Upregulated Expression of Heparanase in the Vitreous of Patients With Proliferative Diabetic Retinopathy Originates From Activated Endothelial Cells and Leukocytes

Abstract

Purpose: To determine and interrelate the levels of heparanase, syndecan-1, and VEGF in proliferative diabetic retinopathy (PDR), and to study the production of heparanase by human retinal microvascular endothelial cells (HRMEC) and its effect on HRMEC barrier function.

Methods: Vitreous samples from 33 PDR and 27 nondiabetic patients, epiretinal membranes from 16 patients with PDR and HRMEC were studied by enzyme-linked immunosorbent assay, immunohistochemistry, and Western blot analysis. The effect of heparanase on HRMEC barrier function was evaluated by transendothelial electrical resistance.

Results: We showed a significant increase in the expression of heparanase, syndecan-1, and VEGF in vitreous samples from PDR patients compared with nondiabetic controls (P < 0.0001 for all comparisons). Significant positive correlations were found between the levels of heparanase and the levels of syndecan-1 (r = 0.75, P < 0.0001) and VEGF (r = 0.91, P < 0.0001) and between the levels of syndecan-1 and the levels of VEGF (r = 0.78, P < 0.0001). In epiretinal membranes, heparanase was expressed in vascular endothelial cells and CD45-expressing leukocytes. High-glucose, tumor necrosis factor alpha (TNF-α), and the combination of TNF-α and interleukin (IL)-1β, but not cobalt chloride induced upregulation of heparanase in HRMEC. Heparanase-reduced transendothelial electrical resistance of HRMEC.

Conclusions: Our findings suggest a link between heparanase, syndecan-1, and VEGF in the progression of PDR and that heparanase is a potential target for therapy of diabetic retinopathy.

Diabetic retinopathy (DR) is characterized by progressive retinal vasculopathy with endothelial cell dysfunction, breakdown of the blood–retinal barrier and ischemia-induced angiogenesis. Vascular endothelial growth factor, an endothelial cell mitogen that also enhances vascular permeability, is thought to be the major angiogenic factor in DR.¹ Several studies have shown the overexpression of proinflammatory and proangiogenic factors in the ocular microenvironment of patients with proliferative diabetic retinopathy (PDR)^2–6 suggesting that persistent inflammation and neovascularization are critical for PDR initiation and progression. Recently, the causal relationship between inflammation and angiogenesis is widely accepted.^7,8 An emerging issue in DR research is the focus on the mechanistic link between chronic, low-grade inflammation and angiogenesis. Heparanase was recently identified to play critical roles in the modulation of inflammation and angiogenesis and might provide a mechanistic link between chronic, low-grade inflammation and angiogenesis in PDR.^9–13 

Heparanase is an endo-β-D-glucuronidase responsible for heparan sulphate degradation, yielding heparan sulphate fragments with an appreciable size (5–7 kDa) and biological potency.^9–13 Heparan sulphate, the carbohydrate structure of heparan sulphate proteoglycans, is a major constituent of the ECM, basement membranes of retinal capillaries and cell surface molecules. Heparan sulphate, therefore, is an essential component responsible for the basement membrane barrier function and in growth factor activity and cellular adhesion.¹⁴ Human heparanase is initially produced as an inactive preproenzyme which undergoes post-translational processing to yield a 65 kDa proenzyme for secretion. Proteolytic cleavage of proheparanase by the cysteine protease cathepsin L leads to the formation of catalytically active heparanase, a heterodimer consisting of a 50 kDa polypeptide noncovalently bound to a 8 kDa peptide.^10–13 Heparan sulphate–degrading activity is strongly implicated in fundamental biological processes associated with remodeling of the ECM such as inflammation, angiogenesis, and cancer metastasis.^9–17 Heparanase is also involved in activation of the coagulation cascade.⁹ 

