- Email: [email protected]
The role of prostaglandins and allied substances in uterine haemostasis
CONTRACEPTION
THE ROLE OF PROSTAGLANDINS AND ALLIED SUBSTANCES IN UTERINE HAKMOSTASIS
Kerstin Hagenfeldt, M.D. Department of Obstetrics and Gynaecology Karolinska Institute Stockholm, Sweden
AHSTRACT This review addresses itself to summarizing the more recent studies published on the bioconversion of arachidonic acid in the human endoand myometrium during the normal menstrual cycle and in women with increased menstrual blood loss. The data indicate an increased ability of the endo- and myometrium from women with menorrllagiato produce prostaglandins with vasodilator and platelet anti-aggregatory properties, viz., prostacycline and prostaglandin E2. Tne data on prostaglandin production in endometria of IUD wearers is reviewed and discussed in relation to present knowledge on morphological findings in IUD-influenced endometrium. INTRODUCTION The interest in studying prostaglandins in the human uterus was primarily evoked by their involvement in myornetrialphysiology; in abortion and parturition and in the clinical condition of dysmenorrhoea. The role of prostaglandius in the control of menstrual blood loss has received considerable attention after Wiqvist -et al. (1) reported that the administration of PGF2Aduring the liltealphase induced menstruation. Furthermore, treatment with prostaglandin synthetase inhibitors during menstruation resulted in reduction of menstrual blood loss in women with menorrhagia during IUD use (2,3). Between 1973 and 197Y several reports were published in rhich the amount of prostaplandins of the E and F type were determined in endometrial specimens obtained at various phases of the normal menstrual cycle, during certain pathogenic conditions and uuring IUD use. The results of these studies nave been discussed in a review paper by Hagenfeldt at the WHO meeting in 1979: Endometrial bleeding and steroidal contraception (4). The results of the studies published until lY79 seem to indicate a higher production of prostaglandin F2&(PGFa) from mid and late secretory endometrium and during menstruation compared with the proliferative phase. Prostaglandin E2 (PGE2) levels did not show cyclic variation in the normal endometrium, but both PGF2&and PGE2 were increased in endometria from women with dysfunctional uterine bleeding.
JULY
1987 VOL. 36 NO. 1
23
CONTRACEPTION
Three reports address themselves to determining PGE2 and/or PGFZhin endometrial specimens obtained from women using intrauterine devices. Green b Hagenfeldt (5) did not find any change in PGP2 or its main metabolite 13,14-dihydro-15-keto PGFzAafter three months use of a large inert device (Dalkon Shield) or a Copper-T ZOO. Hillier h Kasonde (6) made the same observation on PGFz,Xlevels but reported an increase in PGE levels in 3 out of 5 women using a Copper-7 device and in 7 out of 3 women using various inert devices. A third study by Scommegna -et al. (7) reported on PGF levels in endometrial specimens obtained in women using either inert devices, a progesterone-releasing IUD (Progestasert) or a dydrogesterone-releasing IUD. Compared with levels in tne normal endometrium, no change was found in the first tu7o groups; PGF levels in specimens obtained from dydrogesterone influenced endometrium were slightly lower than before iusertion. Prostaglandins are synthesized and metabolized locally and are not stored in tissues. Therefore tne technique of ootaining endometrial specimens by biopsy or after hysterectomy would invariably evoke a trauma which in itself triggers the release of arachidonic acid from cell membranes. Hence the studies discussed above do not report on in vivo levels of the prostaglandins but rather measure the efficacy of?he cyclooxygenase pathway. Over the past few years, more sophisticated methods have been used to study arachidonic acid metabolism in endometrium and myometrium. Furthermore, in additiou to the so-called classical Prostaglandins, PGF2:7(andPGE2, other arachidonic acid metabolites such as prostacyclln (PGI2) and thromboxane A2 (TXA2) have been studied. Research in platelet and vessel wall arachidonic acid metabolism indicates that the metabolites could be important in haemostasis (8). In addition to the known vasoconstrictor effect of PGF2dand tHe vasodilation effect of PGE2, the effect of the other two metabolites, being a potent vasodilator such as PGI2 and a potent platelet aggregator such as TXA2 would probably be of fundamental importance in the regulation of menstrual blood loss. Recently the importance of another pathway of the arachidonic acid metabolism has been stressed. Through tne enzyme lipoxygenase (see Fig. 1) arachidonic acid can be converted to another class of substances now called the leukotrienes (9). These substances have potent biological activities including cytotoxicity, chemotaxis, vasoconstriction in some vascular systems and increasing vascuiar permeability. These properties could add positiveiy or negatively to the haemostatic mechanism in certain tissues, e.g. the uterus. The regulation of prostaglandin synthesis and metabolism in the normal human uterus In a series of papers Published between 197Y and 1986, scientists using in vitro systems such as organ culture of the human endometrium, monolayer cultures of separated endometrial cells or superfusion techniques of endometrial tissue, have drawn attention to several important aspects of the arachidonic acid metabolism in the human uterus.
24
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
5-HPETE
IS-HPETE
LTA,, yg
INHIBNK CELL CYTO10x & 0; -GEN &DEGRkNULAT.
INHIB. DEGRANULAT. IO; -GEN.
$q
ADHESION CHEMOTAX DEGRANULAT. 0; -GEN.
VASC. PERM BRONCHOCONSTR VASDCONST R Lti RELEASE
CONTR PULM STR
Figure 1. Formation a:lclbiological effects of uerivotives of (Samuelsson -et al. Bdv. Prostagland. Tnromboxan. arachidonic acid. Leu,
Thus, the influence of ovarian steroids on the prostaglandins syntheslj in the numan endonetrium has been studied by Baird's rese.lrch group in Edinburgh. Aioel (r Baird (10) snowed that the human endometrium maintained in or;an culture produced more PGFz&than PGE2. 17B-tistradiol increased PGFuproduction in the secretory but not in the Llroliferative endometrium. Progesterone significantly reduce,i PGF~~aand PGE2 in both phases of the cycle and prevented the estradiol-induced increase in PGF2, of the secretory endometriuia. The results suggest that there is an inherent Difference in the reductase to isomerase enzymes or in the ability of the enzymes to respond to hormones . Mternatively PGE2 level.; in the endometrium :Alaynot be under enzymatic control but may reflect a non-eizyaatic conversion of the endoperoxide PGl12 which fails to act as a substrate for the PGFck reductase. In a study by i1oel & Kelly was investigated in the normal are involved in the regulation concentration of the compounds
JULY
(11) tile metaboiism of human endometrium. If of menstrual bleeding, vittii,l the uterus will
1987 VOL. 36 NO. 1
PGEL and PCP2 PGFl+and PGE2 the effective tlepend upon t!le
25
CONTRACEPTION
. relative rates of synthesis and metabolism urithinor removal from the tissue. Both 15-hydroxy-prostaglandin dehydrogenase and the A13- redUC t.dse enzyme are present in both endometrium and myometrium. The dehydrogenasc is NAD+-dependent while the reductase requires NADPH+ as a co-factor. No evidence was found for the existence of a 9-keto reductase enzyme converting PGE2 to PGFz~ The major metabolites were invariably 13,14-dihydro-15-keto compounds. Tile netabolisiaof PGE2 exceeded that of PGF2,,_bothin endometrium and myometrium. A decrease in the metabolizing capacity was seen during the pre-menstrual phase. This could be contributing to the increased concentration of PGF2&and PGE2 seen dt that time. Further studies on the metabolism of 14C-arachidonic acid in separated glandular epithelium and stromal ceils from the human endometrium have been published by Lumsden et al. (12). Glandular epithelial cells produced more PGF2dt"an did stromal cells both in the proliferative and secretory endometrium. The production of PGE2 from Labelled arachidonic acid was variable with no difference between the two cell types or between the two phases of tne cycle. Conversion to b-oxo-PGFlswas negligible in both glandular epithelial and stromal cell. Thus the cyclooxygenase activity was more active in tileglands. This result is at variance with a report from Gal -et al. (13) who studitidthe same ceil types in noilolayercultures for up to 21 days. In their system more PGE2 than PGF2Awas produced and the synthesis was more active in stromai than in glandular epithelial cells. It was, however, pointed out by Lumsden --et al. that the viability of the stronal cells in the Imonolayersystem of Gal -et al. was very low after 21 days, while the viability in the cell preparation used by themselves was satisfactory. Using immunohistochemicdl methods for the localization of the cyclooxydenase enzyme, Rees -et al. (14) found the highest activity in the glanduldr epithelidl celiu of the luteal phase endometrium. No specific staining was found in stroma, myometrium or blood vessels. In an earlier study, Casey -et al. (15) found the highest activity of the prostaglandin degrading enzyme, 15-OH prostaglandin dehydrogenase, in the glanduldr epithelial cells of the iuteal phase endometrium. The failure to demonstrate specific staining of the cyclooxygenase enzyme in non-epithelial cells in the human uterus does not exclude the presence of the enzyme in other cells than glandular epithelium. It could be a scientists in Gurpicie's question of sensitivity of the method used. 'l'ne group at Mount Sinai in New York have used organ culture of the human endonetrium and tissue experiments with superfusion and wouolayer cultures of separated endometrial cells to study tne regulation of prostaglandin synthesis and metabolism. Schatz h Gurpide (16) confirmed the results of the Edinburgh group that estradiol increases the PGF2,p reduction of glandular epithelial cells of the secretory endometrium. Botilepithelial and stromdl cells of the proliferative endometrium release PGF2&but only the glandular cells increase the PGF2Aproduction with addition of estradiol.
