OUP user menu

Impaired vascular function in small resistance arteries of LOX-1 overexpressing mice on high-fat diet

Birgit Eichhorn , Gregor Muller , Anja Leuner , Tatsuya Sawamura , Ursula Ravens , Henning Morawietz
DOI: http://dx.doi.org/10.1093/cvr/cvp089 493-502 First published online: 16 March 2009


Aims LOX-1 is a major vascular receptor for oxidized low-density lipoprotein (oxLDL). In this study, we analysed the impact of LOX-1 overexpression and high dietary fat intake on vascular function in small resistance arteries.

Methods and results Relaxation of mesenteric arteries was measured using a wire myograph. Compared with the control group, mice overexpressing LOX-1 on a high-fat diet (FD) had preserved vascular smooth muscle relaxation, but impaired endothelium-dependent relaxation via NO. Vascular NO availability was decreased by exaggerated formation of reactive oxygen species and decreased endothelial NO synthase expression. Endothelium-derived hyperpolarizing factor (EDHF)-mediated relaxation via cytochrome P450 metabolites was increased in LOX-1 + FD animals, but did not completely compensate for the loss of NO. Currents of calcium-activated potassium channels with large conductance (BKCa channels) were measured by the voltage-clamp method. The BKCa current amplitudes were not altered in endothelial cells, but highly increased in vascular smooth muscle cells from resistance arteries of LOX-1-overexpressing mice on FD. BKCa currents were activated by low-dose H2O2 and cytochrome P450 metabolites 11,12-EET and 14,15-EET as EDHF in control mice.

Conclusion LOX-1 overexpression and FD caused functional changes in endothelial and vascular smooth muscle cells of small resistance arteries.

  • Endothelial function
  • K+-channel
  • Lipoproteins
  • Smooth muscle
  • Lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1)

1. Introduction

Atherosclerosis, with its clinical manifestation in cardiovascular diseases, is the major cause of death in industrialized countries. Functional changes in endothelial cells (ECs) and vascular smooth muscle cells (VSMCs) contribute to the initiation and early progression of cardiovascular diseases like atherosclerosis.13 Resistance arteries do not show morphological alterations in response to high-fat diet (FD) or oxLDL, but rather develop functional impairment.4 Several changes in the early phase of endothelial dysfunction are associated with high plasma levels of lipoproteins. Circulating low-density lipoproteins (LDL) can be modified to oxidized LDL (oxLDL). The major receptor of oxLDL is the lectin-like oxLDL receptor-1 (LOX-1) in the vessel wall.57 LOX-1 mediates endocytosis of oxLDL in ECs,8 VSMCs, and monocytes.9 Basal LOX-1 expression is low, but several pathophysiological conditions like hypertension, diabetes mellitus, and hyperlipidaemia and the development of atherosclerotic lesions have been linked with an increased vascular LOX-1 expression.10 The G501C mutation in the lectin-like oxidized LDL receptor gene (LOX-1/OLR1) has been associated with the risk of myocardial infarction,11 but not with the risk for stroke.12 LOX-1 expression is increased in human atherosclerotic lesions (in early lesions mainly in ECs, in advanced lesions also in VSMCs and macrophages).13 Moreover plaque formation is enhanced in coronary arteries of mice overexpressing LOX-1 against a genetic background of apolipoprotein E deficiency.14 Cell-culture studies have shown that the endothelial generation of reactive oxygen species (ROS) by NAD(P)H oxidase complexes in response to oxLDL is mediated by LOX-1.15 Furthermore, in vitro studies indicate that activation of LOX-1 also initiates a reduction in NO release.15 However, little is known about the contribution of LOX-1 to vascular homeostasis and endothelial dysfunction in small vessels. The intact endothelium plays an important role in vascular function by synthesizing and releasing vasodilating factors.16 Major vasodilating factors in arteries are the endothelium-derived hyperpolarizing factor (EDHF), nitric oxide (NO), and prostacyclin. The reduced NO availability in atherosclerosis can be mediated via decreased expression of the endothelial NO synthase (eNOS) or via inactivation of NO by ROS.2 In addition to eNOS expression, there is also evidence of reduced eNOS activity by lack of cofactors (e.g. BH4) and increased formation of endogenous inhibitors.17 The impaired NO-mediated relaxation in different vessels of hypercholesterolaemic and atherosclerotic animal models or patients18,19 can be compensated by EDHF. Several components for EDHF signalling have been proposed including electrical coupling through gap junctions, certain ROS as for instance H2O2, cytochrome P450 metabolites, and vascular Ca2+-activated K+ channels (KCa)20 in particular those with large conductance (BKCa channels).2123 Activation of BKCa channels facilitates relaxation by cell membrane hyperpolarization. Nevertheless their role in endothelial and smooth muscle dysfunction is not completely understood. Binding of oxLDL to LOX-1 was shown to modulate BKCa-channels in ECs in vitro.15

In this study, we have used mice overexpressing bovine LOX-1 that were fed an FD to examine a potential functional impairment of small resistance arteries (mesenteric arteries). Based on our experimental findings, we provide evidence that LOX-1 receptors can cause vascular dysfunction in resistance vessels.