Heparanase is rarely expressed in normal tissues, but becomes highly expressed in an increasing number of human tumors and plays an important role in the growth, invasion, angiogenesis, and metastatic potential promoting an aggressive phenotype in many tumor types.^{9,11,13,18,19} Much of this activity in tumors can be attributed to the fact that heparanase acts as a potent stimulator of tumor angiogenesis.^9,11,13 This effect on angiogenesis is thought to be mediated by several mechanisms. Heparanase enzyme activity has been associated with destruction of the basement membrane before cell invasion, an event that may enhance endothelial cell migration. Heparanase can also release heparan sulphate–bound growth factors such as VEGF. In addition, heparanase stimulates syndecan-1 (a transmembrane heparan sulphate proteoglycan) shedding from the cell surface. Vascular endothelial growth factor forms a complex with shed syndecan-1 that potentiates interaction of VEGF with its high affinity signaling receptor leading to enhanced endothelial cell invasion and angiogenesis.^{9,11,13,20–23} Shed syndecan-1 in addition to presenting VEGF to endothelial cells can also activate ανβ₃ integrin, a key regulator of endothelial cell activation and angiogenesis.²⁰ Furthermore, via nonenzymatic activity, heparanase can stimulate upregulation of VEGF, activation of intracellular signaling molecules and endothelial cell invasion and migration, key early steps in angiogenesis.^15,24 

Given the key roles of heparanase in inflammation and angiogenesis, we hypothesized that heparanase may be involved in the pathogenesis of PDR. To test this hypothesis, we investigated the expression of heparanase in the vitreous fluid and epiretinal membranes from patients with PDR and correlated heparanase levels with the levels of the soluble (shed) transmembrane heparan sulphate proteoglycan syndecan-1 and the angiogenic factor VEGF. To corroborate a functional link between heparanase and diabetes, we investigated the expression of heparanase in human retinal microvascular endothelial cells (HRMEC) following exposure to the proinflammatory cytokines tumor necrosis factor-α (TNF-α) and interleukin (IL)-1β, high-glucose to replicate hyperglycemia observed following diabetes and the hypoxia mimetic agent cobalt chloride (CoCl₂). In addition, we evaluated the effect of heparanase on HRMEC barrier function. 

Materials and Methods

Vitreous Samples and Epiretinal Membranes Specimens

Undiluted vitreous fluid samples (0.3–0.6 mL) were obtained from 33 patients with PDR during pars plana vitrectomy. The indications for vitrectomy were tractional retinal detachment, and/or non-clearing vitreous hemorrhage. The control group consisted of 27 patients who had undergone vitrectomy for the treatment of rhegmatogenous retinal detachment with no proliferative vitreoretinopathy. Controls were free from systemic disease. Vitreous samples were collected undiluted by manual suction into a syringe through the aspiration line of vitrectomy, before opening the infusion line. The samples were centrifuged (700g for 10 minutes, 4°C) and the supernatants were aliquoted and frozen at −80°C until assay. In addition, paired serum samples were obtained from 16 patients with PDR. Epiretinal fibrovascular membranes were obtained from 16 patients with PDR during pars plana vitrectomy for the repair of tractional retinal detachment. Membranes were fixed for 2 hours in 10% formalin solution and embedded in paraffin. 

The study was conducted according to the tenets of the Declaration of Helsinki. All the patients were candidates for vitrectomy as a surgical procedure. All patients signed a preoperative informed written consent and approved the use of the excised epiretinal membranes and vitreous fluid for further analysis and clinical research. The study design and the protocol were approved by the Research Centre and Institutional Review Board of the College of Medicine, King Saud University. 

Cell Culture

Primary HRMECs were purchased from Cell Systems Corporation (Kirkland, WA, USA) and maintained in complete serum-free media (Cat. No. SF-4Z0-500, Cell System Corp.) supplemented with recombinant RocketFuel (Cat No. SF-4ZO-500, Cell System Corp.), CultureBoost (Cat. No. 4CB-500, Cell System Corp.) and antibiotics (Cat. No. 4ZO-643, Cell System Corp.) at 37°C in a humidified atmosphere with 5% CO₂. We used HRMEC cells up to passage 8 for all the experiments. Approximately 80% confluent HRMECs were starved in medium without growth factors (minimal media; RocketFuel and CultureBoost) overnight to eliminate any residual effects of growth factors. Following starvation, HRMEC Cells were either left untreated or treated either with 30 ng/mL TNF-α (Cat. No. 210-TA-020, R&D Systems, Minneapolis, MN, USA), 10 ng/mL IL-1β (Cat. No. 201-LB, R&D Systems), TNF-α plus IL-1β, or 100 μM cobalt chloride (CoCl₂; AVONCHEM Ltd., Macclesfield, Cheshire, UK) for 24 hours or with 30 mM glucose (Scharlau Chemie, Sentmenat, Spain) or 30 mM mannitol (Scharlau Chemie; as an osmotic control) for 72 hours. Each experiment was done at least three times in duplicate. Agonist concentrations were based on literature data and on preliminary experiments. 

