- Email: [email protected]
Magnesium sulfate inhibits the oxytocin-induced production of inositol 1,4,5-trisphosphate in cultured human myometrial cells
Magnesium sulfate inhibits the oxytocin-induced production of inositol 1,4,5-trisphosphate in cultured human myometrial cells William W. Hurd, MD,a Viswanathan Natarajan, PhD,b John R. Fischer, MD,a Dawn M. Singh, MS,c Shawn G. Gibbs, MS,a and Victor P. Fomin, PhDc Dayton, Ohio, Baltimore, Md, and Indianapolis, Ind OBJECTIVE: The purpose of this study was to determine the effects of magnesium sulfate on inositol trisphosphate production and the mechanism of these effects. STUDY DESIGN: Myometrium was obtained at the time of cesarean delivery from women before labor at term. Inositol trisphosphate was measured in the primary myometrial cell cultures after stimulation with oxytocin, sodium fluoride, or Bay K 8644 with or without preincubation with magnesium sulfate or nifedipine. Experiments were performed in either calcium-containing or calcium-free medium that contained egtazic acid and after preincubation with the intracellular calcium chelator BAPTA-acetoxymethylester. Inositol trisphosphate production was measured by radioreceptor assay. In separate experiments, changes in intracellular calcium concentrations ([Ca2+]i) were measured with the use of Fura-2 and spectrophotofluorometry. RESULTS: Oxytocin, sodium fluoride, and Bay K 8644 increased inositol trisphosphate production 2- to 4fold. Preincubation with magnesium sulfate (3 10–3 mol/L) for ≥5 minutes decreased oxytocin-, sodium fluoride-, and Bay K 8644–induced inositol trisphosphate production in either calcium-containing or calciumfree media. Preincubation with BAPTA-acetoxymethylester decreased oxytocin-stimulated inositol trisphosphate production by 78% in calcium-containing media and completely prevented the oxytocin response in calcium-free media. Magnesium sulfate decreased inositol trisphosphate production in calcium-containing media but had no additional effect in calcium-free media. Oxytocin and Bay K 8644 increased [Ca2+]i in either calcium-containing or calcium-free media, and magnesium sulfate reduced this in both cases. CONCLUSION: Magnesium sulfate appears to inhibit phosphatidylinositol-4, 5-bisphosphate–specific phospholipase C activity and subsequent calcium release in cultured myometrial cells by a direct effect on phospholipase C. (Am J Obstet Gynecol 2002;187:419-24.)
Key words: Magnesium sulfate, phospholipase C, myometrium, pregnancy, calcium, Fura-2
Premature birth remains the largest cause of perinatal morbidity and death in the United States.1 One of the most commonly used drugs to stop premature labor is magnesium sulfate (MgSO4), primarily because of its initial effectiveness, lack of tachyphylaxis, and minimal side effects in most patients.2 Although the ability of MgSO4 to inhibit myometrial contractility has been recognized for >40 years, the mechanisms by which magnesium inhibits contractility are not entirely understood.3-6 Myometrial contractility depends on phasic changes in intracellular calcium concentration ([Ca++]i).7 Membrane receptor agonists, such as oxytocin, result in myo-
From the Department of Obstetrics and Gynecology, Wright State University School of Medicine,a the Department of Internal Medicine, Johns Hopkins School of Medicine,b and the Department of Obstetrics and Gynecology, Indiana University School of Medicine.c Supported by National Institutes of Health grant HD36692. Received for publication November 5, 2001; revised December 18, 2001; accepted January 30, 2002. Reprint requests: William W. Hurd, MD, 128 E Apple St, CHE 3800, Dayton, OH 45409-2793. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/123897 doi:10.1067/mob.2002.123897
metrial contraction at least in part by the release of calcium from intracellular stores through a multistep process. Membrane receptors, linked by G proteins, activate phosphoinositol-specific phospholipase C, which converts phosphatidylinositol-4,5-bisphosphate into inositol-1,4,5-trisphosphate (IP3) and diacylglycerol. IP3 in turn stimulates the release of calcium from intracellular stores, which results in both a transient increase in [Ca++]i and a muscle contraction. The system is complex, in that receptor-mediated generation of IP3 is at least partially dependent on calcium influx through voltage-gated calcium channels.8 Considering the central role that phospholipase C appears to play in calcium-mediated contractions, an effect of magnesium on phospholipase C could explain its inhibitory effect on myometrial contractility. This study was designed to determine the effect of magnesium on IP3 production and the changes in intracellular calcium concentration in cultured human myometrial cells. Material and methods Chemicals. MgSO4, Bay K 8644, sodium fluoride, anti–α-smooth muscle actin (clone 1A4), nifedipine, col419
420 Hurd et al
lagenase (type IA), deoxyribonuclease, N-hydroxyethylpiperazine-N´-ethanesulfonic acid, trichloroacetic acid (TCA), egtazic acid (EGTA), penicillin, streptomycin, and bovine serum albumin were purchased from Sigma Chemical Company (St Louis, Mo); Fura-2-acetoxymethylester and BAPTA–acetoxymethylester (BAPTAAM) were purchased from Molecular Probes (Eugene, Ore); fetal bovine serum was purchased from JRH Biosciences (Lenexa, Kan). Tissue acquisition and preparation. All tissue was obtained in accordance with a protocol that was approved by the Institutional Review Boards at Indiana University and Wright State University. Pregnant myometrium was obtained with written consent from 16 pregnant women at term who would be undergoing cesarean delivery but who were not in labor. Labor was defined as regular, firm contractions that resulted in progressive cervical dilation to ≥4 cm. Women who had any major complications of pregnancy (including hypertension, diabetes mellitus, and premature labor) or who had been treated with any medications other than prenatal vitamins before delivery were excluded. At the time of cesarean delivery, a fullthickness strip of uterine tissue was taken from the upper margin of a lower transverse incision. The tissue was transported to the laboratory on ice in sterile Hanks balanced salt solution. The myometrium was separated from both the endometrium and connective tissue with a scalpel and a dissecting microscope and was immediately processed for cell culturing. Myometrial cell cultures. Primary myometrial cell cultures were established from pregnant myometrium as previously described.9 Briefly, the myometrium was minced into 1-mm3 pieces and digested in a solution of collagenase and deoxyribonuclease for 4 hours with intermittent trituration and filtration. The myometrial cells were separated from fibroblasts and other cells with a discontinuous Percol density gradient, which resulted in a >90% myometrial cell viability by trypan blue exclusion. To verify that our cultured cells were myometrial cells rather than fibroblasts, we used immunostaining to evaluate the ability of the myometrial cells to express smoothmuscle–specific protein α–actin with monoclonal anti–α-smooth muscle actin.10 The cells were plated on 60-mm petri dishes at a concentration of 5 105 cells/cm2 and maintained at 37°C in a 5% carbon dioxide incubator until the cells reached 95% confluency. These primary cell cultures were serum-deprived for 24 hours before use. Cultured human myometrial cells from nonpregnant women that were prepared by the aforementioned protocol have been extensively characterized morphologically, biochemically, and physiologically by other investigators.11 In primary culture, these cells have been shown to retain basic morphologic and physiologic properties of
August 2002 Am J Obstet Gynecol
