Oxidized Low-density Lipoprotein- and Lysophosphatidylcholine-induced Ca²⁺ Mobilization in Human Endothelial Cells

Abstract

The effects of oxidized low-density lipoprotein (OxLDL) and its major lipid constituent lysophosphatidylcholine (LPC) on Ca²⁺ entry were investigated in cultured human umbilical endothelial cells (HUVECs) using fura-2 fluorescence and patch-clamp methods. OxLDL or LPC increased intracellular Ca²⁺ concentration ([Ca²⁺]_i), and the increase of [Ca²⁺]_i by OxLDL or by LPC was inhibited by La³⁺ or heparin. LPC failed to increase [Ca²⁺]_i in the presence of an antioxidant tempol. In addition, store-operated Ca²⁺ entry (SOC), which was evoked by intracellular Ca²⁺ store depletion in Ca²⁺-free solution using the sarcoplasmic reticulum Ca²⁺ pump blocker, 2, 5-di-t-butyl-1, 4-benzohydroquinone (BHQ), was further enhanced by OxLDL or by LPC. Increased SOC by OxLDL or by LPC was inhibited by U73122. In voltage-clamped cells, OxLDL or LPC increased [Ca²⁺]_i and simultaneously activated non-selective cation (NSC) currents. LPC-induced NSC currents were inhibited by 2-APB, La³⁺ or U73122, and NSC currents were not activated by LPC in the presence of tempol. Furthermore, in voltage-clamped HUVECs, OxLDL enhanced SOC and evoked outward currents simultaneously. Clamping intracellular Ca²⁺ to 1 µM activated large-conductance Ca²⁺-activated K⁺ (BK_Ca) current spontaneously, and this activated BK_Ca current was further enhanced by OxLDL or by LPC. From these results, we concluded that OxLDL or its main component LPC activates Ca²⁺-permeable Ca²⁺-activated NSC current and BK_Ca current simultaneously, thereby increasing SOC.

INTRODUCTION

Low-density lipoprotein (LDL) can be oxidized in many types of cells, including monocytes, neutrophils, and fibroblasts, and macrophages, endothelial cells and smooth muscle cells among them are the most likely to contribute to LDL oxidation in vascular wall, thereby inducing atherosclerotic lesions (Young et al., 2001). Thus, oxidized LDL (OxLDL) plays a key role in atherosclerosis (Steinberg et al., 1989; Steinbrecher et al., 1990; Berliner et al., 1996). Taken up by macrophages, OxLDL leads to foam cell formation that is characteristic of the earliest atherosclerotic lesion (Ross, 1993; Steinberg, 1997).

OxLDL is known to evoke endothelial dysfunction by increasing oxidative stress (Ross, 1993; Mehta, 2004; Mehta et al., 2006). OxLDL impairs endothelium-dependent vasodilation (Itagaki et al., 2003), and damages endothelial cells to lose endothelial integrity (Hessler et al., 1983). The major lipid constituent of OxLDL, lysophosphatidylcholine (LPC), plays an important role in OxLDL-induced endothelial dysfunction (Takeshita et al., 2000; Matsubara et al., 2005). LPC increases oxidative stress by generating reactive oxygen species, and thereby, causes endothelial dysfunction (Muller et al., 2002; Zmijewski et al., 2005).

On the other hand, OxLDL or LPC has been reported to change electrophysiological properties of cells (Kuhlmann et al., 2003; Fearon, 2006) and increase intracellular Ca²⁺ concentration ([Ca²⁺]_i) in endothelial cells (Weisser et al., 1992; Thorin et al., 1995; Zhao et al., 1997; van Tits et al., 2000). LPC activates nonselective cation (NSC) current in vascular smooth muscle cells (Jabr et al., 2000; Terasawa et al., 2002), and OxLDL activates large-conductance Ca²⁺-activated K⁺ (BK_Ca) current in human umbilical vein endothelial cells (HUVECs). In addition, OxLDL enhances L-type Ca²⁺ currents via LPC-induced mitochondrial reactive oxygen species production (Fearon, 2006). However, little is known about the mechanisms by which OxLDL or LPC increases [Ca²⁺]_i in endothelial cells.

The present study was designed to elucidate how OxLDL or its main component LPC increases [Ca²⁺]_i in endothelial cells, and evidence indicates that OxLDL or LPC activates Ca²⁺-permeable Ca²⁺-activated NSC current and BK_Ca current simultaneously, thereby increasing Ca²⁺ entry.

