Differential Leukotriene Receptor Expression and Calcium Responses in Endothelial Cells and Macrophages Indicate 5-Lipoxygenase–Dependent Circuits of Inflammation and Atherogenesis
Objective— Inflammatory infiltrates and atherosclerotic lesions emerge when monocytes adhere to endothelial cells (ECs), migrate into the subendothelial space, and become macrophages (MΦs). Leukotrienes (LTs), products of 5-lipoxygenase, are powerful inflammatory mediators. 5-lipoxygenase+ MΦs have been shown to increase during atherogenesis, and LT receptor (LT-R) transcripts were identified in diseased arteries. To investigate LT-Rs in cells involved in inflammation and atherogenesis, we used the in vitro models of human umbilical vein ECs (HUVECs) and monocyte-derived MΦs.
Methods and Results— HUVECs primarily expressed transcripts of the cysteinyl (cys) LT2-R, which was strongly upregulated by interleukin-4. By contrast, MΦs predominantly expressed transcripts of the cysLT1-R. Calcium responses toward LTs revealed differential cysLT-R utilization by both cell types: HUVECs responded to both cysLTs, whereas MΦs preferentially responded to LTD4; HUVECs, but not MΦs, were resistant toward a cysLT1-R antagonist, montelukast; cysLTs generated regular calcium oscillations in HUVECs that lasted >60 minutes, resulting in >500 oscillations per cell. By contrast, calcium elevations in MΦs returned to baseline within seconds and were nonoscillatory.
Conclusions— Our data raise the possibility that MΦ-derived LTs differentially activate cysLT2-Rs via paracrine stimulation and cysLT1-Rs via autocrine and paracrine stimulation during inflammation and atherogenesis.
Leukotriene B4 (LTB4) and the cysteinyl (cys) LTs, LTC4, LTD4, and LTE4, are potent, proinflammatory mediators.1,2⇓ Four LT-activated, putative, 7 transmembrane G protein–coupled receptors have been characterized: 2 receptors for LTB4, namely BLT-Rs,3,4⇓ and 2 receptors for LTD4 and LTC4, namely, cysLT-Rs.5,6⇓ Whereas LTB4 mediates chemotaxis,3 cysLTs trigger a variety of tissue responses, including increases in vascular permeability.7 CysLT1-R is primarily present in the spleen, blood leukocytes, and lung macrophages (MΦs),5 whereas cysLT2-R is expressed in the heart and brain.6 However, the precise cell lineage–specific expression of LT-Rs, their mode of regulation, and the signaling pathways that they trigger remain to be determined.
We observed transcripts of all 4 LT-Rs in human atherosclerotic lesions.8 These data raised the possibility that the 5-lipoxygenase (5-LO) pathway forms foci of inflammation within the arterial wall. Because human umbilical vein endothelial cells (HUVECs) and MΦs have been used to study inflammation and recent studies have associated the 5-LO pathway with experimental atherosclerosis,9–15⇓⇓⇓⇓⇓⇓ we examined their LT-R expression patterns and characterized their calcium responses.