myometrial cells in the tissue. These cells respond to oxytocin with increases in [Ca++]i , produce approximately the same amount of prostaglandins, and are identical to myometrial cells in tissue specimens in terms of actin and myosin distribution. We have cultured human myometrial cells that were obtained from pregnant women with the use of the same protocol and found them to appear similar in terms of both morphologic properties and immunostaining for α-actin and desmin.12 Experimental protocols. To determine the effect of MgSO4 concentration and length of exposure, myometrial cells were preincubated with MgSO4 at concentrations of 3, 5, or 10 10–3 mol/L for 1, 3, 5, 10, 20, or 30 minutes. The cells were then stimulated for 10 seconds with oxytocin (10–7 mol/L) to evaluate the effect of magnesium on receptor-mediated phospholipase C activation, with sodium fluoride (2 10–5 mol/L, a direct G-protein activator) to evaluate the magnesium effects on G-protein activation of phospholipase C, or with Bay K 8644 (5 10–6 mol/L, an L-type calcium channel activator) to evaluate the effect on calcium-mediated phospholipase C activation. To determine the degree to which the magnesium effect is dependent on inhibition of extracellular calcium influx, experiments were repeated after preincubation for 20 minutes with the calcium L–type channel blocker nifedipine (5 10–6 mol/L) in calciumfree Krebs’ buffer that contained the calcium chelator EGTA (10–4 mol/L). To determine whether the effects of magnesium depended on [Ca++]i changes, additional experiments were performed with cells that had been preincubated in the intracellular calcium chelator BAPTA-AM (10–5 mol/L) for 45 minutes then washed for 30 minutes with Krebs’ buffer solution. The amount of IP3 was expressed as picomoles per 106 cells. In a separate set of experiments, the effect of magnesium on agonist-stimulated increases in [Ca++]i was determined by preloading the cells with the Fura-2. Cells in media that contained 1.5 10–3 mol/L calcium were exposed to MgSO4 (3 10–3 mol/L for 5 minutes) and then exposed to oxytocin, Bay K 8644, or sodium fluoride in concentrations as described earlier. Maximal [Ca++]i after agonist exposure was determined for 3 to 4 cell culture preparations. Experiments were repeated in calcium-free media that contained EGTA (10–4 mol/L). IP3 assay. The amount of IP3 that was generated was determined with the use of a tritiated inositol-1,4,5-trisphosphate radioreceptor assay kit (NEK-064; DuPont NEN, Boston, Mass), based on the technique described by Bredt et al.13 Chromatographic analysis and selective enzymatic hydrolysis of IP3 has shown that this assay is highly selective for the 1,4,5-isomer of IP3.13 The buffered, serum-free medium was changed for the cells, and the cells were preincubated with MgSO4 or nifedipine. The cells were then exposed to oxytocin, Bay K 8644, or sodium fluoride for 20 seconds. The generation of IP3 was terminated by
Hurd et al 421
Volume 187, Number 2 Am J Obstet Gynecol
the replacement of the incubation medium in each well with 1 mL of ice-cold TCA (1 mol/L). After 15 minutes, the cells were scraped from the dishes, homogenized, and centrifuged at 1000 revolutions/min for 10 minutes at 4°C. TCA was removed from the supernatant using 1,1,2trichloro-1,2,2-trifluoroethane (3:1, vol/vol). After being shaken for 15 seconds, the mixture was centrifuged (1000 revolutions/min) at 0°C for 10 minutes. The clear aqueous top layer, which contains IP3, was removed and kept on ice. A lyophilized IP3 receptor preparation that contained tritiated IP3 tracer was reconstituted in a buffered solution. IP3 standards (0.12 pmol/0.1 mL to 12 pmol/0.1 mL) were reconstituted with distilled water. Duplicate 100 µL aliquots of the standards and samples were mixed with 400 µL of diluted receptor preparation/tracer and incubated for 1 hour at 2°C to 8°C. After centrifugation (1000 revolutions/min) at 4°C for 10 minutes, the resultant pellets were dissolved in 50 µL of 0.15 mol/L sodium hydroxide, and 5 mL of scintillation fluid was added to each sample. Radioactivity was measured in duplicate samples and averaged, and the concentration of IP3 was determined by interpolation from a standard curve. Measurement of changes in [Ca++]i. [Ca++]i was measured as a function of Fura-2 fluorescence as previously published.14 Myocytes that were isolated from pregnant myometrium were plated on glass cover slips (9 22 mm) and grown to confluency. The cells were deprived of serum for 24 hours before [Ca++]i measurement. The cells were loaded for 30 minutes with 2 µmol/L acetoxymethyl ester of Fura-2 in buffer that contained (10–3 mol/L) 4.8 potassium chloride, 130 sodium chloride, 1.0 magnesium chloride, 1.5 calcium chloride, 1.0 sodium phosphate, 15 glucose, and N-hydroxyethylpiperazine-N’ethanesulfonic acid (pH 7.4). The coverslips were washed 3 times and incubated for an additional 15 minutes in buffer that was supplemented with 0.1% human serum albumin at room temperature. Experiments were performed within 1 hour of loading the cells. Fura-2 fluorescence was recorded with a ratio fluorescence spectrometer (model C-44; Photon Technology International, Monmouth Jct, NJ). After loading with Fura-2-acetoxymethylester, the coverslips that contained the cells were rinsed with albumin-free buffer and mounted diagonally in a 1-cm acryl cuvette that was filled with 2 mL of the buffer. After the cells were preincubated with MgSO4 (1-30 minutes), agonist was added to the cuvette while the buffer was continuously stirred. Fura-2 fluorescence was measured at 1-second intervals at excitation wavelengths of 340 nm and 380 nm and an emission wavelength of 510 nm. The fluorescence-intensity signals were calibrated after autofluorescence was subtracted. The calibration procedure involved the use of 50 10–5 mol/L ionomycin to obtain the maximal fluorescence value. [Ca++]i was estimated according to the
method of Grynkiewicz et al,15 based on the following formula: [Ca++]i = Kd(R – Rmin)/(Rmax – R)(Fo/Fs), where R is the ratio of the fluorescence of the sample at 340 and 380 nm, Rmax and Rmin represent the ratios for Fura-2 acid at the same wavelengths in the presence of saturating Ca++ and zero Ca++, respectively; Fo/Fs is the ratio of Fura-2 at 380 nm in zero and saturating Ca++; and Kd is the dissociation constant of Fura-2 for Ca++, which was assumed to be 225 nmol/L at 37°C. Data were expressed as maximal [Ca++]i within 30 seconds of agonist exposure. Data analysis. IP3 production and maximal [Ca++]i are presented as the mean ± SEM. Treatment groups were compared by 1-factor analysis of variance followed by the Fisher protected least significant difference test with the use of StatView software (Abacus Concepts, Inc, Berkeley, Calif). Probability values of <.05 were considered significant. Results Exposure of myometrial cells to oxytocin (10–7 mol/L) increased IP3 production 4-fold (Fig 1, A). Preincubation with MgSO4 for 10 minutes decreased oxytocin-induced IP3 production by 49% to 67% at all concentrations that were evaluated. In calcium-free media with EGTA, oxytocin-induced IP3 production was increased by >2-fold; MgSO4 preincubation decreased this production by 20% to 60%. When the effect of the preincubation time was examined, MgSO4 (3 10–3 mol/L) significantly inhibited oxytocin-induced IP3 production after 5 minutes of exposure (Fig 1, B). A similar effect was seen in calciumfree media with EGTA. Stimulation of the cells with sodium fluoride increased IP3 production by >2-fold (Fig 2, A). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased this response by 76%. In calcium-free media with EGTA, oxytocin increased IP3 production by 4-fold, and MgSO4 preincubation decreased this response by 75%. Stimulation of the cells with the calcium channel activator Bay K 8644 increased IP3 production by almost 4-fold (Fig 2, B). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased this by 63%. Preincubation with nifedipine (5 10–6 mol/L for 5 minutes) completely prevented the Bay K 8644 response. In calcium-free media, the Bay K 8644 was still able to increase IP3 production by 2-fold. Both MgSO4 and nifedipine preincubation completely prevented the Bay K 8644 response. When the cells were preincubated with the calcium chelator BAPTA-AM to prevent changes in [Ca++]i, oxytocin-stimulated IP3 production was decreased by 78% (Fig 3). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased IP3 production by an additional 18%. In calcium-free media, preincubation with BAPTAAM completely prevented the oxytocin response, and MgSO4 preincubation had no additional effect.