2. Methods

2.1 Animals

Male wild-type mice (C57BL/6, WT) and mice overexpressing bovine LOX-1 in a C57BL/6 background under control of the murine preproendothelin-1 promoter (LOX-1 mice14), aged 8 weeks, were fed standard chow EF R/M CD88137 or FD EF R/M TD88137 (Ssniff Spezialitäten GmbH, Soest, Germany) for 10 weeks. LOX-1 mice were kindly provided by T.S., Department of Vascular Physiology, National Cardiovascular Center Research Institute Fujishirodai, Suita, Osaka, Japan. The generation and characterization of the mice has been recently described by Inoue et al.14 The LOX-1 mice carry 24 copies of the transgene, resulting in an approximately eight-fold higher mRNA-expression and a marked upregulation of endothelial LOX-1 protein expression. A similar upregulation of the LOX-1 protein has been described in ECs of human carotid arteries, covering early atherosclerotic lesions.13 Overexpression of the bovine LOX-1 was verified by PCR in mesenteric arteries (see Supplementary material online, Figure S1). All performed experiments are in accordance 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). The animal research Ethics Committee of the Dresden University of Technology and the Regierungspräsidium Dresden approved the animal facilities and the experiments according to institutional guidelines and German animal welfare regulations (AZ: 24-9168.24-1-2003-13, 24D-9168.24-1/2006-16).

2.2 Serum lipid measurements

Serum lipids were measured at the Institute of Clinical Chemistry and Laboratory Medicine (University of Technology Dresden) using kits for triglycerides, cholesterol, HDLs and LDLs (Roche Diagnostics GmbH, Mannheim, Germany).

2.3 Preparation of mesenteric arteries for in vitro studies

Arteries (third-order branch) were dissected and maintained in Ca2+-free physiological salt solution [PSS; mmol/L: NaCl 119.0; KCl 4.7; MgSO4 1.17; NaHCO3 25.0; KH2PO4 1.18; glucose 5.5, and ethylenediaminetetraacetic acid (EDTA) 0.027; pH 7.4].

2.4 Superoxide anions

Dissected vessels were incubated in Krebs–Henseleit buffer (mmol/L: NaCl 115.0; NaHCO3 25.0; KCl 4.0; KH2PO4 0.9; MgSO4×7H2O 1.1; CaCl2 2.6; glucose 5.5; pH: 7.4) for 30 min at 37°C. Lucigenin (5 µmol/L — a concentration below the threshold of redox cycling24) and NADPH (100 µmol/L) were dissolved in Krebs–Henseleit buffer for determination of ROS. Lucigenin solution containing additional 200 U/mL superoxide dismutase (SOD) and 380 U/mL catalase was used to examine superoxide anions and the resulting H2O2 formation.25 Solution without tissue served as control. Photoemission was detected every second for 30 min in a Fluorescence Microplate Reader Fluorimeter FLUOstar OPTIMA (BMG LABTECH, Jena, Germany). The length of the blood vessels was measured with an eye piece scale (ZEISS, Jena, Germany) and used for data normalization. The increase in ROS production in animals fed with an FD was normalized as 100% of the corresponding control.

2.5 Measurement of contractile function

Small sections of mesenteric arteries (length 2 mm) were mounted in microvascular myographs for isometric tension recordings as described previously26 and maintained in oxygenated PSS (5% CO2 in 95% O2; 1.6 mmol/L CaCl2) at 37°C. During equilibration of the vessels, tension (T) corresponding to a pressure (P) of 70 mm Hg according to the equation P = T2πU−1 (U = inner circumference) was adjusted. In all experiments cyclooxygenase-mediated relaxation was blocked with cyclooxygenase inhibitor diclofenac (0.1 mmol/L; Sigma, Taufkirchen, Germany). The vessel rings were contracted with cumulatively increasing concentrations of phenylephrine (PE). Relaxation was measured by increasing concentrations of acetylcholine (ACh) or sodium nitroprusside (SNP) in PE-precontracted (10 µmol/L) vessels. Relaxing effects of ACh were studied in the absence and presence of NO synthase inhibitor nitro-L-arginine-methylester (L-NAME; 30 µmol/L), BKCa inhibitor paxilline (1 µmol/L), cytochrome P450 inhibitor N,N-diethyl-aminoethyl-2,2-diphenylvalerate (proadifen; 50 µmol/L), and epoxygenase inhibitor 6-(2-propargyloxyphenyl)hexanoic acid (PPOH; 30 µmol/L). The effects of ACh or SNP are expressed in percent of the response to PE (=100%).