METHODS

Endothelial cell culture

Human umbilical vein endothelial cells (HUVECs, CRL-1730) were purchased from the American Type Culture Collection and cultured as monolayer in Medium 199 (Hyclone, Logan, UT) supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin, 100 µg/ml streptomycin and 15 µg/ml endothelial cell growth supplement, or Eagle basal medium 2 (EBM2, Clonetics, Walkersville, MD) supplemented with 2% FBS, 0.1% vascular endothelial growth factor, 0.1% ascorbic acid, 0.1% gentamycin sulfate amphotericin-B, 0.04% hydrocortisone and 0.1% heparin. All cells were maintained at 37℃ in humidified condition under 5% CO₂. The cells were then detached by exposure to trypsin, re-seeded on gelatine-coated cover slips, and maintained in culture for 2~4 days before use. Measurements were performed on non-confluent cells.

Electrophysiology

Electrophysiological methods and Ca²⁺ measurements have previously been described in detail (Nilius et al., 1994). Membrane potential was monitored in current clamp mode or controlled in voltage clamp mode with an EPC-9 (HEKA Elektronik, Lambrecht, Germany), using a nystatin-perforated patch (100 mg/ml). Whole-cell currents were measured using ruptured patches. Voltages were monitored in voltage clamp mode with an EPC-9 (HEKA Elektronik, Lambrecht, Germany, sampling rate 1 ms, 8-Pole Bessel filter 2.9 kHz). Holding potential of whole cell experiment was 0 mV or -60 mV. We applied a voltage ramp from -100 mV or -150 mV to +100 mV every 10 seconds with duration of 650 ms. Currents were recorded at a sampling rate of 1 kHz to 4 kHz.

Ca²⁺ measurement

Cells were loaded with Fura-2/AM (the acetoxymethyl ester form), and [Ca²⁺]_i was measured using a microfluorimeter consisting of an inverted microscope (DM IRB, Leica, Germany) and a PTI filter scan power illuminator system (Photon Technology International Inc, Canada). Fura-2/AM (2 µM) was added to the bath, and the cells were incubated for 25 min at 37℃. The cells were illuminated alternatively at wavelengths of 340 and 380 nm through a chopper wheel (frequency of 50 Hz). Fluorescence was measured at 510 nm and autofluorescence was subtracted from the signals obtained. The free Ca²⁺ concentration was calculated from the ratio of the fluorescence signals emitted at each excitation wavelength. The calibration procedure was identical to that described previously (Nilius et al., 1993; Nilius et al., 1994).

Solutions

The standard extracellular solution was an iso-osmolar Krebs solution, containing (in mM) 150 NaCl, 6 KCl, 1 MgCl₂, 1.5 CaCl₂, 10 glucose, 10 HEPES, pH 7.4 with NaOH. The osmolality of this solution, as measured with a vapour pressure osmometer (Wescor 5,500, Schlag, Gladbach, Germany), was 320±5 mOsm. For simultaneous measurement of membrane current and [Ca²⁺]_i, we used this external solution and an internal high K⁺ solution, unless mentioned otherwise. The internal high K⁺ pipette solution contained (in mM) 40 KCl, 100 K-aspartate, 2 MgCl₂, 0.1 EGTA, 4 Na₂ATP, 10 HEPES, pH 7.2 with KOH (290 mOsm).

In some experiments, we abolished the activity of large conductance Ca²⁺ activated K⁺ (BK_Ca) channels by substituting extra- and intracellular K⁺ with Cs⁺. The internal high Cs⁺ pipette solution contained 145 mM Cs-glutamate, 8 mM NaCl, 2 mM MgCl₂, 10 mM HEPES, and 1 mM Na₂ATP, buffered with CsOH to 7.2. The osmolality of all external solutions was increased to 420±5 mosmol/kg with 100 mM mannitol (Voets et al., 1997) to eliminate volume activated Cl^- currents. For buffering free Ca²⁺, the appropriate amount of Ca²⁺ (calculated by the program CaBuf, G. Droogmans) was added in the presence of 5 mM EGTA.