422 Hurd et al
Fig 1. Effect of MgSO4 concentration and duration on oxytocin (10–7 mol/L) induced IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calcium-free media that contained 10–4 mol/L EGTA (hatched bars). A, Cells were preincubated in MgSO4 at concentrations of 3, 5, or 10 10–3 mol/L for 10 min. B, Cells were preincubated in MgSO4 (3 10–3 mol/L) for 1 to 30 minutes. Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .0001 compared with unstimulated cells; two asterisks indicate P < .001 compared with oxytocin alone; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
To determine the effect of some of our pharmacologic manipulations on changes in [Ca++]i, we repeated some of the protocols with cells that were loaded with fura-2acetoxymethylester. Stimulation with oxytocin (10–7 mol/L) increased [Ca++]i by almost 2-fold in calcium-containing media (Fig 4). After exposure to MgSO4 (3 10–3 for 5 minutes), the increase in [Ca++]i was reduced by 75%. In calcium-free media that contained EGTA, oxytocin increased [Ca++]i by only 41%, and MgSO4 exposure reduced this by 50%. Stimulation with Bay K 8644 (5 10–6 mol/L) increased [Ca++]i by only 51% in Ca++-containing media (Fig 5). Exposure to MgSO4 (3 10–3 for 10 minutes) decreased this response by 40%. In calcium-free media that contained EGTA, oxytocin increased [Ca++]i by 45%, and MgSO4 exposure decreased this response by more than 50%. Comment The results of this study indicate that magnesium inhibits the oxytocin-induced phosphatidylinositol-4, 5-bisphosphate–specific phospholipase C activity in myometrial cells by a direct effect on phospholipase C. This effect was observed at MgSO4 concentrations as low as 3 10–3 mol/L, which is well within the therapeutic serum concentration of 4 to 6 mEq/L (2-3 10–3 mol/L) that is used clinically.16 This ability of MgSO4 to
August 2002 Am J Obstet Gynecol
Fig 2. Effect of MgSO4 (3 10–3 mol/L for 10 minutes) on reception-independent stimulation of IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calciumfree media that contained 10–4 mol/L EGTA (hatched bars). A, Cells were stimulated with sodium fluoride (NaF; 10–3 mol/L). B, Cells were stimulated with Bay K 8644 (5 10–6 mol/L). Some cells were preincubated with nifedipine (5 10–6 mol/L for 10 minutes) rather than MgSO4. Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .001 compared with unstimulated cells; two asterisks indicate P < .01 compared with sodium fluoride or Bay K 8644 alone; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
inhibit IP3 production is relatively rapid, reaching a near maximal effect within 5 minutes, which suggests that these effects must be taking place at or within the cell membrane, because it takes >20 minutes of exposure to high concentrations of MgSO4 to significantly increase intracellular magnesium concentrations.17 The ability of MgSO4 to decrease IP3 in calcium-free media indicates that this effect is not dependent on the prevention of influx of extracellular calcium. To eliminate the possibility that our findings were the result of residual extracellular calcium, we added the calcium-chelator EGTA to our calcium-free media. The effect of magnesium appears to be distal to the oxytocin membrane receptor. When we used sodium flu-
Volume 187, Number 2 Am J Obstet Gynecol
Fig 3. Effect of MgSO4 (3 10–3 mol/L for 10 minutes) and BAPTA-AM (10–3 mol/L for 10 minutes) on oxytocin-induced (10–7 mol/L) IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calcium-free media that contained 10–4 mol/L EGTA (hatched bars). Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .0001 compared with control; two asterisks indicate P < .0001 compared with oxytocin alone; three asterisks indicate P < .03 compared with BAPTA plus oxytocin; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
oride to activate phospholipase C by G-protein stimulation, MgSO4 almost completely inhibited this response, which suggests that magnesium has either a direct effect on phospholipase C or an effect at the G-protein/phospholipase C interface. Further experiments indicated that the inhibitory effect of magnesium is directly on phospholipase C, perhaps at the Ca++ binding site. When Bay K 8644 (an L-type calcium channel agonist) was used to stimulate phospholipase C through nonreceptor mechanisms by increasing [Ca++]i, MgSO4 was still able to inhibit IP3 production, which indicates that the MgSO4 effect was distal to the Gprotein/phospholipase C interface. We did not expect Bay K 8644 to stimulate IP3 production in calcium-free media with EGTA, because Bay K 8644 is believed to activate phospholipase C by inducing extracellular calcium influx, thereby bypassing the membrane receptors and G-protein–related mechanisms.18 However, the ability of MgSO4 to block Bay K 8644–mediated IP3 production in calcium-free media makes it unlikely that the magnesium effect that we observed involves the prevention of calcium influx. This ability of Bay K 8644 to stimulate IP3 production in calcium-free media suggests that, under these conditions, Bay K 8644 stimulates phospholipase C by a mechanism other than the induction of calcium influx. One possibility is that Bay K 8644 has a direct stimulatory effect on phospholipase C. This appears to be unlikely, because nifedipine (a specific L-type calcium channel blocker) effectively blocked the Bay K 8644 effect on IP3
Hurd et al 423
Fig 4. Effects of MgSO4 (3 10–3 for 10 minutes) on [Ca2+]i that were evoked by oxytocin (OT; 10–7 mol/L) in media that contained 1.5 10–3 mol/L Ca++ (A) or in calcium-free media that contained 10–4 mol/L EGTA (B). Tracings are from representative cells. Bar heights represent mean ± SEM of duplicate experiments that were performed on cells that were derived from 4 different patients. Asterisk indicates P < .05 compared with no MgSO4; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
Fig 5. Effects of MgSO4 (3 10–3 for 10 minutes) on [Ca++]i that were evoked by Bay K 8644 (5 10–6 mol/L) in media that contained 1.5 10–3 mol/L Ca++ (A) and Ca++-free media that contained 10–4 mol/L EGTA (B). Tracings are from representative cells. Bar heights represent mean ± SEM of duplicate experiments that were performed on duplicate cells that were derived from 4 different patients. Asterisk indicates P < .05 compared with no MgSO4; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
production in calcium-free media with EGTA. The ability of Bay K 8644 to increase [Ca++]i under these conditions suggests that L-type calcium channels may be linked to either phospholipase C or intracellular calcium stores.
