser with isotype-matched IgG, and laser with different doses of infliximab [10, 80, 320 μg]) were killed 3 weeks after laser induction, and retinal/choroidal tissues were dissected from the enucleated eyes. Retina and choroid homogenates were prepared initially by one cycle of freeze and thaw. Afterward the material was suspended in cold PBS containing 20 mM Tris-HCl, pH 8.0, 150 mM NaCl, protease inhibitors (Complete Protease Inhibitor Cocktail Tablets; Roche, Basel, Switzerland) and phosphatase inhibitor (PhosSTOP Phosphatase Inhibitor Cocktail Tablets; Roche) and was homogenized using Dounce tissue grinder. Finally, the suspension was sonicated for 3 minutes in pulses of 30 seconds per minute in the cold. Aliquots of total proteins (50 μg) from each sample were assayed in triplicate for the quantification of TGF-β and VEGF proteins using ELISA kits for rat VEGF and rat TGF-β1 (Quantikine; R&D Systems, Minneapolis, MN), according to manufacturer's instructions. 

Cell Culture and Cytotoxicity Assay

Adult human retinal pigment epithelial cells (ARPE-19) were grown in basal DMEM/F12 medium (Invitrogen, San Diego, CA) containing 10% fetal bovine serum (Cultilab, Campinas, Brazil), 15 mM HEPES, 2.0 mM l-glutamine, 0.5 mM sodium pyruvate, and 20 mM sodium bicarbonate. Rabbit aorta endothelial cells were grown in F12 medium (Invitrogen) containing 10% fetal bovine serum (Cultilab) and 20 mM sodium bicarbonate. ²⁶  

Cytotoxicity was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assay. For this assay, 10⁵ ARPE-19 cells or 2 × 10⁴ endothelial cells were seeded in 96-well plates and cultured for 5 days. Afterward, the medium was removed and fresh medium containing different amounts of infliximab (10, 20, 40, 80, and 320 μg in 200 μL/well) were added to the wells, and the cells were maintained for 24 hours at 37°C, 5% CO₂ in a humidified incubator. Afterward, the cells were washed with PBS, and serum-free medium containing MTT (0.5 mg/mL) was added. After 2 hours of incubation, isopropanol extraction was performed, and the absorbance was measured at 570 nm with an ELISA reader (ELx800; BioTek Instruments, Winooski, VT). 

Statistical Analysis

Data are expressed as mean ± SEM. Statistical analyses were performed using one-way ANOVA with Bonferroni's post test and statistical software (Prism 5.0; Graph Pad, San Diego, CA). A 95% confidence interval and a 5% level of significance were adopted; therefore, results with P ≤ 0.05 were considered significant. 

Results

Measurement of CNV Using Fluorescein Angiography Analysis

Quantitative assessment of CNV was performed using fluorescein angiography. An important and significant reduction in leakage after laser photocoagulation compared with the untreated group was observed for rats treated with 10, 20, and 40 μg infliximab (Fig. 1A). In these groups, there was reduction in the early hyperfluorescence and a reduction in area and intensity of the late fluorescein leakage compared with the BSS or isotype-matched IgG1 group. For the animals treated with 80 and 320 μg, no significant effect on leakage area was observed. 

Figure 1.

 

Choroidal neovascularization analyses. (A) Area of the CNV leakage evidenced by fluorescein angiography using retina angiograph. (B) Measurement of CNV area by immunofluorescence using anti-vWF antibody. (C) Representative images of confocal analysis of flatmount choroidal tissue using anti-vWF antibody stained with Alexa Fluor 546 (red) in CNV. Scale bar, 100 μm. P values represent the results of statistical analyses (ANOVA). Ctr, no-laser no-infliximab control; BSS (laser and no-infliximab); 10 μg, 80 μg, and 320 μg, represent doses of infliximab; IgG represents isotype-matched IgG1 antibody control.