2.6 Isolation of vascular cells

Mesenteric arteries were stored in low Ca2+-containing PSS (0.16 mmol/L Ca2+) at 4°C. Enzymatic dissociation was carried out in two steps. The first solution (1 mL PPS) contained 0.7 mg papain; 1.5 mg dithioerythritol (Roth, Karlsruhe, Germany), bubbled with O2, 20 min, 37°C. The second solution (1 mL PPS) contained 1.2 mg collagenase type F, 1.5 mg trypsin inhibitor, 0.5 mg elastase (Serva, Heidelberg, Germany), and 1.0 mg bovine albumin fraction V (Serva, Heidelberg, Germany) gassed with O2, 12 min, 37°C. Single VSMCs and ECs were obtained by trituration in PSS.

2.7 Electrophysiological experiments

Potassium outward current through the BKCa channels (IBK,Ca) was measured using a HEKA-EPC8 amplifier (HEKA Elektronik, Lambrecht, Germany) in voltage-clamp mode. Bath superfusion buffer (mmol/L): NaCl 127.0; KCl 5.9; CaCl2 2.4; MgCl2 1.2; glucose 11.0; HEPES 10.0, pH 7.4. Pipette solution (mmol/L): KCl 134.0; NaCl 6.0; MgCl2 1.2; CaCl2 4.2, EGTA 5.0; glucose 11.0; Mg-ATP 3.0; HEPES 10.0; pH 7.4. Pipette tip resistance was 3.0–4.0 MΩ. Experiments were carried out at 21°C. Effects of the following compounds were studied: 1,3-Dihydro-1-[2-hydroxy-5- (trifluoromethyl)phenyl]-5-(trifluoromethyl)-2H-benzimidazol-2-one (NS1619; 30 µmol/L), paxilline (1 µmol/L), H2O2 (1 µmol/L), 11,12-Epoxy-(5Z,8Z,14Z)-eicosatrienoic acid (11,12-EET; 300 nmol/L), and 14(R),15(S)-Epoxy-(5Z,8Z,11Z)-eicosatrienoic acid (14,15-EET; 300 nmol/L). The cell capacity was used for data normalization. Unless stated otherwise, all substances were purchased from Sigma, Taufkirchen, Germany.

2.8 RNA isolation and real-time PCR

Total RNA of mesenteric arteries was isolated using the EZNA Total RNA-Kit (Peqlab, Erlangen, Germany). For quantification of mRNA expression real-time PCR was performed using the QuantiTect SYBR Green RT–PCR kit (Qiagen, Hilden, Germany) in a thermal cycler (Corbett Research, Mortlake, Australia). Primers: bovine LOX-1 (sense: 5′-CCAGGAGAACTGCTTGTCTT-3′, antisense: 5′-GTGCTCTCAATAGATTCGCC-3′), eNOS (sense: 5′-TTCCGGCTGCCACCTGATCCTAA-3′, antisense: 5′-AACATATGTCCTTGCTCAAGGCA-3′), KCNMA (sense: 5′-CGATAAGCTGTGGTTCTGGC-3′, antisense: 5′-AAGAAGACCATGAAGAGGCGTC-3′), and KCNMB1 (sense: 5′-AGAAGCGCGGAGAGACACGA-3′, antisense: 5′-CAGCTCTTCCTGGTCCTTGATA-3′). Quantification by one-step RT real-time PCR included 50°C for 30 min, 95°C for 15 min, subsequent cycles of 94°C for 30 s, 60°C for 30 s, and finally 72°C for 50 s. Internal RNA standards were produced as previously described.27

2.9 Data analysis and statistics

Potency of agonists was determined as −logEC50 [mmol/L]. All results are expressed as mean±standard error of the mean (SEM). Unless stated otherwise, number of experiments is given as arteries or cells from n mice. Student's t-test (unpaired) was used for statistical analysis, differences with P < 0.05 were considered significant. Multiple comparisons were done by one-way ANOVA followed by Bonferroni post hoc test.

3. Results

3.1 Body weight and serum lipids

Table 1 summarizes body weight and serum lipid concentrations in the four investigated animal groups. LOX-1 mice on normal diet were heavier than WT mice, but did not differ in serum lipid concentrations. FD significantly increased body weight in WT and in transgenic mice when compared with control diet. While serum concentrations of triglycerides were not different between the groups, free cholesterol, LDL, and HDL concentrations were significantly higher in WT and LOX-1 animals on high-fat compared with standard diet. The ratio of LDL:HDL increased during FD from 1:7 to 1:4 in WT mice and from 1:8 to 1:4 in LOX-1 mice. Although FD caused significantly larger weight gain in LOX-1 mice compared with WT animals, no statistically significant differences were found in serum lipid parameters.