424 Hurd et al
However, our experiments did not address these possibilities. Regardless, the ability of MgSO4 to completely block the Bay K 8644 effect suggests magnesium has a direct effect on phospholipase C. Because both oxytocin and Bay K 8644 increase [Ca++]i even in calcium-free media with EGTA, we repeated our experiments after loading the cells with BAPTA (an intracellular calcium buffer) to determine whether the MgSO4 effects on IP3 production were dependent on a rise in [Ca++]i. The ability of MgSO4 to decrease the oxytocin-evoked IP3 production in the presence of BAPTA is consistent with a direct effect of magnesium on phospholipase C. The ability of BAPTA to completely block the oxytocin response in calcium-free media suggests that calcium (from either an extracellular or intracellular source) is required for oxytocin to stimulate phospholipase C. The finding that MgSO4 could not decrease IP3 production further than BAPTA in calcium-free media suggests that magnesium may inhibit the oxytocin stimulation of phospholipase C by interfering with the calcium–phospholipase C interaction. The results of this study have potential clinical implications. MgSO4 is very effective in temporarily stopping premature labor in many cases.16 However, its clinical usefulness is limited by the need to administer it at relatively high concentrations by continuous intravenous infusion in an intensive care setting. Several unrelated compounds are known to act as phospholipase C antagonists in the laboratory, including neomycin, compound 48/80, U-73122 (1-[6-([17β-3-methoxyestra-1,3,5(10)trien-17-yl]amino)hexyl]-1H-pyrrole-2,5-dion), NCDC (2nitro-4-carboxyphenyl-N,N,-diphenylcarbamate), and D609 (tricyclodecan-9-yl-xanthogenate).19-22 Not surprisingly, several of these phospholipase C antagonists have been shown to inhibit agonist induced myometrial contractility in vitro.23,24 The ability of these compounds to inhibit premature labor in humans and their safety for use during pregnancy remains to be investigated. Regardless as to whether these phospholipase C inhibitors are subsequently found to have clinical usefulness, a better understanding of the mechanism of action of 1 of our most effective tocolytics might bring us another step closer to the effective prevention of premature delivery.
REFERENCES
1. Stevenson DK, Wright LL, Lemons JA, Oh W, Korones SB, Papile LA, et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1993 through December 1994. Am J Obstet Gynecol 1998;179:1632-9. 2. Simchen MJ, Dulitzky M, Mashiach S, Friedman SA, Schiff E. Adjustment of magnesium sulfate infusion rate in patients with preterm labor. Am J Obstet Gynecol 1998;179:994-8. 3. Hall DG, McGaughey HS, Corey EL, Thornton WN. The effects of magnesium therapy in the duration of labor. Am J Obstet Gynecol 1959;78:27-32.
August 2002 Am J Obstet Gynecol
4. Kumar D, Zourlas PA, Barnes AC. In vitro and in vivo effects of magnesium sulfate on human uterine contractility. Am J Obstet Gynecol 1963;86:1036-40. 5. Phillippe M. Cellular mechanisms underlying magnesium sulfate inhibition of phasic myometrial contractions. Biochem Biophys Res Commun 1998;252:502-7. 6. Morishita F, Kawarabayashi T, Sakamoto Y, Shirakawa K. Role of the sodium-calcium exchange mechanism and the effect of magnesium on sodium-free and high-potassium contractures in pregnant human myometrium. Am J Obstet Gynecol 1995;172: 186-95. 7. Phillippe M, Basa A. Effects of sodium and calcium channel blockade on cytosolic calcium oscillations and phasic contractions of myometrial tissue. J Soc Gynecol Invest 1997;4:72-7. 8. Khac LD, Arnaudeau S, Lepretre N, Mironneau J, Harbon S. Beta adrenergic receptor activation attenuates the generation of inositol phosphates in the pregnant rat myometrium: correlation with inhibition of Ca2+ influx, a cAMP-independent mechanism. J Pharmacol Exp Ther 1996;276:130-6. 9. Fomin VP, Singh DM, Brown HL, Natarajan V, Hurd WW. Effect of cocaine on intracellular calcium regulation in myometrium from pregnant women. J Soc Gynecol Invest 1999;6:147-52. 10. Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen D, Gabbiani G. A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation. J Cell Biol 1986;103:2787-96. 11. Casey ML, MacDonald PC, Mitchell MD, Snyder JM. Maintenance and characterization of human myometrial smooth muscle cells in monolayer culture. In Vitro 1984;20:396-403. 12. Hurd WW, Fomin VP, Natarajan V, Brown H, Bigsby R, Singh DM. Expression of protein kinase C isozymes in nonpregnant and pregnant human myometrium. Am J Obstet Gynecol 2000;183:1525-31. 13. Bredt DS, Mourey RJ, Snyder SH. A simple, sensitive, and specific radioreceptor assay for inositol 1,4,5-trisphosphate in biological tissues. Biochem Biophys Res Commun 1989;159:976-82. 14. Fomin VP, Cox BE, Word RA. Effect of progesterone on intracellular Ca2+ homeostasis in human myometrial smooth muscle cells. Am J Physiol 1999;276:C379-85. 15. Grynkiewicz G, Ponenie M, Tsien Y. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 1985;260:3440-50. 16. Cunningham FG, MacDonald PC, Gant NF, Leveno KJ, Gilstrap LC III. Preterm and postterm pregnancy and fetal growth retardation. In: Cunningham FG, MacDonald PC, Gant F, Leveno KJ, Gilstrap LC, editors. Williams’ obstetrics. 19th ed. Norwalk (CT): Appleton & Lange; 1993. p. 853-89. 17. Mizuki J, Tasaka K, Masumoto N, Kasahara K, Miyake A, Tanizawa O. Magnesium sulfate inhibits oxytocin-induced calcium mobilization in human puerperal myometrial cells: possible involvement of intracellular free magnesium concentration. Am J Obstet Gynecol 1993;169:134-9. 18. Chien E, Saunders T, Phillippe M. The mechanisms underlying Bay K 8644-stimulated phasic myometrial contractions. J Soc Gynecol Invest 1996;3:106-12. 19. Reggiani R, Laoreti P. Evidence for the involvement of phospholipase C in the anaerobic signal transduction. Plant Cell Physiol 2000;41:1392-6. 20. Saetrum Opgaard O, Nothacker H, Ehlert FJ, Krause DN. Human urotensin II mediates vasoconstriction via an increase in inositol phosphates. Eur J Pharmacol 2000;406:265-71. 21. Krakauer T. Suppression of endotoxin- and staphylococcal exotoxin-induced cytokines and chemokines by a phospholipase C inhibitor in human peripheral blood mononuclear cells. Clin Diagn Lab Immunol 2001;8:449-53. 22. Ferretti ME, Nalli M, Biondi C, Colamussi ML, Pavan B, Traniello S, et al. Modulation of neutrophil phospholipase C activity and cyclic AMP levels by fMLP-OMe analogues. Cell Signal 2001;13:233-40. 23. Phillippe M. Neomycin inhibition of hormone-stimulated smooth muscle contractions in myometrial tissue. Biochem Biophys Res Commun 1994;205:245-50. 24. Wassdal I, Nicolaysen G, Iversen JG. Bradykinin causes contraction in rat uterus through the same signal pathway as oxytocin. Acta Physiol Scand 1998;164:47-52.