Enzyme-Linked Immunosorbent Assay

An enzyme-linked immunosorbent assay kit for human heparanase (Cat No: KA1230) was purchased from Abnova GmbH, Heidelberg, Germany. We purchased ELISA kits for human VEGF (Cat No: SVE00) and human syndecan-1 (Cat No: DY2780) from R&D Systems. The minimum detection limit for heparanase and VEGF ELISA kits were 1 U/L and 9 pg/mL, respectively. The plate readings of ELISA were done using a microplate reader (Stat Fax-4200; Awareness Technology, Inc., Palm City, FL, USA). The quantification of human heparanase, VEGF, and syndecan-1 in vitreous fluids was determined using ELISA kits according to the manufacturer's instructions. 

Western Blot Analysis

After harvesting, HRMEC were lysed with RIPA buffer (50 mM Tris/HCl [pH 7.5]), 150 mM NaCl, 1% ([vol/vol] Nonidet P40 [NP40]), 0.5% sodium deoxycholate, 0.1% SDS, and protease inhibitor “complete without EDTA” (Roche, Mannheim, Germany). Whole Cell extracts from different groups were centrifuged at 12,000g for 10 minutes at 4°C. Protein concentrations in the supernatants were measured using DC protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA). Equal amounts of protein (50 μg) were subjected to SDS-PAGE with 10% gel and transferred onto a nitrocellulose membrane (Bio-Rad Laboratories). To confirm the expression of heparanase in the vitreous samples, equal volumes of vitreous samples were boiled in Laemmli's sample buffer (1:1, vol/vol) under reducing conditions for 10 minutes. separated on 10% SDS-PAGE and transferred onto nitrocellulose membranes.² Nonspecific binding sites were blocked (1.5 hours, room temperature) with 5% nonfat milk made in Tris-buffered saline containing 0.1% Tween-20. Blots were then incubated at 40°C overnight with antiheparanase 1 (HPA1; 1:500; sc-25826; Santa Cruz Biotechnology, Inc., Santa Cruz, TX, USA) as previously described.² 

Transendothelial Electrical Resistance (TER) using Electric Cell-substrate Impedance Sensing (ECIS)

Normalized TER was measured using ECIS in HRMEC treated with 12/15 HETEs (0.1 μM) as previously described.²⁵ Briefly, 96W20idf arrays were used. These arrays were coated with cysteine for 30 minutes then with gelatin for 30 minutes before seeding the HRMEC at a density of 10⁵/well in 300 μL of complete media. Cells were left undisturbed until fully attached and forming a confluent monolayer to arrays indicated by capacitance below 20 nF. Cells were then starved for 6 hours in 200 μL serum-free media and treated with vehicle or heparanase (100 ng/mL, Cat. No. 7570-GH-005, R&D systems). Different treatments were prepared (as 3X in serum-free media) and added to the corresponding wells in 100 μL without removing the existing media covering the confluent monolayer of HRMEC. The electric currents, passing through the confluent monolayer of HRMEC, were measured and recorded independently in each well by the ECIS machine (Applied Biophysic, Inc., Troy, NY, USA) at current frequency 64,000 Hz. Then the data were normalized as the ratio of measured resistance at each time point to baseline resistance and plotted as a function of time. 

Immunohistochemical Staining

For heparanase 1 and CD31, antigen retrieval was performed by boiling the sections in citrate based buffer (pH 5.9–6.1; BOND Epitope Retrieval Solution 1; Leica Biosystems, Buffalo Grove, IL, USA) for 10 minutes. For detection of CD45, antigen retrieval was performed by boiling the sections in Tris/EDTA buffer (pH 9, BOND Epitope Retrieval Solution 2; Leica Biosystems) for 20 minutes. Subsequently, the sections were incubated for 60 minutes with mouse monoclonal anti-CD31 (ready-to-use, clone JC70A; Dako Denmark A/S, Glostrup, Denmark), mouse monoclonal anti-CD45 (ready-to-use, clones 2B11+PD7/26; Dako Denmark A/S) and rabbit polyclonal anti-heparanase 1 antibody (1:500; ab109696, Abcam, Cambridge, UK). Optimal working conditions for the antibodies were determined in pilot experiments on placenta and glioblastoma sections. The sections were then incubated for 20 minutes with a post-primary IgG linker followed by an alkaline phosphatase conjugated polymer. The reaction product was visualized by incubation for 15 minutes with the fast red chromogen, resulting in bright-red immunoreactive sites. The slides were then faintly counterstained with Mayer's hematoxylin (BOND Polymer Refine Red Detection Kit; Leica Biosystems). 