Montelukast (Singulair) was a kind gift of Dr R.N. Young, Merck Frosst (Quebec, Canada). BAY u9773 was obtained from Biomol; fura 2-AM, from Calbiochem; fluo 4-AM, from Molecular Probes; transforming growth factor-β1 (TGF β-1), interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), and IL-4, from R&D Systems; LTs, from Cayman Chemicals; CD14 MicroBeads, from Miltenyi Biotech; and other reagents, from Sigma. Monocytes were purified from peripheral blood mononuclear cells by adherence or CD14 microbead adsorption and maintained at 1×106 cells/mL for 5 to 9 days in RPMI-1640 medium plus 20% autologous human serum. HUVECs were cultured as described16 and used at passages 1 to 2. Carotid artery disease lesions were obtained after informed consent of the patients, as described.8 HUVECs or MΦs were loaded with fura 2-AM at 4 μmol/L, and changes in calcium ions were expressed as excitation ratios of 340 nm/380 nm at 510 nm. For single-cell measurements, cells were loaded with 2 μmol/L fluo 4-AM. Responses were recorded on a microscope (Axiovert 200M) equipped with a confocal laser scanning head (LSM510). Quantitative (real-time) reverse transcription–polymerase chain reaction (qRT-PCR) was performed as previously reported.8
Expression Patterns of LT-R Transcripts in HUVECs and MΦs
Determination of sequence-confirmed RT-PCR products of LT-Rs in HUVECs and MΦs indicated differential expression of transcript levels. Preliminary, nonquantitative RT-PCR analyses indicated that HUVECs showed significant cysLT2-R and detectable BLT1-R and BLT2-R but no or low cysLT1-R transcripts. By contrast, MΦs preferentially expressed cysLT1-R transcripts and, to a lesser extent, cysLT2-R, BLT1-R, and BLT2-R transcripts (Figure 1a). qRT-PCR demonstrated the magnitude of differential LT-R transcript expression in both cell types (when expressed as absolute transcripts/105 glyceraldehyde 3-phosphate dehydrogenase [GAPDH] transcripts): HUVECs expressed cysLT2-Rs in large excess when compared with cysLT1-Rs (cysLT2-Rs, 983±527; cysLT1-R, 3±3; n=6). By contrast, MΦs expressed cysLT1-Rs in large excess when compared with cysLT2-Rs (cysLT1-Rs, 6754±4507; cysLT2-R, 156±123; n=8). Interestingly, carotid artery disease lesions that represent advanced atherosclerotic lesions (type V) according to the American Heart Association nomenclature8 expressed cysLT1-Rs that were only 3-fold higher than cysLT2-Rs (cysLT1-R, 439±158; cysLT2-R, 129±53; n=5). The numbers represent mean±SD and are highly significant: P<0.01 for cysLT1R versus cysLT2R in HUVECs and MΦs. These data show that HUVECs expressed cysLT2-R transcripts at >2 orders of magnitude higher levels when compared with cysLT1-R transcripts.
Cytokines implicated in inflammation and atherogenesis were examined for regulation of cysLT2-R transcripts. IL-4, but not IL-1β, TGF-β1, or TNF-α (Figure 1b), caused upregulation of cysLT2-R transcripts in a concentration- and time-dependent way (Figure 1c and 1d). Confluent HUVECs expressed ≈2-fold higher numbers of cysLT2-R transcripts when compared with their rapidly proliferating counterparts (Figure 1d, open columns). Because both IL-4 and cell density contributed to cysLT2-R transcript upregulation in an additive way, this amounted to a 7-fold increase at 48 and 72 hours. Moreover, levels of cysLT2-R transcript decreased with increasing passage numbers.
Calcium Responses Indicate Differential cysLT-R Utilization by HUVECs and MΦs
In receptor-transfected cells, cysLT1-R has been shown to bind LTC4 with lower affinity than LTD4, whereas the cysLT2-R binds LTC4 and LTD4 with similar affinities, but functional studies of cysLT1-R versus cysLT2-R in HUVECs have not yet been reported.5,6⇓ Initially, calcium responses toward LTC4/LTD4 of HUVECs and MΦs were examined. LTC4 or LTD4 (50 nmol/L) triggered robust calcium rises with similar potencies in HUVECs, whereas the selective cysLT1-R antagonist montelukast (Singulair) failed to inhibit calcium transients (Figure 2a and 2b). Moreover, BAY u9773, a partial cysLT2-R agonist,17 induced a moderate response (Figure 2b). By contrast, 50 nmol/L LTD4 triggered robust calcium responses in MΦs, whereas LTC4 at 50 nmol/L was inactive, and LTD4-dependent calcium responses were completely blocked by montelukast (Singulair; Figure 2c). These data are consistent with and extend studies reported by others that cysLTs cause calcium elevations in cultured HUVECs and MΦs.18–20⇓⇓ Thus, our data demonstrate for the first time that the dominant, functional cysLT-R in HUVECs is the cysLT2-R. They establish HUVECs as a culture system in which selective activation of the cysLT2-R for the biology of ECs can be studied.
