JOURNAL OF CELLULAR PHYSIOLOGY 152207-214 (1992)

Platelet-Activating Factor Stimulates Phospholipase C Activity in Human Endometrium A. AHMED* AND S.K. SMITH Department of Obstetrics and Cynaecology, University of Cambridge, Rosie Maternity Hospital, Cambridge CB2 ZSW, United Kingdom

Human preimplantation embryos secrete platelet-activating factor (PAF), which stimulates prostaglandin E, synthesis from secretory endometrium. This study investigated the action of PAF on phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2)-specific phosphol ipase C activity in human endometrium. SI ices of normal endometrium were incubated with 5 pCi/ml myo-[2-’Hl inositol for 3 h at 37°C in 95% 0, and 5% CO, to label tissue phosphoinositides. lnositol phosphates were extracted using trichloroacetic acid precipitation and diethylether neutralization and production was measured using Dowex 1-X8 anion-exchange column chromatography. PAF induced rapid and concentration-dependent accumulation of inositol phosphates (IP) from secretory endometrium, but had no effect on endometrium removed in the proliferative phase of the menstrual cycle. The IP, fraction was significantly elevated from a median value of 14.0 c.p.m. m g ’ dry wt [range: 8-41 c.p.m. mg-’ dry W L ] to 28.0 c.p.m. mg-’ dry wt [range: 11-87 c.p.m. m g ’ dry wt, P < 0.0021 following 1 min exposure of secretory endometrium to PAF-acether, in the presence of 10 rnM LiCI. PAFinduced hydrolysis of Ptdlns(4,S)P2 was inhibited by the specific PAF receptor antagonist WEB 2086, in a dose-dependent manner ( P < 0.021, indicating that in human endometrium Ptdlns(4,5)P, hydrolysis is mediated via a PAF receptor. These results indicate that PAF receptor coupling activates endometrial Ptdlns(4,5)P2-specific phospholipase C only in the secretory phase of the menstrual cycle, suggesting that the PAF response may be under ovarian steroid regulation. It is proposed that the ability of the endometrium to respond to PAF appears to be a feature of the preparation of this tissue for implantation and that the second messengers generated may play a role in cellular processes involved in the maternal recognition of very early human pregnancy. (c) IYW ~ i l e y - ~ i s sInc. ,

One of the first measurable maternal responses to conception is the development of a mild systemic transient thrombocytopenia in mice (O’Neil, 1985a) and women (O’Neil et al., 1985). Subsequent studies demonstrated the production of a n embryonic factor capable of inducing systemic platelet depletion in mice (O’Neil, 1985b,c) and women (Roberts e t al., 1987). This embryo-derived factor exhibited the biochemical and pharmacological characteristics of a group of biologically active ether phospholipids known collectively as platelet-activating factor (PAF) (Collier et al., 1988). PAF is produced not only by the embryo during early pregnancy, but also by rabbit (Angle e t al., 1988a) and human (Alecozay et al., 1989) endometrium. PAF production is restricted to the stromal compartment and is regulated by ovarian steroids in cultured human secretory endometrial cells (Alecozay et al., 1988). PAF causes a dose-dependent increase in the synthesis of prostaglandin E, by enriched glandular, but not stromal, fractions of human endometrium removed in the secretory phase of the menstrual cycle (Smith and Kelly, 1988). Prostaglandin E, is elevated a t the site of implantation in several species, which suggests that Q

1992 WILEY-LISS, INC.

PAF may play a role in the events associated with early implantation. PAF and other calcium (Ca’+)-mobilizing agonists transmit their intracellular messages by binding to specific receptors on the cell surface. The ligand-bound receptors activate effector systems such as phospholipases including phospholipase A2, phospholipase C, and phospholipase D via a receptor-coupled G-protein to generate second messengers. The hydrolysis of PtdIns(4,5)P2 by PtdIns(4,5)P2-specific phospholipase C yields two second messengers, 1,2-diacylglycerol (DAG) and inositol 174,5-triphosphate(Ins(1,4,5)P,) (Berridge and Irvine, 1989). DAG and Ins(1,4,5)P3 are subsequently metabolized, the latter being reincorporated into PtdIns(4,5)P2. Ins(1,4,5)P, mobilizes Ca2+ from intracellular stores (Berridge, 1984) and DAG, in concert with Ca2+, activates the phospholipid-depen-

Received October 7,1991; accepted January 10, 1992. *To whom reprint requestsicorrespondenceshould be addressed.