26
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
Mariciewicz-et al. (17) compared the PGF2Aproduction in proliferative and secretory endometrium under superfusion and organ culture conditions. The basal PGF240utput was greater in proliferative than in secretory endometrium, but the response to estradiol was strolyest in the secretory endometrium. Progesterone lowered basal PGFZo+,outputin both proliferative and secretory endometrium in organ culture. During superfusion, progesterone only affected cells from the proliferative endometrium. The same experimental model was used by Schatz -et al. (18) to study the effect of anti-estrogens on prostaglandin production. Both with cultured endometrial fragments aud tiithmonolayer cultures of separated glandular epithelial and stromal cells, the anti-estrogens (tamoxifen and its metabolite trans-4-monohydroxy-tamoxifen) counteracted the estradiol increase in PGF2dp reduction while failing to affect the basal production of the prostagldndius. This suggests that the basal rate may not be dependent on endogenous estrogens. The inhibition of the estradiol-stimulated increase in PGF2d production suggests that the stimulation is estradiol receptor-mediated. Progesterone inhibits PGF2dproduction in botn proliferative and secretory endometrium maintained in organ culture. The exact mode of action of progesterone is not known. Glucocorticoids are known to exert their dnti-inflammatory effect by increasing tissue production of proteins that inhibit phospholipase activity, thus reducing the amount of arachidonic acid available for prostaglandin synthesis. Toe name lipocortin has been proposed to describe this antiphospholipase. Gurpide --et al. (19) found that lipocortin was produced botil by the proliferative and the secretory endometrium in organ culture. Levels were increased by dexamethazone and decreased by progesteroue. both steroids, however, decreased PGF2,production. The results indicate that lipocortin does not mediate the inhibition of PGF2Aevoked by progesterone. The effect of antiprogesterone on the prostaglandin production of human endometrial cells has been studied by Kelly -et al. (20). Glandular epithelial and strowal cells were cultured for 24 hours in the presence of the progesterone antagonists RU 486 and ZK 98734. Both steroids stimulated PGF2Aproduction by stromal cells in a dose-dependent manner. Added progesterone innibited the RU 486-stimulated increase in PGF24 PGE2, wirenmeasurable, also increased in the presence of KU 486. No effect of KU 486 was seen in tne slandular epithelial cells. Preincubation of stromal cells with arachidonic acid markedly increased the effect of KU 486. Pretreatment wicn progesterone allowed a L;reaterconversion to 13,14-dihydro-1%keto PGF2& The results indicate that antiprogesterone can act as menstrual regulators by stimulating eudogenous prostaglandin F2w production in stromal cells or by inhibiting prostaglandin metabolism. Prostaglandin binding to different cell types has been studied by Hoffman -et al. (21). PGE2 binding sites were significantly higher in proliferative than in secretory endometrium. Tinediagnosis of abnormal uterine bleeding was associated with higher %I-prostaglandin E2
JULY 1987 VOL. 36 NO. 1
27
CONTRACEPTION
binding than diadnosis of dysmenorrhoea or uterine prolapse. regardless of 3H-PGF2Abinding wds low in all endo8aetridlapeci!llens the pudse of the cycle. These results are supported in a recent study by Che;ini et al. (22), who used a quantitative light microscopic autoradiographic -method to analyse prostaglandin binding sites in different cell types of the human endometrium. Stromal cells, glandular epithelial cells, elongated and circular smooth muscles and arterioles contained numerous PGE2 and very few OK no detectable PGF2dbinding sites. Sinding sites in all cell types were higher in the proliferative than in the secretory endometrium while binding sites in myometrium did not vary with the cycle. The reasons for the very few or no detectable PGF2 binding sites are discussed in the paper by Hoffman -et al. (21). &skin& of binding sites by endogenously produced PGF2&, rapid metabolism of jH-PGF2&iin tireincubation system, laaskingof binding sites durint;the tlomogenizationand centrifugation $KOcedUKe, were all ruled out. Studies on prostaglandin productiou in the human uterus in relation to increased menstrual blood loss Tne discovery of tue potent anti-a&reijatory aud vasodilatory prostacyclin, PGI2 (23), triggered the interest to study new metauolites of arachidonic acid in the human uterus in an effort to explain pathological menstrual bleeding. Thus, Abel & Kelly (24) regorted that while the human endometrium mainly produced PGE2 and PGF2&from labelled arachidonic acid, the myometrium produced 6-ke to-PGFl, the stablr metabolite of prostacyclin (PGI2). ihen endometrium and myometrium .derecombined, a fUKtheK enhancement of the PGI2 production was seen with less PGE2 and PGF2hsynthesized. This indicates chat precursor PGG2 and PGB2 produced in the endonetriurncan be used by the myometrium fOK PG12 production. Endometrial PtiF2Aand PGE2 produced from labelled arachidonic acid was studied in women with menoKrhd&id (menstrual olood loss Jreater than 50 ml per cycle) and compared with that in women with normal menstruation (25). PGFa/PGF2 ratio was siynificalrtlylotrerin the eIldOuetKiUu of wonen LJith high menstrual blood loss, i.e. more than 90 ml per cycle. There was an inverse correlation between PGFk/PGE2 ratio and blood loss (K = 0.36, ~(0.025). The result suggests that excessive lnenstrualolood loss may be aasocidted with a shift in the endomctrial conversion of pKOSta&ldndin endoperoxides from PGF2d;to PGE2. In another study, Smith -et al. (.26)measured the synthesis of prosta&landins from labelled arachidonic acid in endometrium and uyometrium obtaineo from women with excessive menstrual blood loss (range 57-186, median 86 ml). The endometrium of women with high MBL was more effective than that of tiomenwitn normal NtiLat enhancing the production of ~-oxo-PGF~~, the staole netabolite of PGI2, in the myomecrium. These results further confirm the hypothesis of an abnormal prostaglandiu synthesis in women with excessive bleeding. The ,)atteru
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
of prostaglandin synthesis in endometrium obtained from women witu persistent anovulatory cycles and increased MUL was publisned by the same group (21). PGF2Alevels t+erelower in the anovulatory endometrium tnan in normal secretory endometrium, but the clmountof PGE2 and PFG2c,,synthetizedin this tissue from exogenous arachidonic acid were similar to those found in normal sacrrtory endometrium. T,lese findings suggest that the persistent proliferative endometrium has a in vivo because it iacks tile reduced capacity to synthetiLe PGF20,,_ essential precursor for PG synthesis proLably because of lack of progesterone influence. Tnis raduced capacity to synthetize PGF24may result in excessive MBL as suggested by the inverse correlation of PGFw/PGE2 ratio and menstrual blood loss (r =-0.7, p
