1 Introduction

Under physiological conditions, cardiac tissue acquires most of its energy from metabolism of two major substrates, glucose and fatty acid (FA), the latter being the preferred substrate consumed [1,2]. Because the heart has no potential to synthesize FA, it is dependent upon supply, with hydrolysis of triglyceride (TG)-rich lipoproteins by lipoprotein lipase (LPL) positioned at the endothelial surface of the coronary lumen [3] being suggested to be the principal source of FA for cardiac utilization [4]. Endothelial cells do not synthesize LPL and hence the enzyme is synthesized in cardiomyocytes [5,6]. LPL secreted as an active enzyme binds to myocyte cell surface heparan sulphate proteoglycans (HSPGs), before it is translocated onto comparable HSPG binding sites on the luminal side of the vessel wall [7–9]. In perfused guinea pig hearts, LPL can move from myocytes to the vascular lumen within 30 min [10] by mechanisms that are not clearly understood. At least in co-culture experiments using adipocytes and endothelial cells, heparanase like compounds secreted from endothelial cells have been suggested to release subendothelial HS bound proteins and specific HS oligosaccharides that serve as extracellular chaperones allowing LPL to be transported across the interstitial space [11].

TG-rich lipoproteins and oleic acid release adipocyte cell surface LPL [12,13]. The authors concluded (but were unable to confirm) that this released LPL from adipocyte cell surface would then be transported to the luminal endothelial cell surface. With adipose tissue (as compared to heart) it is difficult to study LPL movement from underlying parenchymal cells to the vascular lumen. In addition to FA, lysophosphatidylcholine (Lyso-PC), a component of lipoprotein phospholipids, is also released during LPL mediated lipolysis of TG rich lipoproteins and would be expected to be augmented under conditions of extensive lipoprotein TG breakdown [14–16]. Endothelial cells exposed to lyso-PC release heparanase like compounds that cleave cell surface HSPG-bound LPL, at least in adipocytes [11]. The present study was designed to investigate the influence of circulating TG and their lipolysis in facilitating translocation of LPL from the underlying cardiomyocyte cell surface to the coronary lumen.

2 Methods

2.1 Experimental animals

The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Animals were cared for in accordance with the principles promulgated by the Canadian Council on Animal Care and the University of British Columbia. Adult male Wistar rats (270–290 g) were obtained from the UBC Animal Care Unit and maintained under a 12-h light (07:00–19:00 h)/dark cycle and supplied with a standard laboratory diet (PMI Feeds, Richmond, VA, USA) and water ad libitum.

2.2 Diazoxide induced acute hyperglycemia

Diazoxide (DZ), a selective K⁺_ATP channel opener decreases insulin secretion and causes hyperglycemia within 1 h [17–19]. More importantly, changes in glucose are also associated with rapid elevation of serum FA and TG. DZ (25–100 mg/kg) was administered i.p. at 10:00 h and tail vein blood samples obtained at various times after DZ injection. Animals were euthanized 4 h after DZ, and hearts removed.

2.3 Isolated heart perfusion

Rats were anesthetized with 65 mg/kg sodium pentobarbital i.p., the thoracic cavity opened, and the heart carefully excised. The heart was immersed in cold (4°C) Krebs–Ringer 4-(2-hydroxyethyl)-1-piperazine-ethanesulphonic acid (HEPES) buffer containing 10 mM glucose (pH 7.4). After the aorta was cannulated and tied below the innominate artery, hearts were perfused retrogradely by the nonrecirculating Langendorff technique as described previously [20]. Perfusion fluid was continuously gassed with 95% O₂/5% CO₂ in a double-walled, water-heated chamber maintained at 37°C with a temperature-controlled circulating water bath. The flow rate was controlled at 7–8 ml/min.

2.4 Coronary lumen LPL activity

To measure endothelium-bound LPL, Langendorff perfusion solution was changed to Krebs–Ringer HEPES buffer containing 1% fatty acid free bovine serum albumin (BSA) and heparin (5 U/ml). This concentration of heparin can maximally release cardiac LPL from its binding sites. The coronary effluent was collected in timed fractions over 10 min and assayed for LPL activity by measuring the hydrolysis of a sonicated [³H]triolein substrate emulsion [21]. LPL activity is expressed as nanomoles oleate released per hour per milliliter. Subsequent to LPL displacement with heparin, hearts were blotted dry and wet weights recorded (0.56±0.03 g).