Chemicals

Amino-ethoxybiphenyl borane (2-APB), 2,5-di-t-butyl-1,4-benzohydroquinone (BHQ), L-α-lysophosphatidyl choline (LPC) from egg yolk, Tempol and U73312 were purchased from Sigma (St. Louis, MO). OxLDL was obtained from Intracel Inc (Frederick, MD), and Fura-2 AM from Molecular Probe (Eugene, OR). U73312, LPC, 2-APB and nystatin were applied from a stock solution in DMSO. The final concentration of DMSO was less than 0.05%.

All experiments were performed at 37℃. Pooled data are presented as means±SEM and significant differences were detected using the Student's t test (p<0.05).

RESULTS

OxLDL- or LPC-induced increase of intracellular Ca²⁺

Effect of OxLDL or of LPC on [Ca²⁺]_i is illustrated in Fig. 1. OxLDL (20 µg/ml) or LPC (3 µgM) increased [Ca²⁺]_i in the presence of extracellular Ca²⁺ (Fig. 1A and C). OxLDL reversibly increased [Ca²⁺]_i, whereas increased [Ca²⁺]_i by LPC did not return to a resting level for more than one hours (Fig. 1A). This increase in [Ca²⁺]_i by OxLDL or LPC was inhibited by 10 µM La³⁺ (Fig. 1B), 50 IU/ml heparin (Fig. 1C) or depleting extracellular Ca²⁺ (data not shown). In addition, LPC failed to increase [Ca²⁺]_i in the presence of an antioxidant tempol (Fig. 1E). In Ca²⁺-free solution, the increase of [Ca²⁺]i by OxLDL (Fig. 1D) or by LPC (data not shown) was transient and there was no sustained Ca²⁺ plateau, indicating that the Ca²⁺ plateau is mainly due to influx of Ca²⁺ and that Ca²⁺-entry pathway (s) is (are) activated by OxLDL or by LPC. Thus, we next examined whether OxLDL or LPC activated store-operated Ca²⁺ entry. Fig. 1F shows that store-operated Ca²⁺ entry was activated by emptying intracellular Ca²⁺ stores using BHQ, an inhibitor of sarcoplasmic reticulum Ca²⁺ pump. Exposure of HUVECs in Ca²⁺-free solution to 20 µM BHQ transiently increased [Ca²⁺]_i and then [Ca²⁺]_i decreased slowly to a steady state within 10~15 min after BHQ application. Then, we added Ca²⁺ to the Ca²⁺-free solution. The subsequent reapplication of extracellular Ca²⁺ caused an increase of [Ca²⁺]_i (Fig. 1E), and this increase of [Ca²⁺]_i, dependent on extracellular Ca²⁺, obviously proceeds via Ca²⁺ entry pathways which was activated during this protocol (store-operated Ca²⁺ entry). When OxLDL (Fig. 1E) or LPC (Fig. 1F) was applied after store-operated Ca²⁺ entry was evoked, store-operated Ca²⁺ entry was further enhanced by OxLDL or LPC (Fig. 1F).

Fig. 1.

Effect of OxLDL or LPC on intracellular Ca²⁺ concentration ([Ca²⁺]_i) in HUVECs. (A, C and D) OxLDL-induced increase of [Ca²⁺]_i in the presence of (A, C) or in the absence of extracellular Ca²⁺ (D). (B) in the presence of an antioxidant tempol, LPC failed to increase [Ca²⁺]_i. (E and F) The increase of store-operated Ca²⁺ entry by OxLDL (E) or LPC (F).

OxLDL- or LPC-induced activation of nonselective cation current

The increase of [Ca²⁺]_i via Ca²⁺ entry pathways, which were activated by the above protocol, could be modulated by changing membrane potential, which affects the driving force for Ca²⁺ entry. Thus, we clamped membrane potential of cells to 0 mV to eliminate the effect of membrane potential on Ca²⁺ entry, and then examined whether OxLDL enhances store-operated Ca²⁺ entry (Fig. 2). In voltage-clamped cells, we simultaneously recorded membrane current and [Ca²⁺]_i. As shown in Fig. 2, store-operated Ca²⁺ entry was evoked by the reapplication of extracellular Ca²⁺ after exposure of HUVECs in Ca²⁺-free solution to 20 µM BHQ. The reapplication of extracellular Ca²⁺ increased [Ca²⁺]_i. When an increase of [Ca²⁺]_i reached a steady state, OxLDL was applied. As shown in the figure, OxLDL activated large inward and outward currents and further enhanced the evoked store-operated Ca²⁺ entry simultaneously (Fig. 2). Substitution of external monovalent cations by NMDG⁺ completely abolished the inward currents (Fig. 3A), indicating that the current is carried by cations. These results suggest that OxLDL activates Ca²⁺-permeable NSC currents in HUVECs, thereby increasing store-operated Ca²⁺ entry, independently of membrane potential.