Pees -et al. (30) also studied prostaglandin production in endometria and myometrid from women with menorr~lagiausing a superfusion technique. Their results dre at vdriance With those of the Edinburgh group. No difference Leas found in PGE2, PGP2* or G-oxo-PGFiA pro\iuctionbetween women with or without menorrhagia. The reason for the difference in results could be that of definition of increased ,lBL. Rees -et al. compared a group of women with MBL<80 ml with a group of women having NRL) dU ml per cycle. Smith -et al. (25) found tne difference in prostaglandin synthesis between a group of women witil (50 ml XBL per cycle and a group vitu>YU ml blood loss per cycle. Downing -et al. (31) studied organ culture of endometria from healthy women and from vomen with menorrhagia. The uptake of 3H-arachidonic acid in endo.detrialphospholipids was hi&her in normal secretory than normal proliferative endometrium. In endometrid from women witil menorrha~ia, 3’d- ardchidonic acid wds mainly distributed into neutral lipids, especially triJycerides, and particularly in the proliferative dhase. This difference indicates an asnormality in the ardcnidonic acid metabolism in the menorrhagic endometrium. If the increased amount of ardchidonic acid incorporated into triglyceride is released during
JULY 1987 VOL. 36 NO. 1
29
CONTRACEPTION
menstruation, it could lead to an overproduction of PGH2. This may result in saturation of the endoperoxide reducing factor by whict1 PGF2&is synthesized, thus allowing surplus PCH2 to be couverted to larger than normal quantities of PGE2 and PG12 both with vasodilatdtion properties resulting in an increased menstrual blood loss. Tne data presented mainly by the Edinburgh group indicates that an interaction between endometrial and myometrial prostaglandin production in vitro data do not necessarily prove that is possible. Althout;h the -this is happening also -in vivo the possible consequences are interesting. An increased availability of ardchidonic acid in the endometrium of women with menorrhagia can give an increased pool of precursor endoperoxides. The pathway to PGF2Abecomes saturated leading to a diversion of the synthesis towards nonenzymatically produced PGE2 and also to more PGI2 bein,:formed in the rnyoaetrium. The comb.ineddecrease in the PGF2dPGE2 ratio and the increase in PGL2 would increase vasodilatation of endometrial vessels and augment menstrual blood loss. An interesting aspect of arachidonic acid metabolism in the human endometrium was discussed by Demers -et al. (32). Studying short-time incubation of human endometrium and myometrium with L4C-arachidonic lipoxygenasc as well as cyclooxygenase activity was found. acid, both 'The lipoxyyenase products formed by the tissues included the monohydroxy acids, 5-hydroxy-eicosa-tetracrtoicacid (5-HE%), 12-hydroxy-eicosatretraenoic acid (12-tIETE)and 12-iiydroxy-heptadecatrienoicacid (HHT). The endometriai specimens released PGFk, PGE2 and 12-LIETE,while the products found in the myometrium were 12-HBTE and small amounts of 5-ZTE, HHT and 6-keto-PGFld. Tnis study indicates that t,lelipoxygendse activity may be of interest in menstrual disorders such as dysmenorrhoea and menorrhagia. In a review of prostaglandins and menstrual bleeding, Granstiam et al. (33) suggested that one of the pharmacological effects of non-steroidal anti-inflammatory drugs (NSAIDs), ou menstrual blood loss could be that the inhibition of the cyclooxygenase may increase the arachidonic aciJ pool dvailable for tilelipoxygenase pathway producing products with vasoconstrictor activity, i.e. leukotrienes. Further studies on isolated endometrial cell systems during inhibition with prostaglanain synthetase inhibitors could elucidate this possible mechanism. Prostaglandin synthesis and metabolism in endometria of IUD users No studies have been published using --in vitro techniques Such dS organ culture of endometrial tissue or monolayer culture of separated eodometrial cells obtained from women using inert or medicated devices. All IUDs stimulate an increase in leucocytes and other cells in the endometrium aud uterine fluid as part of the foreign body reactiou to the device. Several studies indicate an adherence of these cells to the IUD surface. iryatt-ec al. (34) reported that although the total amount of white blood cells were greater on the Dalkon Shield, the cells on
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
copper IUUs produced the highest concentration of PGF2,hwith the medium. The authors discussed their findings in relation to the effect on implantation. With the present knowledge on arachidonic acid metabolism in human polymorphonuclear leucocytes, viz., leukotriene production (19), an effect on uterine haemostasis is also possible. A few studies have addressed themselves to the effect of metal ions on the arachidonic acid metabolism in the endometrium. Kelly & Abel (35) investigated the action of copper and zinc on the prostaglandin metabolizing enzymes in the husan endometrium and myometrium. Copper and zinc chloride at a concentration of 10V5 mol/l inhibited the metabolism of PGE2 in the luteal phase endometrium and in the myometrium, copper more effectively than zinc. Also metallic copper in the form of a copper foil with an area of 25 mm2 was sufficient to reduce PGE2 metabolism by 50X. The concentration of copper used in this experiment are in the same range as that discussed by Uagenfeldt in the uterine fluid and cervical mucus of women using the CUT device (36). According to Salaverry -et al. (37) copper is concentrated in the glandular epithelial cells where also the prostaglandin dehydrogenase is located(l%It was suggested by r;ellyand Abel that the finding of an inhibition of the PGE2 metabolism induced by copper could contribute to the contraceptive action of the device. Increasing PGE2 versus PGF2dlevels could also have an effect on menstrual bleeding, i.e. increasing vasodilatation and menstrual blood loss. As very few studies on human endometrial tissue are available, one experimental study on the rat will be mentioned. Peplov -et al. (38) studied the effect of copper ions on the prostaglandin production in rat uterine homogenate. Copper in a concentration of 10s4 to 10q5 mol/l was found to stimulate PGF2++,synthesis. By contrast the synthesis of PGE2 and PGI2 were unaffected at all copper concentrations used. The increased PGF2dsynthesis in the presence of copper was not siynificantly changed by salicyclic acid but was inhibited by indomethacin. CONCiUSIONS The information dained from the studies presented in this review indicate that the human endometrium metabolizes arachidonic acid mainly to PGF2&and PGE2. The synthesis of these two compounds seem to be under control of ovarian steroids. 17%Estradiol stimulate PGF2d production in the secretory endometrium and anti-estrogens can inhibit this increase indicating that the process is estrogen receptordependent. Progesterone inhibits PGF2tiproduction in both proliferative and secretory endometrium and anti-progestins increase PGF2& production mainly in the stromal cells. The highest production of PGF2dand PGE2 is seen in the late luteal phase and during menstruation. Glandular epithelial cells invariably syntnetize more PGF2Athan stroual cells. The rnyometriun produces lnainlyPG12 using endoperoxide precursors from tne andometrium. PGF2qand PGE2 are metabolized in the endometrium and myometrium mainly to 13,14-dihydro-15-keto compounds. The metabolism of PGE2 exceeds that of PGF2& A decrease in the metabolizing capacity
