INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

PULMONARY ENDOTHELIAL CELLS are a nonexcitable cell type in which humoral and neural signaling agents increase the free cytosolic Ca²⁺ concentration ([Ca²⁺]_i) by inducing Ca²⁺ release from intracellular stores and Ca²⁺ entry across the cell membrane (4, 34). Increased [Ca²⁺]_i has been implicated in many endothelial-directed vascular responses including regulation of vascular tone and permeability (2, 23, 36), angiogenesis (20), and leukocyte trafficking (17). Activation of Ca²⁺ entry appears essential for each of these processes, although many modes of Ca²⁺ entry exist and a specific pathway regulating endothelial cell shape has yet to be identified.

It is widely accepted that endothelial cells possess capacitative, or store-operated, Ca²⁺ entry pathways (8, 13, 31, 35, 41, 42). However, specific store-operated Ca²⁺ channels (SOCs) responsible for Ca²⁺ entry into nonexcitable cell types are largely unidentified. Recent cloning and expression of the transient receptor potential (trp) gene product from the Drosophila melanogaster retina reveal that this product forms a Ca²⁺-permeant cation channel that mediates Ca²⁺ entry after intracellular inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] is generated and Ca²⁺ is liberated from intracellular stores (11, 15, 25). Six mammalian homologues of Drosophila Trp are known (5), and mRNAs for these have been reported in bovine aortic endothelial cells (12). Although Trp3 and Trp6 are not SOCs (6, 46), Trp1 may form SOCs based on the following experimental evidence: 1) Trp1 and its splice variant TRPC1A increase store-operated Ca²⁺ entry when expressed in COS cells (45, 47) and 2) expression of antisense trp sequences in murine L(tk) cells greatly attenuates store-operated Ca²⁺ entry evoked by Ins(1,4,5)P₃ (45). Information concerning putative functions for Trp2, -4, and -5 is lacking in the literature.

Because activation of store-operated Ca²⁺ entry is known to increase vascular permeability in isolated lungs (9, 18), thereby suggesting that pulmonary endothelial SOCs are important for regulation of endothelial barrier integrity, we designed studies to characterize the store-operated Ca²⁺ entry pathway in rat (R) pulmonary arterial endothelial cells (PAECs). We hypothesized that a functional consequence of activating endothelial SOCs is a change in cell shape, leading to interendothelial gap formation and cytoskeletal rearrangement. To test this hypothesis, we challenged RPAECs with thapsigargin, a plant alkaloid that activates store-operated Ca²⁺ entry independent of ligand-receptor-G protein-coupled processes (40, 43), and monitored the changes in endothelial cell shape and microfilamentous cytoskeletal arrangement. We then determined whether RPAECs express Trp1 in order to address the possible molecular basis for the pulmonary endothelial store-operated Ca²⁺ entry pathway. Our data indicate that store-operated Ca²⁺ entry promotes cell shape change in rat pulmonary endothelial cells expressing Trp1 and further suggest that Ca²⁺ entry through SOCs involves site-specific rearrangement of the microfilamentous cytoskeleton.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Isolation of RPAECs. Male Sprague-Dawley rats (CD strain, 350-400 g; Charles River) were euthanized by an intraperitoneal injection of 50 mg of pentobarbital sodium (Nembutal, Abbott Laboratories, Chicago, IL). After sternotomy, the heart and lungs were removed en bloc, and the pulmonary arterial segment between the heart and lung hili was dissected, split, and fixed onto a 35-mm plastic dish. Endothelial cells were obtained by gentle intimal scraping with a plastic cell lifter and were seeded into a 100-mm petri dish containing 10 ml of seeding medium (~1:1 DMEM-Ham's F-12 + 10% fetal bovine serum) (37). After incubation for 1 wk (21% O₂-5% CO₂-74% N₂ at 37°C), smooth muscle cell contaminants were marked and then removed by pipette aspiration. Cells were verified as endothelial by positive factor VIII staining and uptake of 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI)-labeled acetylated low-density lipoprotein. When the primary culture reached confluence, cells were passaged by trypsin digestion into T-75 culture flasks (Corning), and standard tissue culture techniques were followed until the cells were ready for experimentation ( passages 6-20).