Key words: Magnesium sulfate, phospholipase C, myometrium, pregnancy, calcium, Fura-2
Premature birth remains the largest cause of perinatal morbidity and death in the United States.1 One of the most commonly used drugs to stop premature labor is magnesium sulfate (MgSO4), primarily because of its initial effectiveness, lack of tachyphylaxis, and minimal side effects in most patients.2 Although the ability of MgSO4 to inhibit myometrial contractility has been recognized for >40 years, the mechanisms by which magnesium inhibits contractility are not entirely understood.3-6 Myometrial contractility depends on phasic changes in intracellular calcium concentration ([Ca++]i).7 Membrane receptor agonists, such as oxytocin, result in myo-
From the Department of Obstetrics and Gynecology, Wright State University School of Medicine,a the Department of Internal Medicine, Johns Hopkins School of Medicine,b and the Department of Obstetrics and Gynecology, Indiana University School of Medicine.c Supported by National Institutes of Health grant HD36692. Received for publication November 5, 2001; revised December 18, 2001; accepted January 30, 2002. Reprint requests: William W. Hurd, MD, 128 E Apple St, CHE 3800, Dayton, OH 45409-2793. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/123897 doi:10.1067/mob.2002.123897
metrial contraction at least in part by the release of calcium from intracellular stores through a multistep process. Membrane receptors, linked by G proteins, activate phosphoinositol-specific phospholipase C, which converts phosphatidylinositol-4,5-bisphosphate into inositol-1,4,5-trisphosphate (IP3) and diacylglycerol. IP3 in turn stimulates the release of calcium from intracellular stores, which results in both a transient increase in [Ca++]i and a muscle contraction. The system is complex, in that receptor-mediated generation of IP3 is at least partially dependent on calcium influx through voltage-gated calcium channels.8 Considering the central role that phospholipase C appears to play in calcium-mediated contractions, an effect of magnesium on phospholipase C could explain its inhibitory effect on myometrial contractility. This study was designed to determine the effect of magnesium on IP3 production and the changes in intracellular calcium concentration in cultured human myometrial cells. Material and methods Chemicals. MgSO4, Bay K 8644, sodium fluoride, anti–α-smooth muscle actin (clone 1A4), nifedipine, col419
420 Hurd et al
lagenase (type IA), deoxyribonuclease, N-hydroxyethylpiperazine-N´-ethanesulfonic acid, trichloroacetic acid (TCA), egtazic acid (EGTA), penicillin, streptomycin, and bovine serum albumin were purchased from Sigma Chemical Company (St Louis, Mo); Fura-2-acetoxymethylester and BAPTA–acetoxymethylester (BAPTAAM) were purchased from Molecular Probes (Eugene, Ore); fetal bovine serum was purchased from JRH Biosciences (Lenexa, Kan). Tissue acquisition and preparation. All tissue was obtained in accordance with a protocol that was approved by the Institutional Review Boards at Indiana University and Wright State University. Pregnant myometrium was obtained with written consent from 16 pregnant women at term who would be undergoing cesarean delivery but who were not in labor. Labor was defined as regular, firm contractions that resulted in progressive cervical dilation to ≥4 cm. Women who had any major complications of pregnancy (including hypertension, diabetes mellitus, and premature labor) or who had been treated with any medications other than prenatal vitamins before delivery were excluded. At the time of cesarean delivery, a fullthickness strip of uterine tissue was taken from the upper margin of a lower transverse incision. The tissue was transported to the laboratory on ice in sterile Hanks balanced salt solution. The myometrium was separated from both the endometrium and connective tissue with a scalpel and a dissecting microscope and was immediately processed for cell culturing. Myometrial cell cultures. Primary myometrial cell cultures were established from pregnant myometrium as previously described.9 Briefly, the myometrium was minced into 1-mm3 pieces and digested in a solution of collagenase and deoxyribonuclease for 4 hours with intermittent trituration and filtration. The myometrial cells were separated from fibroblasts and other cells with a discontinuous Percol density gradient, which resulted in a >90% myometrial cell viability by trypan blue exclusion. To verify that our cultured cells were myometrial cells rather than fibroblasts, we used immunostaining to evaluate the ability of the myometrial cells to express smoothmuscle–specific protein α–actin with monoclonal anti–α-smooth muscle actin.10 The cells were plated on 60-mm petri dishes at a concentration of 5 105 cells/cm2 and maintained at 37°C in a 5% carbon dioxide incubator until the cells reached 95% confluency. These primary cell cultures were serum-deprived for 24 hours before use. Cultured human myometrial cells from nonpregnant women that were prepared by the aforementioned protocol have been extensively characterized morphologically, biochemically, and physiologically by other investigators.11 In primary culture, these cells have been shown to retain basic morphologic and physiologic properties of
August 2002 Am J Obstet Gynecol
myometrial cells in the tissue. These cells respond to oxytocin with increases in [Ca++]i , produce approximately the same amount of prostaglandins, and are identical to myometrial cells in tissue specimens in terms of actin and myosin distribution. We have cultured human myometrial cells that were obtained from pregnant women with the use of the same protocol and found them to appear similar in terms of both morphologic properties and immunostaining for α-actin and desmin.12 Experimental protocols. To determine the effect of MgSO4 concentration and length of exposure, myometrial cells were preincubated with MgSO4 at concentrations of 3, 5, or 10 10–3 mol/L for 1, 3, 5, 10, 20, or 30 minutes. The cells were then stimulated for 10 seconds with oxytocin (10–7 mol/L) to evaluate the effect of magnesium on receptor-mediated phospholipase C activation, with sodium fluoride (2 10–5 mol/L, a direct G-protein activator) to evaluate the magnesium effects on G-protein activation of phospholipase C, or with Bay K 8644 (5 10–6 mol/L, an L-type calcium channel activator) to evaluate the effect on calcium-mediated phospholipase C activation. To determine the degree to which the magnesium effect is dependent on inhibition of extracellular calcium influx, experiments were repeated after preincubation for 20 minutes with the calcium L–type channel blocker nifedipine (5 10–6 mol/L) in calciumfree Krebs’ buffer that contained the calcium chelator EGTA (10–4 mol/L). To determine whether the effects of magnesium depended on [Ca++]i changes, additional experiments were performed with cells that had been preincubated in the intracellular calcium chelator BAPTA-AM (10–5 