Measurement of CNV Using Immunofluorescence Analysis

Immunofluorescence analysis of choroidal flatmounts using anti-vWF was performed to assess the area of the CNV lesions on day 21 after laser photocoagulation with or without infliximab treatment. Rats treated with different doses of intravitreous infliximab up to 80 μg showed significant decreases in the lesion area compared with the BSS or isotype-matched IgG1 group (Figs. 1B, 1C). On the other hand, high-dose (320-μg group) infliximab showed no effect in preventing neovascularization. The data show that with vWF as a marker, the reduction of angiogenesis could be detected up to 80 μg infliximab, whereas by FA, the effect of the drug was detected up to 40 μg of the antibody (Fig. 1B). 

Immunofluorescence Analysis of VEGF, TGF-β, and Syndecan-4

Immunofluorescence analysis of choroidal flatmounts using anti-VEGF, anti-TGF-β, or anti-syndecan-4 double stained with anti-vWF was performed to assess the expression and localization of these molecules. Rats treated with the different doses of intravitreous infliximab up to 80 μg showed significant decreases in the expression of VEGF, TGF-β, and syndecan-4 compared with the BSS group. Conversely, the expression of these molecules in the high-dose group (320 μg) showed no decrease in neovascularization area, resembling the BSS group (Figs. 2 3–4). 

Figure 2.

 

VEGF expression in choroidal neovascularization areas. Representative images of confocal immunofluorescence analysis of flatmount choroidal tissue using anti-VEGF antibody stained with Alexa Fluor 488 (green) and anti-vWF antibody stained with Alexa Fluor 546 (red) in CNV area. Merge represents the double staining. BSS (laser and no-infliximab); 10 μg, 80 μg, and 320 μg represent the doses of intravitreous infliximab. Scale bar, 100 μm.

Figure 3.

 

TGF-β expression in choroidal neovascularization areas. Representative images of confocal immunofluorescence analysis of flatmount choroidal tissue using anti-TGF-β antibody stained with Alexa Fluor 488 (green) and anti-vWF antibody stained with Alexa Fluor 546 (red) in CNV area. Merge represents the double staining. BSS (laser and no-infliximab); 10 μg, 80 μg, and 320 μg represent doses of intravitreous infliximab. Scale bar, 100 μm.

Figure 4.

 

Syndecan-4 expression in choroidal neovascularization areas. Representative images of confocal immunofluorescence analysis of flatmount choroidal tissue using anti-syndecan-4 antibody (Syn 4) stained with Alexa Fluor 488 (green) and anti-vWF antibody stained with Alexa Fluor 546 (red) in CNV area. Merge represents the double staining. BSS (laser and no-infliximab); 10 μg, 80 μg, and 320 μg represent doses of intravitreous infliximab. Scale bar, 100 μm.

Quantification of VEGF and TGF-β by ELISA

The expression of VEGF and TGF-β in tissue homogenates of retina and choroid was determined by ELISA. Rats treated with the different doses of intravitreous infliximab up to 80 μg showed significant decreases in the expression of VEGF and TGF-β compared with the BSS or isotype-matched IgG group. Expression of these molecules in the high-dose group (320 μg) showed no decrease in neovascularization area (Fig. 5). 

Figure 5.

 

Effect of CNV and infliximab in the expression of VEGF and TGF-β. VEGF (A) and TGF-β1 (B) expressions were investigated by ELISA in retina/choroid tissue homogenates. Values are mean ± SEM and represent triplicate determinations of six animals from each experimental group. BSS (laser and no-infliximab); IgG represents isotype-matched IgG1 antibody control.

Expression of Sulfated Glycosaminoglycans in Retina and Choroid

Sulfated glycosaminoglycans from retinal and choroid tissues of rats after laser-induced CNV with and without infliximab treatment were identified according to their electrophoretic mobilities in agarose gel and degradation with chondroitinase and heparitinase. Thus, the compound migrating as chondroitin sulfate was totally degraded by chondroitinase AC, whereas the compound migrating as dermatan sulfate was totally degraded by chondroitinase ABC. Neither enzyme had an effect on the compound migrating as heparan sulfate, which in turn was totally degraded by heparitinase. The sulfated glycosaminoglycan distribution is tissue specific. Chondroitin sulfate is the predominant compound in retina, and dermatan sulfate is the predominant compound in choroid tissue (Fig. 6). 

Figure 6.