2.5 Immunolocalization of LPL

Upon excision, control and DZ-treated hearts were retrogradely perfused with non-circulating buffer for 3 min to clear the heart of blood. Perfusion buffer was then changed to fixative (neutral phosphate-buffered 10% formalin) for 2 min. Hearts were stored in 10% formalin for 24 h followed by paraffin processing through graded ethanol and xylene. Blocks were embedded in Paraplast, sectioned at 3 μm and mounted on positively charged glass slides. For immunostaining, sections were deparaffinized, rehydrated, and treated with 5% (v/v) heat inactivated rabbit plasma in Tris-buffered saline (TBS) to block nonspecific background. Sections were then incubated with affinity-purified chicken antibovine LPL polyclonal antibody (1:400 dilution in TBS containing 1%, w/v, BSA) overnight at room temperature in a humid chamber. Samples were then washed with TBS and incubated for 1 h at room temperature with the secondary biotinylated rabbit anti-chicken IgG (Chemicon Corp., 1:150 dilution), followed by incubation for 1 h with streptavidin-conjugated Cy3 fluorescent probe (1:1000 dilution). The unbound fluorescent probe was rinsed with TBS buffer and sections were mounted with DABCO. Slides were visualized using a Bio-Rad 600 Confocal Microscope at 630× magnification. An absence of staining was observed when the primary antibody was omitted or replaced by preimmune chicken serum.

2.6 LPL content of cardiomyocytes

In addition to endothelial-bound LPL, a large pool of enzyme activity is still measurable within the heart, located predominantly within myocytes. Ventricular calcium-tolerant myocytes were prepared by a previously described procedure [20]. Briefly, myocytes were made calcium-tolerant by successive exposure to increasing concentrations of calcium. Our method of isolation yields a highly enriched population of calcium-tolerant myocardial cells that are rod-shaped in the presence of 1 mM Ca²⁺ with clear cross striations. Intolerant cells are intact but hypercontract into vesiculated spheres. Yield of myocytes (cell number was approximately similar in both control and DZ hearts, ∼4.8×10⁶) was determined microscopically using an improved Neubauer hemocytometer. Myocyte viability (generally between 75 and 85% in control and DZ hearts) was assessed as the percentage of elongated cells with clear cross striations that excluded 0.2% trypan blue. Cardiac myocytes were suspended at a final cell density of 0.4×10⁶ cells/ml, incubated at 37 °C, and basal LPL activity in the medium and cell pellet (after centrifugation) measured. To release surface-bound LPL activity, heparin (5 U/ml) was added to the myocyte suspension. After incubation for 10 min, an aliquot of cell suspension was removed, medium separated by centrifugation in an Eppendorf microcentrifuge (1 min, 10,000 g), and assayed for LPL activity. The total basal cellular LPL activity was measured by sonicating (Vibra Cell™ sonicator at a frequency of 40 Hz for 2×30 s) the cell pellets after resuspending them in 0.2 ml of 50 mmol/l NH₄Cl buffer (pH 8.0) containing 0.125% (v/v) Triton X-100. After sonication, the volume was adjusted to 1 ml using a sucrose buffer (0.25 mol/l sucrose, 1 mmol/l EDTA, 1 mmol/l dithiothreitol, 10 mmol/l HEPES, pH 7.4). Assay for cell sonicate LPL activity was done using 20 μl of the cell sonicate and heparin (2 U/ml) was included in the assay.

2.7 Treatments

Table 1 summarizes the experimental design for the indicated treatments given below.

2.7.1 Insulin

In the indicated experiment, rats were injected into the tail vein with a rapid acting insulin (HumulinR, 8 U), 1 h after DZ administration (following verification of hyperglycemia). The animals were killed after 90 min (time required for establishment of sustained euglycemia), and cardiac heparin-releasable LPL activity determined.

2.7.2 Lyso-PC

In the indicated experiment control rat hearts were perfused with l-α-lysophosphatidylcholine palmitoyl (Lyso-PC) (0.1–100 nM) for 60 min. Cardiac heparin-releasable LPL activity was subsequently determined.

2.7.3 Triton WR 1339

WR 1339, a non-ionic detergent, physically alters lipoproteins making them inaccessible for LPL mediated hydrolysis [22,23]. When injected intravenously, newly synthesized TGs accumulate in the plasma. Rats were injected (i.v.) with WR 1339 (25%, w/v, solution in normal saline to give a dose of 600 mg/kg body weight). WR 1339 was injected 1 h prior to DZ administration and blood samples were collected at 4 h after the injection. Serum was separated and the TG concentration was measured using Sigma Infinity diagnostic kit.

