Basic Science

Endothelial Heparanase Regulates Heart Metabolism by Stimulating Lipoprotein Lipase Secretion From CardiomyocytesSignificance

Ying Wang, Dahai Zhang, Amy Pei-Ling Chiu, Andrea Wan, Katharina Neumaier, Israel Vlodavsky, Brian Rodrigues
https://doi.org/10.1161/ATVBAHA.113.301309
Arteriosclerosis, Thrombosis, and Vascular Biology. 2013;33:894-902
Originally published April 10, 2013

Jump to

Abstract

Objective—After diabetes mellitus, transfer of lipoprotein lipase (LPL) from cardiomyocytes to the coronary lumen increases, and this requires liberation of LPL from the myocyte surface heparan sulfate proteoglycans with subsequent replenishment of this reservoir. At the lumen, LPL breaks down triglyceride to meet the increased demand of the heart for fatty acid. Here, we examined the contribution of coronary endothelial cells (ECs) toward regulation of cardiomyocyte LPL secretion.

Approach and Results—Bovine coronary artery ECs were exposed to high glucose, and the conditioned medium was used to treat cardiomyocytes. EC-conditioned medium liberated LPL from the myocyte surface, in addition to facilitating its replenishment. This effect was attributed to the increased heparanase content in EC-conditioned medium. Of the 2 forms of heparanase secreted from EC in response to high glucose, active heparanase released LPL from the myocyte surface, whereas latent heparanase stimulated reloading of LPL from an intracellular pool via heparan sulfate proteoglycan–mediated RhoA activation.

Conclusions—Endothelial heparanase is a participant in facilitating LPL increase at the coronary lumen. These observations provide an insight into the cross-talk between ECs and cardiomyocytes to regulate cardiac metabolism after diabetes mellitus.

Introduction

After diabetes mellitus, the heart has an increased reliance on fatty acids (FAs) for generation of energy.1 The majority of these FAs come from breakdown of circulating triglyceride-rich lipoproteins, a process catalyzed by lipoprotein lipase (LPL) located at the luminal side of vascular endothelial cells (ECs).2–4 ECs do not express LPL. In the heart, LPL is synthesized in cardiomyocytes and secreted to ECs.5 We have previously reported an increase of coronary LPL in animal models of type 1 diabetes mellitus, an effect that was evident in the absence of any change in myocyte LPL gene expression.6,7 We concluded that the augmented coronary LPL was a result of increased secretion of the enzyme from myocytes toward the coronary lumen. Regarding secretion, LPL is first transported from an intracellular pool in the myocyte to the cell surface, where it binds to heparan sulfate proteoglycans (HSPGs).8 We have shown that this intracellular transport depends on actin cytoskeleton polymerization, a process that is magnified after diabetes mellitus.9,10 The subsequent process by which myocyte surface LPL is translocated to the coronary lumen, in addition to its replenishment after this onward movement, has not been completely elucidated.

Myocyte surface HSPGs serve as a temporary docking site and an auxiliary reservoir of LPL. HSPGs are proteoglycans bearing HS side chains attached to specific serine residues of a protein core.11 Core proteins can be attached to the cell surface through a glycosylphosphatidyl inositol anchor in case of glypican, or can traverse the membrane as observed with the syndecan family.12,13 The HS side chains are polymers of repeating disaccharides which interact with multiple ligands, including antithrombin, fibroblast growth factor, and LPL.14–16 Thus, for LPL to translocate from myocytes to the vascular lumen, cleavage of myocyte surface HSPGs is required to release the sequestered LPL. It should be noted that in addition to its ligand binding property, any change in the conformation of HSPGs can also induce intracellular signals. For example, clustering of syndecan-4 can stimulate protein kinase C-α (PKCα),17 whereas shear stress results in ERK1/2 activation through HSPG-mediated mechanotransduction.18 Thus, it is possible that with cleavage of myocyte surface HSPGs and the resultant release of LPL, intracellular signals could be generated in myocytes to help reload this surface LPL pool.

Heparanase is an endoglucuronidase of special interest because it can cleave HS at low-sulfation sites, liberating sequestered ligands from surface HSPGs.19 It has also been shown that through surface HSPGs, heparanase can trigger intracellular signal pathways including Src, Akt, and p38 MAPK.20–22 Hence, heparanase could help LPL secretion by (1) releasing sequestered LPL from myocyte surface for onward movement to the vascular lumen and (2) provoking signals in myocytes to move LPL from an intracellular pool to replenish the surface reservoir. In the heart, heparanase is synthesized in ECs as a latent 65-kDa form, and processed in lysosomes to become a 50-kDa active enzyme.23 Interestingly, increased secretion of active heparanase is evident from EC in response to high glucose.24 Thus, after diabetes mellitus, when ECs are exposed to hyperglycemia, these cells could promote LPL secretion from cardiomyocytes via heparanase. The present study investigated the role of endothelial heparanase in mediating the cross-talk between EC and cardiomyocytes to increase LPL secretion after hyperglycemia.

Materials and Methods

Materials and Methods are available in the online-only Supplement.