mol/L) for 45 minutes then washed for 30 minutes with Krebs’ buffer solution. The amount of IP3 was expressed as picomoles per 106 cells. In a separate set of experiments, the effect of magnesium on agonist-stimulated increases in [Ca++]i was determined by preloading the cells with the Fura-2. Cells in media that contained 1.5 10–3 mol/L calcium were exposed to MgSO4 (3 10–3 mol/L for 5 minutes) and then exposed to oxytocin, Bay K 8644, or sodium fluoride in concentrations as described earlier. Maximal [Ca++]i after agonist exposure was determined for 3 to 4 cell culture preparations. Experiments were repeated in calcium-free media that contained EGTA (10–4 mol/L). IP3 assay. The amount of IP3 that was generated was determined with the use of a tritiated inositol-1,4,5-trisphosphate radioreceptor assay kit (NEK-064; DuPont NEN, Boston, Mass), based on the technique described by Bredt et al.13 Chromatographic analysis and selective enzymatic hydrolysis of IP3 has shown that this assay is highly selective for the 1,4,5-isomer of IP3.13 The buffered, serum-free medium was changed for the cells, and the cells were preincubated with MgSO4 or nifedipine. The cells were then exposed to oxytocin, Bay K 8644, or sodium fluoride for 20 seconds. The generation of IP3 was terminated by
Hurd et al 421
Volume 187, Number 2 Am J Obstet Gynecol
the replacement of the incubation medium in each well with 1 mL of ice-cold TCA (1 mol/L). After 15 minutes, the cells were scraped from the dishes, homogenized, and centrifuged at 1000 revolutions/min for 10 minutes at 4°C. TCA was removed from the supernatant using 1,1,2trichloro-1,2,2-trifluoroethane (3:1, vol/vol). After being shaken for 15 seconds, the mixture was centrifuged (1000 revolutions/min) at 0°C for 10 minutes. The clear aqueous top layer, which contains IP3, was removed and kept on ice. A lyophilized IP3 receptor preparation that contained tritiated IP3 tracer was reconstituted in a buffered solution. IP3 standards (0.12 pmol/0.1 mL to 12 pmol/0.1 mL) were reconstituted with distilled water. Duplicate 100 µL aliquots of the standards and samples were mixed with 400 µL of diluted receptor preparation/tracer and incubated for 1 hour at 2°C to 8°C. After centrifugation (1000 revolutions/min) at 4°C for 10 minutes, the resultant pellets were dissolved in 50 µL of 0.15 mol/L sodium hydroxide, and 5 mL of scintillation fluid was added to each sample. Radioactivity was measured in duplicate samples and averaged, and the concentration of IP3 was determined by interpolation from a standard curve. Measurement of changes in [Ca++]i. [Ca++]i was measured as a function of Fura-2 fluorescence as previously published.14 Myocytes that were isolated from pregnant myometrium were plated on glass cover slips (9 22 mm) and grown to confluency. The cells were deprived of serum for 24 hours before [Ca++]i measurement. The cells were loaded for 30 minutes with 2 µmol/L acetoxymethyl ester of Fura-2 in buffer that contained (10–3 mol/L) 4.8 potassium chloride, 130 sodium chloride, 1.0 magnesium chloride, 1.5 calcium chloride, 1.0 sodium phosphate, 15 glucose, and N-hydroxyethylpiperazine-N’ethanesulfonic acid (pH 7.4). The coverslips were washed 3 times and incubated for an additional 15 minutes in buffer that was supplemented with 0.1% human serum albumin at room temperature. Experiments were performed within 1 hour of loading the cells. Fura-2 fluorescence was recorded with a ratio fluorescence spectrometer (model C-44; Photon Technology International, Monmouth Jct, NJ). After loading with Fura-2-acetoxymethylester, the coverslips that contained the cells were rinsed with albumin-free buffer and mounted diagonally in a 1-cm acryl cuvette that was filled with 2 mL of the buffer. After the cells were preincubated with MgSO4 (1-30 minutes), agonist was added to the cuvette while the buffer was continuously stirred. Fura-2 fluorescence was measured at 1-second intervals at excitation wavelengths of 340 nm and 380 nm and an emission wavelength of 510 nm. The fluorescence-intensity signals were calibrated after autofluorescence was subtracted. The calibration procedure involved the use of 50 10–5 mol/L ionomycin to obtain the maximal fluorescence value. [Ca++]i was estimated according to the
method of Grynkiewicz et al,15 based on the following formula: [Ca++]i = Kd(R – Rmin)/(Rmax – R)(Fo/Fs), where R is the ratio of the fluorescence of the sample at 340 and 380 nm, Rmax and Rmin represent the ratios for Fura-2 acid at the same wavelengths in the presence of saturating Ca++ and zero Ca++, respectively; Fo/Fs is the ratio of Fura-2 at 380 nm in zero and saturating Ca++; and Kd is the dissociation constant of Fura-2 for Ca++, which was assumed to be 225 nmol/L at 37°C. Data were expressed as maximal [Ca++]i within 30 seconds of agonist exposure. Data analysis. IP3 production and maximal [Ca++]i are presented as the mean ± SEM. Treatment groups were compared by 1-factor analysis of variance followed by the Fisher protected least significant difference test with the use of StatView software (Abacus Concepts, Inc, Berkeley, Calif). Probability values of <.05 were considered significant. Results Exposure of myometrial cells to oxytocin (10–7 mol/L) increased IP3 production 4-fold (Fig 1, A). Preincubation with MgSO4 for 10 minutes decreased oxytocin-induced IP3 production by 49% to 67% at all concentrations that were evaluated. In calcium-free media with EGTA, oxytocin-induced IP3 production was increased by >2-fold; MgSO4 preincubation decreased this production by 20% to 60%. When the effect of the preincubation time was examined, MgSO4 (3 10–3 mol/L) significantly inhibited oxytocin-induced IP3 production after 5 minutes of exposure (Fig 1, B). A similar effect was seen in calciumfree media with EGTA. Stimulation of the cells with sodium fluoride increased IP3 production by >2-fold (Fig 2, A). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased this response by 76%. In calcium-free media with EGTA, oxytocin increased IP3 production by 4-fold, and MgSO4 preincubation decreased this response by 75%. Stimulation of the cells with the calcium channel activator Bay K 8644 increased IP3 production by almost 4-fold (Fig 2, B). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased this by 63%. Preincubation with nifedipine (5 10–6 mol/L for 5 minutes) completely prevented the Bay K 8644 response. In calcium-free media, the Bay K 8644 was still able to increase IP3 production by 2-fold. Both MgSO4 and nifedipine preincubation completely prevented the Bay K 8644 response. When the cells were preincubated with the calcium chelator BAPTA-AM to prevent changes in [Ca++]i, oxytocin-stimulated IP3 production was decreased by 78% (Fig 3). Preincubation with MgSO4 (3 10–3 mol/L for 10 minutes) decreased IP3 production by an additional 18%. In calcium-free media, preincubation with BAPTAAM completely prevented the oxytocin response, and MgSO4 preincubation had no additional effect.