 

Effect of CNV and infliximab on the expression of sulfated glycosaminoglycans in choroid and retina. Ctr, no-laser and no-infliximab; BSS (laser and no-infliximab); HS, heparan sulfate; DS, dermatan sulfate; CS, chondroitin sulfate. P values represent the results of statistical analyses (ANOVA).

The total amount of each sulfated glycosaminoglycan was assessed after densitometry of the toluidine blue-stained bands in the gel slab and the newly synthesized compounds by [³⁵S]-sulfate incorporation. The ratio of radioactivity to the amounts of the compound present in the tissue indicates the turnover of each sulfated glycosaminoglycan after exposure of the animals to the different treatments (Fig. 6). The expression of each type of sulfated glycosaminoglycan varied among the different groups of treatment for each compound, both in retina and choroid tissue, as depicted in the figure. 

Bonferroni post test showed a significant increase in the expression of heparan sulfate after laser shot when compared with no-laser no-infliximab control eyes, both in choroid tissue (P = 0.025) and in retina (P = 0.05). It is evident that doses of infliximab up to 80 μg reduced the expression of heparan sulfate to the level of the no-laser no-infliximab control group. Interestingly, a higher dose of infliximab (320 μg) led to an increase in the expression of heparan sulfate in both tissues (P = 0.01). 

Among the galactosaminoglycans, dermatan sulfate was the compound present in choroid tissues, whereas chondroitin sulfate was the compound expressed in the retina. Again, Bonferroni post test showed significant differences in the expression of dermatan sulfate in choroid tissues when comparing no-laser no-infliximab control eyes with those treated with high doses of infliximab (320 μg; P = 0.01). This difference in the expression of dermatan sulfate was also observed comparing the doses of 10 and 320 μg infliximab (P = 0.005). High-dose (320 μg) infliximab increased the expression of chondroitin sulfate in retina (P = 0.045). 

Cytotoxic Effect of Infliximab

The potential cytotoxic effect of infliximab was investigated using both retinal pigmented epithelial (ARPE-19) and endothelial cells. Figure 7A shows no cytotoxic effects on in ARPE-19 cells up to 320 μg monoclonal antibody. A 20% decrease in the viability of endothelial cells was detected when the cells were exposed to 320 μg infliximab. 

Figure 7.

 

Effect of infliximab on the viability of ARPE-19 cells (A) and endothelial cells (B). Cell viability was assessed by MTT assay. ARPE-19, retinal pigment epithelial cells. Zero in the x-axis represents the results from cells treated with BSS and no infliximab. *P < 0.05.

Discussion

Age-related macular degeneration (AMD) with CNV is one of the main causes of blindness in the world. The development of new therapies against the angiogenic component of CNV could have a significant impact on the health and quality of life of patients with AMD. The identification of modulators of angiogenesis has helped in the understanding of these pathologic processes and, as a consequence, in the search for new treatments. ²⁹  

Cytokines, growth factors, and ECM molecules orchestrate cell behavior in the development of choroidal neovascularization. TNF-α is a macrophage/monocyte-derived polypeptide that modulates the expression of various genes in vascular endothelial cells. ⁵ TNF-α is clearly involved in angiogenesis and shows a dual effect depending on the concentration, either inhibiting or stimulating angiogenesis. ^8,14,30 TNF-α-induced angiogenesis is mediated both directly by TNF-α because of its inflammatory activity and indirectly by other TNF-α-induced angiogenesis-promoting factors. It is known that TNF-α increases the expression of VEGF, FGF-2, IL-8, and their receptors in human endothelial cells. ^4–6,8 Therefore, angiogenesis induced by TNF-α results from the combined effect of various angiogenic factors. ^31–33 The data here presented are in agreement with these observations. Our results show that with low doses of anti-TNF-α, the expression of proangiogenic growth factors, such as VEGF and TGF-β, is decreased. Similar observations were found for syndecan-4, a heparan sulfate proteoglycan that is present at the endothelial cell surface. 

