- Email: [email protected]
Prostaglandin production and stimulation by angiotensin II in the isolated perfused human placental cotyledon
BASIC SCIENCE SECTION
Prostaglandin production and stimulation by angiotensin II in the isolated perfused human placental cotyledon D. G. Glance, B.Sc., M.G. Elder, M.D!, and L. Myatt, Ph.D. London, England Levels of prostaglandins E and F2a, thromboxane 8 2 , and 6-oxo-prostaglandin F,. were measured by radioimmunoassay in the maternal and fetal effluents of isolated human placental cotyledons perfused in vitro~ All prostaglandins measured were released in greater amounts by the maternal side than by the fetal side of the perfused cotyledon although there were no consistent concentration gradients between the two sides. The approximate rank order of prostaglandin release into the maternal ~ide was thromboxane 8 2 >prostaglandin F2."" prostaglandin E ""6-oxo-prostaglandin F,., and that into the fetal side was thromboxane 8 2 "" prostaglandin F2a"" prostaglandin E"" 6-oxo-prostaglandin F, •. Injection of angiotensin II (0.5 ,...g) into the fetal circulation stimulated prostaglandin E and 6-oxo-prostaglandin F,. but not thromboxane 8 2 and prostaglandin F2• release into the fetal circulation and had no effect on maternal release. Angiotensin II (0.5 ,...g) had no effect on either side of the perfused cotyledon when injected into the maternal circulation. It is proposed that prostaglandin release into both maternal and fetal circulations may be flow-dependent and that the angiotensin 11-stimulated release of prostaglandin E and 6-oxoprostaglandin F,. may serve to modulate the vasoactive actions of angiotensin II on the fetal vasculature. (AM J 0BSTET GYNECOL 19!35;151:387-91.)
Key words: Placenta, prostaglandin production, angiotensin II
There has been much speculation concerning the role of prostaglandins released by the placenta. It has been suggested that prostaglandin E2 (PGE 2) may regulate inequalities of maternofetal perfusion by acting as a vasoconstrictor of the fetal-placental vessels and as a vasodilator of the uterine vasculature. 1 Prostacyclin (PGI 2) of placental origin may contribute to the elevated plasma PGI 2 levels and the refractoriness to angiotension II (All) found in pregnancy! It may also safeguard maternal blood supply to the placenta by preventing platelet aggregation although this process may also be influenced by placental thromboxane A2 (TxA2) synthesis since TxA 2 is a potent platelet ;1ggregator. Prostaglandip. production by the fetal-placental vasculature could be involved in controlling blood flow through these vessels since it has been shown that prostaglandin F2• (PGF2.) and PGE 2contract the fetal vessels From the Institute of Obstetrics and Gynaecology, Hammersmith Hospital. Action Research for the Crippled Child provided financial support of this work. Received for publication january 30, 1984; accepted September 20, 1984. Reprint requests: Dr. L. Myatt, Institute of Obstetrics and Gynaecology, Hammersmith Hospital, Du Cane Road, London, England W12 OHS. .
of the placenta whereas PGI 2 relaxes them.' Release of prostaglandins into the fetal circulation may also give rise to systemic effects within the fetus contributing to the low vascular resistance of the fetal circulation4 and the patency of the ductus arteriosus. 5 · Inpregnant rabbits, All plays an important part in the maintenance of systemic blood pressure, 6 and the same may be the case in the human since plasma All levels correlate with diastolic blood pressure in pregnant women. 7 All may play an important rote in the maintenance of uterine blood flow which is seen to decrease when the conversion of angiotensin I (AI) to All is blocked by captopril, an inhibitor of angiotensinconverting enzyme. Control of uterine blood flow may involve PGE 2 since All increases the production of PGE2 in the pregnant uterus 6 • 8 although the exact site of production within the uteroplacental complex is unclear. We have previously shown that components of the renin-angiotensin system act as very potent vasoconstrictors of the fetal-placental vasculature of the isolated perfused human placental cotyledon and it was proposed that, in vivo, the renin-angiotensin system may participate in the control of fetal-placental blood flow.~ Prostaglandins may modulate the actions of this system in the fetal-placental vasculature. Meclofena-
387
388 Glance, l;lder, and Myatt
mate, an inhibitor of prostaglandin synthesis, decreases the contractile response of fetal-placental vessels to All, suggesting that prostaglandins normally potentiate the contractile response. 3 To examine further the possible role of prostaglandins in the control of placental blood flows, we have studied the release of prostaglandin E (PGE), PGF2a, 6-oxo-prostaglandin F 1a (6-oxo-PGF 1a, the stable metabolite of PGI 2 ), and thromboxane B2 (TxB 2 , the stable metabolite of TxA2) from both sides of the isolated perfused human placental cotyledon preparation and looked at the modification of this production by All administered into either the fetal or the maternal circulation. Method Placentas from women with uncomplicated pregnancies were collected within 5 minutes of vaginal delivery or cesarean section at term. Placental cotyledons, free of infarcts or tears, were perfused by a method which has been described in detail elsewhere.