208

AHMED AND SMITH

dent protein kinase C (Nishizuka, 1989). In turn, protein kinase C phosphorylates cellular proteins and thereby controls a host or cellular processes (Nishizuka, 1989). Although PAF has been shown to stimulate the hydrolysis of phosphatidylinositol in the rat myometrium (Varol et al., 1989), its action on inositol phosphate turnover in endometrium has not been studied. The aim of this study was to investigate the action of PAF on PtdIns(4,5)P2-specific phospholipase C activity as measured by the accumulation of inositol phosphates in human endometrium. MATERIALS A N D METHODS Materials My0-[2-~H]-inositol (specific activity 1.9 Ci/mmol) was obtained from Amersham International plc (Amersham, Bucks, U.K.). WEB 2086 was kindly provided by Dr. C.H. Weber (Boehringer Ingelheim Limited, Ingelheim AM Rhine, Germany). Dulbecco’s Modified Eagles Medium (DMEM) and Ham’s F12 nutrient medium were from Gibco Ltd (Gibco, Uxbridge, U.K.) and fetal calf serum substitute, Ultroser was obtained from IBF Biotechnics (Life Technology Ltd., Paisley, U.K.). Hank‘s Balance Salt Solution (HBSS), nonessential amino acids for minimum essential medium, Hepes buffer, amphotericin-B, and L-glutamine were from Flow Laboratories (Flow, Irvine, U.K.). The scintillation fluid used was OptiPhase (Hisafe) 3 from Fisons Scientific Apparatus (Loughborough, U.K.). PAFacether and all other reagents were obtained from Sigma Chemical Co. (Poole, Dorset, U.K.).

Patient selection Endometrium was obtained from the uterine cavity of women undergoing dilation and curettage or total abdominal hysterectomy performed for nonmalignant disease. Informed consent was obtained and the study approved by the District Ethical Committee of the Cambridge Health Authority. Endometrium was obtained throughout the cycle classified into proliferative endometrium (n = 6) and secretory endometrium (n = 24) using histological criteria (Noyes et al., 1950). Tissue preparation Endometrium was collected in sterile HBSS and transported to the laboratory. Explants of endometrium (1 mm3 size) were prepared, and 4 to 5 pieces placed in incubation media on squares of sterile capillary matting in a multiwell incubating plate. The incubation medium (medium-I) consisted of a n equal mixture (1:l v/v) of DMEM and Ham’s F12 nutrient medium which contained fetal calf serum substitute, Ultroser (2%), nonessential amino acids for minimum essential medium (1961, Hepes buffer (20 mM), L-glutamine (10 mM), insulin (10 Fg ml-’), gentamycin (50 Fg ml-’), and amphotericin-B (5 pg ml-’1. Flint et al. (1986) have shown t h a t maximal labelling in endometrium using 10 p.Ci/ml was reached after 2 h incubation. In this study, the conditions used to label endometrium were based on those described by Flint et al. (1986) with minor modifications. Tissue was prelabelled for 3 h a t 37°C in 95% O2 and 5% CO, at 95% humidity with 5 FCi ml-l my~-[Z-~H]-inositol containing incubation medium.

P A F stimulation After removal of the labelling medium, tissue was washed by incubation with 1ml of a mixture of DMEM and Ham’s F12 nutrient medium (1:l v/v) for 30 min followed by two consecutive incubations with fresh medium-I for 15 min. During the third incubation, tissue explants were exposed to 10 mM LiC1. All incubations were performed at 37°C in 95% 0, and 5% CO, a t 95% humidity, and the tissue was rinsed with 1ml of HBSS between each incubation. Following the steps to remove the unincorporated label, stimulations were initiated by the addition of 1.0 ml of medium-I containing 10 mM LiCl or PAF-acether in medium-I containing 10 mM LiCl or PAF-acether in the presence of the PAF antagonist WEB 2086. The reaction was terminated by aspirating the medium and replacing with 1ml of 15%(wiv) trichloroacetic acid at 4°C. The multiwell plates were kept at 4°C for 15 min to extract cellular ”-labelled inositol phosphates. Measurements of inositol phosphates Inositol phosphates were extracted and the inositol phosphate fractions analyzed by anion-exchange chromatography on small columns of Dowex 1-X8 by methods described previously (Berridge et al., 1983; Heslop et al., 1986). The ether-extracted and neutralized tissue extracts were applied to the columns, which were washed with 3 0 ml of water to remove [3Hlinositol and glycerophosphoinositol eluted with 5 ml of 60 mM sodium formate. 13H11nositol mono-, bis-, and trisphosphate fractions UPl, IP,, and IP, fractions respectively) were eluted with sequential 5 ml batches of 0.1 M formic acid/02. M ammonium formate, 0.1 M formic acid/ 0.5 M ammonium formate, and 0.1 M formic acidil.0 M ammonium formate. Elution positions were confirmed by using tritiated standards. In some experiments, the IP,, IP,, and IP, fractions were eluted together from the columns with 0.1 M formic acidil.0 M ammonium formate immediately after the elution of glycerophosphoinositol and are referred to as “total inositol phosphates” in this paper. The column eluates were collected directly into scintillation vials, to which were added 10 ml of scintillation fluid, and counted in a 1500-TRICARB liquid scintillation analyzer (Packard Instrument International S.A., Zurich, Switzerland).

Statistical analysis Data was expressed as countsiminlmg dry weight (c.p.m. mg-l dry wt) or as a percentage of control. Statistical analysis used was nonparametric Wilcoxon’s signed rank test on inositol phosphate profiles showing non-Gaussian distribution. Student’s paired t test was used on log normalized data where indicated and the values were expressed a s mean (tsem). The non-Gaussian data were expressed a s median with range. Figures presented as percentage of control are expressed as mean (* sem) with statistical analysis performed on c.p.m./mg-l dry wt. RESULTS The mean ( 5 sem) age of the patients was 41 ? 5 years (range: 25-49 years), and all specimens revealed normal endometrium.