View this table:
Table 1

Characterization of body weight and serum parameters

Body weight, g (n)27.3 ± 0.8 (8)33.1 ± 1.1** (22)30.7 ± 1.2# (7)36.3 ± 1.5*# (18)
Triglycerides, mmol/L (n)1.1 ± 0.1 (7)1.3 ± 0.1 (24)1.1 ± 0.03 (7)1.3 ± 0.1 (18)
Cholesterol, mmol/L, (n)3.4 ± 0.2 (7)5.6 ± 0.3*** (24)3.3 ± 0.1 (7)5.9 ± 0.3*** (18)
HDL, mmol/L, (n)2.7 ± 0.2 (7)4.5 ± 0.2*** (24)2.4 ± 0.1 (7)4.8 ± 0.3*** (18)
LDL, mmol/L, (n)0.3 ± 0.1 (7)1.1 ± 0.1*** (24)0.3 ± 0.04 (7)1.1 ± 0.1*** (18)
  • *P < 0.05.

  • **P < 0.01.

  • ***P < 0.001 standard diet vs. high-fat diet.

  • #P < 0.05 WT vs. LOX-1.

3.2 Vascular function

Basal tone of mesenteric arteries was unaffected by FD (compare pre-PE control values in Supplementary material online, Figure S2A).

Endothelium-dependent and -independent relaxation in mesenteric arteries was studied in PE-preconstricted vessels. In order to determine the optimum concentration for precontraction, we measured contractile responses to cumulatively increasing PE concentrations. All vessels contracted in the same concentration range [average −logEC50 (mol/L) values between 5.5 and 5.7]. However, the maximum contractile response to PE was significantly lower in the LOX-1 + FD mice than in the other groups (LOX-1 + FD: 1.4 ± 0.1 mN/mm vs. WT: 1.7 ± 0.3 mN/mm; WT + FD: 1.8 ± 0.1 mN/mm; LOX-1: 2.0 ± 0.2 mN/mm; P < 0.05; see Supplementary material online, Figure S2A). Subsequently, arteries in all further experiments were pre-contracted with 10 µmol/L PE. Values of maximum contraction produced by 80 mmol/L KCl were similar in the four groups (see Supplementary material online, Figure S2C). Absolute values of the constriction induced by 100 µM PE and 80 mmol/L KCl in WT and LOX-1 mice are consistent with previously published results.28,29

Endothelium-dependent relaxation was studied by exposing the arteries to increasing concentrations of ACh. The resulting concentration–response curves (CRC; Figure 1A and B) revealed incomplete relaxation by ACh with significantly reduced relaxation in LOX-1 + FD animals (maximum relaxation, Effmax: 67.8 ± 3.4%; P < 0.01) compared with WT, WT + FD, and LOX-1 (Effmax: 89.3 ± 5.0%; 88.1 ± 4.7%; and 88.1 ± 3.6%; Figure 1C). Interestingly, potencies were similar in all groups [average −logEC50 (mol/L) values between 6.8 and 7.0]. The relaxing response to ACh was reduced in the presence of the NO synthase (NOS) blocker L-NAME (30 µmol/L). Relaxation was reduced in WT (42.2 ± 5.6%), WT + FD (39.1 ± 7.5%), and LOX-1 mice (40.1 ± 6.1%), but not in LOX-1 + FD arteries (5.5 ± 5.2%). These results suggest that NO-mediated relaxation was significantly impaired in LOX-1 + FD animals compared with the other groups (Figure 1D). In addition the basal tone during L-NAME incubation increased only in arteries from WT, WT + FD, and LOX-1 animals, but not in LOX-1 + FD mice. In the presence of L-NAME, the differences in PE-induced contractions between LOX-1 + FD mice and the other three animal groups persisted (see Supplementary material online, Figure S2D and E).

Figure 1

Effects of acetylcholine in mesenteric arteries of WT and LOX-1 mice. (A) Concentration–response curves (CRCs) for acetylcholine in arteries of WT mice on standard and high-fat diet without and with L-NAME. (B) CRCs for acetylcholine in arteries of LOX-1 mice on standard and high-fat diet without and with L-NAME. Maximum effects of acetylcholine-induced (C), NO-mediated relaxation (D), and EDHF-mediated relaxations (E) of arteries from WT and LOX-1 mice on standard and high-fat diet. ***P < 0.001 control vs. L-NAME of mice on standard diet. ###P < 0.001 control vs. L-NAME of mice on high-fat diet.

