Published online before print November 2, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The Department of Surgery, Division of Trauma, University of Medicine and Dentistry of New Jersey-New Jersey Medical School, Newark
1 Correspondence: Department of Surgery, UMDNJ-New Jersey Medical School, MSB G-523, 185 South Orange Avenue, Newark, NJ 07103. E-mail: itagakki{at}umdnj.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: bioactive lipid • calcium influx • G-protein-coupled receptors
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Lysophosphatidic acid (LPA) is a bioactive lipid, which is structurally similar to S1P. Xu et al. [10 ] reported that LPA is present in the ascites of ovarian cancer patients, where it has been noted to act as an ovarian cancer-activating factor. In addition, they found the elevated plasma LPA concentrations in patients with ovarian and other gynecological cancers ranging from 1.0 to 43.1 µmol/l compared with the healthy volunteers ranging from less than 0.1 to 6.3 µmol/l [10 ]. The ability of LPA to act as a tumor growth factor has since been related to its ability to increase intracellular calcium ([Ca2+]i) [11 , 12 ]. Like S1P, LPA can mobilize [Ca2+]i through GPC receptors of the Edg family. At least three such LPA Edg receptors have been isolated to date (Edg-2/LPA1, Edg-4/LPA2, Edg-7/LPA3) [13 14 15 ]. Noguchi et al. [16 ] have also recently identified a fourth LPA receptor (LPA4), which seems to be structurally distinct from the Edg receptors [16 ]. Moreover, recent studies also suggest that S1P and LPA can be produced by the same exoenzyme, autotaxin (ATX)/lysophospholipase D (lysoPLD). ATX/lysoPLD in plasma can hydrolyze many glycerophospholipids, including lysophosphatidylethanolamine, lysophosphatidylinositol, lysophosphatidylserine, and lysophosphatidylcholine (LPC), to produce LPA [17 ]. Last, Clair et al. [18 ] have shown that ATX/lysoPLD can also hydrolyze sphingosylphosphorylcholine to produce S1P.
Thus, although there are no reports of LPA-mediated Ca2+ influx independence of LPA1–4, LPA clearly plays a role similar to S1P in many circumstances, and we have shown that in PMN, S1P induces Ca2+ influx as a receptor-independent second messenger [1 ]. We therefore hypothesized that LPA might have receptor-independent Ca2+ influx effects on PMN, similar to those of S1P, and that direct Ca2+ entry in response to LPA might have the physiologic characteristics of SOCE.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Neutrophil preparations
Studies were performed in compliance with the Institutional Review Board of the University of Medicine and Dentistry of New Jersey (UMDNJ)-New Jersey Medical School (Newark). Informed consent was obtained for blood withdrawal from healthy volunteers, and human PMN were isolated from the blood immediately after venipuncture. A detailed protocol can be found elsewhere [19
]. In brief, PMN were isolated from heparinized whole blood by a one-step purification procedure using Polymorphoprep gradient solution (Robbins Scientific, Sunnyvale, CA). PMN fractions were collected, and the osmolarity was adjusted. PMN were then washed and resuspended in HEPES-buffered physiological saline solution {120 mM NaCl, 5.4 mM KCl, 0.8 mM Mg2SO4, 20 mM HEPES, 1 mM CaCl2, 10 mM glucose, pH 7.4 [Hanks’ balanced saline solution (HBSS)]}. The purity was examined by flow cytometry. Typically, we observed more than 95% purity.
HL60 cells
HL60-G cells, a subclone of the human promyelocytic leukemia line HL60, were a generous gift from Dr. George Studzinski (UMDNJ) [20
]. Cells were maintained in 10% fetal bovine serum (heat-inactivated; Invitrogen, Carlsbad, CA), containing RPMI medium supplemented with L-glutamine (Invitrogen) at 37°C with 95% air/5% CO2.
[Ca2+]i concentration
[Ca2+]i of PMN and HL60 cells was determined by measurement of fura-2 fluorescence in cuvette with constant stirring, as described elsewhere [1
, 21
]. In brief, cells were incubated with 2 µM Fura-2AM for 30 min at 37°C, divided into 2 x 106 cells/Eppendorf tube, and kept on ice until use. Just prior to use, cells were warmed to 37°C and rapidly centrifuged to remove any extracellular fura. Cells were then transferred to a cuvette containing 3 ml HBSS without calcium and with 0.3 mM EGTA added. Fura-2 fluorescence was measured by excitation at 340 and 380 nm with emissions recorded at 505 nm. The autofluorescence of a typical PMN aliquot after treatment with digitonin and fura quenching with Mn2+ was subtracted from all fluorescence values. After assessing Rmax using digitonin and Rmin using 5 mM EGTA for each individual cuvette, the [Ca2+]i was finally calculated as described by Grynkiewicz et al. [22
].