JULY
1987 VOL. 36 NO. 1
CONTRACEPTION
is seen during menstruation, contrioutiny to hi&i1concentrations of the prostaglandins seen at that time. The cyclooxygenase enzyme as well as the prostaglandin degrading enzyinesare mainly identified in the glandular epithelial cells. PGE2 but very few PGFL&binding sites are found in all cell types of the human endornetriutl with higher concentrations in tne proliferative phase. The noriaalhurtian endometriuladrld ;nyotnetrium netaoolize arachidonic acid both by the cyclooxygenase and the lipoxydase pathway. The former is dominating but the possiaiiity of leukotriene formation could be important in pnysiological and pathological conditions. Li the uterine tissues of vlomensufr'erin&from increased iaenstrudl blood loss, there is a decrease in the PGP&PGE2 ratio and an increased capacity of the myometrium to produce PGI2. The relative decrease of the prostaglandin with vasoconstrictor properties and increase in compounds with vasodilatatiou and anti-aggrrgatory properries would be consistent with a less effective haemostasis at the tilile of lnenstruation. Altnough the situation in IUD-influenced ~ndometriurndnd I.iyonetrium has not been investigated, the findings in women with menorrhagia due to other reaso11scould probably be also relevilntin IUD users bearin& in mind the uorphological findings in the studies on uterine haemostasis in IUD users +blished by several investigators supgorted uy the :JHOTask Force on Intrauterine i4ethodsof Fertility Regulation (39,40). l'nusthe findings of fewer itaemostdticplugs in IUD-i,lfluenced tilenstrual endometrium could oe due to a shift in the balance between PGI2 and TXA2. Tile increased amount of non-occluded sudll vessels in the endometrium could oe a result of a decrease in PGF2d/PGE2 ratio favouring a less effective hdemostasis. Furthermore, an increased dmount 3f +lymor~honuclear leucocytes adhering to the device, especially the copper devices, indicates the possibility of leukotriene formation at++enting the increase in vascuiar permeability typical for the IUD-influenced endometrium.
1.
Viqvist, N., Bygdeman, d. and Kirton. K. Nonstrroidal infertility agents i;zthe female, in Nobel Syspos. 15. Control of tIumau Fertility (E. Diczfalusy, and 0. Borell, Editors). AlrnqvistWiksell, Sweden, lY71, p. 137-143.
2.
Andersson, A.B.M., Haynes, P.J., Guillebaud, I. and Turnbull A.C. Reduction of Inenstrualblood loss by prostaglandin sjnthetase inhibitors. Lancet i: 774-776 (1976).
3.
Guillebaud, J., Anderson, A.8. and Turnbull, A.C. Reduction by mefenaaic acid of increased menstrual blood loss associated with intrauterine contraception. Br J. Obstet Gynaucol 85: 53-62 (1978).
32
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
4.
Dadenfeldt, K. Prostaglandins and related compounds dnd taeir metabolism in steroid exposed endometrium, in Endometrial Bleeding and Steroidal Contraception (E. Dicefalusy, I.S. Frazer, and F.T.ti. Webb, Editors). Pitman Press Ltd, dath, lY80, p. 222-245.
5.
Grren, K. and endometrium.
’ o.
Hillier, K. and Kasonde, J.iI. Prostaklandin E and F concentrations in human endometrium after insertion of intrauterine COntraCeptiVe devices. Lancet i: 1%lo (1376).
7.
Scommegna, A., Tlekis, J., Rae, K., Dmowski, W.P., Rezai, P. and hluetta, F.J. Zndometrial prosta~landiu f content in women wearin& non-medicated or progestin-releasing intrauterine devices. Fertil Steril 29: 500-504 (197Y).
8.
t:oncada,5. and Vane J.K. ilrachidonicacid metaboiites and iuteraction between platelets and blood vessels. N. Engl J. tied 300: 1142-1147 (1979).
Y.
Sdmuuisson, d. Leukotrienes and related compounds. Adv Prostaglandin Thromboxane Leukotriene Res. 15: 1-Y (1985).
Hagenfeldt, K. Prostaglandins in t!lehuman Am J. Obstet Gynecol 122: 611-614 (1975).
iU. Abel, [I.D.and Baird, D.T. The effect of i7b-Estradiol and pro&esterone on prostagldndin production by human endometrium uaihtained in orsan cuiture. Endocrinol 106: 15YY-luU6 (lY8Oj. 11. Abel, M.H. and Kelly, R.W. noripre;nantnumah uterus.
Metabolism of prostaglandins by the J Clin End tietab56: 673+85 (lYr13).
12. Lumsden, Ii.A.,Brown, A. and Baird D.T. Prostaglandin production from homogenates of separdted glanduldr epithelium and strol;la prom human endometrium. Prostaglandins 28: 425-496 (1984). 13. Gal, D., Casey, L.M., Jonnston, J.M. and :IcDonaldP.C. desenchyme-epithelial interactions in human endometrium. Invest 7U: 798~803 (1332).
J Clin
14. Kees, M.C.P., Pdrry, IJ.!~., Anderson, A.B.tI.and Turnbull, B.C. Immunohistochemical localisation of cyclooxygenase in the human uterus. Prostaglandins 23: 2U7-214 (1982). 15. Casey, M.L., Hemsell, P.C., llcDonald,P.C. and Johnston, J.M. NADf dependent 15-hydroxy prostagldndiu dehydrodenaso activity in lluman endonetrium. Prostaglandins 19: 115-120 (lY80). 16. Schatz, F. and Gurpide, E. Effects of estrddiol 0)) prostaglandin P&evels in primary monolayer cultures of epithelial cell tram human proliferative endometrium. Endocrinoi 113: 1274-127Y (lYti3).
17. Narkiewicz, L., Schatz, G., Barg, P. and Gurpide, E.
Prostaglandin F2,outyut by human endometrium under superfusion auJ orga0 culture conditions. .J Steroid Biochem 22: 231-235 (19dSj.
JULY 1987 VOL. 36 NO. 1
33
CONTRACEPTION
18. Schatz, F., tlarkieuicz,L., Barg, P., and Gurpide, E. -In vitro inhibition with antiestrogens of estradiol effects on prostaglandin F20cproduction by human endometrium and endometrial rpithelidl cells. Endocri;lol118: 408-412 (1386). 19. Gurpide, E., Markiewicz, L., Schatz, F. and Hirata, F. Lipocortin output oy numan endometrium -in vitro. J Clin Endocr Metab 53: 162-166 (lY36). 23. i
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
30. Kees, X.C.P., Anderson, A.B.d., Deters, L.M. and Turnbull, A.C. Endometrial and myometrial prostaglandin release during the menstrual cycle in relation to menstrual tiloodloss;. J Ciin End Metab 58: 813-818 (1984). Of 31. Downing, I., Hutchon, D.J.K. and Poyser, N.L. ,,ptdKz [ 3H]-ardchidonic acid by human endometrium. Difference between normal and menorrhagic tissue. Prostaglandins 26: 55-67 (1933).