• Briefly, this involved perfusing tissue culture medium 199 (Gibco) containing polyvinylpyrrolidone (5% w/v) (Sigma Chemical Co.) through a cannulated chorionic artery on the fetal surface of the placenta and collecting the effluent from the associated chorionic vein. Perfusion of the intervillous space was achieved by the insertion of two intravenous 21-gauge butterfly needles through the maternal surface of the perfused cotyledon. The whole preparation was maintained at 37° C in a controlled temperature cabinet. The flow rates of the fetal and maternal perfusates were maintained at 4 and 10 ml· min- 1 , respectively. The resulting perfusion pressures from both sides were between 30 and 60 mm Hg. In a preliminary series of experiments (three placentas) fractions of fetal and maternal effluent (1.5 ml) were collected at 20, 30, 40, 50, and 60 minutes after the establishment of the maternal circulation and immediateiy placed on ice at 4° C. Medium flowing thro~gh the perfusion tubing withoutthe cotyledon in place was collected to serve as a blank. The samples were centrifuged (1700 X g) immediately after each experiment to remove any blood contamination and the supernatants (1 ml) stored at- 20° C for subsequent extraction and analysis of prostaglandins. The samples were extracted with the use of Sep-Pak Cts cartridges (Waters) prior to radioimmunoassay. This removed interference from polyvinylpyrrolidone which was evident when direct radioimmunoassay of perfusates was used. The cartridges were first wetted with 20 ml of methanol/water (4: 1, v/v) and residual methanol removed by washing with deionized/distilled water (20 ml). The perfusate sample (1 ml) was acidified to pH 3.0 with formic acid, loaded on the cartridge, and
February 1, 1985 Am J Obstet Gynecol
eluted with diethyl ether (10 ml), and the fractions were collected in two siliconized glass test tubes. The solvent was evaporated under nitrogen and the residue resuspended in phosphate-buffered saline (Dulbecco) (1 ml). The extracted samples were assayed for prostaglandins with radioimmunoassays developed in this laboratory for PGE, 10 PGF2a/ 0 6-oxo~PGF ta> u and TxB 2 • The TxB 2 antisera cross reacted < 1% with 6-oxo-PGF 1a, prostaglandin E~> PGE 2 , PGF2", prostaglandin A2 , and arachidonic acid and <5% with prostaglandin D2 • The preliminary experiments revealed a large amount of variability in the results and that the levels of prostaglandins were sometimes very close to the level of detection of the assays. The experiments were repeated (three placentas) with collection of larger fractions of the fetal and maternal effluents (5 ml on the fetal side and 15 ml on the maternal side). Samples were treated and extracted in the same way as before, except that 4 ml of the fetal perfusate and 10 ml of the maternal perfusate were extracted and resuspended in 2 ml of phosphate-buffered saline, which would theoretically concentrate the prostaglandins two and five times, respectively. The samples were assayed as before for PGE, PGF 2u, 6-oxo-PGF 1"' and TxB 2 • To assess the effect of All on the production of prostaglandins from the perfused cotyledon, control samples (5-minute duration of flow) were collected from both sides of three placentas beginning 25 minutes after the establishment of maternal perfusion. All (Sigma Chemical Co., 0.5 IJ.g in 0.5 ml of gassed perfusate) was injected into the maternal circulation and fetal effluent collected for a further 5 minutes. Further control samples of maternal and fetal effluent were taken (5 minutes' duration) 10 minutes after baseline perfusion pressures had been reestablished. A second dose of All (0.5 IJ.g in 0.5 ml) was then injected into the fetal circulation and fetal and maternal effluents collected for 5 minutes. All medium collected was treated and extracted as described above. The residue was resuspended in phosphate-buftered saline (2 ml) and assayed for PGF2u, PGE, 6-oxo-PGF 1u, and TxB 2 • All results of prostaglandin assays were expressed as concentration of the prostaglandin per milliliter of medium after correction for any blank reading obtained. Using labeled prostaglandins, we found that recoveries for all prostaglandins were approximately equal (range, 82% to 94%) and thus no corrections were made to the prostaglandin levels obtained after the extraction procedure. As the PGE antisera could not distinguish PGE 1 and PGE 2 , results from this assay are taken to indicate total E series prostaglandin, expressed in this paper as PGE. Statistical significance was calculated by Student's t test.
Prostaglandin production by placental cotyledon
Volume 151 Number 3
10
- - Fetal Effluent ...._Maternal Effluent 6-oxo-PGF
2
0
0
9
1~
**
[:3 Maternal Effluent
c:
i
Ol
c: 6
"' "'~
5 4
c::
3
"
2
-;:; a.
2
Fetal Effluent
389
* ... 121
1 0
6-oxo-
**
*
PGF 2a
PGE
6~ 0~
Cl TxB 2
PGFla
c 3 E 01
c
2
Q)
Vl
~ 0
PGE
j~
Fig. 2. Mean prostaglandin release during 60 minutes of perfusion into fetal (open columns) and maternal (hatched columns) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of six experiments. Significant difference from fetal circulation: * = p < 0.01 and ** = p < 0.005).
QJ
c::
a: 11
0
10
Fetal Effluent
[:3 Maternal Effluent 6
9
8
e5
7
Ol
c: c: 4
6
0
-.;::
5
;;"' c:
4
"'c:u
3
u
2
"a.
0
0
20
30
40
50
60
Perfusion time (min)
6-oxoPGF1a
PGE
Fig. I. Prostaglandin release into fetal (..----.) and maternal (____.) circulations of the isolated perfused human placental cotyledon during 60 minutes of perfusion (results are mean ± SEM of six experiments).
Fig. 3. Mean prostaglandin concentration during 60 minutes of perfusion in fetal (open columns) and maternal (hatched columns) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of six experiments).