Fig. 2.

Effect of OxLDL on store-operated Ca²⁺ entry in a voltage-clamped cell. Membrane currents and [Ca²⁺]_i were measured simultaneously at a holding potential of 0 mV in nystatin-perforated mode. (A) further activation of store-operated Ca²⁺ entry (SOC) by OxLDL. SOC was activated by depletion of the intracellular Ca²⁺ store and re-exposure to Ca²⁺. (B) time course of membrane currents at +50 and -50 mV activated by a rise in [Ca²⁺]_i (A). Membrane currents were activated by repetitive ramps from -100 to +100 mV. (C) current-voltage relations obtained at points indicated 1~3 in (B).

Fig. 3.

Effect of LPC on nonselective cation (NSC) currents in HUVECs. Membrane potential was held at 0 mV. (A, C and E) time courses of membrane currents obtained at +50 and -50 mV during repetitive ramps from -100 to +100 mV. (B, D and F) current-voltage relations obtained at points indicated 1~9 in (A), (C) and (E), respectively. (G) bar graph showing current density at +50 mV. n=8, ^*p<0.001.

LPC also evoked inward and outward currents (Fig. 3). The current was time-independent at each potential and was not inactivated during the voltage steps (data not shown). The reversal potentials of these currents were close to 15 mV, and the current-voltage relationships showed no rectification (Fig. 3). This LPC-activated NSC current was decreased by the inositol 1,4,5-triphosphate (IP₃) -gated Ca²⁺ channel blocker 2-APB (Fig. 3A), La³⁺ (Fig. 3C) or the phospholipase C inhibitor U-73122 (Fig. 3C). Furthermore, LPC failed to activate NSC currents in the presence of an antioxidant tempol (5 µM, Fig. 3E).

OxLDL activates BK_Ca current

In some cells, OxLDL or LPC augmented store-operated Ca²⁺ entry and evoked outward currents simultaneously (Fig. 4). In voltage-clamped cells, store-operated Ca²⁺ entry was activated by emptying intracellular Ca²⁺ stores using BHQ, and [Ca²⁺]_i was increased by adding Ca²⁺ to the Ca²⁺-free solution. When the increase of [Ca²⁺]_i reached a steady state and OxLDL was applied, the increased [Ca²⁺]_i was further augmented by OxLDL application and outward currents was simultaneously developed (Fig. 4).

Fig. 4.

Effect of OxLDL on store-operated Ca²⁺ entry and membrane currents in a voltage-clamped cell. Membrane currents (B) and [Ca²⁺]_i (A) were measured simultaneously at a holding potential of 0 mV in nystatin-perforated mode.

Since large conductance Ca²⁺-activated K⁺ (BK_Ca) current produces outward currents in HUVECs, we examined whether OxLDL or LPC modulates BK_Ca current. Loading cells with a Ca²⁺ solution buffered at 1 µM activated outwardly rectifying current (Fig. 5). This outwardly rectifying current was inhibited by iberiotoxin (data not shown), suggesting that this outwardly rectifying current is BK_Ca current. Then, OxLDL or LPC was applied. With the application of OxLDL, this activated BK_Ca current of 62.7±4.8 pA/pF was further enhanced to 126.4±34.2 pA/pF (Fig. 5A-C). Like OxLDL, LPC also augmented BK_Ca currents (Fig. 5D).

Fig. 5.

Effect of OxLDL and LPC on BK_Ca currents. BK_Ca currents were activated by clamping [Ca²⁺]_i at 1 (A-C) or 0.5 µM (D). Membrane potential was held at 0 mV. (A and D) time course of membrane current change by OxLDL (A) or LPC (D) measured at -50 and +100 mV. (B) current-voltage relationship obtained from the voltage ramps from -100 to +100 mV labeled as a and b in (A). (C) bar graph showing current density at +100 mV. (C) control. OxLDL, maximal current density obtained during OxLDL (20 µg/ml) application; w/o, wash-out. n=17, ^*p<0.001.

