Heparanase inhibition as a systemic approach to protect the endothelial glycocalyx and prevent microvascular complications in diabetes

Abstract

Background: Diabetes mellitus is a chronic disease which is detrimental to cardiovascular health, often leading to secondary microvascular complications, with huge global health implications. Therapeutic interventions that can be applied to multiple vascular beds are urgently needed. Diabetic retinopathy (DR) and diabetic kidney disease (DKD) are characterised by early microvascular permeability changes which, if left untreated, lead to visual impairment and renal failure, respectively. The heparan sulphate cleaving enzyme, heparanase, has previously been shown to contribute to diabetic microvascular complications, but the common underlying mechanism which results in microvascular dysfunction in conditions such as DR and DKD has not been determined.

Methods: In this study, two mouse models of heparan sulphate depletion (enzymatic removal and genetic ablation by endothelial specific Exotosin-1 knock down) were utilized to investigate the impact of endothelial cell surface (i.e., endothelial glycocalyx) heparan sulphate loss on microvascular barrier function. Endothelial glycocalyx changes were measured using fluorescence microscopy or transmission electron microscopy. To measure the impact on barrier function, we used sodium fluorescein angiography in the eye and a glomerular albumin permeability assay in the kidney. A type 2 diabetic (T2D, db/db) mouse model was used to determine the therapeutic potential of preventing heparan sulphate damage using treatment with a novel heparanase inhibitor, OVZ/HS-1638. Endothelial glycocalyx changes were measured as above, and microvascular barrier function assessed by albumin extravasation in the eye and a glomerular permeability assay in the kidney.

Results: In both models of heparan sulphate depletion, endothelial glycocalyx depth was reduced and retinal solute flux and glomerular albumin permeability was increased. T2D mice treated with OVZ/HS-1638 had improved endothelial glycocalyx measurements compared to vehicle treated T2D mice and were simultaneously protected from microvascular permeability changes associated with DR and DKD.

Conclusion: We demonstrate that endothelial glycocalyx heparan sulphate plays a common mechanistic role in microvascular barrier function in the eye and kidney. Protecting the endothelial glycocalyx damage in diabetes, using the novel heparanase inhibitor OVZ/HS-1638, effectively prevents microvascular permeability changes associated with DR and DKD, demonstrating a novel systemic approach to address diabetic microvascular complications.

Fig. 1

HS forms part of the eGlx barrier in retinal and glomerular microvasculature. a Left: low magnification representative TEM image of retinal vessel from mouse that had been cardiac perfused with Alcian blue and glutaraldehyde. Lumen (Lm), retinal endothelial cell (REC), and pericyte (Pr) indicated. Right: high magnification image showing eGlx staining on luminal side of REC (arrow heads). Basement membrane (BM) indicated. REC vesicles (*) were also visible. b Retinal flat mounts stained with anti-HS in green. DAPI in blue. Retinal endothelial cell (REC) and lumen (Lm) indicated. Note the green signal on luminal side of REC nucleus (arrowhead), indicating presence of HS in the eGlx. c Left: low magnification representative EM image of glomerular capillary from mouse that had been cardiac perfused with Alcian blue and glutaraldehyde. Lumen (Lm) and glomerular filtration barrier (GFB) indicated. Right panel: High magnification image showing eGlx staining (arrow heads) on luminal glomerular endothelial cell (GEnC). Glomerular basement membrane (GBM) and podocyte (P) indicated. d Mouse kidney tissue stained with anti-HS in green, DAPI in blue, membranes in red (R18). Inset shows zoomed in image of eGlx HS staining (white arrow head) on luminal (LM) side of vessel. *GBM.

Fig. 2

Enzymatic removal of HS results in reduced eGlx and loss of barrier function in the retina and glomerulus. a Representative TEM images of retinal vessels from mice perfused with inactive or active heparinase III and then Alcian Blue to visualise eGlx and glutaraldehyde for EM preparation. Basement membrane (BM) and retinal endothelial cell (E) indicated. Arrowheads point to eGlx on endothelial cell. b Glycocalyx depth measured. Mouse averages shown (n = 6 mice (inactive) and n = 7 mice (active), unpaired t-test indicated, *P < 0.05). c Representative images of lectin-stained retina sections from mice treated with inactive or active heparinase III. Lectin (FITC-LEL) staining in green, R18 cell membrane stain in red, and DAPI in blue. Inset images show green signal on the luminal side of the vessel, indicating eGlx staining. d Example of confocal fluorescence peak-to-peak (P-P) measurement for measurements of eGlx depth using LEL and cell membrane (R18) staining. e Confocal fluorescence profile peak-to-peak measurements in capillaries were taken for a minimum of three vessels per animal. Mouse averages shown (n = 6 mice/group, *P < 0.05, unpaired t-test indicated). f Image of mouse retina perfused with sodium fluorescein angiography. Within the selected ROI, the area in main vessel (V₁) and in the tissue next to vessel (V₂) measured over time to calculate solute flux. g Solute flux after treatment with inactive or active enzyme. (n = 6 mice/group, *P < 0.05, unpaired t-test indicated) h Representative glomerular filtration barrier TEM images of glomerular capillary in inactive and active heparinase III treated mice showing podocyte (P), podocyte slit diaphragm (SD), endothelial cell (E), basement membrane (BM), podocyte Glx (open arrowhead), and eGlx (solid arrowhead). i Quantification of EM images measuring eGlx depth (n = 5 mice/group,*P < 0.05, unpaired t-test). j Glomerular albumin permeability (Ps’alb) measured for both groups. Glomeruli analysed shown on graph and in parentheses. Stats performed on mouse number (n = 5 mice/group,*P < 0.05, unpaired t-test).