ACh-induced relaxation in the presence of NOS and cyclooxygenase inhibitors is mediated by EDHF. The EDHF-mediated fraction of relaxation was largest in LOX-1 + FD mice (62.6 ± 3.5%) compared with the other groups (WT: 47.0 ± 3.2%; WT + FD: 49.0 ± 4.0%; LOX-1: 48.0 ± 4.1%; P < 0.01; Figure 1E). Cytochrome P450 enzymes are a substantial source of EDHF, they are responsible for the transformation of arachidonic acid into epoxyeicosatrienoic acids (EETs).30,31 The role of these enzymes in ACh-induced relaxation was examined with proadifen, a non-specific blocker of cytochrome P450 isoenzymes and PPOH, a specific epoxygenase blocker. Proadifen and PPOH did not significantly block EDHF-mediated relaxation in WT, WT + FD, and LOX-1 animals, but significantly reduced EDHF-mediated relaxation in LOX-1 + FD (proadifen 41.3 ± 6.3%; P < 0.01 and PPOH 28.7 ± 6.3%; P < 0.05; Figure 2AD). Both compounds changed the efficacy of ACh to a larger extent in LOX-1 + FD mice compared with the other groups, leading to similar levels of relaxation in all four groups.

Figure 2

Concentration–response curve for acetylcholine in the presence of BKCa channel blocker and cytochrome P450 blocker in mesenteric arteries of WT and LOX-1 mice. Concentration–response curves (CRCs) for acetylcholine in the presence of cytochrome P450 blocker proadifen and epoxygenase blocker PPOH in arteries of (A) WT; (B) WT + FD; (C) LOX-1; (D) LOX-1 + FD in combination with L-NAME. CRCs for acetylcholine in the presence of BKCa-channel blocker paxilline on the arteries of (E) WT; (F) WT + FD; (G) LOX-1, and (H) LOX-1 + FD and in the presence and absence of L-NAME. *P < 0.05 proadifen + L-NAME vs. L-NAME; **P < 0.01 PPOH + L-NAME vs. L-NAME.

The contribution of BKCa channels as potential EDHF targets in LOX-1 + FD was tested by blocking the channels with the specific BKCa blocker paxilline (Figure 2EH). Paxilline reduced ACh-mediated relaxation only in LOX-1 + FD (18.42 ± 3.6%; P = 0.0026), but not in WT (0.1 ± 3.3%), WT + FD (1.6 ± 2.5%), and LOX-1 animals (1.1 ± 3.9%), indicating that BKCa channels were activated only in LOX-1 + FD mice. Even in the presence of L-NAME, paxilline was able to inhibit relaxation, indicating a BKCa channel involvement in the EDHF-mediated relaxation. During paxilline incubation the basal tone significantly elevated in arteries from LOX-1 + FD mice compared with WT, WT + FD, and LOX-1 animals (see Supplementary material online, Figure S2F). PE-induced contraction in the presence of paxilline was similar in all four groups (see Supplementary material online, Figure S2G).

Endothelium-independent relaxation was measured using the NO-donor SNP. Potency and efficacy of SNP were calculated from cumulative CRCs (see Supplementary material online, Figure S3) and were found to be similar in all groups [−logEC50 (mol/L): WT: 7.4 ± 0.2; WT + FD: 7.4 ± 0.1; LOX-1: 7.2 ± 0.2; LOX-1 + FD: 7.2 ± 0.1; Effmax: WT: 0.6 ± 2.5%; WT + FD: 1.2 ± 2.6%; LOX-1: 2.8 ± 2.9%; LOX-1 + FD 1.7 ± 0.5%].

3.3 Reactive oxygen species

Since vascular dysfunction is associated with increased ROS production, we have examined the formation of ROS by chemiluminescence in mesenteric arteries. ROS formation was increased in WT + FD and LOX-1 + FD, and the increase was even higher in LOX-1 + FD than in WT + FD animals (see Supplementary material online, Figure S4). We have also determined superoxide anions, because this particular fraction of ROS has been previously shown to result in H2O2 formation and act directly as an EDHF component.32 We found that H2O2 levels were significantly increased in LOX-1 + FD mice (56.2 ± 5.8%) compared with the other groups (WT: 33.2 ± 3.4%; WT + FD: 40.6 ± 8.3%; LOX-1: 39.8 ± 5.7%; see Supplementary material online, Figure S4).

3.4 BKCa currents

Our results provide evidence that EDHF formation is involved in the contractile responses and that EDHF has the largest impact in the LOX-1 + FD group. Since EDHF activates BKCa channels, we measured BKCa currents in smooth muscle and ECs in order to elucidate their contribution to the observed EDHF-mediated response in LOX-1 + FD mice.