Results

RhoA Activation Is Involved in Increasing Cardiomyocyte LPL Secretion

Animals injected with diazoxide (DZ) developed hyperglycemia within 1 hour after injection, and blood glucose remained high at 4 hours (control, 6.5±0.6; DZ, 19.2±2.7 mmol/L, P<0.01). LPL at the coronary lumen determines FA delivery to the underlying myocytes. We perfused hearts with heparin to release LPL from this location, and consistent with our previous observations,25 LPL activity was increased in hyperglycemic hearts (Figure 1A). Within 30 minutes, 4 U insulin effectively inhibited the development of hyperglycemia in DZ animals (DZ+In, 7.6±1.4; DZ, 21.5±4.2 mmol/L, P<0.01), and normal glycemia remained until the animals were killed. Insulin attenuated the increase of coronary LPL activity in these DZ hyperglycemic animals (Figure 1A). As RhoA has been reported to regulate actin cytoskeleton remodeling, an event that could affect LPL secretion,26 we tested RhoA activation in diabetic hearts. On activation, RhoA shifts from the cytosolic to the particulate fraction and binds to GTP.27 In DZ hearts, an increased RhoA in the particulate fraction was observed, which was inhibited by insulin treatment (Figure 1B). Similar to DZ animals, D55 animals also have higher coronary LPL activity, which was accompanied by RhoA activation in these hearts (Figure I in the online-only Data Supplement). To determine whether RhoA activation can induce LPL secretion from myocytes in vitro, cells were incubated with lysophosphatidic acid (LPA). After 2 hours of LPA, an increased amount of LPL was released into the medium (Figure 1C). RhoA activation by LPA was confirmed in cardiomyocyte as GTP-RhoA increased immediately 1 minute after LPA, and declined with time (Figure 1C, inset). One effect of RhoA activation is actin cytoskeleton polymerization, which we observed as an increase in the formation of F actin in the presence of LPA (Figure 1D). Given that LPL secretion relies on stress fibers to move from an intracellular pool to the myocyte surface,1,28 we inhibited actin cytoskeleton polymerization using Cy and found that the impact of LPA on LPL secretion was abolished (Figure 1E). As the effect of LPA on LPL secretion was reproducible in mycoytes in which protein synthesis was inhibited (Figure 1E, inset, right panel), our data suggest that the increased LPL secretion observed with LPA is not a consequence of augmented protein synthesis, but likely attributable to increased LPL trafficking. Thus, RhoA activation could contribute toward augmented LPL secretion, possibly via actin cytoskeleton remodeling.

Figure 1.
Figure 1.

Increased lipoprotein lipase (LPL) activity in diabetic hearts involves RhoA-mediated actin cytoskeleton remodeling. Animals were made hyperglycemic with diazoxide (DZ) and kept for 4 hours. One hour after DZ injection, some animals were given 4 U of rapid-acting insulin intravenously (DZ+In), and kept for another 3 hours. Animals were killed, hearts removed, and perfused with heparin (5 U/mL) to release coronary LPL. Coronary effluents were collected (for 10 seconds) at different time points over 5 minutes, and LPL activity in each fraction was determined. The results are presented as area under the curve (AUC) for heparin-released LPL activity over 5 minutes (A). In a separate experiment, particulate fractions were also isolated from whole heart homogenates of control (CON), DZ and DZ+In animals, and RhoA recruitment to the particulate fraction determined by Western blot. A plasma membrane protein, Na+-K+-ATPase-α subunit was used as a loading control (B). *P<0.05, compared with CON, #P<0.05, compared with DZ, n= 3. Isolated cardiomyocytes from CON were incubated with increasing concentrations of lysophosphatidic acid (LPA) for 2 hours, and LPL protein secreted into the medium measured by Western blot (C). Using 1 µmol/L LPA, RhoA activation in cardiomyocytes was determined at the indicated times using G-LISA assay (C, inset). *P<0.05, compared with CON, n=3. Myocytes were also pretreated with 1 µmol/L cytochalasin D for 30 minutes before incubation with 1 µmol/L LPA (LPA+Cy). Actin cytoskeleton polymerization (D) and LPL secretion in these cells (E) were determined after 1 and 2 hours, respectively. In a different experiment, LPL secretion in the presence of LPA was tested in myocytes preincubated with 50 μmol/L of the protein synthesis inhibitor cycloheximide (CHX; E, inset, right) *P<0.05, **P<0.01, compared with CON, #P<0.05, compared with LPA, n= 3 to 4.

ECCM Stimulates LPL Secretion From Cardiomyocytes

After diabetes mellitus, ECs are the first cells exposed to hyperglycemia and could potentially release multiple factors affecting cardiomyocyte metabolism. For this reason, we incubated myocytes with high glucose–treated EC–conditioned medium (ECCM). Interestingly, ECCM released myocyte LPL within 30 minutes, an effect that was more significant after 1 hour. High glucose itself had no impact on LPL (Figure 2A). As this increase in medium LPL was accompanied by a reciprocal decrease in LPL activity remaining at the myocyte surface (Figure 2A, inset, upper panel), we concluded that ECCM is capable of releasing myocyte surface LPL. Importantly, this ECCM-released LPL was catalytically active and able to breakdown exogenous very low density lipoprotein-triglyceride to FA (Figure 2A, inset, bottom panel). In addition to its ability to release LPL, we also tested whether ECCM could stimulate the replenishment of LPL when the enzyme at the myocyte surface is depleted. Myocytes were pretreated with ECCM or heparin (10 mU/mL), followed by a bolus dose of heparin (8 U/mL) to deplete surface LPL, as confirmed in Figure 2B (left panel). After a 2-hour recovery, myocytes pretreated with ECCM were able to recruit significantly more LPL activity to the surface (Figure 2B, right panel), an effect not observed with 10 mU/mL heparin. Considering that RhoA is involved in augmented LPL secretion from myocytes, we tested the effect of ECCM on RhoA activation. An increase in GTP-bound RhoA was observed within 5 minutes in response to ECCM, reaching a peak at 15 minutes, and declined to basal levels after 30 minutes (Figure 3A). We did not see a similar effect with 25 mmol/L glucose or 10 mU/mL heparin (data not shown). The downstream effect of RhoA activation, actin cytoskeleton polymerization, was also augmented with ECCM, an effect that was abolished in the presence of the Rho-associated protein kinase inhibitor Y-27632 (Figure 3B). Hence, our data suggested that ECCM can stimulate both the release, and likely through RhoA-mediated actin cytoskeleton polymerization, replenishment of LPL at the myocyte surface.