32. Demers, L.li.,Kees, i1.C.P.and Turllbull,A.C. Aracnidouic acid metabolism by the nonpregnant uterus. Prostagland Leukotriene Med 14: 175-180 (1934). 33. Granstrijm,E., Swahn, X-L. and Lundstrom, V. The pos;;iblcroles of prostaglandins and related compounds in endometrial uleeding. Acta Obstet Gynaecol Stand Suppl 113: 31-99 (lYU3). 34. Myatt, L., Bray, M.A., Gordon, D. and Morley, J. Macrophages on intrauterine contraceptive devices produce 2rostaglandias. :Jattirc 257: 227-228 (1975). 35. Kelly, 1I.U.an.lAbel, m.H. Coyger dnd zinc inhibit tne metabolism of prostaglandin by the human uterus. Biol Reprod 28: 883-88Y (1983). 36. tia&enfeldt,%. Studies of tilemode of action of the copper-T device. Acta Endocrin (Cph) Suppl 169: l-37 (lY72). 37. Salaverry, G., Mendes, C., Ziyper, J. and rledel,?I. Copper determination and localization in different morphoiogical components of human endometrium during tileiuenstrualcycle in copper intrauterine contraceptive device wearers. Am J Obstet Gynecol 115: 103-158 (lY73). 38. Peplov, P.V. and Hurst, P.R. Prostailandin synthesis by rdt uterine homogenate in the presence of copper ions, and the effects of 21: indomethacin and salicyclic acid. Prosta~laad Leukotriene lleil 15-22 (1986). 3r. Christaens, G.C.M.L., Siama, J.J. and Haspels, A.A. Haemostasis in menstrual endometrium in the presence of an intrauterine device. Br J Obstet Gynaecol 88: 32j-835 (lY81). 40. Sileppard,B.L. and Bonnar, J. The effects of intrauterine contraceptive devices on the ultrastructure of endometrium in relation to bleeding complications. Am J Obstet Gynecol 146: 829-534,(1983).
JULY
1987 VOL. 36 NO. 1
35
THE ROLE OF PROSTAGLANDINS AND ALLIED SUBSTANCES IN UTERINE HAKMOSTASIS
Kerstin Hagenfeldt, M.D. Department of Obstetrics and Gynaecology Karolinska Institute Stockholm, Sweden
AHSTRACT This review addresses itself to summarizing the more recent studies published on the bioconversion of arachidonic acid in the human endoand myometrium during the normal menstrual cycle and in women with increased menstrual blood loss. The data indicate an increased ability of the endo- and myometrium from women with menorrllagiato produce prostaglandins with vasodilator and platelet anti-aggregatory properties, viz., prostacycline and prostaglandin E2. Tne data on prostaglandin production in endometria of IUD wearers is reviewed and discussed in relation to present knowledge on morphological findings in IUD-influenced endometrium. INTRODUCTION The interest in studying prostaglandins in the human uterus was primarily evoked by their involvement in myornetrialphysiology; in abortion and parturition and in the clinical condition of dysmenorrhoea. The role of prostaglandius in the control of menstrual blood loss has received considerable attention after Wiqvist -et al. (1) reported that the administration of PGF2Aduring the liltealphase induced menstruation. Furthermore, treatment with prostaglandin synthetase inhibitors during menstruation resulted in reduction of menstrual blood loss in women with menorrhagia during IUD use (2,3). Between 1973 and 197Y several reports were published in rhich the amount of prostaplandins of the E and F type were determined in endometrial specimens obtained at various phases of the normal menstrual cycle, during certain pathogenic conditions and uuring IUD use. The results of these studies nave been discussed in a review paper by Hagenfeldt at the WHO meeting in 1979: Endometrial bleeding and steroidal contraception (4). The results of the studies published until lY79 seem to indicate a higher production of prostaglandin F2&(PGFa) from mid and late secretory endometrium and during menstruation compared with the proliferative phase. Prostaglandin E2 (PGE2) levels did not show cyclic variation in the normal endometrium, but both PGF2&and PGE2 were increased in endometria from women with dysfunctional uterine bleeding.
JULY
1987 VOL. 36 NO. 1
23
CONTRACEPTION
Three reports address themselves to determining PGE2 and/or PGFZhin endometrial specimens obtained from women using intrauterine devices. Green b Hagenfeldt (5) did not find any change in PGP2 or its main metabolite 13,14-dihydro-15-keto PGFzAafter three months use of a large inert device (Dalkon Shield) or a Copper-T ZOO. Hillier h Kasonde (6) made the same observation on PGFz,Xlevels but reported an increase in PGE levels in 3 out of 5 women using a Copper-7 device and in 7 out of 3 women using various inert devices. A third study by Scommegna -et al. (7) reported on PGF levels in endometrial specimens obtained in women using either inert devices, a progesterone-releasing IUD (Progestasert) or a dydrogesterone-releasing IUD. Compared with levels in tne normal endometrium, no change was found in the first tu7o groups; PGF levels in specimens obtained from dydrogesterone influenced endometrium were slightly lower than before iusertion. Prostaglandins are synthesized and metabolized locally and are not stored in tissues. Therefore tne technique of ootaining endometrial specimens by biopsy or after hysterectomy would invariably evoke a trauma which in itself triggers the release of arachidonic acid from cell membranes. Hence the studies discussed above do not report on in vivo levels of the prostaglandins but rather measure the efficacy of?he cyclooxygenase pathway. Over the past few years, more sophisticated methods have been used to study arachidonic acid metabolism in endometrium and myometrium. Furthermore, in additiou to the so-called classical Prostaglandins, PGF2:7(andPGE2, other arachidonic acid metabolites such as prostacyclln (PGI2) and thromboxane A2 (TXA2) have been studied. Research in platelet and vessel wall arachidonic acid metabolism indicates that the metabolites could be important in haemostasis (8). In addition to the known vasoconstrictor effect of PGF2dand tHe vasodilation effect of PGE2, the effect of the other two metabolites, being a potent vasodilator such as PGI2 and a potent platelet aggregator such as TXA2 would probably be of fundamental importance in the regulation of menstrual blood loss. Recently the importance of another pathway of the arachidonic acid metabolism has been stressed. Through tne enzyme lipoxygenase (see Fig. 1) arachidonic acid can be converted to another class of substances now called the leukotrienes (9). These substances have potent biological activities including cytotoxicity, chemotaxis, vasoconstriction in some vascular systems and increasing vascuiar permeability. These properties could add positiveiy or negatively to the haemostatic mechanism in certain tissues, e.g. the uterus. The regulation of prostaglandin synthesis and metabolism in the normal human uterus In a series of papers Published between 197Y and 1986, scientists using in vitro systems such as organ culture of the human endometrium, monolayer cultures of separated endometrial cells or superfusion techniques of endometrial tissue, have drawn attention to several important aspects of the arachidonic acid metabolism in the human uterus.
24
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
5-HPETE
IS-HPETE
LTA,, yg
INHIBNK CELL CYTO10x & 0; -GEN &DEGRkNULAT.
INHIB. DEGRANULAT. IO; -GEN.
$q
ADHESION CHEMOTAX DEGRANULAT. 0; -GEN.