To identify the phenotype of cells expressing heparanase 1, sequential double immunohistochemistry was performed. The sections were incubated with the first primary antibody (anti-CD45) and subsequently treated with peroxidase conjugated secondary antibody. The sections were visualized with 3, 3′-diaminobenzidine tetrahydrochloride. Incubation of the second primary antibody (anti-heparanase 1) was followed by treatment with alkaline phosphatase conjugated secondary antibody. The sections were visualized with fast red. No counterstain was applied. 

Omission or substitution of the primary antibody with an irrelevant antibody from the same species (rabbit monoclonal anti-human estrogen receptor α; ready-to-use; clone EPI, Dako Denmark A/S) and staining with chromogen alone were used as negative controls. 

Quantitation

Immunoreactive blood vessels and cells were counted in five representative fields, using an eyepiece calibrated grid in combination with the ×40 objective. These representative fields were selected based on the presence of immunoreactive blood vessels and cells. With this magnification and calibration, immunoreactive blood vessels and cells present in an area of 0.33 × 0.22 mm were counted. 

Statistical Analysis

Data are presented as the mean ± standard deviation. The nonparametric Mann-Whitney test was used to compare means from two independent groups. The χ² test was used for comparing proportions when analyzing data for two categorical variables. Pearson correlation coefficients were computed to investigate correlation between variables. A value of P < 0.05 indicated statistical significance. We used commercial software (SPSS version 20.0 for Windows; IBM Corp., Chicago, IL, USA) for statistical analysis. 

Results

ELISA Levels of Heparanase, Syndecan-1, and VEGF in Vitreous Samples

With the use of ELISA, we demonstrated that heparanase was detected in 6 of 27 (22.2%) vitreous samples from nondiabetic control patients (range, 1.9–3.1 U/L), as well as in 29 of 33 (87.5%) samples from patients with PDR (range, 7.6–80.7 U/L; P < 0.0001; χ² test). The mean heparanase level in vitreous samples from PDR patients (17.0 ± 15.6 U/L) was significantly higher than the mean level in nondiabetic control patients (2.32 ± 0.49 U/L, P < 0.0001; Mann-Whitney test; Fig. 1). Heparanase was detected in all serum samples from patients with PDR (n = 16). The mean heparanase level in vitreous samples from patients with PDR (n = 16; 23.6 ± 18.7 U/L) was significantly higher than that in paired serum samples (8.9 ± 2.9 U/L, P < 0.001; Mann-Whitney test; Fig. 2). Syndecan-1 was detected in 18 of 22 (81.8%) vitreous samples from nondiabetic control patients (range, 107–311 pg/mL), and in 31 out of 32 (96.9%) samples from patients with PDR (range, 121–1285 pg/mL; P = 0.160; χ² test). Mean syndecan-1 level in vitreous samples from PDR patients (314.6 ± 231.6 pg/mL) was significantly higher than the mean level in nondiabetic control patients (149.5 ± 98.4 pg/mL, P < 0.0001; Mann-Whitney test; Fig. 1). Vascular endothelial growth factor was detected in 11 of 22 (50%) vitreous samples from nondiabetic control patients (range, 4–60 pg/mL), and in all vitreous samples from patients with PDR (n = 24; range, 30–3527 pg/mL; P = 0.0003; χ² test). Mean VEGF level in vitreous samples from PDR patients (650.6 ± 986.3 pg/mL) was significantly higher than the mean level in nondiabetic control patients (25.6 ± 59.9 pg/mL, P < 0.0001; Mann-Whitney test; Fig. 1). 

Figure 1

 

Comparison of mean heparanase, syndecan-1, and VEGF levels between PDR patients and control patients. *The difference between the two means was statistically significant at 5% level of significance.

Figure 2

 

Detectable levels of heparanase in paired serum and vitreous fluid samples from 16 patients with proliferative diabetic retinopathy.