Figure 2.
Figure 2.

Endothelial cell–conditioned media (ECCM) increases lipoprotein lipase (LPL) secretion from cardiomyocytes. ECCM was applied to cardiomyocytes, and at the indicated times, LPL activity released into the medium determined. Cardiomyocytes were also incubated with 5.5 mmol/L glucose (normal glucose [NG]) and 25 mmol/L glucose (high glucose [HG]) as controls (CON; A). After 30 minutes incubation with NG or ECCM, LPL remaining on the myocyte surface was released by incubating cells with 8 U/mL heparin for 3 minutes, and LPL activity was determined (A, inset, top). Cardiomyocytes were treated with 5.5 mmol/L glucose DMEM (CON) or ECCM for 30 minutes, and the medium was collected. This medium was then incubated with increasing concentrations of very low density lipoprotein-triglyceride (0–0.8 mmol/L) and the concentration of released free fatty acids determined after 30 minutes (A, inset, bottom). *P<0.05, **P<0.01, compared with NG, n=3 to 5. Isolated cardiomyocytes were treated with 5.5 mmol/L glucose (CON), ECCM, or 10 mU/mL heparin for 30 minutes. After this, surface LPL were depleted by incubating these cells with 8 U/mL heparin for 3 minutes. To validate depletion of this surface LPL, 8 U/mL heparin was given and surface LPL activity determined (B, left). Following a wash with PBS, cells were allowed to recover in 5.5 mmol/L glucose DMEM for 2 hours. LPL activity recruited to the myocyte surface was then determined (B, right). *P<0.05, compared with CON, n=3.

Figure 3.
Figure 3.

Endothelial cell–conditioned medium (ECCM) induces RhoA-mediated actin cytoskeleton polymerization. Myocytes were incubated with ECCM for 5, 15, or 30 minutes, and RhoA activation was evaluated by measuring GTP-bound RhoA. Results were compared with myocytes incubated with 5.5 mmol/L glucose DMEM (control [CON], A). In a different experiment, F-to-G actin ratio was measured 1 hour after incubation with ECCM (B). Ten micromoles per liter Y-27632 was also added to myocytes 30 minutes before ECCM (ECCM+Y), and actin cytoskeleton polymerization was evaluated (B). *P<0.05, **P<0.01, compared with CON; #P<0.05, compared with ECCM, n=3 to 4.

Effect of ECCM on LPL Is Related to the Presence of Heparanase

As LPL at the cardiomyocyte surface resides on HSPGs, its release by ECCM could be a consequence of cleavage of these binding sites by heparanase. As anticipated, ECCM contained a higher amount of both latent (65 kDa) and active (50 kDa) heparanase (Figure 4A), with a reciprocal decrease in the intracellular content of this enzyme (Figure IIIC in the online-only Data Supplement). The increase in active heparanase protein mirrored the higher heparanase activity in this medium (Figure II in the online-only Data Supplement). In vivo, both latent and active heparanase also increased in the interstitial space of hearts from DZ animals (Figure 4B), where coronary ECs are exposed to high glucose. When heparanase in the ECCM was immunoprecipitated by an antiheparanase antibody, thereby reducing the amount of both latent and active heparanase (Figure 4C, inset), the LPL releasing effect of ECCM was compromised (Figure 4C). When bovine coronary artery ECs were exposed to high glucose for 2 consecutive periods of 30 minutes, the amount of heparanase released into the medium diminished during the second incubation (Figure IIIA in the online-only Data Supplement). In addition, the ability of this ECCM from the second incubation to release myocyte surface LPL also decreased (Figure IIIB in the online-only Data Supplement).

Figure 4.
Figure 4.

Latent and active heparanase have divergent effects on lipoprotein lipase (LPL) secretion. Bovine coronary artery endothelial cells (bCAECs) were incubated with 5.5 mmol/L (normal glucose [NG]), or 25 mmol/L (high glucose [HG]) glucose DMEM for 30 minutes. The amount of latent and active heparanase released into the medium was determined by Western blot (A). Control (CON) and diazoxide (DZ) hearts were perfused using a modified Langendorff perfusion technique to separate interstitial fluid from coronary perfusate. Latent (65 kDa) and active heparanase (50 kDa) were measured by Western blot in the interstitial fluid (B). Heparanase was immunoprecipitated from ECCM by an antiheparanase antibody and the amount of latent and active heparanase remaining in this ECCM was determined and compared with the original ECCM using Western blot (C, inset). This immunoprecipitated ECCM (IP-ECCM) was applied to myocyte for 30 minutes, and LPL activity was released into medium tested (C). *P<0.05, compared with NG or CON; #P<0.05, compared with ECCM, n=3 to 4. Isolated cardiomyocytes were incubated with medium 199 (CON) and 1 µg/mL purified active (A-HEPA) or latent heparanase (L-HEPA) for 30 minutes, and LPL activity was released into the medium determined (D). In a different experiment, myocytes were preincubated with 8 U/mL heparin for 3 minutes before 1 µg/mL purified active heparanase (Hep+A-HEPA), and LPL activity was released after 30 minutes determined (E). Results were compared with CON and myocytes without heparin preincubation (A-HEPA). RhoA activation was measured in cardiomyocytes in the presence of 1 µg/mL A-HEPA or L-HEPA for 15 minutes (F). *P<0.05, **P<0.01, compared with CON; #P<0.05, compared with A-HEPA, n=3.