[Ca²⁺]_i measurement by fura 2 fluorescence. RPAECs were seeded onto chambered glass coverslips (Nalge Nunc, Naperville, IL) and grown to confluence. [Ca²⁺]_i was estimated with the Ca²⁺-sensitive fluorophore fura 2-AM (Molecular Probes, Eugene, OR) according to methods previously described by our laboratory (38). Because this is the first report of [Ca²⁺]_i measurements in cultured RPAECs, a summary of the technique will be presented. RPAECs were washed with 2 ml of a HEPES (Fisher Scientific, Atlanta, GA)-buffered physiological salt solution (PSS) containing (in g/l) 6.9 NaCl, 0.35 KCl, 0.16 KH₂PO₄, 0.141 MgSO₄, 2.0 D-glucose, and 2.1 NaHCO₃. The loading solution (1 ml) consisted of PSS plus 3 µM fura 2-AM, 3 µl of a 10% pluronic acid solution, and 2 mM or 100 nM CaCl₂. RPAECs were fura loaded for 20 min in a CO₂ incubator at 37°C. After this loading period, the cells were again washed with PSS (2 ml) and treated with deesterification medium (PSS + 2 mM or 100 nM CaCl₂) for an additional 20 min. After deesterification, [Ca²⁺]_i was assessed with an Olympus IX70 inverted microscope at ×400 with a xenon arc lamp photomultiplier system (Photon Technologies, Monmouth Junction, NJ), and data were acquired and analyzed with PTI Felix software. Epifluorescence (signal averaged) was measured from three to four endothelial cells in a confluent monolayer, and the changes in [Ca²⁺]_i are expressed as the fluorescence ratio of the Ca²⁺-bound (340-nm) to Ca²⁺-unbound (380-nm) excitation wavelengths emitted at 510 nm.

Electrophysiological determination of store-operated Ca²⁺ entry. Whole cell patch clamp was utilized to measure transmembrane ion flux in thapsigargin-stimulated RPAECs. Confluent RPAECs were enzyme dispersed, seeded onto 35-mm plastic culture dishes, and then allowed to reattach for at least 24 h before patch-clamp experiments were performed. Single RPAECs exhibiting a flat, polyhedral morphology were studied. These cells were chosen for study because their morphology was consistent with RPAECs from a confluent monolayer. The extracellular and pipette solutions were composed of the following (in mM): 1) extracellular: 110 tetraethylammonium aspartate, 10 calcium aspartate, 10 HEPES, and 0.5 3,4-diaminopyridine; and 2) pipette: 130 N-methyl-D-glucamine, 1.15 EGTA, 10 HEPES, 1 Ca(OH)₂, and 2 Mg²⁺-ATP. Both solutions were adjusted to 290-300 mosM with sucrose and pH 7.4 with methane sulfonic acid, and [Ca²⁺]_i was estimated as 100 nM (10). The pipette resistance was 2-5 M. Data were obtained with a HEKA EPC9 amplifier (Lambrecht/Pfaltz) and sampled on-line with Pulse + Pulsefit software (HEKA). All recordings were made at room temperature (22°C). To generate current-voltage (I-V) relationships, voltage pulses were applied from 100 to +100 mV in 20-mV increments, with a 200-ms duration during each voltage step and a 2-s interval between steps. The holding potential between each step was 0 mV.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Thapsigargin activates store-operated Ca²⁺ entry in RPAECs. We monitored fura 2 epifluorescence, and as shown in Fig. 1A and summarized in Fig. 1C (open bars), RPAECs incubated in 2 mM [Ca²⁺]_o had baseline fluorescence ratios averaging 0.91 ± 0.02. Thapsigargin produced a gradual increase in [Ca²⁺]_i to a peak level followed by a modest decline, producing a new steady state, or plateau, in [Ca²⁺]_i. Figure 1, B (dashed line) and C (solid bars), illustrates that the thapsigargin-induced response was dependent on [Ca²⁺]_o. When experiments were repeated in PSS containing 100 nM [Ca²⁺]_o, the baseline fluorescence ratio value decreased slightly, and both the peak and sustained plateau phases of the thapsigargin-induced response were significantly attenuated. Subsequent readdition of 2 mM [Ca²⁺]_o produced an immediate increase in [Ca²⁺]_i, thereby illustrating functional store-operated Ca²⁺ entry pathways.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Store-operated Ca2+ entry in rat (R) pulmonary arterial endothelial cells (PAECs). A: individual traces from cells challenged with thapsigargin (TG; 1 µM) in presence of 2 mM extracellular Ca2+ concentration ([Ca2+]o). B: representative traces comparing TG-induced cytosolic Ca2+ concentration ([Ca2+]i) response in presence (solid line) and absence (dashed line) of [Ca2+]o. [Ca2+]o was increased from 100 nM to 2 mM at time indicated by Ca2+. Arrows, time of addition. C: averaged data from all experiments conducted in presence (n = 5 cells; open bars) and absence (n = 5 cells; solid bars) of [Ca2+]o. t, Time. * Significant difference compared with respective baseline fluorescence ratio of Ca2+-bound (340-nm) to Ca2+-unbound (380-nm) excitation wavelengths emitted at 510 nm (340/380), P < 0.05.