PMN chemotaxis
PMN were incubated with 3 µg/ml calcein-AM (Molecular Probes) for 30 min at 37°C in the dark and were used for assays. Detailed protocols can be found elsewhere [4
]. In brief, a modified Boyden chamber was created by applying a thin layer of Biomatrix (Biomedical Technologies, Stoughton, MA) onto Transwell systems (Corning, Corning, NY) with polycarbonate membranes and 3 µm pores. Labeled PMN were placed in the upper wells, and LPA was placed in the lower wells. fMLP was used as a positive control. For each chemotaxis condition, a control transwell was set up to measure nondirectional motility (chemokinesis). This was done by placing identical concentrations of the chemoattractant (LPA or fMLP) in the upper and the lower wells. Wells were incubated for 90 min at 37°C in the dark. The upper wells were removed without scraping. Calcein fluorescence in the lower chamber was measured using a FL500 microplate reader (Bio-Tek Instruments, Winooski, VT) at an excitation of 485 nm and an emission of 530 nm. The absolute number of PMN migrating into the bottom well was then calculated from a standard curve created using reserved, calcein-labeled PMN.
PMN respiratory burst
Respiratory burst was measured spectrofluorometrically using an excitation wavelength of 488 nm and an emission wavelength of 530 nm. A detailed description can be found elsewhere [23
]. Briefly, freshly isolated PMN (2x106 cells/reaction) were applied into the cuvette containing DHR 123 (15 ng/ml). PMN respiratory burst induced by 10 nM phorbol 12-myristate 13-acetate (PMA) or various concentrations of LPA was examined. The slope of the linear portion of the trace was analyzed by a curve-fitting program of SigmaPlot for a relative respiratory burst value.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (25K): [in a new window] |
Figure 1. LPA induces Ca2+ influx in human PMN, which were loaded with fura-2 and used for experiments. Under nominal calcium-free conditions, various concentrations of LPA were applied at t = 30 s, and 1 mM calcium was readded at t = 200 s. Water (15 µl; vehicle) was applied for the zero-LPA control. No store-depletion transients were detected, but LPA induced Ca2+ influx in human PMN in a dose-dependent manner or recalcification. Data are representative of at least three to four experiments using multiple PMN donors.
|
 View larger version (24K): [in a new window] |
Figure 2. (A and B) LPA causes calcium influx without store release in PMN. (A) LPA (15 µM) or vehicle (9 µl; water) was added at t = 30 s. Ionomycin (100 nM) was added at t = 50 s. The cuvettes are recalcified with 1 mM CaCl2 at t = 250 s. LPA has no effect on ionomycin-depletable stores but causes a marked increase in calcium entry above that produced by ionomycin alone. (B) An enlargement of the first portion of A, demonstrating that ionomycin-releasable stores were completely indistinguishable in the presence and absence of LPA. Mean and SE values are shown. N = 3 experiments. (C) LPA-induced Ca2+ influx is not mediated by SOCE channels. Ca2+ influx was indicated as area under curve (AUC), which was calculated for 100 s (250–350 s). Blank, Water (9 µl) at t = 30 s, 3 µl dimethyl sulfoxide (DMSO) at t = 50 s, and 1 mM calcium at t = 250 s. Iono, Ionomycin; 9 µl water at t = 30 s, 100 nM ionomycin at t = 50 s, and 1 mM calcium at t = 250 s. LPA, LPA (15 µM) at t = 30 s, 3 µl DMSO at t = 50 s, and 1 mM calcium at t = 250 s. LPA/Iono, LPA (15 µM) at t = 30 s, 100 nM ionomycin at t = 50 s, and 1 mM calcium at t = 250 s. Iono + LPA, Mathematical addition of values from Iono and LPA. AUC, Area between traces, and basal line was calculated and used as Ca2+ influx. Details can be found elsewhere [21
]. N = 3 experiments for each. Mean and SE are shown.
|
LPA does not trigger PMN chemotaxis
Edg/LPA receptors are chemotactic in a wide variety of cells at picomolar or nanomolar concentrations [24
25
26
27
]. Moreover, PMN display chemotaxis toward all calcium-mobilizing GPC receptors we have tested [4
]. Yet, as seen in Figure 3
, PMN did not show any detectable chemotaxis toward LPA at any concentration, whereas fMLP caused typically brisk chemotaxis. We also noted that there was no PMN chemotaxis toward LPA at LPA1–4-active nanomolar concentrations or at the micromolar LPA concentrations associated with direct Ca2+ entry. This lack of chemotaxis further supports the absence of functional GPC LPA receptors on human PMN.