Latent and Active Heparanase Play Different Roles in LPL Secretion

Because high glucose stimulates the release of both latent and active heparanase from EC, we tested the roles of these 2 forms on LPL secretion using purified heparanase. Active, but not latent, heparanase caused the release of LPL (Figure 4D), and this effect of active heparanase was dose dependent (Figure IV in the online-only Data Supplement). When heparin was added to remove LPL from the myocyte surface, active heparanase was unable to release LPL (Figure 4E), suggesting that LPL released by active heparanase is from the myocyte surface. Unexpectedly, RhoA activation was not observed with active heparanase, which only responded to the latent form of the enzyme (Figure 4F). Our data imply that active heparanase releases LPL from the myocyte surface, whereas latent heparanase may move LPL from an intracellular store to the surface.

Activation of RhoA by Latent Heparanase Depends on HSPGs and PKCα

To examine the mechanism of RhoA activation by latent heparanase, we considered whether the integrity of the myocyte surface HSPGs is required for this signal mechanotransduction. As expected, removal of HS by heparinase III blocked RhoA activation by latent heparanase (Figure 5A). Interestingly, the effect of latent heparanase on RhoA activation and cytoskeleton polymerization was abolished by the PKCα/β inhibitor Gö6976 (Figure 5B and 5C), whereas phorbol 12-myristate 13-acetate, a conventional PKC activator, had effects similar to that seen with latent heparanase (Figure V in the online-only Data Supplement). Furthermore, the RhoA activation effect observed with latent heparanase was only attenuated in mycoytes with reduced PKCα expression, but not in cells in which PKCβ was specifically inhibited (Figure VI in the online-only Data Supplement). In normal myocytes, syndecan-4 is distributed on myocyte surface in a dispersed manner. On addition of latent heparanase and antisyndecan-4 antibody, syndecan-4 appeared clustered on the surface (Figure 5D, arrow), an effect not seen with active heparanase.

Figure 5.
Figure 5.

Activation of RhoA by latent heparanase requires heparan sulfate proteoglycans (HSPGs) and protein kinase C-α. RhoA activation in myocytes was determined 15 minutes after incubation with 1 µg/mL latent heparanase pretreated with (III+HEPA) or without (HEPA) 10 IU/L heparinase III for 30 minutes. The effect of heparinase III per se on RhoA activation was also evaluated (III; A). A similar experiment was repeated, except that 5 mmol/L Gö6976 was used to pretreat the myocytes (B). In this experiment, F-to-G actin ratio was evaluated 1 hour after addition of latent heparanase (C). Isolated cardiomyocytes were treated with 1 µg/mL active (A-HEPA) or latent heparanase (L-HEPA), or 3 µg/mL antisyndecan-4 antibody (Ab) for 30 minutes, and syndecan-4 distribution visualized by immunofluorescence (D). *P<0.05, compared with control (CON), n=3. #P<0.05, compared with HEPA, n=3.

LPL Secretion Is Increased When Myocytes Are Cocultured With ECs Exposed to High Glucose

To simulate diabetes mellitus in vitro, myocytes were cocultured with ECs. As a limited number of myocytes can be seeded in the coculture system, the intrinsic LPL activity at the surface of these myocyte is low. Hence, we used purified LPL added exogenously to amplify this surface pool. Two hundred micrograms of purified LPL was sufficient to saturate HSPGs binding sites on the myocyte surface (Figure 6A, inset). Using this amount of LPL, we exposed the coculture system to normal and high glucose. A significantly higher medium LPL activity was detected from the lower chamber in the presence of high glucose (Figure 6A). To examine whether the released LPL from myocytes is ultimately recruited onto the apical side of EC, we tested LPL activity on the EC surface after 2 and 4 hours with high glucose. In the normal glucose coculture, we did not observe any change in EC surface LPL activity at 2 or 4 hours. With high glucose, a robust increase in EC surface LPL activity was evident after 4 hours (Figure 6B). Notably, high-glucose coculture also caused actin polymerization (within 1 hour; Figure 6C, inset), that was preceded by RhoA activation (within 30 minutes) in myocytes (Figure 6C).

Figure 6.
Figure 6.

Heparanase is key for regulating lipoprotein lipase (LPL) secretion from myocytes to endothelial cells (ECs) after diabetes mellitus. A coculture was performed using EC in the insert (top) and cardiomyocytes in the well (bottom). Cardiomyocytes were loaded with increasing amount of purified LPL for 30 minutes, and LPL activity bound to the myocyte surface was determined (A, inset). Myocytes were loaded with 200 µg purified LPL for 30 minutes and washed with PBS to remove unbound LPL. Inserts with EC were placed on top, and the coculture was exposed to normal glucose (NG, 5.5 mmol/L) or high glucose (HG, 25 mmol/L). Medium was collected from the lower chamber after 30 minutes, and LPL activity was determined (A). After 2 and 4 hours, LPL activity at the EC surface was also released and tested by incubating these cells with 8 U/mL heparin for 3 minutes (B). In this coculture, activation of RhoA was measured in myocytes after 30 minutes, and compared with cells that did not have EC in the top chamber (−EC; C). F-to-G actin ratio was also tested at 30 and 60 minutes (C, inset). *P<0.05, compared with NG, n=3. D, How EC regulates cardiac metabolism after diabetes mellitus is summarized. TG indicates triglyceride.