All VSMCs exhibited robust BKCa currents. The current amplitude was completely suppressed by the selective BKCa channel-blocker paxilline (1 µmol/L).33 Current densities of IBK,Ca in VSMCs at a test potential of +60 mV were significantly higher in LOX-1 + FD mice (623 ± 93 pA/pF) compared with WT (356 ± 54 pA/pF), WT + FD (346 ± 47 pA/pF), and LOX-1 (324 ± 54 pA/pF; Figure 3A and B). In LOX-1 + FD IBK,Ca voltage dependence was shifted to lower voltages suggesting that the current is already activated at more negative potentials. Moreover, opening of BKCa-channels with the selective channel activator NS1619 (30 µmol/L)34 did not further increase current density in LOX-1 + FD mice, but increased the current in the other groups to almost the same levels as previously observed in LOX-1 + FD (Figure 3C). Interestingly, the IBK,Ca current densities in ECs were similar in all four mice groups (WT: 64 ± 31 pA/pF; WT + FD: 61 ± 12 pA/pF; LOX-1: 63 ± 27 pA/pF; LOX-1 + FD: 68 ± 32 pA/pF), but much smaller than in VSMCs (Figure 3D and E). Next, we examined the effects of three supposed EDHFs, i.e. H2O2, 11,12-EET, and 14,15-EET, on BKCa channel activity in WT mice. At + 60 mV H2O2 (1 µmol/L) significantly increased current density, but not to the same value as observed with NS1619 or in LOX-1 + FD mice. 11,12-EET and 14,15-EET (300 nmol/L each) activated IBK,Ca to the same maximum as NS1619 (Figure 4AC) in WT mice and as the observed amplitude in LOX-1 + FD animals.

Figure 3

Characterization of electrophysiological properties in vascular smooth muscle cells (VSMCs) and endothelial cells (ECs) of mesenteric arteries from WT and LOX-1 mice on standard and high-fat diet. (A) Current density in pA/pF of BKCa currents in VSMCs from WT, WT + FD, LOX-1, and LOX-1 + FD. (B) Current–voltage relationships of BKCa currents of VSMC from WT, WT + FD, LOX-1, and LOX-1 + FD. **P < 0.01 standard diet vs. high-fat diet. (C) Current density in pA/pF of BKCa channels in VSMCs in response to clamp steps to + 60 mV under control condition, with NS1619 (30 µmol/L) and paxilline (1 µmol/L) in WT, WT + FD, LOX-1, and LOX-1 + FD. Please note the different scaling factors for the ordinate. **P < 0.01 control vs. NS1619. (D) Current density in pA/pF of BKCa channels in ECs from WT, WT + FD, LOX-1, and LOX-1 + FD. (E) Comparison of the BKCa currents in VSMCs and ECs from WT. (F) (Above) Pulse protocol; (Below) Typical BKCa current tracings of a VSMC from WT.

Figure 4

EDHF-mediated activation of BKCa currents in vascular smooth muscle cells (VMSCs) of mesenteric arteries from WT. (A) Hydrogen peroxide (1 µmol/L) activated the BKCa currents in VSMCs from WT, and even further increased with additional NS1619 (30 µmol/L). (B) 11,12-EET (300 nmol/L) significantly activated BKCa currents in VSMCs from WT to the same maximum like NS1619 (30 µmol/L). (C) Activation of BKCa currents by 14,15-EET (300 nmol/L) in VSMCs from WT to the similar values like NS1619 (30 µmol/L).

3.5 mRNA expression of BKCa channel subunits and eNOS

Differences in current densities could be due to different expression levels of the relevant channel subunits. Therefore, we have measured mRNA expression of the BKCa-channels using real-time PCR (Figure 5A and B). Both the pore-forming α-subunits and the accessory β1-subunits were expressed in similar amounts in all groups. In contrast mRNA expression of eNOS was lower in LOX-1 + FD mice compared with WT, WT + FD, and LOX-1 animals (Figure 5C). This finding is in agreement with the reduced NO-dependent vascular relaxation in this group of animals. In heart and kidney, we did not detect statistically significant differences in the mRNA and protein expression of the BKCa channel α- and β1-subunits and eNOS between all groups.

Figure 5

Expression of BKCa channel subunits and eNOS in mesenteric arteries. (A) BKCa channel α-subunit (KCNMA) mRNA expression was analysed by real-time PCR using an internal standard. (B) Amount of BKCa channel β1-subunit (KCNMB1) mRNA was quantified by real-time PCR using an internal standard. (C) Expression of eNOS mRNA determined by real-time PCR using an internal standard. WT, wild-type; LOX-1, mice overexpressing LOX-1; FD, high-fat diet.

4. Discussion

The main findings of our study show that mice overexpressing the LOX-1 in combination with FD have markedly increased body weight, display impaired NO-mediated, and enhanced EDHF-dependent relaxation in mesenteric arteries, and have increased vascular ROS production in LOX-1 + FD compared with WT + FD animals. Furthermore, BKCa channel activity in VSMCs was higher in LOX-1 + FD, which could be increased by the EDHFs H2O2 and EETs. In contrast, endothelial BKCa channel activity was unchanged.