The dual role of TNF-α in angiogenesis is clearly pointed out in the present study. Our data show that a decrease in angiogenesis is already observed at low doses (10 and 20 μg) of the antibody. Increasing the dose to 80 μg did not increase the antiangiogenic effect. Furthermore, CNV in the presence of higher doses (320 μg) showed a tendency to increase above the nontreated group (BSS). Intravitreous injection of isotype-matched IgG1 showed no effect on neovascularization. Thus, the combined data suggest that low doses of infliximab render the cytokine available to modulate angiogenesis, as demonstrated by the decrease in the expression of the VEGF and TGF-β in the neovascularization area shown by both ELISA quantification and immunofluorescence confocal microscopy. Our results show a dose-response relationship for the primary end point and contrast with those previously reported by Olson et al. ³⁴ These authors were able to demonstrate a reduction of CNV lesion with only 75 μg infliximab. Nevertheless, it should be pointed out that our method measured CNV using a specific marker of endothelial cells, whereas in their work the measurements reflected the vessels stained after perfusion of dextran labeled with a fluorophore. 

Glycosaminoglycans participate in many biological functions, including cell adhesion, growth control, cellular signaling, organogenesis, inflammation, tumorigenesis, and angiogenesis. ^35–40 It has been demonstrated that certain cytokines require interactions with glycosaminoglycans for their in vivo function. ⁴¹ The interaction is thought to provide a mechanism for retaining chemokines on cell surfaces, facilitating the formation of chemokine gradients. ⁴²  

Heparan sulfate plays a role in vascular endothelial cell function, modulating the activities of angiogenic growth factors. ²² Heparan sulfate acts in collaboration with members of the heparin-binding growth factors (as FGF, VEGF, and TGF-β) and their receptors to control various aspects of vascular development and angiogenesis. The signaling pathways triggered by FGFs and VEGFs are essentially heparan sulfate dependent. The potentiation of angiogenic growth factors mediated by heparan sulfate occurs via distinct pathways: at the cell surface contributing to the formation of a ternary complex with the growth factor and its receptor or serving as important reservoirs for these angiogenic factors at the ECM. ^2,38,39,43 The differences in heparan sulfate expression both in choroid tissue and in retina followed the same pattern of the CNV area observed for the different experimental groups showing a correlation with the process of angiogenesis. Laser exposure enhanced the expression of heparan sulfate, which returned to control levels after administration of low doses of infliximab. Again, using higher doses of the antibody, an increase in the expression of heparan sulfate was observed, reflecting the stimulus in angiogenesis. 

Dermatan sulfate has been implicated in wound repair and fibrosis. ⁴⁴ After laser exposure, there was a tendency toward overexpression of this glycosaminoglycan (P = 0.1), but, more important, when using a high dose of the antibody, a significant increase in the expression of dermatan sulfate was observed (P = 0.01). Because CNV is a fibrovascular lesion, the increase in dermatan sulfate synthesis could be related to the role of this glycosaminoglycan in fibrosis related to macrophage function. ^45,46  

Injury to the nerve tissue of vertebrates leads to the formation of a glial scar and the production of inhibitory molecules, including chondroitin sulfate proteoglycans. ^47–49 As observed for dermatan sulfate in choroid, high doses of the antibody lead to an increase in the expression of chondroitin sulfate in retina, which could be explained by injury to the nerve tissue. 

These changes in wound healing observed with high doses of the antibody could be related to cytotoxic effects. However, no toxic effects were detected in RPE cells when measuring the activity of mitochondrial reductase, and endothelial cells showed a 20% decrease in viability when treated with infliximab. These results are in accordance with the findings of ocular safety of intravitreal infliximab in a rabbit model. ⁵⁰ Thus, the effects on the expression of cell surface and ECM components on CNV lesion apparently are not the result of cytotoxicity of the antibody. 

The results presented here raise important questions regarding the effective dose that should be used to obtain the maximum response without collateral adverse effects and point out that inhibition of angiogenesis is obtained with low doses of anti-TNF-α. 

Footnotes

 Supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Vision Institute from UNIFESP Ophthalmology Department. CVSR is a recipient of a fellowship from CNPq, and JLD is a recipient of a fellowship from FapUnifesp (Jairo Ramos Award).

Footnotes

 Disclosure: C.V. Regatieri, None; J.L. Dreyfuss, None; G.B. Melo, None; D. Lavinsky, None; M.E. Farah, None; H.B. Nader, None

Footnotes

 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.

The authors thank Acácio A. Lima (Ophthalmos) for providing the antibody (infliximab) used in the present study.