Discussion

Cardiac muscle has a high demand for provision of energy which is obtained from oxidation of an assortment of different substrates including lactate, ketone bodies, glucose, and FAs.29,30 Of these substrates, FA is the favorite and major fuel of the heart.31 The dominant source of FA is from breakdown of circulating triglyceride-rich lipoproteins (chylomicrons and very low density lipoprotein) by LPL at the coronary lumen.32 Hence, cardiac LPL is of crucial importance for regulating energy supply in the heart. This enzyme is synthesized in myocytes and secreted to the vascular lumen. After diabetes mellitus, the heart increases its demand for FAs as a result of impaired glucose utilization.1,33 As a consequence, LPL activity at the coronary lumen is expected to increase. Accordingly, in our animal model of hyperglycemia (DZ) and type 1 diabetes mellitus (D55), we observed increased LPL activity at this location, an event that we have previously shown to occur in the absence of any change in (1) LPL-specific activity, (2) LPL gene expression, or (3) the number of coronary lumen binding sites.7 We concluded that this augmented LPL at the vascular lumen is likely a consequence of increased secretion from myocytes.

LPL secretion requires trafficking from an intracellular pool to myocyte surface HSPGs, followed by translocation to the vascular lumen. Our previous studies have indicated that the trafficking component requires actin cytoskeleton polymerization, which can be achieved by RhoA activation.10,26 In fact, when isolated hearts were perfused with LPA, a RhoA activator, increased coronary LPL activity was detected.26 RhoA is a small GTP-binding protein, with inactive RhoA associated with GDP and sequestered in the cytosol. On activation, GDP is replaced by GTP and GTP-RhoA shifts to the plasma membrane. Subsequently, through its downstream effector Rho-associated protein kinase, activation of RhoA will ultimately induce actin cytoskeleton polymerization. We observed that the increase in LPL activity at the coronary lumen in hyperglycemic DZ and diabetic D55 hearts were both accompanied by RhoA activation. As RhoA activation by LPA also promoted LPL secretion in cardiomyocytes through actin cytoskeleton polymerization, our data suggest that this GTP-binding protein is an important contributor to move LPL to the myocyte surface.

The mechanism by which LPL leaves the myocyte surface to move to the coronary lumen is not completely understood. At the myocyte surface, LPL is sequestered to HSPGs via an ionic interaction. Thus, its release is possible by negatively charged heparin. However, cleavage of HS, or shedding of the extracellular part of HSPGs is a more likely event for LPL to leave the surface in vivo. Interestingly, on exposure of cardiomyocytes to ECCM, LPL is released into the medium with a reciprocal decrease in surface LPL activity, suggesting that ECCM is likely releasing LPL from the myocyte surface. An additional observation noticed with ECCM was its ability to assist in LPL replenishment when the surface pool was depleted. We attributed this effect to its ability to activate RhoA in cardiomyocytes in a pattern similar to that seen with LPA. In both cases, after reaching a peak, the amount of GTP-RhoA declined as it is normally converted to a GDP-bound form after activation of downstream effectors.34 Intriguingly, heparin had no effect on RhoA or LPL replenishment, implying that simple displacement of LPL is not sufficient to trigger the replenishment process, and that ECCM is likely altering the conformation of HSPGs.

An enzyme of special relevance that is known to cleave HS of HSPGs is heparanase. This enzyme is synthesized in the EC as a latent 65-kDa form, which possesses no catalytic activity. After its secretion, it binds to EC surface HSPGs and is internalized into lysosomes.35 There it is cleaved into a 50-kDa active form and stored until it is secreted in response to stimulation such as adenosine, ADP, and ATP.23,36 In the present study, increased secretion of both latent and active heparanase into the medium was also observed when ECs were exposed to high glucose. As both forms of heparanase were augmented in the interstitial space of hearts from DZ animals, it suggests that hyperglycemia is effective in triggering coronary endothelial heparanase release. We have previously demonstrated that high glucose–induced secretion of active heparanase from EC occurs through activation of P2Y receptors that initiates stress fiber formation (across which heparanase-containing cargos are transported) and disruption of cortical actin cytoskeleton (to allow heparanase to be released into the extracellular space).37 Whether the same mechanism applies to release of latent heparanase is unknown because we cannot rule out the possibility that high glucose could also block endocytosis of latent heparanase from the EC surface, or is capable of directly detaching the latent enzyme from surface HSPGs. Irrespective of the release mechanism, our study is the first to demonstrate that it is heparanase within ECCM that is responsible for detachment of LPL from the myocyte surface.

Although active heparanase is predicted to release surface LPL by cleaving HS, latent enzyme could also facilitate LPL release. As expected, only active heparanase initiated release of LPL from the myocyte surface. Surprisingly, this release was not accompanied by RhoA activation. In fact, it was latent heparanase that accounted for the RhoA activation seen with ECCM. Growing evidence has demonstrated that heparanase also possesses activity-independent effects, for example, activation of Src during cell adhesion.38 Hence, our data suggest that active heparanase detaches LPL from the myocyte surface HSPGs, whereas latent heparanase activates RhoA to move intracellular LPL to replenish this released reservoir. Studies on HSPG-mediated signaling have focused on syndecan-4, a transmembrane HSPG with its cytoplasmic domain capable of activating and stabilizing PKCα, and syndecan-4–dependent PKCα activation has been indicated in fibroblast growth factor signaling.39,40 Moreover, neuronal Thy-1–induced cell spreading requires RhoA-mediated actin cytoskeleton polymerization, a suggested downstream event after syndecan-4–dependent PKCα activation.41 Our study also indicated that the RhoA activation induced by latent heparanase is mediated by PKCα. In U87 MG human glioma cells, clustering of syndecan-4 by latent heparanase or an antiheparanase antibody was sufficient to induce cell spreading, a process involving PKC activation.17 In addition, on binding of phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5P)(2)) to its cytoplasmic domain, oligomerization of syndecan-4 was promoted with a subsequent upregulation of PKCα.42,43 In this study, it is likely that clustering of syndecan-4 with latent heparanase triggers RhoA activation. It is possible that latent heparanase does this by acting as a bridging factor by binding to HS. This idea was supported by our results using heparinase III that eliminates HS and attenuated RhoA activation by latent heparanase. Altogether, results from our study suggest that active and latent heparanase work cooperatively to promote LPL secretion after diabetes mellitus (Figure 6D).