2.7.4 N⁶-cyclopentyladenosine (CPA)

Adenosine by inhibiting adipose tissue lipolysis has been demonstrated to lower serum FA and circulating TG [24,25]. In the indicated experiment, adenosine (8 mg/kg, i.p.) was administered 1 h after DZ administration. The animals were killed after 4 h and cardiac heparin-releasable LPL activity determined.

2.8 Plasma measurements

Plasma samples were stored at −20°C until assayed. Diagnostic kits were used to measure glucose, TG (Sigma), non esterified fatty acid (NEFA; Wako), and insulin (Linco).

2.9 Materials

[³H]Triolein was purchased from Amersham Canada. Heparin sodium injection (Hapalean; 1000 USP U/ml) was obtained from Oraganon Teknika. All other chemicals were obtained from Sigma Chemical.

2.10 Statistical analysis

Values are means±S.E. LPL activity in response to heparin perfusion over time was analyzed by multivariate analysis of variance (two-way ANOVA) using the NCSS. Wherever appropriate, one-way ANOVA followed by the Tukey or Bonferroni tests or the unpaired and paired Student’s t-test was used to determine differences between group mean values (as indicated in the specific figure legends). The level of statistical significance was set at P<0.05.

3 Results

3.1 Effects of DZ on plasma parameters

To evaluate the dose dependent effect on serum glucose, varying doses of DZ (25, 50, and 100 mg/kg) were administered. Although doses of 25 and 50 mg/kg increased glucose, the extent and duration of hyperglycemia were not as substantial as that seen with 100 mg/kg (Fig. 1, inset), which caused a rapid decline in serum insulin within 1 h (Fig. 1). Notably, at this dose, serum glucose reached a maximum level after 2 h, and was comparable to the hyperglycemia produced by administration of 55 mg/kg STZ [20]. Therefore, all subsequent experiments were carried out using DZ 100 mg/kg. Changes in plasma parameters with DZ also included significant and rapid increases in FA and TG (Fig. 2). Other characteristics normally associated with hyperglycemia, such as polydipsia, was also observed in DZ-treated animals.

3.2 LPL activity, immunolocalization and effects of insulin treatment

Retrograde perfusion of control hearts with heparin resulted in the release of LPL into the coronary perfusate (Fig. 3A. This heparin-mediated LPL discharge was rapid, and peak activity, thought to represent LPL located at or near the endothelial surface, was observed within 1 min. Compared to control rat hearts, there was a substantial increase in coronary LPL activity (∼300%, Fig. 3A) and immunofluroscence (Fig. 3B) at the vascular lumen following 4 h of DZ. To determine the kinetics of LPL up regulation at the vascular lumen, some DZ-treated hearts were isolated at 1–4 h, and LPL activity measured. Interestingly, increase in LPL activity became apparent as early as 1 h subsequent to injection of DZ, and was maintained for an additional 3 h (Fig. 3A, inset).

One caveat associated with DZ is that in addition to inhibiting insulin secretion, it also lowers blood pressure [19]. In some anesthetized rats, DZ produced a significant fall in mean carotid artery blood pressure (∼20 mmHg). This effect was transient, and blood pressure returned to normal within 3 h (data not shown). As this hypotensive effect was not modified by a dose of insulin (8 U) that normalizes cardiac LPL (Fig. 4) and serum glucose (Fig. 4, inset), control of LPL by DZ is probably dependent on its lowering of insulin rather than its effects on blood pressure. Given that DZ is a K⁺_ATP channel opener with potential direct effects on the heart, we examined whether DZ could directly influence cardiac heparin releasable LPL activity. Isolated hearts when perfused in the absence or presence of DZ (100 μM, 1 h), followed by heparin, did not significantly change peak LPL activity (control, 390; DZ, 432 nmol/ml/h). Thus, the DZ-induced increase in LPL observed in vivo could not be duplicated in vitro, and suggests that the increased enzyme observed in vivo occurs subsequent to drop in insulin.