Results None of the prostaglandins measured, apart from TxB 2, showed any consistent trend in the amount released from either side of the placenta during the 60 minutes of perfusion (Fig. 1). TxB 2 release showed a steady decline on the maternal side which tended to parallel the gradual washout of blood from the intervillous space. When the means of all values obtained at each time point throughout the course of perfusion were calculated, no statistical differences were found
between the mean amounts of prostaglandin released in the preliminary experiments and the repeat experiments, and so the two sets of results were combined (Fig. 2). The amount of each prostaglandin released into the maternal circulation was significantly greater than the amount released into the fetal circulation (TxB2, p < 0.0005; PGF 2., p < 0.005; PGE, p < 0.01; and 6-oxo-PGF 1. , p < 0.01). When the results were expressed in terms of concentrations of prostaglandins per milliliter of perfusate (Fig. 3) no difference was
390 Glance, Elder, and Myatt
February 1, 1985 Am J Obstet Gyneco1
[J Fetal Effluent E=:1 Matern~! Effluent 6
PGF 2a
TxB 2
1:
!
5 0>4
1:
~
cu 3
"'~
'ii
2
Cl:: 1
~
0
ct!JO FC MA FA
MC MA FA
ts6l5 FC MA FA
~Iii~ MC MA FA
Fig. 4. TxB 2 and PGF2• release into fetal (open columns) and maternal (hatched columns) circulations before (FC andMC) and after injection of All (0.5 ~J.g) into the fetal (FA) and maternal (MA) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of three experiments).
CJ
Fetal Effluent
E=:1
Maternal Effluent
...
? 1:
·e Cl
1:
-
.. cu
5
4
cu
D.
0
1a
~j -- 0 ~c~ .....
~ 2 1.-'1
6-oxo-PGF
PGE
6
150 FC MA FA
MC MA FA
FC MA FA
MC MA FA
Fig. 5. PGE and 6-oxo-PGF 1• release into the fetal (open columns) and maternal (hatched columns) circulations before (FC and MC) and after injection of All (0.5 ~J.g) into the fetal (FA) and maternal (MA) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of three experiments. Significant difference from fetal control: ** = p < 0.005).
seen between the two sides for any of the prostaglandins. The amounts of prostaglandins released into the maternal circulation had an approximate rank order of TxB 2 > PGF2• ::::o PGE ::::o 6-oxo-PGFI•• and those released into the fetal circulation had a rank order of TxB 2 ::::o PGE ::::o PGF2• ::::o 6-oxo-PGF 1•. After injection of All into the fetal circulation the rank orders of release were TxB 2 ::::o PGE > PGF 2• ::::o 6-oxo-PGF1• for the maternal side and PGE > TxB 2 ::::o PGF2• ::::o 6-oxoPGF1. for the fetal side. Injection of All into either the maternal or the fetal circulation did not alter TxB 2 or PGF2• release on either side of the placenta (Fig. 4) when compared with the control preinjection levels (the control samples taken before injection of All into the fetal and maternal sides showed no statistical differences and so they were combined). However, injection of All into the fetal circulation significantly increased the amount of PGE (p <
0.005) and 6-oxo-PGF1• (p < 0.005) (Fig. 5) released into the fetal perfusate but did not significantly increase the release into the maternal perfusate. PGE and 6oxo-PGF1. levels were not significantly affected by injection of All into the maternal circulation of the placenta. The amounts of PGE and 6-oxo-PGF1• released into the fetal circulation by All stimulation were found to be significantly greater (PGE, p < 0.01; 6-oxo-PGF 1., p < 0.01) than the average amounts released during the time-course experiments and so were not the result of variability of the preparation. The increase in fetal perfusion pressure which resulted from the injection of All into the fetal side was 16 ± 3 mm Hg, and when All was injected into the maternal side the resulting increase in tetal perfusion pressure was 3 ± 1 mm Hg. These pressure increments are in agreement with our previously reported data. 9
Comment The use of the isolated perfused human placental cotyledon preparation for the study of prostaglandin release offers many distinct advantages over conventional techniques such as tissue culture or superfusion of tissue and tissue slices, the main one being that production can be assessed from both sides of the placenta in a dynamic and more physiologic approximation of in vivo conditions. With this system it was found that there was a greater release of all prostaglandins measured into the maternal circulation than into the fetal circulation (Fig. 2) although there were no consistent concentration gradients between the two sides (Fig. 3). One exception to this was that the concentration of TxB 2 in the maternal effluent was generally much higher than its concentration in the fetal effluent but this changed with time. The release of TxB 2 on the maternal side tended to parallel the progressive washout of blood elements from the intervillous space, including platelets, which are known to produce large amounts of TxA2 • Thus the high levels of TxB 2 found in the maternal effluent probably reflect platelet rather than placental production. Prostaglandins are highly lipid-soluble and so they would be expected to cross the placenta with relative ease. The amounts of prostaglandin released into either side of the placenta would thus reflect local tissue production and release directly into the perfusate and also the relative amounts being transferred from one circulation into the other. This transfer would in turn depend on the concentration gradient and the flow rates on either side of the placenta. As no consistent concentration gradients were observed for the prostaglandins across the placenta, the fact that significantly more prostaglandin was released into the maternal circulation than into the fetal circulation was probably due to the greater flow rate on the maternal side and suggests that in this preparation the release of prostaglan-
Volume 151 Number 3