Correlations

Significant positive correlations were found between vitreous fluid levels of heparanase and levels of syndecan-1 (r = 0.75; P < 0.0001) and VEGF (r = 0.91; P < 0.0001). A significant positive correlation was observed between vitreous fluid levels of syndecan-1 and VEGF (r = 0.78; P < 0.0001; Fig. 3). As is usually the case in clinical studies of patients with diverse genetic backgrounds, we also observed outliers. To ascertain that these outliers did not skew the correlations, we performed correlation analyses after omitting the outliers. Significant positive correlations were still found between vitreous fluid levels of heparanase and levels of syndecan-1 (r = 0.62; P < 0.0001) and VEGF (r = 0.79; P < 0.0001) and between vitreous fluid levels of syndecan-1 and VEGF (r = 0.75; P < 0.0001). 

Figure 3

 

Significant positive correlations between vitreous fluid levels of heparanase and levels of (A) soluble syndecan-1, and (B) VEGF, and (C) between vitreous fluid levels of soluble syndecan-1 and levels of VEGF.

Western Blot Analysis of Vitreous Samples

With the use of Western blot analysis, we demonstrated that heparanase 1 protein migrated as two protein bands on SDS-PAGE when immunoblotted and analyzed with a specific antibody. The upper band (65 kDa) corresponded to the proenzyme, whereas the lower protein band corresponded to the activated enzyme (50 kDa; Fig. 4). 

Figure 4

 

The expression of heparanase 1 in vitreous samples from patients with PDR and control patients without diabetes (C) was determined by Western blot analysis. Heparanase 1 protein migrated as two protein bands on SDS-PAGE when immunoblotted and analyzed with the specific antibody. The upper band corresponded to the proenzyme (65 kDa), whereas the lower protein band corresponded to the activated enzyme (50 kDa). A representative set of samples is shown. R, recombinant heparanase protein (Abnova GmbH, Heidelberg, Germany).

Immunohistochemical Analysis of Epiretinal Membranes

To identify possible cell sources of vitreous fluid heparanase, epiretinal membranes from patients with PDR were studied by immunohistochemical analysis. No staining was observed in the negative control studies (Fig. 5A). The level of vascularization and proliferative activity in epiretinal membranes were determined by immunodetection of the endothelial cell maker CD31. All membranes showed blood vessels that were positive for the endothelial cell marker CD31 (Fig. 3B) with a mean of 37.6 ± 22.9 (range, 15–95). Strong immunoreactivity for heparanase 1 was present in all membranes and was noted in stromal cells, intravascular leukocytes and vascular endothelial cells (Figs. 5C, 5D). The majority of heparanase 1-positive stromal cells were monocytes/macrophages and neutrophils. In serial sections, the distribution and morphologies of stromal cells expressing heparanase 1 were similar to those of cells expressing the leukocyte common antigen CD45 (Fig. 5E). Double staining confirmed that stromal cells and intravascular leukocytes (Figs. 5F–H) expressing heparanase 1 coexpressed CD45. The number of blood vessels that were immunoreactive for heparanase 1 ranged from 14 to 95, with a mean of 35.3 ± 19.1. The number of heparanase 1–positive immunoreactive stromal cells ranged from 16 to 220, with a mean of 73.4 ± 59.3. Significant positive correlations were detected between the numbers of blood vessels expressing CD31 and the numbers of blood vessels (r = 0.72; P = 0.013) and stromal cells (r = 0.69; P = 0.019) expressing heparanase 1. 

Figure 5

 

Proliferative diabetic retinopathy epiretinal membranes immunostainings. (A) Negative control slide that was treated with an irrelevant antibody showing no labeling (original magnification ×40). (B) Immunohistochemical staining for CD31 (original magnification ×40). (C) Immunohistochemical staining for heparanase 1 showing immunoreactivity in intravascular leukocytes (arrows; original magnification ×25), in the vascular endothelium (arrows) and in (D) stromal cells (arrowheads; original magnification ×40). (E) Immunohistochemical staining for CD45 showing stromal cells positive for CD45 (original magnification, ×40). (F–H) Double immunohistochemistry for CD45 (brown) and heparanase 1 (red) showing stromal cells and intravascular leukocytes (arrows) coexpressing CD45 and heparanase 1 (original magnification ×40).