3.3 Cardiac myocyte cell surface and intracellular LPL activity

To determine whether the increase in luminal LPL was a consequence of augmented synthesis from cardiac cells, myocytes were isolated from control and 4 h DZ-treated rats. There was no difference in myocyte viability or yield following DZ. Total cellular LPL activity (surface bound+intracellular) remained unchanged between control and DZ-treated myocytes (Fig. 5, inset). Myocytes from the two groups were also incubated with heparin. Basal LPL activity released into the medium remained unchanged between the two groups. Interestingly, myocytes from DZ-treated rats demonstrated a 50% reduction in heparin releasable surface bound LPL activity at a time when luminal LPL activity was augmented suggesting an accelerated transfer of the enzyme from the myocyte cell surface to the coronary lumen (Fig. 5). It should be noted that establishment of a temporal relationship between the rise in luminal with a drop in myocyte cell surface LPL activity in STZ diabetic rats was never possible given the triphasic pattern of changes in glucose and insulin profile in the 24-h period subsequent to STZ administration [26].

3.4 Lysophosphatidylcholine effects on luminal LPL

To investigate the role of lipolytic byproducts on translocation of LPL from the underlying myocyte cell surface to the coronary lumen, we examined the effects of varying doses of Lyso-PC on luminal LPL activity in control rat hearts. Graded concentrations of Lyso-PC (0.1–100 nM) when perfused for 60 min increased heparin-releasable LPL activity (Fig. 6, inset). Interestingly, LPL activity following 1 nM Lyso-PC (Fig. 6) was identical to that seen subsequent to in vivo DZ administration. Lyso-PC (0.1–100 nM) by itself was unable to release luminal LPL (data not shown). Additionally, following titration with these doses of Lyso-PC, hearts did not demonstrate any change in rate and pattern of contraction.

3.5 Manipulation of circulating TG

As a lipolytic byproduct was able to influence LPL, and given the increase in TG in DZ-treated animals, we hypothesized that manipulation of circulating TG in these animals would influence luminal LPL. Treatment of control animals with WR 1339 brought about a 30-fold increase in circulating TG compared to control within 4 h, with no change in LPL activity (data not shown). Although DZ administered to WR 1339-treated animals did not affect the elevated serum TG (Fig. 7, left panel) or glucose (Table 2), there was a considerable decline in serum FA (Table 2) and luminal LPL activity (Fig. 7, right panel).

Using an alternative strategy, we attempted to prevent increases in circulating TG by inhibiting adipose tissue lipolysis. In preliminary experiments, CPA demonstrated a reduction in circulating FA and TG (data not shown). Interestingly, CPA prevented the DZ induced augmentation of circulating TG (Fig. 7, left panel) and FA (Table 2) with normalization of luminal LPL activity (Fig. 7, right panel).

4 Discussion

Impaired glucose utilization following hypoinsulinemia increases the requirement of the heart for FA [27,28]. This excess demand is partly achieved by up regulation of LPL at the coronary luminal surface, as described in our previous studies using the STZ rat heart [20]. Although we concluded that accelerated translocation of LPL from the myocyte cell surface to the vascular lumen could be responsible for augmenting luminal LPL activity, verification of this mechanism was technically difficult due to the triphasic pattern of blood glucose following STZ. This pattern comprises of an initial brief hyperglycemia followed by a period of hypoglycemia before noticeable hyperglycemia is attained within 12–16 h [26]. As metabolic switching in the heart from glucose to predominantly FA likely occurs rapidly following hypoinsulinemia, we pursued an alternate model of acute diabetes to study LPL translocation. In the present study, injection of DZ caused a rapid drop in insulin and hyperglycemia within 1 h, which was sustained for 4 h. Our data for the first time demonstrate a rapid DZ induced increase in coronary luminal LPL activity and protein indicative of posttranslational processing. In support of this idea, changes in luminal LPL activity following 4 days of STZ diabetes was independent of shifts in mRNA levels or rates of synthesis (unpublished observation). Furthermore, in adipose tissue, following nutritional changes short-term change in LPL is not associated with changes in LPL mRNA or rate of biosynthesis [29,30].

Agents that increase intracellular cAMP enhance luminal LPL [31,32]. As DZ, a phosphodiesterase inhibitor also increases cAMP [33,34], we were concerned about the direct effects of this agent on cardiac LPL. However, as perfusion of the isolated heart with DZ did not increase LPL, the increased enzyme observed in vivo likely occurred subsequent to a drop in insulin. Confirmation of the inhibitory role of in vivo insulin on cardiac LPL was supported by its ability to reverse DZ induced augmentation of luminal LPL.