din into either side may be flow-dependent. If this is the case in vivo, one consequence of this would be that if the flow was reduced on the maternal side more prostaglandins would be released into the fetal circulation where they could act to reduce blood flow and thus preserve the maternal-fetal perfusion ratio. In this context, it is interesting to note that in chronically catheterized sheep, embolization of the uterine vascular bed with microspheres produced parallel decreases in umbilical blood flow. 12 In this study, prostaglandin release showed a rank order (except for TxB 2 into the maternal circulation) of TxB 2 :::= PGE :::= PGF2a :::= 6-oxo-PGF 1a. This compares with the results reported by MitchelP' for superfused placentas where the approximate rank order was TxB 2 :::= PGE > PGF 2 :::= 6-oxo-PGF 1a and by DembeleDuchesne et a!. 14 for placental homogenates where the order was TxB 2 - PGE 2 :::= 6-oxo-PGF 1a. This contrasted with the production of prostaglandins after incubation of the placental homogenates with endoperoxide (PGH 2 ) where PGE 2 was the predominant prostaglandin produced. 14 Caution should be expressed when one is interpreting the relative importance of placental production of one prostaglandin over another as it can be seen that the relative amounts of prostaglandin produced change with the preparation used. Also the expression of the relative rank order of production does not take into account the differences in the biologic activity of the various prostaglandins. In our preparation, the pattern of prostaglandin production changed after injection of All into the fetal circulation, which stimulated PGE and 6-oxo-PGF 1a but not TxB 2 and PGF 2a release into the effluent. All has been shown to stimulate PGE release from the pregnant uterus, 8 kidney, 15 and lung 16 and thereby to modulate the pressor action of All. PGI 2 , a potent vasodilator, may act in a similar manner since All has been shown to stimulate PGI 2 release from fetal rabbit lungs. 17 Thus the release of PGI 2 (measured as 6-oxo-PGF 1a) and PGE from the fetal vasculature in response to All may serve as an important regulatory process maintaining the fetal-placental blood flow in the face of increased fetal peripheral resistance due to circulating All. All failed to stimulate prostaglandin release significantly into the maternal effluent, an<;! this suggests that the release of prostaglandins into the two circulations may have certain independent control mechanisms. The absence of maternal release of prostaglandins in response to stimulation by All also suggests that the observed All-induced increase of PGE 2 release from the pregnant uterus of other animals may not be of placental origin. All increases fetal perfusion pressure when injected into the maternal circulation. This response is probably
Prostaglandin production by placental cotyledon
391
not mediated by any of the prostaglandins measured in this study since All had no effect on the amounts released into the fetal circulation when it was injected on the maternal side. Work is now in progress in this laboratory to examine the effects of endogenously produced and exogenously added prostaglandins with or without AI or All on the fetal vasculature of this system. This work will expand on the interactions of prostaglandins and All reported here. REFERENCES 1. RankinJHG, McLaughlin MK. The regulation of the placental blood flow. J Dev Physiol 1979; I :3. 2. Goodman RP, Killam AP, Brash AR, Branch RA. Prostacyclin production during pregnancy: comparison of production during normal pregnancy and pregnancy complicated by hypertension. AM J 0BSTET GYNECOL 1982;142:817. 3. Tulenko TN. The actions of prostaglandins and cyclooxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 1981;21:1033. 4. Misiani R, Remuzzi G, Mecca G. Prostacyclin generation by human umbilical and placental vessels in normal and hypertensive pregnancy. In: Lewis PJ, O'Grady J, eds. Clinical pharmacology of prostacyclin. New York: Raven Press, 1981. 5. Rudolph AM. The effects of nonsteroidal anti-inflammatory compounds on fetal circulation and pulmonary function. Obstet Gynecol 1981;58:635. 6. Ferris JF, Weir EK. Effect of captopril on uterine blood flow and prostaglandin E synthesis in the pregnant rabbit. J Clin Invest 1983;71:809. 7. Symonds EM, Broughton-Pipkin F. Pregnancy hypertension, parity, and the renin-angiotensin system. AM J OBSTET GYNECOL 1978;132:473. 8. Terragno NA, Terragno DA, Pacholczyk D, McGiff JC. Prostaglandins and the regulation of uterine blood flow in pregnancy. Nature 1974;249:57. 9. Glance DG, Elder MG, Bloxam DL, Myatt L. The effects of components of the renin-angiotensin system on the isolated perfused human placental cotyledon. AM J OBSTET GYNECOL 1984;149:450. 10. Myatt L, Bray MA, Gordon D, Morley J. Macrophages on intrauterine contraceptive devices produce prostaglandins. Nature 1975;257:227. 11. Hensby CN, Jogee M, Elder MG, Myatt L. A comparison of the quantitative analysis of 6-oxo-PGF 1a in biological fluids by gas chromatography, mass spectrometry and radioimmunoassay. Biomed Mass Spectrom 1981;8:111. 12. Kunze! W. Umbilical circulation-physiology and pathology. J Perinat Med 1981;9(suppl 1):68. 13. Mitchell MD. Prostaglandins during pregnancy and the perinatal period. J Reprod Fertil 1981;62:305. 14. Dembele-Duchesne MJ, Thaler-Dao H, Chavris C, Crastes de Paulet A. Some new prospects in the mechanism of control of arachidonate metabolism in human placenta and amnion. Prostaglandins 1981 ;22 :979. 15. Aiken JW, Vane JR. Intrarenal prostaglandin release attenuates the renal vasoconstrictor activity of angiotensin. J Pharmacol Exp Ther 1973;184:678. 16. Piper PJ, Vane JR. The release of prostaglandin from lung and other tissues. Ann NY Acad Sci 1971;180:363. 17. Omini C, Vigani T, Marini A, Pasargiklian R, Fano M, Maselli MA. Angiotensin II: a releaser of PGI 2 from fetal and newborn rabbit lungs. Prostaglandins 1983;25:901.