VASC. PERM BRONCHOCONSTR VASDCONST R Lti RELEASE
CONTR PULM STR
Figure 1. Formation a:lclbiological effects of uerivotives of (Samuelsson -et al. Bdv. Prostagland. Tnromboxan. arachidonic acid. Leu,
Thus, the influence of ovarian steroids on the prostaglandins syntheslj in the numan endonetrium has been studied by Baird's rese.lrch group in Edinburgh. Aioel (r Baird (10) snowed that the human endometrium maintained in or;an culture produced more PGFz&than PGE2. 17B-tistradiol increased PGFuproduction in the secretory but not in the Llroliferative endometrium. Progesterone significantly reduce,i PGF~~aand PGE2 in both phases of the cycle and prevented the estradiol-induced increase in PGF2, of the secretory endometriuia. The results suggest that there is an inherent Difference in the reductase to isomerase enzymes or in the ability of the enzymes to respond to hormones . Mternatively PGE2 level.; in the endometrium :Alaynot be under enzymatic control but may reflect a non-eizyaatic conversion of the endoperoxide PGl12 which fails to act as a substrate for the PGFck reductase. In a study by i1oel & Kelly was investigated in the normal are involved in the regulation concentration of the compounds
JULY
(11) tile metaboiism of human endometrium. If of menstrual bleeding, vittii,l the uterus will
1987 VOL. 36 NO. 1
PGEL and PCP2 PGFl+and PGE2 the effective tlepend upon t!le
25
CONTRACEPTION
. relative rates of synthesis and metabolism urithinor removal from the tissue. Both 15-hydroxy-prostaglandin dehydrogenase and the A13- redUC t.dse enzyme are present in both endometrium and myometrium. The dehydrogenasc is NAD+-dependent while the reductase requires NADPH+ as a co-factor. No evidence was found for the existence of a 9-keto reductase enzyme converting PGE2 to PGFz~ The major metabolites were invariably 13,14-dihydro-15-keto compounds. Tile netabolisiaof PGE2 exceeded that of PGF2,,_bothin endometrium and myometrium. A decrease in the metabolizing capacity was seen during the pre-menstrual phase. This could be contributing to the increased concentration of PGF2&and PGE2 seen dt that time. Further studies on the metabolism of 14C-arachidonic acid in separated glandular epithelium and stromal ceils from the human endometrium have been published by Lumsden et al. (12). Glandular epithelial cells produced more PGF2dt"an did stromal cells both in the proliferative and secretory endometrium. The production of PGE2 from Labelled arachidonic acid was variable with no difference between the two cell types or between the two phases of tne cycle. Conversion to b-oxo-PGFlswas negligible in both glandular epithelial and stromal cell. Thus the cyclooxygenase activity was more active in tileglands. This result is at variance with a report from Gal -et al. (13) who studitidthe same ceil types in noilolayercultures for up to 21 days. In their system more PGE2 than PGF2Awas produced and the synthesis was more active in stromai than in glandular epithelial cells. It was, however, pointed out by Lumsden --et al. that the viability of the stronal cells in the Imonolayersystem of Gal -et al. was very low after 21 days, while the viability in the cell preparation used by themselves was satisfactory. Using immunohistochemicdl methods for the localization of the cyclooxydenase enzyme, Rees -et al. (14) found the highest activity in the glanduldr epithelidl celiu of the luteal phase endometrium. No specific staining was found in stroma, myometrium or blood vessels. In an earlier study, Casey -et al. (15) found the highest activity of the prostaglandin degrading enzyme, 15-OH prostaglandin dehydrogenase, in the glanduldr epithelial cells of the iuteal phase endometrium. The failure to demonstrate specific staining of the cyclooxygenase enzyme in non-epithelial cells in the human uterus does not exclude the presence of the enzyme in other cells than glandular epithelium. It could be a scientists in Gurpicie's question of sensitivity of the method used. 'l'ne group at Mount Sinai in New York have used organ culture of the human endonetrium and tissue experiments with superfusion and wouolayer cultures of separated endometrial cells to study tne regulation of prostaglandin synthesis and metabolism. Schatz h Gurpide (16) confirmed the results of the Edinburgh group that estradiol increases the PGF2,p reduction of glandular epithelial cells of the secretory endometrium. Botilepithelial and stromdl cells of the proliferative endometrium release PGF2&but only the glandular cells increase the PGF2Aproduction with addition of estradiol.
26
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
Mariciewicz-et al. (17) compared the PGF2Aproduction in proliferative and secretory endometrium under superfusion and organ culture conditions. The basal PGF240utput was greater in proliferative than in secretory endometrium, but the response to estradiol was strolyest in the secretory endometrium. Progesterone lowered basal PGFZo+,outputin both proliferative and secretory endometrium in organ culture. During superfusion, progesterone only affected cells from the proliferative endometrium. The same experimental model was used by Schatz -et al. (18) to study the effect of anti-estrogens on prostaglandin production. Both with cultured endometrial fragments aud tiithmonolayer cultures of separated glandular epithelial and stromal cells, the anti-estrogens (tamoxifen and its metabolite trans-4-monohydroxy-tamoxifen) counteracted the estradiol increase in PGF2dp reduction while failing to affect the basal production of the prostagldndius. This suggests that the basal rate may not be dependent on endogenous estrogens. The inhibition of the estradiol-stimulated increase in PGF2d production suggests that the stimulation is estradiol receptor-mediated. Progesterone inhibits PGF2dproduction in botn proliferative and secretory endometrium maintained in organ culture. The exact mode of action of progesterone is not known. Glucocorticoids are known to exert their dnti-inflammatory effect by increasing tissue production of proteins that inhibit phospholipase activity, thus reducing the amount of arachidonic acid available for prostaglandin synthesis. Toe name lipocortin has been proposed to describe this antiphospholipase. Gurpide --et al. (19) found that lipocortin was produced botil by the proliferative and the secretory endometrium in organ culture. Levels were increased by dexamethazone and decreased by progesteroue. both steroids, however, decreased PGF2,production. The results indicate that lipocortin does not mediate the inhibition of PGF2Aevoked by progesterone. The effect of antiprogesterone on the prostaglandin production of human endometrial cells has been studied by Kelly -et al. (20). Glandular epithelial and strowal cells were cultured for 24 hours in the presence of the progesterone antagonists RU 486 and ZK 98734. Both steroids stimulated PGF2Aproduction by stromal cells in a dose-dependent manner. Added progesterone innibited the RU 486-stimulated increase in PGF24 PGE2, wirenmeasurable, also increased in the presence of KU 486. No effect of KU 486 was seen in tne slandular epithelial cells. Preincubation of stromal cells with arachidonic acid markedly increased the effect of KU 486. Pretreatment wicn progesterone allowed a L;reaterconversion to 13,14-dihydro-1%keto PGF2& The results indicate that antiprogesterone can act as menstrual regulators by stimulating eudogenous prostaglandin F2w production in stromal cells or by inhibiting prostaglandin metabolism. Prostaglandin binding to different cell types has been studied by Hoffman -et al. (21). PGE2 binding sites were significantly higher in proliferative than in secretory endometrium. Tinediagnosis of abnormal uterine bleeding was associated with higher %I-prostaglandin E2
JULY 1987 VOL. 36 NO. 1
27
CONTRACEPTION