PAF-INDUCED ENDOMETRIAL PHOSPHOLIPASE C ACTIVATION

25C

209

250

(4 **

T

200

200

i

150

100

50

Control IPi

IP2

IP3

.

Y

.

Control IPi

I P ~ IPS

INOSITOL PHOSPHATES Fig. 1. Effect of PAF on accumulation of inositol phosphates in human secretory endometrium. After the addition of 1.8 p M PAF, incubations were determined in the presence of 10 mM LiCl, at 1 min (A) and at 30 min (B). Results are presented as mean (tsem) percentage of basal activity (solid columns), which is expressed as 100. IP, (dashed columns), IP, (crossed columns), and IP, (hatched columns) are inosi-

to1 monophosphate, bisphosphate, and trisphosphate fractions respectively. Statistical analysis using nonparametric Wilcoxon's signed rank test was performed on c.p.m. mg dry wt of inositol phosphates. n = 12, *P < 0.05, **P < 0.002,***P i 0.003 when compared to basal values of inositol phosphates.

150 N.S

100

100

50

50

0

n

Control IP?

IPz

IR

Control

IPi

I P ~IP~

INOSITOL PHOSPHATES Fig. 2. Effect of PAF on accumulation of inositol phosphates in human proliferative endometrium. After the addition of 1.8 pM PAF', incubations were determined in the presence of 10 mM LiCl, at 1min (A)and at 30 min (3). Results are presented as mean I( sem) percentage of basal activity (solid columns), which is expressed as 100. IP, (dashcd columns), IP, (crossed columns), and IP, (hatched columns)

are inositol monophosphate, bisphosphate, and trisphosphate fractions respectively. Statistical analysis using nonparametric Wilcoxon's signed rank test was performed on c.p.m. mg-' dry wt of inositol phosphates. n = 6, P < N.S. when compared to basal values of inositol phosphates.

PAF-induced hydrolysis of PtdIns(4,5)P2 a s measured by accumulation of inositol phosphates was significantly increased from endometrium obtained during the secretory phase (Fig.l), but had no effect on endometrium removed in the proliferative phase of the cycle (Fig. 2). No difference was demonstrated in the basal values of inositol phosphates between proliferative and secretory endometrium (Table 1). The IP, fraction was significantly elevated from a median value of 14.0 c.p.m. mg-' dry wt [range: 8-41

c.p.m. mg-' dry wt] to 28.0 c.p.m. mg-' dry wt [11-87 c.p.m. mg-I dry wt, P < 0.0021 following 1 min exposure of secretory endometrium to 1.8 pM PAF-acether, in the presence of 10 mM LiC1. Levels of IP, were also increased ( P < 0.05), whilst IP, levels remained constant (Fig. 1A). After 30 min incubation in the absence of PAF (basal release), the median levels of IP,, IP,, and IP , in the presence of 10 mM LiC1, were 353.5 c.p.m. mg-' dry wt, 59.5 c.p.m. mg-I dry wt, and 21.5 c.p.m. mg-' dry wt

AIIMED AND SMITH

210

TABLE 1. Accumulation of basal inositol phosphates in secretory and proliferative human endometrium after 1and 30 min incubation in the uresence of 10 mM LiCl Secretory endometrium

Proliferative endometrium

-

1rnin

Median Range

30 min

1 min

30 min

IPl

IP2

1p3

IPI

IP2

IP3

IP,

IP,

1P.S

IPl

IP*

IPS

201.0 68-989

45.0 21-162

14.0 1147

353.5* 14S1074

59.5 21-240

21.5* 11-99

551.0 14S809

62.5 22-151

21.0 1150

767.0 24S1207

71.5

25.5 13-70

29-240

Strips of endometrium were prepared, labeled with 13H1-inositola s described in Materials and Methods. Data shown are median values of means of triplicate determinations of 12 experiments from secretory and six experiments from proliferatlve endometrium. Data expressed as c.p.m. mg dry wt; median with ranges. IP1, IPz, IP, arc inositol monophosphate, bisphosphate, and trisphosphate fractions respectively. *P c 0.01 when compared to basal inositol phosphate values a t 1 min incubation using Wilcuxon'b signed rank Lest.

350

** ***

300

****

250

* 200

150

100

50

0

I -r

0.0

-r

0.009

7

0.09

7

1 .a

0.9

5.4

Platelet activating factor / pM

Fig. 3. Effect of increasing concentrations of PAF on total inositol phosphates accumulation in human secretory endometrium. Values are presented as means (2sem) and expressed as e.p.m. mg-' dry wt of tissue. Student's paired t test was used on log normalized data. n = 6, "P < 0.004, **P 0.002, ***P< 0.0006, and **-*P < 0.0001 when compared to basal values of inositol phosphates. :s