View larger version (23K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Fig. 2.   Currents in RPAECs were measured under whole cell configuration with bath and pipette solutions given in METHODS. Average (±SE) currents were calculated from last 20 ms of each 200-ms voltage step. A: summary of current-voltage recordings for unstimulated (n = 4) and thapsigargin-treated (1 µM thapsigargin in pipette; n = 8) RPAECs normalized to membrane capacitance to yield current density. Insets: sample sets of current pulses. B: net inward current density generated in response to thapsigargin.

Store-operated Ca²⁺ entry evokes endothelial cell shape change and F-actin cytoskeletal rearrangement in RPAECs. To determine a functional consequence of SOC activation in RPAECs, we assessed changes in endothelial cell shape and formation of intercellular gaps in thapsigargin-treated confluent RPAEC monolayers. Because we determined that SOC activation was apparent 3-5 min after thapsigargin treatment, we studied endothelial morphology at this fixed time point. Figure 3 shows scanning electron micrographs of RPAECs after different treatments. Untreated RPAECs (Fig. 3A) in 2 mM [Ca²⁺]_o exhibited a characteristic "cobblestone" morphology essentially devoid of intercellular gaps. Thapsigargin produced endothelial cell retraction and intercellular gap formation (Fig. 3B). The changes in endothelial cell morphology were dependent on [Ca²⁺]_o because RPAECs incubated in 100 nM [Ca²⁺]_o and challenged with thapsigargin displayed little change in morphology and a lack of interendothelial gaps (Fig. 3C). The subsequent readdition of 2 mM [Ca²⁺]_o had a dramatic effect on endothelial cell shape, causing pronounced cell retraction and gap formation (Fig. 3D). Thus Ca²⁺ entry through activated SOCs sufficiently promoted endothelial cell shape alterations and interendothelial gap formation.

View larger version (159K): [in this window] [in a new window]   Fig. 3.   Scanning electron micrographs of RPAEC monolayers (×2,000). A: control (2 mM [Ca2+]o). B: thapsigargin treatment. C: RPAECs in 100 nM [Ca2+]o challenged with thapsigargin. D: readdition of 2 mM [Ca2+]o to RPAECs treated as in C.