Prostaglandin production and stimulation by angiotensin II in the isolated perfused human placental cotyledon D. G. Glance, B.Sc., M.G. Elder, M.D!, and L. Myatt, Ph.D. London, England Levels of prostaglandins E and F2a, thromboxane 8 2 , and 6-oxo-prostaglandin F,. were measured by radioimmunoassay in the maternal and fetal effluents of isolated human placental cotyledons perfused in vitro~ All prostaglandins measured were released in greater amounts by the maternal side than by the fetal side of the perfused cotyledon although there were no consistent concentration gradients between the two sides. The approximate rank order of prostaglandin release into the maternal ~ide was thromboxane 8 2 >prostaglandin F2."" prostaglandin E ""6-oxo-prostaglandin F,., and that into the fetal side was thromboxane 8 2 "" prostaglandin F2a"" prostaglandin E"" 6-oxo-prostaglandin F, •. Injection of angiotensin II (0.5 ,...g) into the fetal circulation stimulated prostaglandin E and 6-oxo-prostaglandin F,. but not thromboxane 8 2 and prostaglandin F2• release into the fetal circulation and had no effect on maternal release. Angiotensin II (0.5 ,...g) had no effect on either side of the perfused cotyledon when injected into the maternal circulation. It is proposed that prostaglandin release into both maternal and fetal circulations may be flow-dependent and that the angiotensin 11-stimulated release of prostaglandin E and 6-oxoprostaglandin F,. may serve to modulate the vasoactive actions of angiotensin II on the fetal vasculature. (AM J 0BSTET GYNECOL 19!35;151:387-91.)
Key words: Placenta, prostaglandin production, angiotensin II
There has been much speculation concerning the role of prostaglandins released by the placenta. It has been suggested that prostaglandin E2 (PGE 2) may regulate inequalities of maternofetal perfusion by acting as a vasoconstrictor of the fetal-placental vessels and as a vasodilator of the uterine vasculature. 1 Prostacyclin (PGI 2) of placental origin may contribute to the elevated plasma PGI 2 levels and the refractoriness to angiotension II (All) found in pregnancy! It may also safeguard maternal blood supply to the placenta by preventing platelet aggregation although this process may also be influenced by placental thromboxane A2 (TxA2) synthesis since TxA 2 is a potent platelet ;1ggregator. Prostaglandip. production by the fetal-placental vasculature could be involved in controlling blood flow through these vessels since it has been shown that prostaglandin F2• (PGF2.) and PGE 2contract the fetal vessels From the Institute of Obstetrics and Gynaecology, Hammersmith Hospital. Action Research for the Crippled Child provided financial support of this work. Received for publication january 30, 1984; accepted September 20, 1984. Reprint requests: Dr. L. Myatt, Institute of Obstetrics and Gynaecology, Hammersmith Hospital, Du Cane Road, London, England W12 OHS. .
of the placenta whereas PGI 2 relaxes them.' Release of prostaglandins into the fetal circulation may also give rise to systemic effects within the fetus contributing to the low vascular resistance of the fetal circulation4 and the patency of the ductus arteriosus. 5 · Inpregnant rabbits, All plays an important part in the maintenance of systemic blood pressure, 6 and the same may be the case in the human since plasma All levels correlate with diastolic blood pressure in pregnant women. 7 All may play an important rote in the maintenance of uterine blood flow which is seen to decrease when the conversion of angiotensin I (AI) to All is blocked by captopril, an inhibitor of angiotensinconverting enzyme. Control of uterine blood flow may involve PGE 2 since All increases the production of PGE2 in the pregnant uterus 6 • 8 although the exact site of production within the uteroplacental complex is unclear. We have previously shown that components of the renin-angiotensin system act as very potent vasoconstrictors of the fetal-placental vasculature of the isolated perfused human placental cotyledon and it was proposed that, in vivo, the renin-angiotensin system may participate in the control of fetal-placental blood flow.~ Prostaglandins may modulate the actions of this system in the fetal-placental vasculature. Meclofena-
387
388 Glance, l;lder, and Myatt
mate, an inhibitor of prostaglandin synthesis, decreases the contractile response of fetal-placental vessels to All, suggesting that prostaglandins normally potentiate the contractile response. 3 To examine further the possible role of prostaglandins in the control of placental blood flows, we have studied the release of prostaglandin E (PGE), PGF2a, 6-oxo-prostaglandin F 1a (6-oxo-PGF 1a, the stable metabolite of PGI 2 ), and thromboxane B2 (TxB 2 , the stable metabolite of TxA2) from both sides of the isolated perfused human placental cotyledon preparation and looked at the modification of this production by All administered into either the fetal or the maternal circulation. Method Placentas from women with uncomplicated pregnancies were collected within 5 minutes of vaginal delivery or cesarean section at term. Placental cotyledons, free of infarcts or tears, were perfused by a method which has been described in detail elsewhere.