In summary, heparanase is a key mediator of the cross-talk between EC and cardiomyocyte to increase LPL secretion after hyperglycemia. This idea was confirmed in DZ-induced hyperglycemia and the high-glucose coculture system, with both experimental models exhibiting increased heparanase secretion together with LPL amplification at the apical side of EC. Overall, our data provide evidence for a novel role of EC to control FA delivery to the heart after diabetes mellitus. Pharmaceutical manipulation of this process, for instance, inhibition of heparanase secretion/activity, could potentially provide an additional strategy to limit FA delivery to the heart, and eventually minimize cardiac lipotoxicity observed with chronic diabetes mellitus (Figure 6D).

Acknowledgments

Y. Wang conceived the idea, generated most of the data, and wrote the manuscript. D.H. Zhang, A.P.L.C., A. Wan, and K. Neumaier helped with obtaining some of the data. B. Rodrigues helped with writing the manuscript. I. Vlodavsky(Cancer and Vascular Biology Research Center, Israel) assisted with valuable suggestions and the preparation of highly purified latent and active heparanase.

Sources of Funding

This work was supported by an operating grant from the Canadian Institutes of Health Research (MOP-98012 to Dr Brian Rodrigues), and in part by a grant to Dr Israel Vlodavsky from the Israel Ministry of Health. Y. Wang is a recipient of a Doctoral Student Research Award from the Canadian Diabetes Association.

Disclosures

None.

Footnotes

  • Received October 3, 2012.
  • Accepted February 19, 2013.