View larger version (101K): [in this window] [in a new window]   Fig. 4.   Effect of [Ca2+]o on F-actin distribution. Confocal microscopy was performed on 0.3-µm sections. Three micrographs/treatment are shown. A: unchallenged RPAEC monolayers in 2 mM [Ca2+]o. B: RPAECs incubated in 100 nM [Ca2+]o. I: cross section through tops of cells. In 2 mM [Ca2+]o, staining appeared as a peripheral band with apparent cell-to-cell contact sites. In low [Ca2+]o, diffuse punctate staining was observed throughout cells, but contact sites between cells were still obvious. II and III: cross sections through middle and lower aspects of cells, respectively. In 2 mM [Ca2+]o, F-actin aligned in radiating strands, with obvious F-actin-containing focal contact sites. Low [Ca2+]o was characterized by diffuse staining throughout, with cell junction integrity still intact.

View larger version (145K): [in this window] [in a new window]   Fig. 5.   Effect of thapsigargin on F-actin organization in presence of 2 mM (A), 100 nM (B), and 100 nM [Ca2+]o followed by restoration of [Ca2+]o to 2 mM (C). Confocal microscopy was performed on 0.3-µm sections. Three micrographs/treatment are shown. I: cross section through upper portion of cells. In presence of 2 mM [Ca2+]o, diffuse perinuclear staining is evident. In low [Ca2+]o, a peripheral actin band with cell-to-cell contact sites is prominent. This peripheral band retracted after Ca2+ was readded, and intercellular actin projections are discernible. II and III: cross sections of middle and lower portions of cells, respectively. In presence of 2 mM [Ca2+]o, peripheral (cortical) actin band is absent, and F-actin appears to align in stress fibers. In low [Ca2+]o, diffuse punctate staining is observed, but cortical actin band is still present. On readdition of [Ca2+]o, stress fiber formation is obvious.

View larger version (55K): [in this window] [in a new window]   Fig. 6.   A: sequence comparison (nucleotides 1-195) between human Trp1 (hTrp1) and RT-PCR products from human PAECs (HPAECs) and RPAECs. * Differences between RPAEC and hTrp1 sequences. B: deduced amino acid sequence (amino acids 27-91) of RPAEC and HPAEC RT-PCR products with sequence alignment.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Although activation of Ca²⁺ entry is sufficient to induce the interendothelial cell gap formation necessary for the transit of macromolecules and cells from blood into tissue, the mode of Ca²⁺ entry responsible for changing cell shape is unknown. Nonexcitable cells possess store-operated Ca²⁺ entry pathways. Store-operated Ca²⁺ entry is activated in response to agonist-induced stimulation of membrane phospholipases, generation of Ins(1,4,5)P₃, Ca²⁺ release from intracellular stores, and subsequent lowering of store Ca²⁺ concentrations (4, 8, 13, 16, 31, 34, 35, 41, 42). Presently, there are three prevailing questions regarding store-operated Ca²⁺ entry pathways. 1) What specific cellular functions are regulated by Ca²⁺ influx through this pathway? 2) What is the molecular identity of the membrane channels responsible for mediating store-operated Ca²⁺ entry? 3) What is the nature of the signal linking Ca²⁺ store depletion to store-operated Ca²⁺ entry? Our present study addressed the first two of these three important questions.

Thapsigargin was utilized to test store-operated Ca²⁺ entry pathways because this agent produces intracellular Ca²⁺ store depletion without the confounding influences of ligand-receptor-heterotrimeric G protein activation (40, 43, 47). Fura 2-loaded RPAEC monolayers exhibited an increased [Ca²⁺]_i that was dependent on Ca²⁺ influx in response to thapsigargin, thereby indicating the presence of store-operated Ca²⁺ entry pathways. To begin elucidating the electrophysiological characteristics of RPAEC SOCs, we performed whole cell patch clamp in single cells. We designed intracellular and extracellular patch solutions to isolate thapsigargin-induced Ca²⁺ currents and determine whether thapsigargin activated a channel(s) responsible for Ca²⁺ release-activated current (I_CRAC) (16). However, the thapsigargin-induced current measured under these experimental conditions was not identical to I_CRAC because significant outward current was also measured.