DISCUSSION

Present study suggests that OxLDL or its major component LPC activates Ca²⁺-permeable NSC current and BK_Ca current, thereby activating store-operated Ca²⁺ entry. OxLDL or LPC might activate NSC current by releasing Ca²⁺ from intracellular stores. This Ca²⁺-activated NSC might be Ca²⁺-permeable, therefore acts as Ca²⁺ entry pathway. In addition, OxLDL or LPC activates BK_Ca current. BK_Ca current activation hyperpolarizes membrane, thereby increasing the driving force for Ca²⁺ entry. Thus, OxLDL or LPC promotes Ca²⁺ entry through NSC channel via activating BK_Ca current.

The present results suggest that OxLDL- or LPC-activated NSC channel in HUVECs is virtually permeable to Ca²⁺, inasmuch as an enhancement of store-operated Ca²⁺ entry is accompanied by the activation of this NSC currents. This is consistent with previous findings that Ca²⁺-activated NSC channel in endothelial cell line derived from human umbilical vein (EA cells) is permeable to Ca²⁺ (Kamouchi et al., 1999; Suh et al., 2002). Kamouchi et al. reported that the relative permeability for Ca²⁺ over Na⁺ in the NSC in EA cells is 0.07±0.01. In addition, Nilius et al. (1990) reported that the relative permeability in histamine-activated NSC current is 0.2. Thus, it is highly likely that Ca²⁺ entry pathway to evoke store-operated Ca²⁺ entry is NSC channel.

Our present results suggest that NSC current activation by OxLDL or LPC is induced by a [Ca²⁺]_i-dependent mechanism. NSC current in EA cells is activated by an increase in [Ca²⁺]_i (Suh et al., 2002). Thus, vasoactive agonists, such as ATP, histamine and bradykinin, which can increase [Ca²⁺]_i, activate NSC current in EA cells (Kamouchi et al., 1999). Like NSC current in EA cells, NSC current in HUVECs was also activated by increasing [Ca²⁺]_i and by vasoactive agents to increase [Ca²⁺]_i. Therefore, OxLDL, LPC and ATP might activate NSC current via releasing Ca²⁺ from intracellular stores in HUVECs.

OxLDL or LPC increases [Ca²⁺]_i in HUVECs and the increase of [Ca²⁺]_i activate BK_Ca current, because BK_Ca currents are activated dependently of intracellular Ca²⁺. This explains the mechanism of how OxLDL (Kuhlmann et al., 2003) or LPC activates BK_Ca current. In addition, OxLDL or LPC might increase the sensitivity of BK_Ca channels to Ca²⁺, because OxLDL or LPC augmented BK_Ca current further which was already activated by clamping [Ca²⁺]_i to 1 µM. BK_Ca current activation can hyperpolarize membrane, thereby increasing the driving force for Ca²⁺ (Kamouchi et al., 1999; Kim et al., 2006). Thus, OxLDL or LPC might facilitate Ca²⁺ entry through Ca²⁺-permeable Ca²⁺-activated NSC channel via activating BK_Ca current.

Furthermore, OxLDL and LPC might increase the sensitivity of NSC channel to Ca²⁺. In the present study, the increase of Ca²⁺ entry might have been caused by increased driving force for Ca²⁺ and/or by further activation of Ca²⁺-activated NSC channel. Since OxLDL increased Ca²⁺ entry in voltage-clamped cells, further activation of Ca²⁺-activated NSC channel by OxLDL or by LPC might evoke increased Ca²⁺ entry, suggesting that OxLDL or LPC increases the sensitivity of Ca²⁺-activated NSC channel to intracellular Ca²⁺.

In conclusion, we have shown that OxLDL activates NSC current and BK_Ca current, thereby increasing [Ca²⁺]_i. These effects of OxLDL on HUVECs are mimicked by its main lipid constituent LPC, and therefore, LPC mediates these responses on endothelial cells. These data explain the mechanisms involved in physiological and pathological regulation of endothelial cells when exposed to OxLDL.

ABBREVIATIONS

oxLDL

oxidized low-density lipoprotein

LPC

lysophosphatidylcholine

HUVECs

human umbilical endothelial cells

[Ca²⁺]_i

intracellular Ca²⁺ concentration

SOC

store-operated Ca²⁺ entry

BHQ

2,5-di-t-butyl-1, 4-benzohydroquinone

NSC currents

non-selective cation currents

BK_Ca current

large-conductance Ca²⁺-activated K⁺ current

2-APB

amino-ethoxybiphenyl borane