422 Hurd et al
Fig 1. Effect of MgSO4 concentration and duration on oxytocin (10–7 mol/L) induced IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calcium-free media that contained 10–4 mol/L EGTA (hatched bars). A, Cells were preincubated in MgSO4 at concentrations of 3, 5, or 10 10–3 mol/L for 10 min. B, Cells were preincubated in MgSO4 (3 10–3 mol/L) for 1 to 30 minutes. Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .0001 compared with unstimulated cells; two asterisks indicate P < .001 compared with oxytocin alone; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
To determine the effect of some of our pharmacologic manipulations on changes in [Ca++]i, we repeated some of the protocols with cells that were loaded with fura-2acetoxymethylester. Stimulation with oxytocin (10–7 mol/L) increased [Ca++]i by almost 2-fold in calcium-containing media (Fig 4). After exposure to MgSO4 (3 10–3 for 5 minutes), the increase in [Ca++]i was reduced by 75%. In calcium-free media that contained EGTA, oxytocin increased [Ca++]i by only 41%, and MgSO4 exposure reduced this by 50%. Stimulation with Bay K 8644 (5 10–6 mol/L) increased [Ca++]i by only 51% in Ca++-containing media (Fig 5). Exposure to MgSO4 (3 10–3 for 10 minutes) decreased this response by 40%. In calcium-free media that contained EGTA, oxytocin increased [Ca++]i by 45%, and MgSO4 exposure decreased this response by more than 50%. Comment The results of this study indicate that magnesium inhibits the oxytocin-induced phosphatidylinositol-4, 5-bisphosphate–specific phospholipase C activity in myometrial cells by a direct effect on phospholipase C. This effect was observed at MgSO4 concentrations as low as 3 10–3 mol/L, which is well within the therapeutic serum concentration of 4 to 6 mEq/L (2-3 10–3 mol/L) that is used clinically.16 This ability of MgSO4 to
August 2002 Am J Obstet Gynecol
Fig 2. Effect of MgSO4 (3 10–3 mol/L for 10 minutes) on reception-independent stimulation of IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calciumfree media that contained 10–4 mol/L EGTA (hatched bars). A, Cells were stimulated with sodium fluoride (NaF; 10–3 mol/L). B, Cells were stimulated with Bay K 8644 (5 10–6 mol/L). Some cells were preincubated with nifedipine (5 10–6 mol/L for 10 minutes) rather than MgSO4. Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .001 compared with unstimulated cells; two asterisks indicate P < .01 compared with sodium fluoride or Bay K 8644 alone; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
inhibit IP3 production is relatively rapid, reaching a near maximal effect within 5 minutes, which suggests that these effects must be taking place at or within the cell membrane, because it takes >20 minutes of exposure to high concentrations of MgSO4 to significantly increase intracellular magnesium concentrations.17 The ability of MgSO4 to decrease IP3 in calcium-free media indicates that this effect is not dependent on the prevention of influx of extracellular calcium. To eliminate the possibility that our findings were the result of residual extracellular calcium, we added the calcium-chelator EGTA to our calcium-free media. The effect of magnesium appears to be distal to the oxytocin membrane receptor. When we used sodium flu-
Volume 187, Number 2 Am J Obstet Gynecol
Fig 3. Effect of MgSO4 (3 10–3 mol/L for 10 minutes) and BAPTA-AM (10–3 mol/L for 10 minutes) on oxytocin-induced (10–7 mol/L) IP3 production in media that contained 1.5 10–3 mol/L Ca++ (filled bars) or in calcium-free media that contained 10–4 mol/L EGTA (hatched bars). Bar heights represent mean ± SEM of duplicate experiments that were performed on 3 to 4 cultures that were derived from individual patients. Asterisk indicates P < .0001 compared with control; two asterisks indicate P < .0001 compared with oxytocin alone; three asterisks indicate P < .03 compared with BAPTA plus oxytocin; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
oride to activate phospholipase C by G-protein stimulation, MgSO4 almost completely inhibited this response, which suggests that magnesium has either a direct effect on phospholipase C or an effect at the G-protein/phospholipase C interface. Further experiments indicated that the inhibitory effect of magnesium is directly on phospholipase C, perhaps at the Ca++ binding site. When Bay K 8644 (an L-type calcium channel agonist) was used to stimulate phospholipase C through nonreceptor mechanisms by increasing [Ca++]i, MgSO4 was still able to inhibit IP3 production, which indicates that the MgSO4 effect was distal to the Gprotein/phospholipase C interface. We did not expect Bay K 8644 to stimulate IP3 production in calcium-free media with EGTA, because Bay K 8644 is believed to activate phospholipase C by inducing extracellular calcium influx, thereby bypassing the membrane receptors and G-protein–related mechanisms.18 However, the ability of MgSO4 to block Bay K 8644–mediated IP3 production in calcium-free media makes it unlikely that the magnesium effect that we observed involves the prevention of calcium influx. This ability of Bay K 8644 to stimulate IP3 production in calcium-free media suggests that, under these conditions, Bay K 8644 stimulates phospholipase C by a mechanism other than the induction of calcium influx. One possibility is that Bay K 8644 has a direct stimulatory effect on phospholipase C. This appears to be unlikely, because nifedipine (a specific L-type calcium channel blocker) effectively blocked the Bay K 8644 effect on IP3
Hurd et al 423
Fig 4. Effects of MgSO4 (3 10–3 for 10 minutes) on [Ca2+]i that were evoked by oxytocin (OT; 10–7 mol/L) in media that contained 1.5 10–3 mol/L Ca++ (A) or in calcium-free media that contained 10–4 mol/L EGTA (B). Tracings are from representative cells. Bar heights represent mean ± SEM of duplicate experiments that were performed on cells that were derived from 4 different patients. Asterisk indicates P < .05 compared with no MgSO4; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
Fig 5. Effects of MgSO4 (3 10–3 for 10 minutes) on [Ca++]i that were evoked by Bay K 8644 (5 10–6 mol/L) in media that contained 1.5 10–3 mol/L Ca++ (A) and Ca++-free media that contained 10–4 mol/L EGTA (B). Tracings are from representative cells. Bar heights represent mean ± SEM of duplicate experiments that were performed on duplicate cells that were derived from 4 different patients. Asterisk indicates P < .05 compared with no MgSO4; plus and minus signs indicate presence or absence of the corresponding hormone or chemical, respectively.
production in calcium-free media with EGTA. The ability of Bay K 8644 to increase [Ca++]i under these conditions suggests that L-type calcium channels may be linked to either phospholipase C or intracellular calcium stores.