In the present study, the contribution of this augmented LPL towards FA utilization is unclear, given that serum FA are also elevated subsequent to DZ. In a previous study that compared the utilization of exogenous FA (NEFA) and lipoprotein-TG, alone or in combination, using the isolated working control rat heart, a cardiac preference for NEFA was suggested [35]. More recently, studies examining the in vivo mechanism of FA uptake showed that LPL mediated TG hydrolysis was the major supplier of FA for cardiac utilization [4]. Whether the latter circumstance occurs in a setting of compromised glucose utilization by the heart is uncertain. Our data suggest that following acute hypoinsulinemia and impaired glucose utilization, manifold changes in lipoprotein-TG, serum FA and luminal LPL occur to guarantee FA supply to the heart.

LPL is a distinctive protein that migrates towards the lumen from underlying parenchymal cells [36,37]. During excessive luminal requirement for LPL, an anticipated translocation of the enzyme from the myocyte cell surface to the vascular lumen would be predicted. Indeed subsequent to DZ administration, a 50% drop in heparin releasable surface bound myocyte LPL was observed. Our results suggest that this enzyme pool is a key participant for supplementing LPL at the luminal surface following an increased cardiac demand for FA. However, as smooth muscle derived LPL can translocate to the endothelial surface of vascular tissue, the contribution of this LPL pool in explaining the effect of DZ cannot be completely disregarded [38]. Similarly, at least in human myocardium, LPL protein was detected on interstitial cells and could be equally contributive in explaining the effects of DZ [39].

Mediators responsible for the cleavage and transfer of LPL from myocyte cell surface to the lumen have not been identified. At least in adipocytes, Lyso-PC (50–100 μM) formed from lipoprotein breakdown has been shown to either directly [12] or indirectly [11], through the release of heparanase like compounds, displace surface bound LPL from adipocytes. In the current study, 100 μM Lyso-PC induced a severe cardiodepressant action in the isolated heart. However, exposure of control hearts to an extremely low concentration of Lyso-PC (1 nM) was able to enhance heparin releasable luminal LPL to levels observed following DZ. Whether this effect of Lyso-PC in the heart is related to its ability to release heparanases or directly cleave myocyte LPL merits further investigation. At least at the luminal surface Lyso-PC was unable to release LPL directly. The contribution of TG lipolysis in vivo towards generation of equivalent amounts of Lyso-PC is currently unknown. Although it would be useful to measure plasma Lyso-PC following DZ, it is unlikely that if DZ increases Lyso-PC in vivo, we would be able to detect this trivial amount in the plasma. More importantly, Lyso-PC release in vivo may not be retained in the plasma due to rapid transport into the adjacent tissues.

LPL possesses both TG hydrolase and phospholipase activity [40]. Thus following DZ, an increase in serum TG could augment formation of lysophophospholipids like Lyso-PC releasing endothelial heparanase like compounds that enable LPL translocation from the cardiomycoyte cell surface to the luminal surface. Given the potential importance of circulating TG breakdown in influencing cardiac substrate switching, we attempted to manipulate circulating TG lipolysis using WR 1339. This agent, by virtue of its inhibitory effect on LPL mediated TG hydrolysis brought about a 30-fold increase in serum TG. Cardiac LPL activity remained unaffected in these animals. Interestingly, although serum TG remained elevated following DZ, its ability to enhance luminal LPL was dramatically restricted. These data emphasize the importance of luminal lipolysis with its potential release of Lyso-PC, and not absolute TG, in regulating cardiac LPL. Under conditions of impaired glucose utilization, inhibiting breakdown of circulating TG would compel the heart to utilize FA. Indeed in DZ-treated animals pretreated with WR 1339 we observed a threefold decrease in serum FA. Prevention of TG accumulation by adenosine mediated inhibition of adipose tissue lipolysis was equally effective in lowering DZ induced hypertriglyceridemia, and the increase in cardiac LPL. Taken together our data suggest that TG hydrolysis is a key trigger for LPL transfer to the luminal surface.

In summary, following hypoinsulinemia, there is a rapid amplification of LPL at the coronary luminal surface. As lipolytic byproducts like Lyso-PC could allow transfer of LPL from the myocyte surface to the vascular lumen, luminal TG hydrolysis could be regarded as an important regulator of “functional” LPL. Transgenic mouse lines overexpressing human LPL in skeletal and cardiac muscle demonstrated cardiomyopathy characterized by muscle fiber degeneration, extensive lipid deposition and proliferation of mitochondria and peroxisomes [41]. In a more recent study, mice that specifically overexpressed cardiomyocyte surface bound functional LPL, exhibited lipid oversupply and impaired contractile function [42]. Whether the increase in luminal LPL during diabetes is a major cause of metabolic, morphological, and mechanical changes observed in the heart requires further investigation.