binding than diadnosis of dysmenorrhoea or uterine prolapse. regardless of 3H-PGF2Abinding wds low in all endo8aetridlapeci!llens the pudse of the cycle. These results are supported in a recent study by Che;ini et al. (22), who used a quantitative light microscopic autoradiographic -method to analyse prostaglandin binding sites in different cell types of the human endometrium. Stromal cells, glandular epithelial cells, elongated and circular smooth muscles and arterioles contained numerous PGE2 and very few OK no detectable PGF2dbinding sites. Sinding sites in all cell types were higher in the proliferative than in the secretory endometrium while binding sites in myometrium did not vary with the cycle. The reasons for the very few or no detectable PGF2 binding sites are discussed in the paper by Hoffman -et al. (21). &skin& of binding sites by endogenously produced PGF2&, rapid metabolism of jH-PGF2&iin tireincubation system, laaskingof binding sites durint;the tlomogenizationand centrifugation $KOcedUKe, were all ruled out. Studies on prostaglandin productiou in the human uterus in relation to increased menstrual blood loss Tne discovery of tue potent anti-a&reijatory aud vasodilatory prostacyclin, PGI2 (23), triggered the interest to study new metauolites of arachidonic acid in the human uterus in an effort to explain pathological menstrual bleeding. Thus, Abel & Kelly (24) regorted that while the human endometrium mainly produced PGE2 and PGF2&from labelled arachidonic acid, the myometrium produced 6-ke to-PGFl, the stablr metabolite of prostacyclin (PGI2). ihen endometrium and myometrium .derecombined, a fUKtheK enhancement of the PGI2 production was seen with less PGE2 and PGF2hsynthesized. This indicates chat precursor PGG2 and PGB2 produced in the endonetriurncan be used by the myometrium fOK PG12 production. Endometrial PtiF2Aand PGE2 produced from labelled arachidonic acid was studied in women with menoKrhd&id (menstrual olood loss Jreater than 50 ml per cycle) and compared with that in women with normal menstruation (25). PGFa/PGF2 ratio was siynificalrtlylotrerin the eIldOuetKiUu of wonen LJith high menstrual blood loss, i.e. more than 90 ml per cycle. There was an inverse correlation between PGFk/PGE2 ratio and blood loss (K = 0.36, ~(0.025). The result suggests that excessive lnenstrualolood loss may be aasocidted with a shift in the endomctrial conversion of pKOSta&ldndin endoperoxides from PGF2d;to PGE2. In another study, Smith -et al. (.26)measured the synthesis of prosta&landins from labelled arachidonic acid in endometrium and uyometrium obtaineo from women with excessive menstrual blood loss (range 57-186, median 86 ml). The endometrium of women with high MBL was more effective than that of tiomenwitn normal NtiLat enhancing the production of ~-oxo-PGF~~, the staole netabolite of PGI2, in the myomecrium. These results further confirm the hypothesis of an abnormal prostaglandiu synthesis in women with excessive bleeding. The ,)atteru
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
of prostaglandin synthesis in endometrium obtained from women witu persistent anovulatory cycles and increased MUL was publisned by the same group (21). PGF2Alevels t+erelower in the anovulatory endometrium tnan in normal secretory endometrium, but the clmountof PGE2 and PFG2c,,synthetizedin this tissue from exogenous arachidonic acid were similar to those found in normal sacrrtory endometrium. T,lese findings suggest that the persistent proliferative endometrium has a in vivo because it iacks tile reduced capacity to synthetiLe PGF20,,_ essential precursor for PG synthesis proLably because of lack of progesterone influence. Tnis raduced capacity to synthetize PGF24may result in excessive MBL as suggested by the inverse correlation of PGFw/PGE2 ratio and menstrual blood loss (r =-0.7, p
Pees -et al. (30) also studied prostaglandin production in endometria and myometrid from women with menorr~lagiausing a superfusion technique. Their results dre at vdriance With those of the Edinburgh group. No difference Leas found in PGE2, PGP2* or G-oxo-PGFiA pro\iuctionbetween women with or without menorrhagia. The reason for the difference in results could be that of definition of increased ,lBL. Rees -et al. compared a group of women with MBL<80 ml with a group of women having NRL) dU ml per cycle. Smith -et al. (25) found tne difference in prostaglandin synthesis between a group of women witil (50 ml XBL per cycle and a group vitu>YU ml blood loss per cycle. Downing -et al. (31) studied organ culture of endometria from healthy women and from vomen with menorrhagia. The uptake of 3H-arachidonic acid in endo.detrialphospholipids was hi&her in normal secretory than normal proliferative endometrium. In endometrid from women witil menorrha~ia, 3’d- ardchidonic acid wds mainly distributed into neutral lipids, especially triJycerides, and particularly in the proliferative dhase. This difference indicates an asnormality in the ardcnidonic acid metabolism in the menorrhagic endometrium. If the increased amount of ardchidonic acid incorporated into triglyceride is released during
JULY 1987 VOL. 36 NO. 1
29
CONTRACEPTION
menstruation, it could lead to an overproduction of PGH2. This may result in saturation of the endoperoxide reducing factor by whict1 PGF2&is synthesized, thus allowing surplus PCH2 to be couverted to larger than normal quantities of PGE2 and PG12 both with vasodilatdtion properties resulting in an increased menstrual blood loss. Tne data presented mainly by the Edinburgh group indicates that an interaction between endometrial and myometrial prostaglandin production in vitro data do not necessarily prove that is possible. Althout;h the -this is happening also -in vivo the possible consequences are interesting. An increased availability of ardchidonic acid in the endometrium of women with menorrhagia can give an increased pool of precursor endoperoxides. The pathway to PGF2Abecomes saturated leading to a diversion of the synthesis towards nonenzymatically produced PGE2 and also to more PGI2 bein,:formed in the rnyoaetrium. The comb.ineddecrease in the PGF2dPGE2 ratio and the increase in PGL2 would increase vasodilatation of endometrial vessels and augment menstrual blood loss. An interesting aspect of arachidonic acid metabolism in the human endometrium was discussed by Demers -et al. (32). Studying short-time incubation of human endometrium and myometrium with L4C-arachidonic lipoxygenasc as well as cyclooxygenase activity was found. acid, both 'The lipoxyyenase products formed by the tissues included the monohydroxy acids, 5-hydroxy-eicosa-tetracrtoicacid (5-HE%), 12-hydroxy-eicosatretraenoic acid (12-tIETE)and 12-iiydroxy-heptadecatrienoicacid (HHT). The endometriai specimens released PGFk, PGE2 and 12-LIETE,while the products found in the myometrium were 12-HBTE and small amounts of 5-ZTE, HHT and 6-keto-PGFld. Tnis study indicates that t,lelipoxygendse activity may be of interest in menstrual disorders such as dysmenorrhoea and menorrhagia. In a review of prostaglandins and menstrual bleeding, Granstiam et al. (33) suggested that one of the pharmacological effects of non-steroidal anti-inflammatory drugs (NSAIDs), ou menstrual blood loss could be that the inhibition of the cyclooxygenase may increase the arachidonic aciJ pool dvailable for tilelipoxygenase pathway producing products with vasoconstrictor activity, i.e. leukotrienes. Further studies on isolated endometrial cell systems during inhibition with prostaglanain synthetase inhibitors could elucidate this possible mechanism. Prostaglandin synthesis and metabolism in endometria of IUD users No studies have been published using --in vitro techniques Such dS organ culture of endometrial tissue or monolayer culture of separated eodometrial cells obtained from women using inert or medicated devices. All IUDs stimulate an increase in leucocytes and other cells in the endometrium aud uterine fluid as part of the foreign body reactiou to the device. Several studies indicate an adherence of these cells to the IUD surface. iryatt-ec al. (34) reported that although the total amount of white blood cells were greater on the Dalkon Shield, the cells on