It was possibile that the total current measured in response to thapsigargin reflected coactivation of both a Ca²⁺-selective cation channel and an anion channel because aspartate was utilized to replace Cl in the extracellular solution, and aspartate has recently been shown to be conducted through Ca²⁺- and/or volume-activated Cl channels(29). In support of this idea, N-phenylanthranilic acid, a potent blocker of Ca²⁺-activated anion channels (27), had little effect on the inward current observed at negative voltages but strongly attenuated the outward current at positive voltages (data not shown). Thus thapsigargin may activate an anion channel capable of conducting large organic anions as previously reported in bovine pulmonary endothelium (27, 29). When the anion conductance contribution to the total thapsigargin-stimulated current is then considered, a current analogous to I_CRAC is apparent. Future electrophysiological studies, including ion-selectivity experiments and single-channel analysis, are necessary to fully characterize the thapsigargin-sensitive Ca²⁺-permeable channels in RPAECs.

Activation of SOCs in RPAECs causes the appearance of intercellular gaps and rounding of endothelial cells. One intracellular target affected by SOC activation is plasmalemmal-associated and centrally located F-actin. It is accepted that changes in [Ca²⁺]_i lead to reconfigurations of the microfilamentous cytoskeleton (21, 22, 30), although it has previously been unclear whether Ca²⁺ release from intracellular stores or Ca²⁺ influx is necessary to produce cytoskeletal changes leading to cell shape change.

Thapsigargin produced a loss of plasmalemmal F-actin staining concurrent with an increase in central F-actin staining. When store depletion alone was produced, i.e., thapsigargin in the absence of [Ca²⁺]_o, rearrangement of cortical actin fibers did not occur and less F-actin staining was observed centrally. Under these conditions, RPAECs did not respond to thapsigargin with a change in cell shape. The readdition of [Ca²⁺]_o caused morphological changes in both the peripheral (loss of dense actin staining) and centrally located (increased actin staining and transcellular filament formation) F-actin pools, indicating that Ca²⁺ influx through SOCs is sufficient to adjust the microfilament system of the cells to produce interendothelial gap formation. It is presently unclear how Ca²⁺ influx through SOCs specifically regulates the endothelial F-actin cytoskeleton, although a possible mediator of the Ca²⁺ influx-induced cytoskeletal rearrangement is Rho, a small-molecular-weight monomeric G protein, the activity of which produces actin polymerization and stress fiber formation (1, 14).

Interestingly, incubation of RPAEC monolayers in low [Ca²⁺]_o alone caused rearrangement of central F-actin but had no apparent effect on peripheral, or cortical, F-actin. Under these conditions, Ca²⁺ release could have been promoted because a more favorable electrochemical gradient for Ca²⁺ to leak from intracellular stores existed. Centrally located F-actin in close proximity to Ca²⁺ stores could have been affected by Ca²⁺ release but not in a manner sufficient to drive an active cell shape change. We speculate that these observations may allude to the mechanism(s) leading to plasmalemmal SOC activation, i.e., through Ca²⁺ release-induced cytoskeletal rearrangement coupled to activation of plasmalemmal SOCs. Another possibility to consider with respect to the F-actin rearrangement is that low [Ca²⁺]_o provided less basal Ca²⁺ influx that was somehow setting the F-actin cytoskeletal architecture. Future studies will be required to address this novel observation of the ability of [Ca²⁺]_o to regulate the endothelial cytoskeleton and to specifically address whether F-actin is a vital component of the SOC activation mechanism.

Although our data clearly demonstrate that activation of SOCs regulate endothelial cell shape via effects on the microfilamentous cytoskeleton, we were unable to perform antagonist studies to specifically block SOC activation and the resulting cell shape change. This is because only nonspecific antagonists of endothelial Ca²⁺ entry pathways exist and the molecular identity of SOCs is unknown. In fact, the collective data from several previous studies (7, 24, 28, 31, 33, 35, 39, 41, 44) that investigated the nature of Ca²⁺ entry pathways in endothelial cells indicate that multiple SOCs and receptor-operated channels may exist, each having specific electrophysiological profiles and modes of optimal activation. We did, however, begin to deduce the identity of pulmonary endothelial SOCs using RT-PCR. Several trp gene products (Trp1 and Trp3-6) have recently been identified in systemic endothelial cells (12), and our findings indicate that at least Trp1, but neither Trp3 nor Trp6, is expressed in pulmonary endothelial cells. It is uncertain how trp gene products may be organized in the membrane to form a functional channel, but it has been proposed that SOCs may be composed of trp homo- and/or heteromultimers (5). Because our data indicate that neither Trp3 nor Trp6 are present in rat or human pulmonary endothelial cells, the SOC is not composed of Trp1-Trp3 or Trp1-Trp6 heteromultimers.