4.1 Body weight and plasma lipids

As expected, WT and LOX-1 overexpressing mice gained weight after FD. However, LOX-1 mice on standard diet were also heavier than their corresponding WT animals suggesting an impact of LOX-1 on weight control or metabolic status. Higher body weight in LOX-1 mice was not associated with a significant shift in plasma lipid parameters to a more harmful lipid profile as observed with FD in both WT and LOX-1 mice. In the latter two groups, elevated total plasma cholesterol, LDL, and HDL are in agreement with previous studies.35,36 Furthermore the LDL/HDL quotient shifted to higher values suggesting an enhanced risk for endothelial dysfunction and atherosclerosis.37

4.2 Vascular constriction and relaxation

None of the animals had grossly visible atherosclerotic lesions. Therefore, we tested for vascular dysfunction as an early sign of cardiovascular disease by measuring contraction and endothelium-dependent as well as endothelium-independent relaxation of PE-preconstricted small resistance arterioles. The impact of oxLDL on vascular tone is not fully resolved. OxLDL enhances the basal tone in rabbit cerebral arteries.38 In contrast, oxLDL has no effect on basal tone in rabbit aorta.39 We observed no differences in basal vascular tone in the animal groups fed standard or FD. However, the maximum contractile response to PE was lower in LOX-1 + FD than in the other animals, but potencies were similar. This difference in the vessel contraction between LOX-1 + FD mice and the other three animal groups was not observed in the presence of high, depolarizing potassium concentration or by the combination of PE and paxilline, suggesting a mechanism that involves potassium channel activity. BKCa channels also participate in ACh-/endothelium-mediated relaxation.2123

ACh-stimulated, NO-mediated dilatation was lower in LOX-1 + FD mice than in WT, WT + FD, and LOX-1 animals, suggesting reduced NO availability in LOX-1 + FD. In accordance basal tone during L-NAME incubation was less increased in LOX-1 + FD mice compared with WT, WT + FD, and LOX-1 animals. Effectivity of L-NAME to eliminate NO production has been questioned because in some species, especially in pig and rabbit, endothelium-dependent/NO-mediated relaxation was resistant to eNOS-inhibitors.40 In our experiments with mouse mesenteric arteries, however, relaxation persisting after inhibition of eNOS with L-NAME was fully suppressed with apamin and charybdotoxin, the combination of which is commonly used for EDHF block (data not shown). Possible explanations for decreased NO-mediated relaxation in mesenteric arteries of LOX-1 + FD mice are (i) that ROS production is elevated, (ii) that expression of eNOS mRNA is reduced, and (iii) that EDHF can compensate for the loss of NO.41 In contrast to the reduced NO-mediated relaxation, the EDHF-dependent response was elevated in LOX-1 + FD mice compared with WT, WT + FD, and LOX-1 animals, but could be abolished by a general cytochrome P450 blocker or by an epoxygenase inhibitor. This finding suggests a potential role of cytochrome P450 and its epoxygenases as a compensatory mechanism that makes up for the loss of NO and subsequent endothelial dysfunction in LOX-1 + FD mice. This same amount of EDHF, which was abolished by the cytochrome P450 inhibitors, was also blocked by paxilline in the presence and absence of L-NAME suggesting an essential role of the BKCa channel in the EDHF-mediated relaxation. Block of cytochrome P450 or BKCa channels in LOX-1 + FD led to similar levels of persisting EDHF-mediated relaxation in all four groups. The nature of the persisting EDHF-mediated relaxation was not investigated in this study. As reported before, release of EDHF from ECs is associated with the endothelial activity of calcium-activated potassium channels of small and intermediate conductance,21 but the activation of BKCa channels is involved in the EDHF-mediated relaxation of VSMCs. No relevant differences in endothelium-independent relaxation were detected, i.e. potency and efficacy of SNP were similar in all groups, suggesting comparable properties of VSMC-relaxation in mice on FD and standard diet. Interestingly, an EETs-mediated relaxation in response to 11,12-EET and 14,15-EET was significantly increased in mesenteric arteries with and without endothelium of LOX-1 + FD mice compared with WT, WT + FD, and LOX-1 animals (see Supplementary material online, Figure S5). This could be mediated by the compensatory enhanced activity of the BKCa channels in VSMCs of LOX-1 + FD mice. However, activation is not sufficient to compensate for the loss of NO in the LOX-1 + FD mice.

4.3 Vascular reactive oxygen species productions

Our results and several previous studies have shown that hypercholesterolaemia is associated with impaired endothelium-dependent relaxation in experimental and clinical studies.1,4244 Rapid degradation of endothelium-derived NO by high levels of ROS is thought to be a major mechanism underlying impaired endothelium-dependent vasodilatation under hypercholesterolaemic conditions. It should also be responsible for decreased endothelium-dependent relaxation.45,46 In our experiments, ROS was indeed increased in LOX-1 + FD and WT + FD mice, and to a larger extent in the former than the latter. ROS leads to oxidation of LDL resulting in increased oxLDL levels. Uptake of oxLDL via LOX-1 into ECs activates NAD(P)H oxidases and enhances production of ROS,47,48 starting a vicious cycle finally leading to endothelial dysfunction. However, exaggerated ROS production has not only deleterious effects. Growing evidence supports an import role of redox-sensitive signalling in vascular function.49 Furthermore Shimokawa and Matoba32 have suggested that the ROS product H2O2 could act as EDHF. H2O2 amount was higher in LOX-1 + FD mice than in WT, WT + FD, and LOX-1 animals, suggesting a potential role in the elevated EDHF-mediated relaxation.