References

  1. 1.↵
    1. An D,
    2. Rodrigues B
    . Role of changes in cardiac metabolism in development of diabetic cardiomyopathy. Am J Physiol Heart Circ Physiol. 2006;291:H1489–H1506.
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    1. Voshol PJ,
    2. Rensen PC,
    3. van Dijk KW,
    4. Romijn JA,
    5. Havekes LM
    . Effect of plasma triglyceride metabolism on lipid storage in adipose tissue: studies using genetically engineered mouse models. Biochim Biophys Acta. 2009;1791:479–485.
    OpenUrlPubMed
  3. 3.↵
    1. Tannock LR,
    2. Chait A
    . Lipoprotein-matrix interactions in macrovascular disease in diabetes. Front Biosci. 2004;9:1728–1742.
    OpenUrlPubMed
  4. 4.↵
    1. Blanchette-Mackie EJ,
    2. Masuno H,
    3. Dwyer NK,
    4. Olivecrona T,
    5. Scow RO
    . Lipoprotein lipase in myocytes and capillary endothelium of heart: immunocytochemical study. Am J Physiol. 1989;256(6 pt 1):E818–E828.
    OpenUrl
  5. 5.↵
    1. Camps L,
    2. Reina M,
    3. Llobera M,
    4. Vilaró S,
    5. Olivecrona T
    . Lipoprotein lipase: cellular origin and functional distribution. Am J Physiol. 1990;258(4 pt 1):C673–C681.
    OpenUrlPubMed
  6. 6.↵
    1. Pulinilkunnil T,
    2. Qi D,
    3. Ghosh S,
    4. Cheung C,
    5. Yip P,
    6. Varghese J,
    7. Abrahani A,
    8. Brownsey R,
    9. Rodrigues B
    . Circulating triglyceride lipolysis facilitates lipoprotein lipase translocation from cardiomyocyte to myocardial endothelial lining. Cardiovasc Res. 2003;59:788–797.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    1. Pulinilkunnil T,
    2. Abrahani A,
    3. Varghese J,
    4. Chan N,
    5. Tang I,
    6. Ghosh S,
    7. Kulpa J,
    8. Allard M,
    9. Brownsey R,
    10. Rodrigues B
    . Evidence for rapid “metabolic switching” through lipoprotein lipase occupation of endothelial-binding sites. J Mol Cell Cardiol. 2003;35:1093–1103.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Severson DL,
    2. Lee M,
    3. Carroll R
    . Secretion of lipoprotein lipase from myocardial cells isolated from adult rat hearts. Mol Cell Biochem. 1988;79:17–24.
    OpenUrlCrossRefPubMed
  9. 9.↵
    1. Ewart HS,
    2. Severson DL
    . Insulin and dexamethasone stimulation of cardiac lipoprotein lipase activity involves the actin-based cytoskeleton. Biochem J. 1999;340(pt 2):485–490.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. Kim MS,
    2. Kewalramani G,
    3. Puthanveetil P,
    4. Lee V,
    5. Kumar U,
    6. An D,
    7. Abrahani A,
    8. Rodrigues B
    . Acute diabetes moderates trafficking of cardiac lipoprotein lipase through p38 mitogen-activated protein kinase-dependent actin cytoskeleton organization. Diabetes. 2008;57:64–76.
    OpenUrlAbstract/FREE Full Text
  11. 11.↵
    1. Nadanaka S,
    2. Kitagawa H
    . Heparan sulphate biosynthesis and disease. J Biochem. 2008;144:7–14.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Filmus J,
    2. Capurro M,
    3. Rast J
    . Glypicans. Genome Biol. 2008;9:224.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Couchman JR
    . Transmembrane signaling proteoglycans. Annu Rev Cell Dev Biol. 2010;26:89–114.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Kolset SO,
    2. Salmivirta M
    . Cell surface heparan sulfate proteoglycans and lipoprotein metabolism. Cell Mol Life Sci. 1999;56:857–870.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Mertens G,
    2. Cassiman JJ,
    3. Van den Berghe H,
    4. Vermylen J,
    5. David G
    . Cell surface heparan sulfate proteoglycans from human vascular endothelial cells. Core protein characterization and antithrombin III binding properties. J Biol Chem. 1992;267:20435–20443.
    OpenUrlAbstract/FREE Full Text
  16. 16.↵
    1. Chuang CY,
    2. Lord MS,
    3. Melrose J,
    4. Rees MD,
    5. Knox SM,
    6. Freeman C,
    7. Iozzo RV,
    8. Whitelock JM
    . Heparan sulfate-dependent signaling of fibroblast growth factor 18 by chondrocyte-derived perlecan. Biochemistry. 2010;49:5524–5532.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Levy-Adam F,
    2. Feld S,
    3. Suss-Toby E,
    4. Vlodavsky I,
    5. Ilan N
    . Heparanase facilitates cell adhesion and spreading by clustering of cell surface heparan sulfate proteoglycans. PLoS ONE. 2008;3:e2319.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Shi ZD,
    2. Abraham G,
    3. Tarbell JM
    . Shear stress modulation of smooth muscle cell marker genes in 2-D and 3-D depends on mechanotransduction by heparan sulfate proteoglycans and ERK1/2. PLoS ONE. 2010;5:e12196.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Pillarisetti S,
    2. Paka L,
    3. Sasaki A,
    4. Vanni-Reyes T,
    5. Yin B,
    6. Parthasarathy N,
    7. Wagner WD,
    8. Goldberg IJ
    . Endothelial cell heparanase modulation of lipoprotein lipase activity. Evidence that heparan sulfate oligosaccharide is an extracellular chaperone. J Biol Chem. 1997;272:15753–15759.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Fux L,
    2. Ilan N,
    3. Sanderson RD,
    4. Vlodavsky I
    . Heparanase: busy at the cell surface. Trends Biochem Sci. 2009;34:511–519.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Gingis-Velitski S,
    2. Zetser A,
    3. Flugelman MY,
    4. Vlodavsky I,
    5. Ilan N
    . Heparanase induces endothelial cell migration via protein kinase B/Akt activation. J Biol Chem. 2004;279:23536–23541.
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    1. Cui H,
    2. Shao C,
    3. Liu Q,
    4. Yu W,
    5. Fang J,
    6. Yu W,
    7. Ali A,
    8. Ding K
    . Heparanase enhances nerve-growth-factor-induced PC12 cell neuritogenesis via the p38 MAPK pathway. Biochem J. 2011;440:273–282.
    OpenUrlAbstract/FREE Full Text
  23. 23.↵
    1. Zetser A,
    2. Levy-Adam F,
    3. Kaplan V,
    4. Gingis-Velitski S,
    5. Bashenko Y,
    6. Schubert S,
    7. Flugelman MY,
    8. Vlodavsky I,
    9. Ilan N
    . Processing and activation of latent heparanase occurs in lysosomes. J Cell Sci. 2004;117(pt 11):2249–2258.
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    1. Wang F,
    2. Kim MS,
    3. Puthanveetil P,
    4. Kewalramani G,
    5. Deppe S,
    6. Ghosh S,
    7. Abrahani A,
    8. Rodrigues B
    . Endothelial heparanase secretion after acute hypoinsulinemia is regulated by glucose and fatty acid. Am J Physiol Heart Circ Physiol. 2009;296:H1108–H1116.
    OpenUrlAbstract/FREE Full Text
  25. 25.↵
    1. Wang Y,
    2. Puthanveetil P,
    3. Wang F,
    4. Kim MS,
    5. Abrahani A,
    6. Rodrigues B
    . Severity of diabetes governs vascular lipoprotein lipase by affecting enzyme dimerization and disassembly. Diabetes. 2011;60:2041–2050.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    1. Pulinilkunnil T,
    2. An D,
    3. Ghosh S,
    4. Qi D,
    5. Kewalramani G,
    6. Yuen G,
    7. Virk N,
    8. Abrahani A,
    9. Rodrigues B
    . Lysophosphatidic acid-mediated augmentation of cardiomyocyte lipoprotein lipase involves actin cytoskeleton reorganization. Am J Physiol Heart Circ Physiol. 2005;288:H2802–H2810.
    OpenUrlAbstract/FREE Full Text
  27. 27.↵
    1. Hakoshima T,
    2. Shimizu T,
    3. Maesaki R
    . Structural basis of the Rho GTPase signaling. J Biochem. 2003;134:327–331.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    1. Kim MS,
    2. Wang Y,
    3. Rodrigues B
    . Lipoprotein lipase mediated fatty acid delivery and its impact in diabetic cardiomyopathy. Biochim Biophys Acta. 2012;1821:800–808.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Bassenge E,
    2. Wendt VE,
    3. Schollmeyer P,
    4. Bluemchen G,
    5. Gudbjarnason S,
    6. Bing RJ
    . Effect of ketone bodies on cardiac metabolism. Am J Physiol. 1965;208:162–168.
    OpenUrl
  30. 30.↵
    1. van der Vusse GJ,
    2. van Bilsen M,
    3. Glatz JF
    . Cardiac fatty acid uptake and transport in health and disease. Cardiovasc Res. 2000;45:279–293.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Lopaschuk GD,
    2. Ussher JR,
    3. Folmes CD,
    4. Jaswal JS,
    5. Stanley WC
    . Myocardial fatty acid metabolism in health and disease. Physiol Rev. 2010;90:207–258.
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    1. Saddik M,
    2. Lopaschuk GD
    . Triacylglycerol turnover in isolated working hearts of acutely diabetic rats. Can J Physiol Pharmacol. 1994;72:1110–1119.
    OpenUrlCrossRefPubMed
  33. 33.↵
    1. Camps M,
    2. Castelló A,
    3. Muñoz P,
    4. Monfar M,
    5. Testar X,
    6. Palacín M,
    7. Zorzano A
    . Effect of diabetes and fasting on GLUT-4 (muscle/fat) glucose-transporter expression in insulin-sensitive tissues. Heterogeneous response in heart, red and white muscle. Biochem J. 1992;282(pt 3):765–772.
    OpenUrlAbstract/FREE Full Text
  34. 34.↵
    1. Loirand G,
    2. Pacaud P
    . The role of Rho protein signaling in hypertension. Nat Rev Cardiol. 2010;7:637–647.
    OpenUrlCrossRefPubMed
  35. 35.↵
    1. Gingis-Velitski S,
    2. Zetser A,
    3. Kaplan V,
    4. Ben-Zaken O,
    5. Cohen E,
    6. Levy-Adam F,
    7. Bashenko Y,
    8. Flugelman MY,
    9. Vlodavsky I,
    10. Ilan N
    . Heparanase uptake is mediated by cell membrane heparan sulfate proteoglycans. J Biol Chem. 2004;279:44084–44092.
    OpenUrlAbstract/FREE Full Text
  36. 36.↵
    1. Shafat I,
    2. Vlodavsky I,
    3. Ilan N
    . Characterization of mechanisms involved in secretion of active heparanase. J Biol Chem. 2006;281:23804–23811.
    OpenUrlAbstract/FREE Full Text
  37. 37.↵
    1. Wang F,
    2. Wang Y,
    3. Kim MS,
    4. Puthanveetil P,
    5. Ghosh S,
    6. Luciani DS,
    7. Johnson JD,
    8. Abrahani A,
    9. Rodrigues B
    . Glucose-induced endothelial heparanase secretion requires cortical and stress actin reorganization. Cardiovasc Res. 2010;87:127–136.
    OpenUrlAbstract/FREE Full Text
  38. 38.↵
    1. Barash U,
    2. Cohen-Kaplan V,
    3. Dowek I,
    4. Sanderson RD,
    5. Ilan N,
    6. Vlodavsky I
    . Proteoglycans in health and disease: new concepts for heparanase function in tumor progression and metastasis. FEBS J. 2010;277:3890–3903.
    OpenUrlCrossRefPubMed
  39. 39.↵
    1. Keum E,
    2. Kim Y,
    3. Kim J,
    4. Kwon S,
    5. Lim Y,
    6. Han I,
    7. Oh ES
    . Syndecan-4 regulates localization, activity and stability of protein kinase C-alpha. Biochem J. 2004;378(pt 3):1007–1014.
    OpenUrlCrossRefPubMed
  40. 40.↵
    1. Kuriyama S,
    2. Mayor R
    . A role for Syndecan-4 in neural induction involving ERK- and PKC-dependent pathways. Development. 2009;136:575–584.
    OpenUrlAbstract/FREE Full Text
  41. 41.↵
    1. Avalos AM,
    2. Valdivia AD,
    3. Muñoz N,
    4. Herrera-Molina R,
    5. Tapia JC,
    6. Lavandero S,
    7. Chiong M,
    8. Burridge K,
    9. Schneider P,
    10. Quest AF,
    11. Leyton L
    . Neuronal Thy-1 induces astrocyte adhesion by engaging syndecan-4 in a cooperative interaction with alphavbeta3 integrin that activates PKCalpha and RhoA. J Cell Sci. 2009;122(pt 19):3462–3471.
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    1. Oh ES,
    2. Woods A,
    3. Lim ST,
    4. Theibert AW,
    5. Couchman JR
    . Syndecan-4 proteoglycan cytoplasmic domain and phosphatidylinositol 4,5-bisphosphate coordinately regulate protein kinase C activity. J Biol Chem. 1998;273:10624–10629.
    OpenUrlAbstract/FREE Full Text
  43. 43.↵
    1. Horowitz A,
    2. Murakami M,
    3. Gao Y,
    4. Simons M
    . Phosphatidylinositol-4,5-bisphosphate mediates the interaction of syndecan-4 with protein kinase C. Biochemistry. 1999;38:15871–15877.
    OpenUrlCrossRefPubMed