 View larger version (14K): [in a new window] |
Figure 3. LPA-induced chemotaxis and chemokinesis. The chemotaxis (CTX; solid bars) of PMN was examined by applying increasing concentrations (0.1, 1, 10, and 100 µM) of LPA to the lower wells of modified Boyden chambers. Chemokinesis (CKS; shaded bars) was assessed by applying the same concentrations of LPA to upper and lower chambers. Unstimulated PMN migration was assessed using wells that contain no stimulant. fMLP (100 nM) was used as a positive control. The data are displayed as the percent of unstimulated migration, and mean and SE values are shown. Data are from four separate experiments.
|
, Gi/o, and G12/13. Thus, to further confirm the independence of LPA-induced Ca2+ entry from GPC receptors, we evaluated the effects of inhibiting these pathways on [Ca2+]i flux after LPA. PTX blocks activation of Gi/o and thus inhibits LPA-induced activation of PLCß [30
], which synthesizes inositol (1,4,5)-trisphosphate (IP3) and releases Ca2+ from the ER, as well as activates the Ras/mitogen-activated protein kinase cascade [31
, 32
]. PMN were therefore treated with PTX (1 µg/ml for 3 h at 37°C) prior to Ca2+ influx experiments. The effect of PTX was confirmed by a complete absence of PMN store-depletion after stimulation by 10 nM fMLP. As shown in Figure 4
, however, PTX pretreatment had no effect on LPA-induced Ca2+ influx. These findings confirm that LPA-induced Ca2+ influx is not mediated by Gi/o-coupled LPA receptors.
 View larger version (25K): [in a new window] |
Figure 4. Effect of PTX on LPA-induced calcium movements. PMN were incubated in HBSS with 0.1% BSA for 3 h at 37°C (A) or incubated in media with 1 µg/ml PTX for 3 h at 37°C (B). PMN were loaded with fura-2AM for 30 min at the end of the 3-h incubation. Various concentrations (0, 5, 10, 20 µM) of LPA were applied at t = 30 s. Calcium (1 mM) was readded at t = 50 s. PTX had no effect on calcium entry in response to LPA. (C) In control experiments using fMLP (10 nM), calcium store-depletion was completely inhibited by PTX (+PTX, dotted line). All data are representative of at least three experiments.
|
 View larger version (28K): [in a new window] |
Figure 5. No inhibitory effect of U73122 on LPA-induced calcium movements. PMN were preincubated with 20 µM U73122 or vehicle (DMSO) for 15 min in the cuvette. The effect of U73122 was determined by applying 10 nM fMLP at t = 30 s. No store-depletion was detected by fMLP. For untreated cells, the fMLP vehicle (DMSO) was added at t = 30 s. In both conditions, 20 µM LPA was added at t = 200 s, and the cuvettes were recalcified (to 1 mM calcium) at t = 250 s. The results displayed are representatives of experiments performed in triplicate using PMN from different donors.
|
 View larger version (21K): [in a new window] |
Figure 6. Effect of Y27632 on LPA-induced calcium movements. PMN were preincubated with 100 µM Y27632 for 15 min or used without any preincubation. LPA (15 µM) was applied at t = 30 s, and 1 mM calcium was readded at t = 200 s. Traces are mean and SE of three experiments per treatment.
|
, and leukotriene B4 [1
]. As shown in Figure 7
, when LPA was applied to HL60 cells in Ca2+-free medium, no store-depletion was detected. As with PMN, however, dose-responsive Ca2+ influx was detected immediately after readdition of external calcium. These observations also suggest that LPA-induced Ca2+ influx is independent of store-depletion and does not require the involvement of GPC receptors.
 View larger version (34K): [in a new window] |
Figure 7. LPA-induced calcium movements on undifferentiated HL60 cells. Various concentrations of LPA were applied to undifferentiated HL60 cells. LPA or water was applied at t = 30 s, and 1 mM calcium was readded at t = 200 s. No store-depletion was detected; however, the magnitude of calcium influx was influenced by LPA in a dose-dependent manner.
|
 View larger version (29K): [in a new window] |
Figure 8. Inhibitory effect of Gd3+. PMN were preincubated with various concentrations of Gd3+ for 1 min in cuvettes in media without added calcium. No EGTA was used (see text). LPA (15 µM) was added at t = 30 s in all cases except for the "no LPA" control. Cuvettes were recalcified (to 1 mM calcium) at t = 50 s. Mean and SE values are shown. N = 3 or 4 experiments.