• Briefly, this involved perfusing tissue culture medium 199 (Gibco) containing polyvinylpyrrolidone (5% w/v) (Sigma Chemical Co.) through a cannulated chorionic artery on the fetal surface of the placenta and collecting the effluent from the associated chorionic vein. Perfusion of the intervillous space was achieved by the insertion of two intravenous 21-gauge butterfly needles through the maternal surface of the perfused cotyledon. The whole preparation was maintained at 37° C in a controlled temperature cabinet. The flow rates of the fetal and maternal perfusates were maintained at 4 and 10 ml· min- 1 , respectively. The resulting perfusion pressures from both sides were between 30 and 60 mm Hg. In a preliminary series of experiments (three placentas) fractions of fetal and maternal effluent (1.5 ml) were collected at 20, 30, 40, 50, and 60 minutes after the establishment of the maternal circulation and immediateiy placed on ice at 4° C. Medium flowing thro~gh the perfusion tubing withoutthe cotyledon in place was collected to serve as a blank. The samples were centrifuged (1700 X g) immediately after each experiment to remove any blood contamination and the supernatants (1 ml) stored at- 20° C for subsequent extraction and analysis of prostaglandins. The samples were extracted with the use of Sep-Pak Cts cartridges (Waters) prior to radioimmunoassay. This removed interference from polyvinylpyrrolidone which was evident when direct radioimmunoassay of perfusates was used. The cartridges were first wetted with 20 ml of methanol/water (4: 1, v/v) and residual methanol removed by washing with deionized/distilled water (20 ml). The perfusate sample (1 ml) was acidified to pH 3.0 with formic acid, loaded on the cartridge, and
February 1, 1985 Am J Obstet Gynecol
eluted with diethyl ether (10 ml), and the fractions were collected in two siliconized glass test tubes. The solvent was evaporated under nitrogen and the residue resuspended in phosphate-buffered saline (Dulbecco) (1 ml). The extracted samples were assayed for prostaglandins with radioimmunoassays developed in this laboratory for PGE, 10 PGF2a/ 0 6-oxo~PGF ta> u and TxB 2 • The TxB 2 antisera cross reacted < 1% with 6-oxo-PGF 1a, prostaglandin E~> PGE 2 , PGF2", prostaglandin A2 , and arachidonic acid and <5% with prostaglandin D2 • The preliminary experiments revealed a large amount of variability in the results and that the levels of prostaglandins were sometimes very close to the level of detection of the assays. The experiments were repeated (three placentas) with collection of larger fractions of the fetal and maternal effluents (5 ml on the fetal side and 15 ml on the maternal side). Samples were treated and extracted in the same way as before, except that 4 ml of the fetal perfusate and 10 ml of the maternal perfusate were extracted and resuspended in 2 ml of phosphate-buffered saline, which would theoretically concentrate the prostaglandins two and five times, respectively. The samples were assayed as before for PGE, PGF 2u, 6-oxo-PGF 1"' and TxB 2 • To assess the effect of All on the production of prostaglandins from the perfused cotyledon, control samples (5-minute duration of flow) were collected from both sides of three placentas beginning 25 minutes after the establishment of maternal perfusion. All (Sigma Chemical Co., 0.5 IJ.g in 0.5 ml of gassed perfusate) was injected into the maternal circulation and fetal effluent collected for a further 5 minutes. Further control samples of maternal and fetal effluent were taken (5 minutes' duration) 10 minutes after baseline perfusion pressures had been reestablished. A second dose of All (0.5 IJ.g in 0.5 ml) was then injected into the fetal circulation and fetal and maternal effluents collected for 5 minutes. All medium collected was treated and extracted as described above. The residue was resuspended in phosphate-buftered saline (2 ml) and assayed for PGF2u, PGE, 6-oxo-PGF 1u, and TxB 2 • All results of prostaglandin assays were expressed as concentration of the prostaglandin per milliliter of medium after correction for any blank reading obtained. Using labeled prostaglandins, we found that recoveries for all prostaglandins were approximately equal (range, 82% to 94%) and thus no corrections were made to the prostaglandin levels obtained after the extraction procedure. As the PGE antisera could not distinguish PGE 1 and PGE 2 , results from this assay are taken to indicate total E series prostaglandin, expressed in this paper as PGE. Statistical significance was calculated by Student's t test.
Prostaglandin production by placental cotyledon
Volume 151 Number 3
10
- - Fetal Effluent ...._Maternal Effluent 6-oxo-PGF
2
0
0
9
1~
**
[:3 Maternal Effluent
c:
i
Ol
c: 6
"' "'~
5 4
c::
3
"
2
-;:; a.
2
Fetal Effluent
389
* ... 121
1 0
6-oxo-
**
*
PGF 2a
PGE
6~ 0~
Cl TxB 2
PGFla
c 3 E 01
c
2
Q)
Vl
~ 0
PGE
j~
Fig. 2. Mean prostaglandin release during 60 minutes of perfusion into fetal (open columns) and maternal (hatched columns) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of six experiments. Significant difference from fetal circulation: * = p < 0.01 and ** = p < 0.005).
QJ
c::
a: 11
0
10
Fetal Effluent
[:3 Maternal Effluent 6
9
8
e5
7
Ol
c: c: 4
6
0
-.;::
5
;;"' c:
4
"'c:u
3
u
2
"a.
0
0
20
30
40
50
60
Perfusion time (min)
6-oxoPGF1a
PGE
Fig. I. Prostaglandin release into fetal (..----.) and maternal (____.) circulations of the isolated perfused human placental cotyledon during 60 minutes of perfusion (results are mean ± SEM of six experiments).
Fig. 3. Mean prostaglandin concentration during 60 minutes of perfusion in fetal (open columns) and maternal (hatched columns) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of six experiments).