respectively. Levels of IP, and IP, were significantly increased compared to basal levels observed after 1min incubation (P < 0.01; P < 0.01, respectively). After 30 min exposure of secretory endometrium to PAF, the levels of IP, and IP, were increased to median values of 452.0 c.p.m. mg-l dry wt [171-1,476 c.p.m. mg-l dry wt, P < 0.0031 and 33.5 c.p.m. mg-l dry wt 19-167 c.p.m. mg-' dry wt, P < 0.0031 respectively. However, the increase in the level of IP, just reached statistical significance a t the 5% levcl (Fig. 1B). PAF caused a n increase in the intracellular levels of tritiated total inositol phosphates in a concentrationdependent manner in secretory endometrium following a 30 min incubation in the presence of 10 mM LiCl (Fig. 3). Maximal stimulation was produced at 1.8 pM PAF and the lowest dose used was 0.009 pM PAF. At a concentration of 0.009 pM PAF, the mean total inositol phosphate level increased from 170.74 5 19.36 c.p.m. mg-l dry wt 1101-219 c.p.m. mg-' dry wt] to 202.44 ? 18.41 c.p.m. mg-l dry wt [13&248 c.p.m. mg-l dry wt, P < 0.0041. After 0.09 pM PAF, the mean total inositol phosphate value further increased to 242.38 2 18.14 c.p.m. mg-l dry wt D79-389 c.p.m. mg-' dry wt1 and

the increase was significant compared to that observed a t 0.009 pM PAF (P < 0.003) using Student's paired t test. With a n increase in PAF concentration still further to 0.9 pM PAF, the mean total inositol phosphate value was 284.92 2 23.49 c.p.m. mg-' dry wt [199-328 c.p.m. mg-' dry wt, P < 0.02 compared with inositol phosphates a t 0.09 pM PAF using paired t test]. The stimulatory effect of 1.8 pM PAF on total inositol phosphates accumulation was attenuated by the PAF receptor antagonist WEB 2086 (Fig. 4). Inhibition by WEB 2086 was dose-dependent and the effect of PAF on total inositol phosphates a t 10 pM WEB 2086 was almost completely abolished. PAF-induced inositol phosphate accumulation as a percentage of WEB 2086 alone was 81.5% 2 8.5%; however, in the presence of 100 nM WEB 2086, the PAF response was only 38.0% 4.2% and further decreased to 8.1% L 5.2% a t 10 pM WEB

*

2086.

DISCUSSION This study demonstrates that PAF stimulates PtdIns(4,5)P2breakdown resulting in the generation of inositol phosphates in endometrium obtained from

211

PAF-INDUCED ENDOMETRIAL PHOSPHOLIPASE C ACTIVATION

300

5

1

i

200

L.

****

-r

'D

. E"

*** __

€

2

100

0

-r

WEB Alone

PAF Alone

T

7

lo-'

10-7

10-6

WEB 2086 Dose (mol/l) Fig. 4. Effect of increasing concentrations of WEB 2086 on accumulation of total inositol phosphates by P A F in human secretory endometrium. The tissue was preincubated for 10 min with WEB 2086 and in the presence of 10 mM LiCl, incubations were determined at 30 min after addition of 1.8 pM-PAF. The open column represents the effect of WEB 2086 alone on inositol phosphates accumulation and the effect of PAF on increasing concentration of WEB 2086 is shown by the

hatched columns. Results are presented as mean ( * sem) and expressed a s c.p.m. mg-' dry wt. Statistical analysis was by Student's paired t test on log normalized c.p.m. mg-' dry wt of total inositol phosphates. n = 6, *P< 0.02, **P < 0.004, ***P < 0.001, and ****P .< 0.0005 when compared to PAF stimulated values of inositol phosphates in the absence of the antagonist (solid column).

women during the secretory phase of the menstrual cycle, but has no significant effect on PtdIns(4,5)P2 hydrolysis in proliferative endometrium. The profile of inositol phosphates accumulation in response to PAF in secretory phase endometrium suggests that the initial event is the hydrolysis of PtdIns(4,5)P2 resulting in the rapid formation of the IP, fraction which in the presence of LiCl undergoes stepwise dephosphorylation to IP, and then to IP,. The response to PAF was rapid, dose-related, and inhibited by the specific PAF receptor antagonist in a dose-dependent manner, indicating a receptor-mediated mechanism. Inositol phosphate within each fraction shows wide variation between women. A possible explanation for the variation in range, apart from the likely variation between individual women, may be that the tissue collected at different stages of the secretory phase of the menstrual cycle respond to PAF to varying degrees. This study not only demonstrates the rapid accumulation of IP, fraction following PAF stimulation, which suggests activation of endometrial PtdIns(4,5)P2-specific phospholipase C, but also shows that the effect of PAF on PtdIns(4,5)P2 hydrolysis is confined to the secretory endometrium, thus implying that the PAF response may be under ovarian steroid regulation. It is interesting to note that PAF production was shown to be regulated by ovarian steroids in endometrium (Alecozay et al., 1991). The rapid formation of IP, fraction has also been reported in cultured rabbit endometrial cells with 1pM PGF,, (Orlicky et al., 1986). These findings are consis-