Significance

After diabetes mellitus, the heart switches to use more fatty acid for energy supply, and this is achieved through rapid augmentation of lipoprotein lipase at the coronary lumen. Previous studies have shown that the amount of coronary lipoprotein lipase depends on secretion from myocytes, a process that is regulated by multiple signaling pathways in these cells. Our data suggest that after hyperglycemia, endothelial cells, through the release of heparanase, can signal to cardiomyocytes to increase lipoprotein lipase secretion from these cells. Given that endothelial cells are the first to be exposed to hyperglycemia, this cross-talk allows the heart to predominantly use fatty acids for energy supply. Chronically, this excess fatty acid delivery to the heart may explain lipotoxicity and ultimately cardiomyopathy after diabetes mellitus.

View Abstract

This Issue

Jump to

Article Tools

Share this Article

  • Email
  • Endothelial Heparanase Regulates Heart Metabolism by Stimulating Lipoprotein Lipase Secretion From CardiomyocytesSignificance
    Ying Wang, Dahai Zhang, Amy Pei-Ling Chiu, Andrea Wan, Katharina Neumaier, Israel Vlodavsky and Brian Rodrigues
    Arteriosclerosis, Thrombosis, and Vascular Biology. 2013;33:894-902, originally published April 10, 2013
    https://doi.org/10.1161/ATVBAHA.113.301309
    Permalink:
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Related Articles

Cited By...