Fig. 3

Conditional knockout of the HS polymerizing enzyme, Ext1, in endothelial cells leads to reduced eGlx and compromised microvascular barrier function. a Schematic of experimental timeline. Age and sex matched littermate controls (LMCs) or Ext1 endothelial specific conditional knockout mice (Ext1^(ECKO)) were given Doxycycline water for three weeks to induce knock-out. At three weeks, fluorescein angiographies were performed for solute flux measurements and urine analysis experiments conducted. Eye and kidney tissue was collected and analysed after three weeks of Doxycycline treatment. b Representative images of lectin-stained retinal tissue from LMC and Ext1^(ECKO) mice. Lectin staining in green, R18 cell membrane stain in red, and DAPI in blue. Inset images show green signal on luminal side of the vessel, indicating eGlx staining. c Confocal fluorescence profile peak-to-peak measurements in capillaries were taken for a minimum of three vessels per animal. Mouse averages shown (n = 6, *P < 0.05, unpaired t-test). d Solute flux measured in LMC (n = 8 mice) and Ext1^(ECKO) (n = 5 mice) after three weeks of doxycycline treatment (**P < 0.01, unpaired t-test). e Representative glomerular filtration barrier TEM images of litter mate control (LMC) and Ext1 endothelial specific conditional knockout mice (Ext1^(ECKO)) showing podocyte (P), podocyte slit diaphragm (SD), endothelial cell (E), basement membrane (BM), podocyte Glx (open arrow heads), and eGlx (solid arrow heads). f Quantification of TEM images measuring eGlx depth (n = 6 mice,**P < 0.01, unpaired t-test). g End point urine albumin creatinine ratios (uACR) for LMC (n = 10) and Ext1^(ECKO) (n = 7) mice (not significant, P = 0.24, unpaired t-test). h Fold change uACR from pre and post L-lysine (Lys) treatment in LMC (n = 12) and Ext1^(ECKO) (n = 9) mice (*P < 0.05, unpaired t-test). i Glomerular albumin permeability (Ps’alb) measured for LMC (n = 6) and Ext1^(ECKO) (n = 5 mice). Number of glomeruli analysed shown on graph and in parentheses. Stats performed on mouse number (***P < 0.001, unpaired t-test).

Fig. 4

Treatment with the heparanase inhibitor OVS/HS-1638 in db/db mice prevents systemic eGlx damage and associated microvascular permeability changes. a Schematic of experimental timeline. Lean or db/db mice were treated with vehicle or OVZ/HS-1638 daily for 14-days by i.p. End-point urine was collected for analysis and eye and kidney tissue collected for analysis. b Retinas from lean and db/db mice treated with vehicle (db/db) or with OVZ/HS-1638 (db/db + HI) were stained with FITC-LEL (green), R18 cell membrane stain (red) and DAPI (blue). Inset images show green staining on luminal side of vessel, indicating eGlx staining. c Glycocalyx depth in lean and diabetic animals was measured using confocal fluorescence profile peak-to-peak (n = 5, mice per group, *P ≤ 0.05, One-way ANOVA with Tukey’s multiple comparison test). d Retinas from lean and db/db mice treated with vehicle (db/db) or with OVZ/HS-1638 (db/db + HI) were stained for albumin (red) and DAPI (blue). Arrows point to vessels filled with albumin and arrowheads point to extravascular albumin. Extravascular albumin staining found near Ganglion cell layer (GCL), inner nuclear layer (INL), and outer nuclear layer (ONL) in db/db retina. e Extravascular corrected total cell fluorescence (CTF_EXV) in arbitrary units (a.u.) (n = 5 mice per group, **P ≤ 0.01, ***P ≤ 0.001, One-way ANOVA with Tukey’s multiple comparison test). f Average CTF_EXV value for each mouse was plotted against corresponding eGlx depth for same animal (n = 15 total mice, **P ≤ 0.01, Pearson correlation analysis). g Representative TEM images of lean, diabetic (db/db) and diabetic mice treated with OVZ/HS-1638 (db/db + HI) mouse glomerular filtration barrier showing podocyte (P), podocyte slit diaphragm (SD), endothelial cell (E), basement membrane (BM), podocyte Glx (open arrow heads), and eGlx (solid arrow heads). h Quantification of TEM images measuring eGlx depth (n = 5 mice,*P < 0.05, Kruskal–Wallis test). i End point urine albumin creatinine ratio for lean (n = 5), db/db (n = 4), and db/db + HI (n = 4) (**P < 0.01, Kruskal–Wallis test). j Glomerular albumin permeability (Ps’alb) measured. Glomeruli analysed shown on graph and in parentheses. Stats performed on mouse number (n = 5, **P < 0.01, ***P < 0.001, One way ANOVA, Tukey’s multiple comparison test). (k,l) Correlation analysis. (k) Ps’alb vs eGlx depth (n = 7,**P < 0.01), (l) Ps’alb vs podocyte foot process width (n = 7, P > 0.05, not significant (ns)).