tent with PAF-mediated response in other tissues. PAF has been shown to stimulate accumulation of inositol phosphates in human platelets (Shukla, 1985), in bovine pulmonary artery endothelial cells (Kawaguchi et al., 1990), in rat and bovine cultures of anterior pituitary cells (Grandison, 1990), and in strips of rat (Varol et al., 1989) and human (Schrey et al., 1988) myometrium. These and other studies indicate that PAF binds to a specific membrane receptor and activates the PtdIns(4,5)P2-specific phospholipase C pathway via a G-protein resulting in the hydrolysis of PtdIns(4,5)P2 which yields two second messengers, DAG and Insi1,4,5)P, (Barzaghi et al., 1989). In addition it has been reported that PAF stimulates protein phosphorylation of pp6OC-"" tyrosine kinase and causes its rapid translocation from cytosol to membranes in rabbit platelets (Dhar and Shukla, 1991). It is not known whether the activation of Ptd1ns(4,5)P2-specific phospholipase C by PAF occurs via a G-protein or through a tyrosine kinase activity in endometrium. Sawyer and Andersen (1989) reported that PAF at concentration 3 4 FM disrupts the barrier properties of the cell membrane and suggested that this might increase intracellular calcium. In this study, the effect of PAF on inositol phosphate accumulation was not due to nonspecific membrane perturbations as the specific PAF receptor antagonist WEB 2086 attenuated the effect of 1.8 pM PAF in a dose-dependent manner, indicating a receptor-specific response. In addition, PAF is rapidly metabolized by a PAF-specific acetylhydrolase (Blank et al., 1981) and it is probable that the endome-

212

AHMED AND SMITH

trial “tissue explants” were exposed to significantly lower concentrations of PAF than were initially added to the wells. Varol et al. (1989) reported that PAF at 1 FM concentration stimulated phosphatidylinositol hydrolysis in strips of rat myometrium and Grandison (1990)reported that PAF stimulated inositol phosphate accumulation in cultures of rat anterior pituitary cells was maximal at 100 to 1,000 nM PAF concentration. The increased basal concentration of IP,, IP2, and IP, with longer incubation time suggests that there is a basal turnover of inositol phosphates in human endometrium in the presence of 10 mM LiCl and is in agreement with that observed in Ishikawa cells (Weiss and Gurpide, 1988). The increased accumulation of inositol phosphates in tissue stimulated with PAF after 30 min incubation is presumably due to Lit-inhibition of inositol-1-phosphatase which hydrolyses IP, to free inositol and phosphate (Joseph and Williams, 1985). Although the “IP, fraction” eluted with 0.1 M formic acidil .O M ammonium formate includes both the isomers Ins(1,4,5)P3and Ins(1,3,4)P, together with inosito1 1,3,4,5-tetrakisphosphate(Ins(1,3,4,5)P4)(Batty et al., 1985; Heslop et al., 1985), it is a valid measure of phospholipase C activity, because Ins(1,4,5)P, formed directly by PtdIns(4,5)P, breakdown appears to be the only precursor for the formation of Ins(1,3,4,5)P4and Ins(1,3,4)P3(Taylor et al., 1986).It may be inferred that PAF induced a rapid and concentration-dependent formation of Ins(1,4,5)P3 in secretory endometrium. As PtdIns(4,5)P2hydrolysis by PtdIns(4,5)P2-specificphospholipase C yields not only Ins(1,4,5jP3but also DAG (Berridge and Irvine, 1989), which is an additional source of free arachidonic acid for eicosanoid synthesis, it is inferred that PAF generates DAG from endometrium. In other systems the biological action of PAF is modulated through binding to specific membrane sites (Hwang et al., 1985; Hwang, 1987; Vallari et al., 1990; Snyder, 1990). Kudolo and Harper characterized highaffinity PAF binding sites on rabbit endometrium, suggesting that binding sites represent the PAF receptor (Kudolo and Harper, 1989). Platelet-activating factor antagonists that were not structural analogs of PAF did not compete to displace [,H]PAF in this system, implying that uterine binding sites differ in their specificity to those in other tissues (Junier et al., 1988; Van Delft et al., 1988). Experiments with L3H1WEB indicates its binding site on human platelets is identical to the specific [3H]PAFbinding sites (Ukena et al., 1988). In this study, the effect of PAF on PtdIns(4,5)P2hydrolysis was inhibited by WEB 2086 in a dose-dependent manner, suggesting the effect was mediated via a specific endometrial PAF receptor. There are several consequences of these actions of PAF. First, PAF-induced phospholipase C activation results in increased production of Ins(1,4,5)P3 which elevates the intracellular calcium ([Ca2+1,)levels required for Ca2+-dependent processes. PAF has been shown t o induce ICa”1, mobilization in human platelets (Hallam et al., 1984). hs(1,4,5)P3 releases Ca2+ from a nonmitochondrial pool which has characteristics to suggest that it is the endoplasmic reticulum, but only part of this pool seems to be Ins(l,4,5)P3-sensitive (Streb et al., 1984). Immunocytochemical studies using