|
 View larger version (38K): [in a new window] |
Figure 9. Inhibitory effect of MRS1845. (A) LPA (15 µM) was applied at t = 30 s. Calcium (1 mM) was readded at t = 50 s. MRS1845 (25 µM) was added at t = 100 s to one trace that dropped calcium influx immediately. For the other trace, no MRS1845 was added. (B) Various concentrations (0, 25, 50 µM) of MRS1845 were applied at t = 0 s. LPA (15 µM) was added at t = 30 s followed by 1 mM calcium at t = 50 s.
|
 View larger version (25K): [in a new window] |
Figure 10. DMS has no inhibitory effect on LPA-induced calcium influx. (A) The effects of DMS on TG-induced calcium movements were shown. TG (500 nM) was applied at t = 30 s, 5 µM DMS was applied at t = 330 s (for +DMS), and 1 mM calcium was readded at t = 400 s. (B) Two traces were shown: PMN were pretreated with 5 µM DMS for 1 min, and then 15 µM LPA was added at t = 30 s, followed by 1 mM calcium at t = 200 s; and no DMS treatment.
|
 View larger version (14K): [in a new window] |
Figure 11. LPA-induced respiratory burst in human PMN. PMA (10 nM; positive control) and various concentrations of LPA were applied at t = 30 s in the presence of 1.5 µg/ml DHR. Reactions were recorded for more than 300 s. Excitation at 488 nm and emission at 530 nm were used. Respiratory burst was presented as a slope of the curve. Data represent mean and SE values from N = 4 experiments, except PMA (n=2).
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Multiple observations made in this study suggest that LPA-mediated PMN Ca2+ entry is independent of the GPC receptors LPA1–4. The marked Ca2+ entry detected in response to LPA, despite a complete absence of store-depletion transients, strongly implicates a mechanism other than GPC receptor-dependent store-depletion. Moreover, direct measurements of ionomycin-depletable stores were completely unaffected by LPA stimulation. LPA induced additional Ca2+ influx when treated with 100 nM ionomycin, which causes maximum SOCE, strongly suggesting that LPA-induced Ca2+ influx comes through routes other than SOCE. However, Ca2+ influx channels for LPA have similar sensitivity to inorganic inhibitors such as La3+, Ni2+, Zn2+, and Gd3+. Further, edg/LPA receptors are active at picomolar to low nanomolar concentrations [11 ] and are uniformly chemotactic. In PMN, we found that LPA-induced Ca2+ entry required exposure to low micromolar concentrations and that LPA failed to cause PMN chemotaxis at any concentration. Last, the signaling pathways downstream from LPA1–4 are now well studied, and pharmacologic interventions to inhibit the relevant G-proteins Gi/o, G12/13, and Rho were unsuccessful in inhibiting LPA-induced Ca2+ influx.
Our results should not be seen as conflicting with recent suggestions by Idzko et al. [41 ], that human eosinophils express functional edg/LPA receptors. First, although eosinophils and PMN are granulocytes, their phenotypes are clearly different. Moreover, although we do find mRNA for LPA receptors in PMN (LPA1–4, data not shown), there is no evidence that these receptors exist as functional proteins (Figs. 1 2 3 4) . mRNA expression also occurs without functional protein expression in the case of PMN edg/S1P receptor expression [1 ]. In addition, although Idzko et al. [41 ] demonstrated increased [Ca2+]i after eosinophil exposure to up to 1 mM LPA concentration, the fura-fluorescence ratio change found was small, and the presence of calcium in the media precluded distinguishing calcium store-depletion from calcium entry.
Thus, PMN Ca2+ entry response to LPA found was similar to Ca2+ influx, as manifested in response to S1P [1
] in terms of sensitivity to heavy metals such as La3+, Ni2+, Zn2+, and Gd3+ and the synthesized dihydropyridine channel inhibitor MRS1845. However, LPA-induced Ca2+ influx comes through a mechanism other than SOCE. Thus, some unidentified Ca2+ influx channels with similar sensitivity to those inhibitors must be involved in LPA-induced Ca2+ influx. It is as unknown at this time how LPA and S1P gate Ca2+ influx channels. Like S1P, LPA is readily synthesized by platelets [42
]. It can also be synthesized at the cell surface by ATX/lysoPLD from LPC [18
]. Work by McIntyre et al. [43
], examining its effect on peroxisome proliferator-activated receptor-
, suggests the possibility that extracellular LPA may act by crossing plasma membranes directly. Similarly, S1P can be synthesized intracellularly in response to agonists or Ca2+ store-depletion or can enter the cell from extracellular sources [1
, 44
, 45
]. There is no proof thus far that LPA can be synthesized intracellularly after Ca2+ store-depletion by GPC chemoattractants. LPA and S1P can each act as second messengers to induce PMN Ca2+ influx by stimulating Ca2+ influx channel proteins. This might occur by direct interactions with the unidentified channel proteins that underlie Ca2+ influx in the PMN [46
] or by a more generalized effect on lipid raft assembly and thus on membrane protein interactions, as has been suggested by Ait Slimane and Hoekstra [47
].