Results None of the prostaglandins measured, apart from TxB 2, showed any consistent trend in the amount released from either side of the placenta during the 60 minutes of perfusion (Fig. 1). TxB 2 release showed a steady decline on the maternal side which tended to parallel the gradual washout of blood from the intervillous space. When the means of all values obtained at each time point throughout the course of perfusion were calculated, no statistical differences were found
between the mean amounts of prostaglandin released in the preliminary experiments and the repeat experiments, and so the two sets of results were combined (Fig. 2). The amount of each prostaglandin released into the maternal circulation was significantly greater than the amount released into the fetal circulation (TxB2, p < 0.0005; PGF 2., p < 0.005; PGE, p < 0.01; and 6-oxo-PGF 1. , p < 0.01). When the results were expressed in terms of concentrations of prostaglandins per milliliter of perfusate (Fig. 3) no difference was
390 Glance, Elder, and Myatt
February 1, 1985 Am J Obstet Gyneco1
[J Fetal Effluent E=:1 Matern~! Effluent 6
PGF 2a
TxB 2
1:
!
5 0>4
1:
~
cu 3
"'~
'ii
2
Cl:: 1
~
0
ct!JO FC MA FA
MC MA FA
ts6l5 FC MA FA
~Iii~ MC MA FA
Fig. 4. TxB 2 and PGF2• release into fetal (open columns) and maternal (hatched columns) circulations before (FC andMC) and after injection of All (0.5 ~J.g) into the fetal (FA) and maternal (MA) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of three experiments).
CJ
Fetal Effluent
E=:1
Maternal Effluent
...
? 1:
·e Cl
1:
-
.. cu
5
4
cu
D.
0
1a
~j -- 0 ~c~ .....
~ 2 1.-'1
6-oxo-PGF
PGE
6
150 FC MA FA
MC MA FA
FC MA FA
MC MA FA
Fig. 5. PGE and 6-oxo-PGF 1• release into the fetal (open columns) and maternal (hatched columns) circulations before (FC and MC) and after injection of All (0.5 ~J.g) into the fetal (FA) and maternal (MA) circulations of the isolated perfused human placental cotyledon (results are mean ± SEM of three experiments. Significant difference from fetal control: ** = p < 0.005).
seen between the two sides for any of the prostaglandins. The amounts of prostaglandins released into the maternal circulation had an approximate rank order of TxB 2 > PGF2• ::::o PGE ::::o 6-oxo-PGFI•• and those released into the fetal circulation had a rank order of TxB 2 ::::o PGE ::::o PGF2• ::::o 6-oxo-PGF 1•. After injection of All into the fetal circulation the rank orders of release were TxB 2 ::::o PGE > PGF 2• ::::o 6-oxo-PGF1• for the maternal side and PGE > TxB 2 ::::o PGF2• ::::o 6-oxoPGF1. for the fetal side. Injection of All into either the maternal or the fetal circulation did not alter TxB 2 or PGF2• release on either side of the placenta (Fig. 4) when compared with the control preinjection levels (the control samples taken before injection of All into the fetal and maternal sides showed no statistical differences and so they were combined). However, injection of All into the fetal circulation significantly increased the amount of PGE (p <
0.005) and 6-oxo-PGF1• (p < 0.005) (Fig. 5) released into the fetal perfusate but did not significantly increase the release into the maternal perfusate. PGE and 6oxo-PGF1. levels were not significantly affected by injection of All into the maternal circulation of the placenta. The amounts of PGE and 6-oxo-PGF1• released into the fetal circulation by All stimulation were found to be significantly greater (PGE, p < 0.01; 6-oxo-PGF 1., p < 0.01) than the average amounts released during the time-course experiments and so were not the result of variability of the preparation. The increase in fetal perfusion pressure which resulted from the injection of All into the fetal side was 16 ± 3 mm Hg, and when All was injected into the maternal side the resulting increase in tetal perfusion pressure was 3 ± 1 mm Hg. These pressure increments are in agreement with our previously reported data. 9
Comment The use of the isolated perfused human placental cotyledon preparation for the study of prostaglandin release offers many distinct advantages over conventional techniques such as tissue culture or superfusion of tissue and tissue slices, the main one being that production can be assessed from both sides of the placenta in a dynamic and more physiologic approximation of in vivo conditions. With this system it was found that there was a greater release of all prostaglandins measured into the maternal circulation than into the fetal circulation (Fig. 2) although there were no consistent concentration gradients between the two sides (Fig. 3). One exception to this was that the concentration of TxB 2 in the maternal effluent was generally much higher than its concentration in the fetal effluent but this changed with time. The release of TxB 2 on the maternal side tended to parallel the progressive washout of blood elements from the intervillous space, including platelets, which are known to produce large amounts of TxA2 • Thus the high levels of TxB 2 found in the maternal effluent probably reflect platelet rather than placental production. Prostaglandins are highly lipid-soluble and so they would be expected to cross the placenta with relative ease. The amounts of prostaglandin released into either side of the placenta would thus reflect local tissue production and release directly into the perfusate and also the relative amounts being transferred from one circulation into the other. This transfer would in turn depend on the concentration gradient and the flow rates on either side of the placenta. As no consistent concentration gradients were observed for the prostaglandins across the placenta, the fact that significantly more prostaglandin was released into the maternal circulation than into the fetal circulation was probably due to the greater flow rate on the maternal side and suggests that in this preparation the release of prostaglan-
Volume 151 Number 3