a specific antibody on Purkinje cells reveal that Ins(1,4,5)P3 receptor is localized on the nuclear envelope and on parts of the endoplasmic reticulum, near the nucleus (Ross et al., 1989). To release Ca2+, Ins(1,4,5)P3must bind to receptors that are linked to Ca2* channels connected with the Ins(1,4,5)P,-sensitive Ca2+ pool. This Ins(1,4,5)P,-induced Caz+ signal can drive a process of Ca2+-induced Ca2+release from the Ins(l,4,5)P3-insensitivepools to produce a spike which might be organized in the form of a wave, so spreading the signal throughout the cell (Berridge and Irvine, 1989). As Ins(1,4,5)P3is a key regulator of the periodic release of internal calcium, it is reasonable t o assume that activation of endometrial PtdIns(4,5)P2specific phos holipase C by PAF may result in the elevation of [Ca%+1,. Secondly, elevated [Ca”], increases the activity of phospholipases. In many tissues, including human (Bonney, 1985) and guinea-pig (Downing and Poyser, 1983) endometrium, phospholipase A, activity is Ca2+dependent and its activity is influenced by intracellular levels of free Ca2’ (Irvine, 1982;Hokin, 1985)which reflect the release of Ca2+ from intracellular stores, cytosolic Ca2+-bindingproteins like calmodulin, and the transport of Ca2+across the plasma membrane (Carafoli, 1987). CaZt stimulation of phospholipase A, activity after exposure to PAF represents an additional potential source of arachidonic acid for prostaglandin synthesis. Smith and Kelly showed that PAF stimulates PGE, production only from secretory phase endometrium (Smith and Kelly, 1988)which is consistent with the pattern of inositol phosphates accumulation in this study. The stimulation of secretory endometrial phospholipase C and PAF may in part account for PAFinduced secretion of PGE,. Thus PAF-induced prostaglandin synthesis may arise either by the release of arachidonic acid from DAG or by increased phospholipase A, activity. The site of implantation is characterized by an inflammatory-type response with expansion of extracellular fluid volume, increased vascular permeability, and vasodilation (Hertig, 1964). Animal studies show that levels of prostaglandin E,, PGF,,, and 6-ketoPGF,, are elevated at the site of implantation (Kennedy and Zamecnik, 1978) and PGE, increases local vascular permeability and stromal oedema (Kennedy, 1983). PAF, like PGE2, is a highly potent inducer of vascular permeability (Humphrey et al., 1984). An increase in vascular permeability in the stroma at the site of impending blastocyst implantation occurring over the 24-h period prior to attachment is an obligate accompaniment of normal implantation and decidualisation in rodents. Mouse embryos in culture produce maximal PAF a t a time corresponding to development of the morula (Angle et al., 198813). Plateletactivating factor is produced not only by the embryo during early pregnancy, but also by the stroma of human secretory endometrium (Alecozay et al., 1989; Harper et al., 1989). Uterine PAF production is highest just prior to implantation and drops in areas of the uterus adjacent to the embryo at the time of implantation (Johnston et al., 1990).These findings suggest that at the site of blastocyst implantation, the localized high concentration of PAF released either by the embryo or

PAF-INDUCED ENDOMETRIAL PHOSPHOLIPASE C ACTIVATION

from the stroma displaces the biologically inactive PAF' precursor, lyso-PAF, from receptors on the endometrium, thus enabling PAF receptor coupling to take place (Harper et al., 1989). In summary, this study shows that the action of PAF is mediated via a specific PAF receptor in human endometrium, which when coupled to the agonist activates endometrial PtdIns(4,5)P2-specific phospholipase C only in the secretory phase of the menstrual cycle, thus suggesting that the PAF response may be under ovarian steroid regulation and that the ability of the endometrium to respond to PAF appears to be a feature of the preparation of this tissue for implantation in women.

ACKNOWLEDGMENTS We thank Dr. C.H. Weber, Boehringer Ingelheim Limited, Ingelheim AM Rhine, Germany, for the generous gift of WEB 2086 and Dr. R.F. Irvine, Dr. K.D. Brown, and Mr. C.J. Littlewood for their helpful suggestions during this study. This study was supported by a grant 03006011.5 from the Wellcome Trust.

213

mechanism by which oxytocin controls prostaglandin synthesis in the ovine endometrium. Biochem. J., 237t797-805. Grandison, L. (1990) Platelet activating factor induces inositol phosphate accumulation in cultures of rat and bovine anterior pituitary cells. Endocrinology,127t17861791. Hallam, T.J., Sanchez, A., and Rink, T.J. (1984) Stimulus-response coupling in human platelets, changes evoked by platelet activating factor in cytoplasmic free calcium monitored with the fluorescent calcium indicator quin 2. Biochem. J.,218r819-827. Harper, M.J.K., Kudolo, G.B., Alecozay, A.A., and Jones, M.A. (1989) Platelet-activating factor (PAF) and blastocyst-endometrial interactions. b o g . Clin. Biol. Res., 294t305-315. Hertig, A.T. (1964) Gestational hyperplasia of endometrium; a morphologic correlation ova, endometrium, and corpora lutea during early pregnancy. Lab Invest., 1311153-1191. Heslop, J.P., Irvine, R.F., Tashjian, A.H. Jr., andBerridge, M.J. (1985) Inositol tetrakis- and pentakisphosphate in GH4 cells. J . Exp. Biol., 119t395401. Heslop, J.P., Blakeley, D.M., Brown, K.D., Irvine, R.F., and Berridge, M.J. (1986) Effect of bombesin and insulin on inositol (1,4,5) trisphosphate and inositol (1,3,4)trisphosphate formation in Swiss 3T3 cells. Cell, 47:703-709. Hokin, L.E. (1985) Receptors and phosphoinositide-generatedsecond messengers. Annu. Rev. Biochem. 54205-235. Humphrey, D.M., McManus, L.M., Hanahan, D.J., and Pinckard, R.N. (1984) Morphologic basis of increased vascular permeability induced by acetyl glyceryl ether phosphorylcholine. Lab. Invest., 50: 1 6 2 5 .