424 Hurd et al
However, our experiments did not address these possibilities. Regardless, the ability of MgSO4 to completely block the Bay K 8644 effect suggests magnesium has a direct effect on phospholipase C. Because both oxytocin and Bay K 8644 increase [Ca++]i even in calcium-free media with EGTA, we repeated our experiments after loading the cells with BAPTA (an intracellular calcium buffer) to determine whether the MgSO4 effects on IP3 production were dependent on a rise in [Ca++]i. The ability of MgSO4 to decrease the oxytocin-evoked IP3 production in the presence of BAPTA is consistent with a direct effect of magnesium on phospholipase C. The ability of BAPTA to completely block the oxytocin response in calcium-free media suggests that calcium (from either an extracellular or intracellular source) is required for oxytocin to stimulate phospholipase C. The finding that MgSO4 could not decrease IP3 production further than BAPTA in calcium-free media suggests that magnesium may inhibit the oxytocin stimulation of phospholipase C by interfering with the calcium–phospholipase C interaction. The results of this study have potential clinical implications. MgSO4 is very effective in temporarily stopping premature labor in many cases.16 However, its clinical usefulness is limited by the need to administer it at relatively high concentrations by continuous intravenous infusion in an intensive care setting. Several unrelated compounds are known to act as phospholipase C antagonists in the laboratory, including neomycin, compound 48/80, U-73122 (1-[6-([17β-3-methoxyestra-1,3,5(10)trien-17-yl]amino)hexyl]-1H-pyrrole-2,5-dion), NCDC (2nitro-4-carboxyphenyl-N,N,-diphenylcarbamate), and D609 (tricyclodecan-9-yl-xanthogenate).19-22 Not surprisingly, several of these phospholipase C antagonists have been shown to inhibit agonist induced myometrial contractility in vitro.23,24 The ability of these compounds to inhibit premature labor in humans and their safety for use during pregnancy remains to be investigated. Regardless as to whether these phospholipase C inhibitors are subsequently found to have clinical usefulness, a better understanding of the mechanism of action of 1 of our most effective tocolytics might bring us another step closer to the effective prevention of premature delivery.
REFERENCES
1. Stevenson DK, Wright LL, Lemons JA, Oh W, Korones SB, Papile LA, et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1993 through December 1994. Am J Obstet Gynecol 1998;179:1632-9. 2. Simchen MJ, Dulitzky M, Mashiach S, Friedman SA, Schiff E. Adjustment of magnesium sulfate infusion rate in patients with preterm labor. Am J Obstet Gynecol 1998;179:994-8. 3. Hall DG, McGaughey HS, Corey EL, Thornton WN. The effects of magnesium therapy in the duration of labor. Am J Obstet Gynecol 1959;78:27-32.
August 2002 Am J Obstet Gynecol
4. Kumar D, Zourlas PA, Barnes AC. In vitro and in vivo effects of magnesium sulfate on human uterine contractility. Am J Obstet Gynecol 1963;86:1036-40. 5. Phillippe M. Cellular mechanisms underlying magnesium sulfate inhibition of phasic myometrial contractions. Biochem Biophys Res Commun 1998;252:502-7. 6. Morishita F, Kawarabayashi T, Sakamoto Y, Shirakawa K. Role of the sodium-calcium exchange mechanism and the effect of magnesium on sodium-free and high-potassium contractures in pregnant human myometrium. Am J Obstet Gynecol 1995;172: 186-95. 7. Phillippe M, Basa A. Effects of sodium and calcium channel blockade on cytosolic calcium oscillations and phasic contractions of myometrial tissue. J Soc Gynecol Invest 1997;4:72-7. 8. Khac LD, Arnaudeau S, Lepretre N, Mironneau J, Harbon S. Beta adrenergic receptor activation attenuates the generation of inositol phosphates in the pregnant rat myometrium: correlation with inhibition of Ca2+ influx, a cAMP-independent mechanism. J Pharmacol Exp Ther 1996;276:130-6. 9. Fomin VP, Singh DM, Brown HL, Natarajan V, Hurd WW. Effect of cocaine on intracellular calcium regulation in myometrium from pregnant women. J Soc Gynecol Invest 1999;6:147-52. 10. Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen D, Gabbiani G. A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation. J Cell Biol 1986;103:2787-96. 11. Casey ML, MacDonald PC, Mitchell MD, Snyder JM. Maintenance and characterization of human myometrial smooth muscle cells in monolayer culture. In Vitro 1984;20:396-403. 12. Hurd WW, Fomin VP, Natarajan V, Brown H, Bigsby R, Singh DM. Expression of protein kinase C isozymes in nonpregnant and pregnant human myometrium. Am J Obstet Gynecol 2000;183:1525-31. 13. Bredt DS, Mourey RJ, Snyder SH. A simple, sensitive, and specific radioreceptor assay for inositol 1,4,5-trisphosphate in biological tissues. Biochem Biophys Res Commun 1989;159:976-82. 14. Fomin VP, Cox BE, Word RA. Effect of progesterone on intracellular Ca2+ homeostasis in human myometrial smooth muscle cells. Am J Physiol 1999;276:C379-85. 15. Grynkiewicz G, Ponenie M, Tsien Y. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 1985;260:3440-50. 16. Cunningham FG, MacDonald PC, Gant NF, Leveno KJ, Gilstrap LC III. Preterm and postterm pregnancy and fetal growth retardation. In: Cunningham FG, MacDonald PC, Gant F, Leveno KJ, Gilstrap LC, editors. Williams’ obstetrics. 19th ed. Norwalk (CT): Appleton & Lange; 1993. p. 853-89. 17. Mizuki J, Tasaka K, Masumoto N, Kasahara K, Miyake A, Tanizawa O. Magnesium sulfate inhibits oxytocin-induced calcium mobilization in human puerperal myometrial cells: possible involvement of intracellular free magnesium concentration. Am J Obstet Gynecol 1993;169:134-9. 18. Chien E, Saunders T, Phillippe M. The mechanisms underlying Bay K 8644-stimulated phasic myometrial contractions. J Soc Gynecol Invest 1996;3:106-12. 19. Reggiani R, Laoreti P. Evidence for the involvement of phospholipase C in the anaerobic signal transduction. Plant Cell Physiol 2000;41:1392-6. 20. Saetrum Opgaard O, Nothacker H, Ehlert FJ, Krause DN. Human urotensin II mediates vasoconstriction via an increase in inositol phosphates. Eur J Pharmacol 2000;406:265-71. 21. Krakauer T. Suppression of endotoxin- and staphylococcal exotoxin-induced cytokines and chemokines by a phospholipase C inhibitor in human peripheral blood mononuclear cells. Clin Diagn Lab Immunol 2001;8:449-53. 22. Ferretti ME, Nalli M, Biondi C, Colamussi ML, Pavan B, Traniello S, et al. Modulation of neutrophil phospholipase C activity and cyclic AMP levels by fMLP-OMe analogues. Cell Signal 2001;13:233-40. 23. Phillippe M. Neomycin inhibition of hormone-stimulated smooth muscle contractions in myometrial tissue. Biochem Biophys Res Commun 1994;205:245-50. 24. Wassdal I, Nicolaysen G, Iversen JG. Bradykinin causes contraction in rat uterus through the same signal pathway as oxytocin. Acta Physiol Scand 1998;164:47-52.