4.4 Role of endothelium-derived hyperpolarizing factor and BKCa channel activity

In contrast to reduced NO-mediated responses, EDHF-mediated relaxation was most pronounced in LOX-1 + FD mice suggesting that this mechanism might partly compensate the impaired NO-mediated relaxation. Since different EDHFs, like EETs and H2O2 are known to activate BKCa channels in VSMCs,21 we studied BKCa currents directly. At any activating potential, the amplitude of IBK,Ca was indeed larger in VSMCs from LOX-1 + FD than in the other groups. Moreover, IBK,Ca amplitude in LOX-1 + FD mice could not be further increased by the BKCa channel opener NS1619, suggesting that increased current amplitude resulted from enhanced channel open probability. This interpretation is supported by the fact that mRNA expression of the BKCa channel α- and β1-subunits was similar in all mice. Because of the limited availability of vascular tissue from the mesenteric arteries, no additional data on protein level could be obtained. However, similar mRNA and protein expression of BKCa channels have been demonstrated in human arteries and veins.50 The IBK,Ca amplitudes in ECs were smaller than in VSMCs and did not reveal any differences between the four groups. Therefore BKCa channels in ECs do not contribute to vascular dysfunction in LOX-1 + FD mice. The activation of endothelial BKCa-channels contributes to the EDHF-mediated relaxation. Endothelial hyperpolarization can be transmitted directly via gap junctions to the VSMCs.51,52 In addition, the K+ outward current through BKCa channels increases the extracellular potassium concentration.53 The subsequent activation of the inward-rectifier potassium channel (KIR) and the Na+/K+ pump overcomes the minor depolarizing effects linked to the K+ increase54,55 and relaxes smooth muscle cells.

The possible EDHF H2O2 significantly increased IBK,Ca in VSMCs, although the level did not reach the maximum current amplitude of NS1619 or as in LOX-1 + FD mice. Therefore, H2O2 plays only a minor compensatory role in EDHF-mediated relaxation of this model. Recently, an inhibiting effect of H2O2 on cytochrome P450 has been suggested, indicating a negative feedback mechanism of EET production.56 This could not be confirmed in our experimental model.

Reduced NO levels have been reported to disinhibit cytochrome P450 and hence elevate two other possible EDHF components, 11,12-EET and 14,15-EET.5759 These two EETs were able to enhance IBK,Ca to similar maximum current amplitude as observed in the presence of NS1619 and in LOX-1 + FD. In agreement with previous findings in isolated renal arteries of hypercholesterolaemic rabbits,18,19 we suggest that enhanced formation of EDHF represents a compensatory mechanism of the decreased NO-mediated vessel relaxation. In contrast, Urakami-Harasawa et al.60 found a significant inhibition of endothelium-dependent hyperpolarization in isolated gastroepiploic arteries from atherosclerotic patients. These contradictory results could be explained by the longer duration of hypercholesterolaemic conditions and the progressive development of vascular diseases.

In the last few years, vascular BKCa channels were considered as potential therapeutic targets in the treatment of hypertension, endothelial dysfunction, and other cardiovascular diseases,61 because aldosterone overexpression induces a nitric-oxide-independent coronary dysfunction with decreased VSMC BKCa expression and coronary BKCa-dependent relaxation.62 However activation of LOX-1 in clinical manifestation of atherosclerosis could activate VSMC BKCa channels. In contrast, enhanced LOX-1 expression does not influence endothelial BKCa channels.

In conclusion, we have consistently detected significant changes in contractile and electrophysiological properties of small resistance vessels only in the combination of LOX-1 overexpression and FD. The endothelium-mediated relaxation via NO release was impaired, but partly compensated by the higher release of EDHF. The consequence of this compensatory mechanism was a higher IBK,Ca due to increased open probability of BKCa channels in VSMCs, but not in ECs. Our results clearly demonstrate that LOX-1 overexpression and FD cause functional changes in ECs and VSMCs of small resistance arteries leading to vascular dysfunction as an early sign of cardiovascular diseases.


This work was supported by the German Federal Ministry of Education and Research and the Dr. Robert Pfleger foundation.


The authors would like to thank Dr Melinda Wuest for helpful discussion and Claudia Bodenstein for excellent technical assistance.

Conflict of interest: none declared.


View Abstract