Hwang, S.B. (1987) Specific receptor sites for platelet activating factor on rat liver plasma membranes. Arch. Biochem. Biophys., 257r339LITERATURE CITED 344. Alecozay, A.A., Casslen, B.G., Riehl, R.M., DeLeon, F.D., Harpcr, Hwang, S.B., Lam, M.H., and Shen, T.Y. (1985) Specific binding sites for platelet activating factor in human lung tissue. Biochem. BioM.J.K., Silva, M., Nouchi, T., and Hanahan, D.J. (1989) Plateletphys. Res. Commun., 18:972-979. activating factor (PAF) in human luteal phase endometrium. Biol. Irvine, R.F. (1982) How is the level of free arachidonic acid controlled Reprod., 41t578-586. in mammalian cells? Binchem. J.,204.3-16. Alecozay, A.A., Harper, M.J.K., Schenken, R.S., and Hanahan, D.J. Johnston, J.M., Noriei, M., Angle, M.J., and Hoffman, D.R. (1990) (1991) Paracrine interaction between platelet-activating factor and Regulation of the arachidonic acid cascade and PAF metabolism in prostaglandins in hormonally-treated human luteal phase endoreproductive tissues. In: Eicosanoids in Reproduction. M.D. Mibchmetrium in vitro. J . Reprod. Fertil., 91t301-312. ell, ed. CRC Press, Boston, pp. 5-37. Angle, M.J., Jones, M.A., McManus, L.M., Pinckard, R.N., and Harper, M.J. (1988a) Platelet-activating factor in the rabbit uterus Joseph, S.K., and Williams, R.J. (1985) Subcellular localization and some properties of the enzymes hydrolysing inositol polyphosphates during early pregnancy. J. Reprod. Fertil., 83t711-722. in rat liver. FEBS Lett., 180t150-154. Angle, M.J., Byrd, W., and Johnston, J.M. (198813)Embryonic production of platelet-activating factor in culture. Fertil. Steril., 566 Junier, P.M., Tiberghien, C., Rougeot, C., Fafeur, V., and Dray, P. (1988)Inhibitory effect of platelet-activating factor (PAF) on lutein(suppl):Abstr. 158. izing hormone-releasing hormone and somatostatin release from rat Barzaghi, G., Sarau, H.M., and Mong, S. (1989) Platelet-activating median eminence in vitro correlated with the characterization of factor-induced phosphoinositide metabolism in differentiated U-937 specific PAF-receptor sites in rat hypothalamus. Endocrinology, cells in culture. J. Pharmacol. Exp. Ther., 2 4 8 5 5 9 6 6 6 , 123:72-80. Batty, I.R., Nahorski, S.R., and Irvine, R.F. (19851Rapid formation of Kawaguchi, H., Sawa, H., and Yasuda, H. (1990) Mechanism of ininositol 1,3,4,5-tetrakisphosphatefollowing muscarinic receptor creased angiotensin-converting enzyme activity stimulated by stimulation of rat cerebral cortical slices. Hiochem. J., 232r211-215. platelet-activating factor. Biochim. Bipohys. Acta, 1052:503-508. Berridge, M.J. (1984) Inositol trisphosphate and diacylglycerol as secKennedy, T.G. (1983) Embryonic signals and the initiation of blastoond messengers. Biochem. J., 220t345360. cyst implantation. Aust. J. Biol. Sci., 3 6 5 3 1 4 4 3 . Berridge, M.J., and Irvine, R.F. (1989) Inositol phosphates and cell Kennedy, T.G., and Zamecnik, J . (1978) The concentration of 6-ketosignalling. Nature, 341:197-205. prostaglandin F1 alpha is markedly elevated at the site ofblastocyst Berridge, M.J., Dawson, R.M.C., Downes, C.P., Heslop, J.P., and Irimplantation in the rat. Prostaglandins, 61 1599-605. vine, R.F. (1983) Changes in the levels of inositol phosphates after agonist-dependent hydrolysis of membrane phosphoinositides. Bio- Kudolo, G.B., and Harper, M.J.K. (1989) Characterization of platelet activating factor binding sites on uterine membranes from pregnant chem. J., 222:195-201. rabbits. Biol. Reprod., 41r587-603. Blank, M.L., Lee, T-C., Fitzgerald, V., and Snyder, F. (1981) A specific Nishizuka, Y. (1989) The molecular heterogeneity of protein kinase C acetylhydrolase for l-0-alkyl-2-acetyl-sn-glycero-3-phosphorylchoand its implication for cellular regulation. Nature, 334:661-665. line (a hypotensive and platelet activating lipid). J. Biol. Chem., Noyes, R.W., Hertig, A.T., and Rock, J. (1950) Dating the endometrial 236t175-178. biopsy. Fertil. Steril., lr3-25. Bonney, R.C. (1985) Measurement of phospholipase A, activity in human endometrium during the menstrual cycle. J. Endocrinol., O"ei1, C. (1985a) Thrombocytopenia is an initial maternal response to fertilization in mice. J . Reprod. Fertil., 73559-566. 107t183-189. Carafoli, E. (1987) Intracellular calcium homeostasis. Annu. Rev. Bio- ONeil, C. (1985b3 Examination of the cause of early pregnancy associated thrombocytopenia in mice. J . Reprod. Fertil., 73567-577. chem., 56:395-443. Collier, M., ONeil, C., Ammit, A.J., and Saunders, D.M. (1988) Bio- ONeil, C. (1985~)Partial characterization of the embryo-derived platelet-activating factor in mice. J. Reprod. Fertil., 75t375-380. chemical and pharmacological characterization of human embryoONeil, C., Gidley-Baird, A.A., Pike, I.L., Porter, R.N., Sinosich, M.J., derived platelet activating factor. Human Reprod., 3t993-998. and Saunders, D.M. (1985) Maternal blood platelet physiology and Dhar, A. and Shukla, S.D. (1991) Involvement of pp60'-"'" in plateletluteal phase endocrinology as a means of monitoring pre- and postactivating factor-stimulated platelets. Hiol. Chem., 266t18797implantation embryo viability following in vitro fertilization. J. In 18801. Vitro Fertil. Embryo Transf., 2r87-93. Downing, I., and Poyser, N.L. (1983) Estimation of phospholipase A, Orlicky, D.J., Silio, M., Williams, C., Gorden, J., and Gerschenson, activity in guinea-pig endometrium on day 7 and 16 of the estrous L.E. (1986) Regulation of inositol phosphate levels by prostaglancycle. Prostaglandms Leukotrienes Med., 22t107-117. dins in cultured endometrial cells. J. Cell. Physiol., 128r105-112. Flint, A.P.F., Leat, W.M.F., Sheldrick, E.L., and Stewart, H.J. (1986) Roberts, T.K., Adamson, L.M., Smart, Y.C., Stanger, J.D., and MurStimulation of phosphoinositide hydrolysis by oxytocin and the