All of these observations suggest downstream mechanisms by which the lysophospholipid generation elicited by activation of intracellular phospholipases can gate Ca2+ entry, as suggested recently by Smani et al. [48 ]. In addition, however, our findings emphasize the possibility that release of lysophospholipids during thrombosis or other relevant extracellular events may activate PMN Ca2+ entry and inflammatory responses directly by acting as a second messenger after passing through plasma membranes. Our global hypotheses as to how intra- and extracellularly generated lysophospholipids may interact in generating PMN calcium influx are summarized in Figure 12 . Note that LPA and S1P can be released into the bloodstream by degranulating platelets [42 ] or synthesized locally at the cell membrane by ATX/lysoPLD [17 , 18 ]. S1P can be generated by SK after Ca2+ store-depletion triggered by GPC receptor-chemoattractant interactions; however, the role of increased [Ca2+]i in SK activation is unknown (dotted lines). In all cases, however, such events lead to PMN activation by Ca2+ influx through the unidentified calcium channels, SOCE channels for S1P and non-SOCE channels for LPA. This occurs by unknown mechanisms without depleting Ca2+ stores. The data suggest that the ability of lysophospholipids, generated extra- and intracellularly, to mediate Ca2+ entry directly, by acting as an intracellular second messenger, may provide a novel point of signaling convergence at which PMN may respond to innate-immune "danger signals."
 View larger version (23K): [in a new window] |
Figure 12. A hypothesis for LPA- and S1P-mediated calcium movements in PMN. LPA can be generated by platelets or by ATX on cell surface of PMN from LPC, and S1P can also be generated by platelets and by SK from SP after Ca2+ store depletion by various chemoattractants in PMN. LPA and S1P can trigger Ca2+ influx by reacting calcium influx channels directly on plasma membranes by an unknown mechanism. IP3, inositol (1,4,5)-triphosphate; IP3R, inositol (1,4,5)-triphosphate receptor; SP, sphingosine.
|
Received July 9, 2004; revised September 23, 2004; accepted October 4, 2004.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
agonist Proc. Natl. Acad. Sci. USA 100,131-136This article has been cited by other articles:
 |
|
 |
|
  S. Brechard and E. J. Tschirhart Regulation of superoxide production in neutrophils: role of calcium influx J. Leukoc. Biol., November 1, 2008; 84(5): 1223 - 1237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Kusner, C. R. Thompson, N. A. Melrose, S. M. Pitson, L. M. Obeid, and S. S. Iyer The Localization and Activity of Sphingosine Kinase 1 Are Coordinately Regulated with Actin Cytoskeletal Dynamics in Macrophages J. Biol. Chem., August 10, 2007; 282(32): 23147 - 23162. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Frasch, K. Zemski-Berry, R. C. Murphy, N. Borregaard, P. M. Henson, and D. L. Bratton Lysophospholipids of Different Classes Mobilize Neutrophil Secretory Vesicles and Induce Redundant Signaling through G2A J. Immunol., May 15, 2007; 178(10): 6540 - 6548. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Steinckwich, J.-P. Frippiat, M.-J. Stasia, M. Erard, R. Boxio, C. Tankosic, I. Doignon, and O. Nusse Potent inhibition of store-operated Ca2+ influx and superoxide production in HL60 cells and polymorphonuclear neutrophils by the pyrazole derivative BTP2 J. Leukoc. Biol., April 1, 2007; 81(4): 1054 - 1064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Katayama, Y. Kusaka, H. Yokota, T. Akao, M. Kojima, O. Nakamura, E. Mekada, and E. Mizuki Parasporin-1, a Novel Cytotoxic Protein from Bacillus thuringiensis, Induces Ca2+ Influx and a Sustained Elevation of the Cytoplasmic Ca2+ Concentration in Toxin-sensitive Cells J. Biol. Chem., March 9, 2007; 282(10): 7742 - 7752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Harfi, F. Corazza, S. D'Hondt, and E. Sariban Differential Calcium Regulation of Proinflammatory Activities in Human Neutrophils Exposed to the Neuropeptide Pituitary Adenylate Cyclase-Activating Protein J. Immunol., September 15, 2005; 175(6): 4091 - 4102. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