High-Glucose and the Proinflammatory Cytokines TNF-α and IL-1β, but not the Hypoxia Mimetic Agent CoCl₂ Induce Upregulation of Heparanase in HRMEC

To confirm the observed expression of heparanase by endothelial cells, and to define possible heparanase regulatory molecules, we performed in vitro experiments on HRMEC with inducers relevant in the context of PDR: TNF-α, IL-1β, TNF-α plus IL-1β or CoCl₂ for 24 hours or high-glucose and mannitol for 72 hours. We found that TNF-α, TNF-α plus IL-1β and high-glucose treatment significantly increased the expression of heparanase compared with untreated control (Fig. 6). However, IL-1β, mannitol (Fig. 6) and CoCl₂ (data not shown) treatment did not affect the expression of heparanase compared with untreated control. 

Figure 6

 

Human retinal microvascular endothelial cells were left untreated or treated either with (A) 30 ng/mL TNF-α, 10 ng/mL IL-1β or TNF-α plus IL-1β for 24 hours or with (B) 30 mM glucose or 30 mM mannitol for 72 hours. Western blots are representative of at least three different experiments, each is performed in duplicate and bar graph is representative of all three experiments. * The difference between the two means was statistically significant at the 5% level. NS, not significant.

Effect of Heparanase on HRMEC Barrier Function

Given the prominent increase in levels of heparanase under hyperglycemic conditions, next we proceeded to investigate its propensity to disrupt barrier function in HRMEC monolayers using real-time analysis of TER, an indicator of monolayer integrity. An increase in the endothelial permeability is accompanied with reduction in TER. Significant decreases in TER were first observed after 50 hours of heparanase treatment compared with vehicle treatment and this effect continued throughout the experiment period (100 hours; Fig. 7). 

Figure 7

 

Heparanase reduces TER in HRMECs. Monolayers of HRMECs were treated with vehicle or herapanase (100 ng/mL) and changes in TER were monitored. * P < 0.05 versus control; n = 5.

Discussion

Heparanase has been implicated in the pathology of diabetes-associated complications.^10,27–29 In the present study, we showed for the first time that heparanase was significantly upregulated in the vitreous fluid from patients with PDR and that heparanase protein was specifically localized in vascular endothelial cells and leukocytes expressing the leukocyte common antigen CD45 in epiretinal fibrovascular membranes from patients with PDR. In addition, heparanase levels in the vitreous fluid from patients with PDR were significantly higher than those in paired serum samples. These findings suggest that local cellular production is the relevant source of heparanase within the ocular microenvironment and that systemic inflow mechanism is rather improbable. Many pathogenic factors in PDR, direct and indirect ones, can contribute to upregulation of heparanase. To corroborate the findings at the cellular level, stimulation with high-glucose to replicate hyperglycemia observed following diabetes caused upregulation of heparanase in HRMEC. Similarly, previous studies demonstrated that high-glucose is a potent stimulator of endothelial heparanase secretion.^26,30–32 In addition to hyperglycemia, the inflammatory ocular microenvironment of patients with PDR may further enhance heparanase production in endothelial cells. In our study, we confirmed that heparanase expression was upregulated in HRMEC treated with the proinflammatory cytokines TNF-α and TNF-α plus IL-1β. These findings are consistent with previous studies.^16,33 On the other hand, the hypoxia mimetic agents CoC1₂ did not affect the expression of heparanase. Taken together, these findings suggest that hyperglycemia and inflammation, but not hypoxia might be involved in diabetes-induced heparanase upregulation, which might be a chief pathophysiologic mechanism for DR It should be emphasized that heparanase and syndecan-1 were detected in 22.2% and 81.8%, respectively, of vitreous samples from patients with rhegmatogenous retinal detachment with no proliferative vitreoretinopathy. These findings are consistent with previous reports that demonstrated upregulated expression of inflammatory mediators in the vitreous fluid from patients with rhegmatogenous retinal detachment,^34–36 and in the detached retina following experimental retinal detachment.³⁷ 