din into either side may be flow-dependent. If this is the case in vivo, one consequence of this would be that if the flow was reduced on the maternal side more prostaglandins would be released into the fetal circulation where they could act to reduce blood flow and thus preserve the maternal-fetal perfusion ratio. In this context, it is interesting to note that in chronically catheterized sheep, embolization of the uterine vascular bed with microspheres produced parallel decreases in umbilical blood flow. 12 In this study, prostaglandin release showed a rank order (except for TxB 2 into the maternal circulation) of TxB 2 :::= PGE :::= PGF2a :::= 6-oxo-PGF 1a. This compares with the results reported by MitchelP' for superfused placentas where the approximate rank order was TxB 2 :::= PGE > PGF 2 :::= 6-oxo-PGF 1a and by DembeleDuchesne et a!. 14 for placental homogenates where the order was TxB 2 - PGE 2 :::= 6-oxo-PGF 1a. This contrasted with the production of prostaglandins after incubation of the placental homogenates with endoperoxide (PGH 2 ) where PGE 2 was the predominant prostaglandin produced. 14 Caution should be expressed when one is interpreting the relative importance of placental production of one prostaglandin over another as it can be seen that the relative amounts of prostaglandin produced change with the preparation used. Also the expression of the relative rank order of production does not take into account the differences in the biologic activity of the various prostaglandins. In our preparation, the pattern of prostaglandin production changed after injection of All into the fetal circulation, which stimulated PGE and 6-oxo-PGF 1a but not TxB 2 and PGF 2a release into the effluent. All has been shown to stimulate PGE release from the pregnant uterus, 8 kidney, 15 and lung 16 and thereby to modulate the pressor action of All. PGI 2 , a potent vasodilator, may act in a similar manner since All has been shown to stimulate PGI 2 release from fetal rabbit lungs. 17 Thus the release of PGI 2 (measured as 6-oxo-PGF 1a) and PGE from the fetal vasculature in response to All may serve as an important regulatory process maintaining the fetal-placental blood flow in the face of increased fetal peripheral resistance due to circulating All. All failed to stimulate prostaglandin release significantly into the maternal effluent, an<;! this suggests that the release of prostaglandins into the two circulations may have certain independent control mechanisms. The absence of maternal release of prostaglandins in response to stimulation by All also suggests that the observed All-induced increase of PGE 2 release from the pregnant uterus of other animals may not be of placental origin. All increases fetal perfusion pressure when injected into the maternal circulation. This response is probably
Prostaglandin production by placental cotyledon
391
not mediated by any of the prostaglandins measured in this study since All had no effect on the amounts released into the fetal circulation when it was injected on the maternal side. Work is now in progress in this laboratory to examine the effects of endogenously produced and exogenously added prostaglandins with or without AI or All on the fetal vasculature of this system. This work will expand on the interactions of prostaglandins and All reported here. REFERENCES 1. RankinJHG, McLaughlin MK. The regulation of the placental blood flow. J Dev Physiol 1979; I :3. 2. Goodman RP, Killam AP, Brash AR, Branch RA. Prostacyclin production during pregnancy: comparison of production during normal pregnancy and pregnancy complicated by hypertension. AM J 0BSTET GYNECOL 1982;142:817. 3. Tulenko TN. The actions of prostaglandins and cyclooxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 1981;21:1033. 4. Misiani R, Remuzzi G, Mecca G. Prostacyclin generation by human umbilical and placental vessels in normal and hypertensive pregnancy. In: Lewis PJ, O'Grady J, eds. Clinical pharmacology of prostacyclin. New York: Raven Press, 1981. 5. Rudolph AM. The effects of nonsteroidal anti-inflammatory compounds on fetal circulation and pulmonary function. Obstet Gynecol 1981;58:635. 6. Ferris JF, Weir EK. Effect of captopril on uterine blood flow and prostaglandin E synthesis in the pregnant rabbit. J Clin Invest 1983;71:809. 7. Symonds EM, Broughton-Pipkin F. Pregnancy hypertension, parity, and the renin-angiotensin system. AM J OBSTET GYNECOL 1978;132:473. 8. Terragno NA, Terragno DA, Pacholczyk D, McGiff JC. Prostaglandins and the regulation of uterine blood flow in pregnancy. Nature 1974;249:57. 9. Glance DG, Elder MG, Bloxam DL, Myatt L. The effects of components of the renin-angiotensin system on the isolated perfused human placental cotyledon. AM J OBSTET GYNECOL 1984;149:450. 10. Myatt L, Bray MA, Gordon D, Morley J. Macrophages on intrauterine contraceptive devices produce prostaglandins. Nature 1975;257:227. 11. Hensby CN, Jogee M, Elder MG, Myatt L. A comparison of the quantitative analysis of 6-oxo-PGF 1a in biological fluids by gas chromatography, mass spectrometry and radioimmunoassay. Biomed Mass Spectrom 1981;8:111. 12. Kunze! W. Umbilical circulation-physiology and pathology. J Perinat Med 1981;9(suppl 1):68. 13. Mitchell MD. Prostaglandins during pregnancy and the perinatal period. J Reprod Fertil 1981;62:305. 14. Dembele-Duchesne MJ, Thaler-Dao H, Chavris C, Crastes de Paulet A. Some new prospects in the mechanism of control of arachidonate metabolism in human placenta and amnion. Prostaglandins 1981 ;22 :979. 15. Aiken JW, Vane JR. Intrarenal prostaglandin release attenuates the renal vasoconstrictor activity of angiotensin. J Pharmacol Exp Ther 1973;184:678. 16. Piper PJ, Vane JR. The release of prostaglandin from lung and other tissues. Ann NY Acad Sci 1971;180:363. 17. Omini C, Vigani T, Marini A, Pasargiklian R, Fano M, Maselli MA. Angiotensin II: a releaser of PGI 2 from fetal and newborn rabbit lungs. Prostaglandins 1983;25:901.