What are the implications of the observation that SOC activation produces changes in PAEC shape? It is possible that endothelial SOCs are integral for regulating pulmonary vascular permeability responses to inflammatory mediators. Whole lung studies (9, 18) have shown that activation of SOCs alone is sufficient to produce increased vascular permeability as assessed by measures of the filtration coefficient. In addition, SOC activation promotes increased flux of macromolecules across RPAEC monolayers (19, 26). However, stimulation of the thapsigargin-sensitive store-operated Ca²⁺ entry pathway in rat pulmonary microvascular endothelial cells promotes neither increased macromolecular permeability nor changes in cell shape (19). These observations suggest that inflammatory processes involving endothelial SOC activation can produce pulmonary edema mediated by the appearance of large-vessel leak sites away from the gas-exchanging microcirculatory bed. Therefore, future studies are needed to determine whether 1) pulmonary conduit-vessel endothelium and microvascular endothelium represent distinct phenotypes having separate regulatory properties, 2) changes in conduit-vessel endothelial cell shape in situ lead to significant, function-compromising pulmonary edema, 3) the precipitating factors for increasing large-vessel and small-vessel (capillary) permeabilities are the same, and 4) interventions to alleviate pulmonary edema can be designed to selectively target conduit-vessel endothelial cells vs. microvascular endothelial cells.

The shape change elicited in response to SOC activation in RPAECs has additional importance for other endothelial-directed physiological processes such as angiogenesis and regulation of leukocyte movement. The angiogenic process requires migration of endothelial cells that, in turn, is dependent on the ability of cells to change shape and decrease their cell-to-cell and cell-to-matrix tethering (3). Inhibition of non-voltage-gated Ca²⁺ channels, presumably including SOCs, inhibits angiogenic factor-induced proliferation, migration, and tube formation of human umbilical venous endothelial cells (20), which are endothelial cells derived from conduit vessels. In addition, a study (17) has shown that human umbilical venous endothelial cell-directed regulation of leukocyte trafficking is [Ca²⁺]_i dependent. Changes in endothelial cell shape and tethering that accompany neutrophil adhesion and migration require increased [Ca²⁺]_i. How the increased [Ca²⁺]_i occurs is not clear, but a transmembrane Ca²⁺ flux is required for certain leukocyte secretory products to increase endothelial [Ca²⁺]_i (32), thereby suggesting a role for SOC-mediated Ca²⁺ entry. Our data, in combination with these findings, suggest that initiation sites for angiogenesis and leukocyte diapedesis in vivo may be located in pulmonary vascular segments lined with endothelial cells possessing SOCs that regulate cell shape.

In summary, we have shown that RPAECs possess thapsigargin-activated SOCs that conduct current similar to I_CRAC. RPAECs respond to this mode of Ca²⁺ entry with changes in cell shape, interendothelial gap formation, and rearrangement of the F-actin cytoskeleton. Cytoskeletal rearrangement may be differentially regulated by the extracellular and intracellular Ca²⁺ pools, with Ca²⁺ influx being necessary to produce a cytoskeleton configured for cell shape change. In addition, pulmonary endothelial cells from rats (and humans) express Trp1, which may be integral for forming native SOCs in these cell types. Finally, pulmonary conduit vessel-derived endothelial SOC activation leading to interendothelial gap formation may be the basis for some forms of pulmonary edema and/or a component of angiogenesis and regulation of leukocyte trafficking to and from the vasculature.