Calcium Elevations in HUVECs and MΦs Reveal Differences in Kinetics, Amplitude, Duration, and Oscillatory Patterns
We noted that the initial rise of calcium ions in HUVECs was followed by a pronounced, sustained component, whereas that in MΦs rapidly returned to near-baseline levels within seconds (compare Figure 2a, 2b, and 2c). We next determined calcium responses in single cells. In HUVECs, the rapid initial LTC4/D4-dependent rise in calcium ions was followed by oscillations that, once established, proceeded at regular intervals (Figure 2d). These data confirmed studies by Datta et al,18 who demonstrated that cysLTs trigger oscillatory calcium rises in HUVECs. By contrast, in MΦs, LTD4 induced an initial increase in calcium ions, which rapidly returned to baseline levels (Figure 2e) and was only rarely followed by single or irregular oscillations. These data show that oscillations are not invariably associated with the action of LTs on target cells, although they appear to depend on both the LT-R subtype and/or the target cell. We next initiated more detailed studies of the pharmacology, LT-R dependence, regularity, duration, intracellular location, variability, and dependence on culture conditions of oscillations in HUVECs. The frequency of oscillations did not vary considerably in individual HUVECs for extended periods of time, although variability was noted between individual HUVECs. LT-dependent oscillations in HUVECs originated in subcellular areas in the cytoplasm (Figure 2f, arrow). Increases in calcium occurred in both the cytosol and the nucleus, whereby the nucleus was affected secondary to cytoplasmic calcium waves (Figure 2f). Once generated in one EC, cysLT-triggered calcium waves appeared to be transferred to other ECs, thus generating global intercellular calcium waves, probably through gap junctions. This interpretation was supported by inspection of slow-motion recordings of oscillations across intact portions of the monolayer that revealed areas of ECs in which coordinated calcium oscillations that originated in calcium “puffs” of a single EC were observed. Importantly, calcium oscillations were a response to cysLT2-R activation, because montelukast failed to eliminate them (not shown). Moreover, calcium oscillations were dependent on extracellular calcium, because EGTA did not prevent the initial rise in calcium ions but led to an immediate cessation of ongoing oscillations. Culture parameters also influenced characteristics of calcium oscillations: a higher percentage of early-passage HUVECs responded to low concentrations of cysLTs when compared with higher-passage HUVECs, and this response was correlated with cysLT2-R transcripts; the frequency of oscillations was correlated with the concentration of the agonist. At 1 μmol/L LTC4, the oscillation periods of HUVECs persisted for >60 minutes, resulting in >500 oscillations/h per HUVEC, whereas the response to another G protein–coupled receptor agonist, thrombin, was short-lived and subsided within seconds.
HUVECs principally expressed cysLT2-Rs, which were upregulated 7-fold by a combination of IL-4 and cell density (Figure 1). By contrast, HUVECs did not express significant numbers of cysLT1-Rs (Figures 1a, 2a, and 2⇑b), yet MΦs principally expressed cysLT1-Rs (Figures 1a and 2⇑c). Because selective cysLT2-R antagonists are not yet available, these data establish HUVECs as the first bona fide cell system in which the signaling pathway of the cysLT2-R can be studied without interfering with cysLT1-R–dependent signaling. The cysLT-R expression patterns in HUVECs and MΦs yielded calcium transients with very different properties: HUVECs and MΦs differed in their sensitivities toward LTC4 and LTD4 (Figure 2); response to LT-R antagonists; kinetics, regularity, and duration of calcium increases; and oscillatory versus nonoscillatory patterns.
When previous and the present data on expression of the 5-LO pathway and LT-Rs in ECs, MΦs, and advanced human atherosclerotic lesions are considered, our data raise the possibility that leukocyte/EC/T-cell interactions might constitute inflammatory “circuits” through at least 2 major mechanisms: MΦ-derived LTs might activate cysLT1-Rs via autocrine mechanisms and EC cysLT2-Rs and T-lymphocyte cysLT1-Rs, via paracrine mechanisms during physiologic leukocyte trafficking, inflammation, and atherogenesis (Figure 3).
This study was supported by grants of the Deutsche Forschungsgemeinschaft (Ha 1083/13–1/13–2), the European Union research network (QLG1-CT-2001–01521), and the Interdisziplinäre Zentrum für Klinische Forschung, Jena. We thank C. Ströhl, M. Franke, and E. Teuscher for expert technical assistance.
↵*These authors contributed equally to this work.
- Received May 21, 2003.
- Accepted June 4, 2003.