214

AHMED AND SMITH

doch, R.N. (1987) An evaluation of peripheral blood platelet enumeration as a monitor of fertilization and early pregnancy. Fertil. Steril., 473484354, Ross, C.A., Meldolesi, J.,Milner, T.A., Satoh, T., Supattapone, S., and Snyder, S.H. (1989) Inositol1,4,5-trisphosphate receptor localized to endopiasmic reticulum in cerebellar Purkinje neurons. Nature, 339t468-470. Sawyer, D.B. and Andersen, O.S. (1989) Platelet-activating factor is a general membrane perturbant. Biochim. Biophys. Acta, 987:129132. Schrey, M.P., Cornford, P.A., Read, A.M., and Steer, P.J. (1988)A role for phosphoinositide hydrolysis in human uterine smooth muscle during parturition. Am. J. Obstet. Gynecol., 159:964-970. Shukla, S.D. (1985) Platlet activating factor stimulated formation of inositol trisphosphate in platelets and its regulation by various agents including Ca'+, indomethacin, CV3988 and forskolin. Arch. Biochem. Biophys., 240t674-681. Smith, S.K., and Kelly, R.W. (1988) Effect of platelet-activating factor on the release of PGF,, and PGE, by separated cells of human endometrium. J. Reprod. Fertil., 82r271-276. Snyder, F. (1990)Platelet-activating factor and related and acetylated lipids as potent biologically active cellular mediators. Am. J. Physiol., 259:C697-C708. Streb, H., Bayerdorffer, E., Hasse, W., Irvine, R.F., and Schulz, I.

(1984) Effect of inositol-1,4,5-trisphosphate in isolated subcellular fractions of rat pancreas. J. Membr. Biol., 81:241-253. Taylor, C.W., Merritt, J.E., Putney, J.W. Jr., and Rubin, R.P. (1986) Effect of Ca"+ on phosphoinositide breakdown in exocrine pancreas. Biochem. J.,238t765-772. Ukena, D., Dent, G . , Birke, F.W., Robaut, C., Sybrecht, G.W., and Barnes, P.J. (1988) Radioligand binding of antagonists of plateletactivating factor to intact human platelets. FEBS Lett., 228t285289. Vallari, D.S., Austinhirst, R., and Snyder, F. (1990) Development of specific functionally active receptors for platelet-activating factor in HL-60 cells following granulocytic differentiation. J. Biol. Chern., 255:42614265. Van Delft, J.L., Van Haeringen, N.J., Verbey, N.L.J., Domingo, M.T., Chabrier, P.E., and Brawuet, P. (1988) Specific receptors sites for PAF in iris and cilary body of the rabbit eye. Curr. Eye Rcs., 7:10631068. Varol, F.G., Hadjiconstantinou, M., Tavers, J.B., and Neff, N.H. (1989) Platelet-activating factor stimulates phosphoinositol hydrolysis in the rat myometrium. Eur. J. Pharmacol., 159t97-98. Weiss, D.J., and Gurpide, E. (1988) Regulation of phosphoinositide hydrolysis in transformed human endometrial cells. Endocrinology, 123t981-990.