Current data show that in addition to its well characterized role in cancer, heparanase activity may represent an important determinant in the pathogenesis of several inflammatory disorders.^11–13 In an animal model of delayed-type hypersensitivity inflammatory reaction, heparanase expression is upregulated at the site of inflammation and is produced locally by endothelial cells. Increase in heparanase levels causes vascular leakage through degradation of heparan sulphate chains,¹⁶ responsible for the structural integrity of the subendothelial basement membrane.¹⁴ Similarly, upregulation of heparanase in the ocular microenvironment of patients with PDR might be implicated in degradation of the subendothelial basement membrane and subsequent vascular leakage—a hallmark of DR To corroborate the findings at the cellular level, stimulation with heparanase induced increase in HRMEC permeability. In addition, a role for heparanase in inflammatory cell trafficking was demonstrated.¹⁷ 

The angiogenic potency of heparanase has been confirmed in several in vitro and in vivo model systems providing a strong clinical evidence for the proangiogenic function of heparanase.^{9,11,13,15,21,38} Using immunohistochemistry, we demonstrated a significant positive correlation between the level of vascularization in PDR epiretinal membranes and the number of blood vessels and stromal cells expressing heparanase. In addition, we found a significant positive correlation between the vitreous fluid levels of heparanase and those of the angiogenic biomarker VEGF, a key angiogenic factor in PDR.¹ This finding is consistent with previous studies that demonstrated active involvement of heparanase via its nonenzymatic functions in the upregulation of VEGF expression in several cell lines.^9,10,15,38 Additionally, degradation of heparan sulphate by heparanase liberates various ECM-resident bioactive angiogenic factors such as VEGF.^9,10,12,13 Substantial evidence supports the association between inflammation and angiogenesis.^7,8 Inflammatory conditions are present in the ocular microenvironment of patients with PDR.^2,4–6 Leukocytes in fibrovascular epiretinal membranes from patients with PDR supply proinflammatory and proangiogenic molecules that foster angiogenesis.^2,6 In the present study, we demonstrate the presence of leukocytes expressing heparanase suggesting that inflammation-stimulated heparanase upregulation in PDR may be mechanistically involved in coupling inflammation and angiogenesis and contributes to PDR initiation and progression. In the current study, we found a significant correlation between the vitreous levels of heparanase and that of syndecan-1. This finding is consistent with previous studies that reported enhanced shedding of the transmembrane heparan sulphate proteoglycan syndecan-1 from the surface of cancer cells induced by the enzymatic activity of heparanase stimulation.^20–23 Mechanistically, this occurs at least in part by heparanase-mediated activation of extracellular signal-regulated kinase signaling that leads to increased expression of matrix metalloproteinase-9 (MMP-9) expression, which then acts as sheddase of syndecan-1.²² In a previous report, we demonstrated increased levels of MMP-9 in the vitreous fluid from patients with PDR.⁶ The heparanase-mediated upregulation of VEGF in cancer cells, together with the enhanced level of syndecan-1 shedding, stimulates endothelial cell invasion and angiogenesis.^20,23 Taken together, our findings suggest that upregulation of the heparanase/syndecan-1 axis within the PDR microenvironment might contribute to PDR angiogenesis and progression. 

In conclusion, our findings suggest a potential link between heparanase, syndecan-1, and VEGF in the progression of PDR and that heparanase is a candidate molecule linking inflammation and angiogenesis in PDR. The elucidation of specific effects of heparanase in PDR progression will accelerate the development of complementary therapeutic interventions designed to disrupt the heparanase/syndecan-1/VEGF axis, in particular in patients who are refractory to anti-VEGF therapy. 

Acknowledgments

The authors thank Mohamed Al-Shabrawey, MD, PhD, and Ahmed Ibrahim, PhD, Department of Oral Biology, Cellular Biology and Anatomy, Culver Vision discovery Institute and Department of Ophthalmology, Georgia Regents University (Augusta, GA, USA) for performing the transendothelial electrical resistance experiment; Wilfried Versin and Nathalie Volders for technical assistance; Michelle B. Rasonabe and Connie Unisa-Marfil for secretarial work. 

Supported by Dr. Nasser Al-Rashid Research Chair in Ophthalmology (AMAE), the Concerted Research Actions of the Regional Government of Flanders (GOA 2013/014) and the Fund for Scientific Research of Flanders (FWO-Vlaanderen; GO). 

Disclosure: A.M. Abu El-Asrar, None; K. Alam, None; M.I. Nawaz, None; G. Mohammad, None; K. Van den Eynde, None; M.M. Siddiquei, None; A. Mousa, None; G. De Hertogh, None; K. Geboes, None; G. Opdenakker, None