- ↵Samuelsson B, Dahlén SE, Lindgren JA, Rouzer CA, Serhan CN. Leukotrienes and lipoxins: structures, biosynthesis, and biological effects. Science. 1987; 237: 1171–1176.
- ↵Funk CD. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science. 2001; 294: 1871–1875.
- ↵Yokomizo T, Kato T, Terawaki K, Izumi T, Shimizu T. A second leukotriene B4 receptor, BLT2, a new therapeutic target in inflammation and immunological disorders. J Exp Med. 2000; 192: 421–432.
- ↵Lynch KR, O’Neill GP, Liu Q, Im DS, Sawyer N, Metters KM, Coulombe N, Abramovitz M, Figueroa DJ, Zeng Z, Connolly BM, Bai C, Austin CP, Chateauneuf A, Stocco R, Greig GM, Kargman S, Hooks SB, Hosfield E, Williams DL Jr, Ford-Hutchinson AW, Caskey CT, Evans JF. Characterization of the human cysteinyl leukotriene cysLT1 receptor. Nature. 1999; 399: 789–793.
- ↵Heise CE, O’Dowd BF, Figueroa DJ, Sawyer N, Nguyen T, Im DS, Stocco R, Bellefeuille JN, Abramovitz M, Cheng R, Williams DL Jr, Zeng Z, Liu Q, Ma L, Clements MK, Coulombe N, Liu Y, Austin CP, George SR, O’Neill GP, Metters KM, Lynch KR, Evans JF. Characterization of the human cysteinyl leukotriene 2 receptor. J Biol Chem. 2000; 275: 30531–30536.
- ↵Spanbroek R, Gräbner R, Lötzer K, Hildner M, Urbach A, Rühling K, Moos MPW, Kaiser B, Cohnert TU, Wahlers T, Zieske A, Plenz G, Robenek H, Salbach P, Kühn H, Rådmark O, Samuelsson B, Habenicht AJR. Expanding expression of the 5-lipoxygenase pathway within the arterial wall during human atherogenesis. Proc Natl Acad Sci U S A. 2003; 100: 1238–1243.
- ↵Libby P, Ridker PM, Maseri A. Inflammation and atherogenesis. Circulation. 2002; 105: 1135–1143.
- ↵Aiello RJ, Bourassa PA, Lindsey S, Weng W, Freeman A, Showell HJ. Leukotriene B4 receptor antagonism reduces monocytic foam cells in mice. Arterioscler Thromb Vasc Biol. 2002; 22: 443–449.
- ↵Mehrabian M, Allayee H, Wong J, Shi W, Wang XP, Shaposhnik Z, Funk CD, Lusis AJ, Shih W. Identification of 5-lipoxygenase as a major gene contributing to atherosclerosis susceptibility in mice. Circ Res. 2002; 91: 120–126.
- ↵Kühn H, Anton M, Gerth C, Habenicht AJR Amino acid exchanges observed in the 5-lipoxygenase of atheroresistant CON6 mice confer impaired activity when introduced into its human homologue. Arterioscler Thromb Vasc Biol. 2003; 23: 1072–1076.
- ↵Nuszkowski A, Gräbner R, Marsche G, Unbehaun A, Malle E, Heller R. Hypochlorite-modified low density lipoprotein inhibits nitric oxide synthesis in endothelial cells via an intracellular dislocalization of endothelial nitric-oxide synthase. J Biol Chem. 2001; 276: 14212–14221.
- ↵Nothacker HP, Wang Z, Zhu Y, Reinscheid RK, Lin SH, Civelli O. Molecular cloning and characterization of a second human cysteinyl leukotriene receptor: discovery of a subtype selective agonist. Mol Pharmacol. 2000; 58: 1601–1608.
- ↵Datta YH, Romano M, Jacobson BC, Golan DE, Serhan CN, Ewenstein BM. Peptidoleukotrienes are potent agonists of von Willebrand factor secretion and P-selectin surface expression in human umbilical vein endothelial cells. Circulation. 1995; 92: 3304–3311.
- ↵Thivierge M, Stanková J, Rola-Pleszczynski M. IL-13 and IL-4 up-regulate cysteinyl leukotriene 1 receptor expression in human monocytes and macrophages. J Immunol. 2001; 167: 2855–2860.