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
copper IUUs produced the highest concentration of PGF2,hwith the medium. The authors discussed their findings in relation to the effect on implantation. With the present knowledge on arachidonic acid metabolism in human polymorphonuclear leucocytes, viz., leukotriene production (19), an effect on uterine haemostasis is also possible. A few studies have addressed themselves to the effect of metal ions on the arachidonic acid metabolism in the endometrium. Kelly & Abel (35) investigated the action of copper and zinc on the prostaglandin metabolizing enzymes in the husan endometrium and myometrium. Copper and zinc chloride at a concentration of 10V5 mol/l inhibited the metabolism of PGE2 in the luteal phase endometrium and in the myometrium, copper more effectively than zinc. Also metallic copper in the form of a copper foil with an area of 25 mm2 was sufficient to reduce PGE2 metabolism by 50X. The concentration of copper used in this experiment are in the same range as that discussed by Uagenfeldt in the uterine fluid and cervical mucus of women using the CUT device (36). According to Salaverry -et al. (37) copper is concentrated in the glandular epithelial cells where also the prostaglandin dehydrogenase is located(l%It was suggested by r;ellyand Abel that the finding of an inhibition of the PGE2 metabolism induced by copper could contribute to the contraceptive action of the device. Increasing PGE2 versus PGF2dlevels could also have an effect on menstrual bleeding, i.e. increasing vasodilatation and menstrual blood loss. As very few studies on human endometrial tissue are available, one experimental study on the rat will be mentioned. Peplov -et al. (38) studied the effect of copper ions on the prostaglandin production in rat uterine homogenate. Copper in a concentration of 10s4 to 10q5 mol/l was found to stimulate PGF2++,synthesis. By contrast the synthesis of PGE2 and PGI2 were unaffected at all copper concentrations used. The increased PGF2dsynthesis in the presence of copper was not siynificantly changed by salicyclic acid but was inhibited by indomethacin. CONCiUSIONS The information dained from the studies presented in this review indicate that the human endometrium metabolizes arachidonic acid mainly to PGF2&and PGE2. The synthesis of these two compounds seem to be under control of ovarian steroids. 17%Estradiol stimulate PGF2d production in the secretory endometrium and anti-estrogens can inhibit this increase indicating that the process is estrogen receptordependent. Progesterone inhibits PGF2tiproduction in both proliferative and secretory endometrium and anti-progestins increase PGF2& production mainly in the stromal cells. The highest production of PGF2dand PGE2 is seen in the late luteal phase and during menstruation. Glandular epithelial cells invariably syntnetize more PGF2Athan stroual cells. The rnyometriun produces lnainlyPG12 using endoperoxide precursors from tne andometrium. PGF2qand PGE2 are metabolized in the endometrium and myometrium mainly to 13,14-dihydro-15-keto compounds. The metabolism of PGE2 exceeds that of PGF2& A decrease in the metabolizing capacity
JULY
1987 VOL. 36 NO. 1
CONTRACEPTION
is seen during menstruation, contrioutiny to hi&i1concentrations of the prostaglandins seen at that time. The cyclooxygenase enzyme as well as the prostaglandin degrading enzyinesare mainly identified in the glandular epithelial cells. PGE2 but very few PGFL&binding sites are found in all cell types of the human endornetriutl with higher concentrations in tne proliferative phase. The noriaalhurtian endometriuladrld ;nyotnetrium netaoolize arachidonic acid both by the cyclooxygenase and the lipoxydase pathway. The former is dominating but the possiaiiity of leukotriene formation could be important in pnysiological and pathological conditions. Li the uterine tissues of vlomensufr'erin&from increased iaenstrudl blood loss, there is a decrease in the PGP&PGE2 ratio and an increased capacity of the myometrium to produce PGI2. The relative decrease of the prostaglandin with vasoconstrictor properties and increase in compounds with vasodilatatiou and anti-aggrrgatory properries would be consistent with a less effective haemostasis at the tilile of lnenstruation. Altnough the situation in IUD-influenced ~ndometriurndnd I.iyonetrium has not been investigated, the findings in women with menorrhagia due to other reaso11scould probably be also relevilntin IUD users bearin& in mind the uorphological findings in the studies on uterine haemostasis in IUD users +blished by several investigators supgorted uy the :JHOTask Force on Intrauterine i4ethodsof Fertility Regulation (39,40). l'nusthe findings of fewer itaemostdticplugs in IUD-i,lfluenced tilenstrual endometrium could oe due to a shift in the balance between PGI2 and TXA2. Tile increased amount of non-occluded sudll vessels in the endometrium could oe a result of a decrease in PGF2d/PGE2 ratio favouring a less effective hdemostasis. Furthermore, an increased dmount 3f +lymor~honuclear leucocytes adhering to the device, especially the copper devices, indicates the possibility of leukotriene formation at++enting the increase in vascuiar permeability typical for the IUD-influenced endometrium.
1.
Viqvist, N., Bygdeman, d. and Kirton. K. Nonstrroidal infertility agents i;zthe female, in Nobel Syspos. 15. Control of tIumau Fertility (E. Diczfalusy, and 0. Borell, Editors). AlrnqvistWiksell, Sweden, lY71, p. 137-143.
2.
Andersson, A.B.M., Haynes, P.J., Guillebaud, I. and Turnbull A.C. Reduction of Inenstrualblood loss by prostaglandin sjnthetase inhibitors. Lancet i: 774-776 (1976).
3.
Guillebaud, J., Anderson, A.8. and Turnbull, A.C. Reduction by mefenaaic acid of increased menstrual blood loss associated with intrauterine contraception. Br J. Obstet Gynaucol 85: 53-62 (1978).
32
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
4.
Dadenfeldt, K. Prostaglandins and related compounds dnd taeir metabolism in steroid exposed endometrium, in Endometrial Bleeding and Steroidal Contraception (E. Dicefalusy, I.S. Frazer, and F.T.ti. Webb, Editors). Pitman Press Ltd, dath, lY80, p. 222-245.
5.
Grren, K. and endometrium.
’ o.
Hillier, K. and Kasonde, J.iI. Prostaklandin E and F concentrations in human endometrium after insertion of intrauterine COntraCeptiVe devices. Lancet i: 1%lo (1376).
7.
Scommegna, A., Tlekis, J., Rae, K., Dmowski, W.P., Rezai, P. and hluetta, F.J. Zndometrial prosta~landiu f content in women wearin& non-medicated or progestin-releasing intrauterine devices. Fertil Steril 29: 500-504 (197Y).
8.
t:oncada,5. and Vane J.K. ilrachidonicacid metaboiites and iuteraction between platelets and blood vessels. N. Engl J. tied 300: 1142-1147 (1979).
Y.
Sdmuuisson, d. Leukotrienes and related compounds. Adv Prostaglandin Thromboxane Leukotriene Res. 15: 1-Y (1985).
Hagenfeldt, K. Prostaglandins in t!lehuman Am J. Obstet Gynecol 122: 611-614 (1975).
iU. Abel, [I.D.and Baird, D.T. The effect of i7b-Estradiol and pro&esterone on prostagldndin production by human endometrium uaihtained in orsan cuiture. Endocrinol 106: 15YY-luU6 (lY8Oj. 11. Abel, M.H. and Kelly, R.W. noripre;nantnumah uterus.
Metabolism of prostaglandins by the J Clin End tietab56: 673+85 (lYr13).
12. Lumsden, Ii.A.,Brown, A. and Baird D.T. Prostaglandin production from homogenates of separdted glanduldr epithelium and strol;la prom human endometrium. Prostaglandins 28: 425-496 (1984). 13. Gal, D., Casey, L.M., Jonnston, J.M. and :IcDonaldP.C. desenchyme-epithelial interactions in human endometrium. Invest 7U: 798~803 (1332).
J Clin
14. Kees, M.C.P., Pdrry, IJ.!~., Anderson, A.B.tI.and Turnbull, B.C. Immunohistochemical localisation of cyclooxygenase in the human uterus. Prostaglandins 23: 2U7-214 (1982). 15. Casey, M.L., Hemsell, P.C., llcDonald,P.C. and Johnston, J.M. NADf dependent 15-hydroxy prostagldndiu dehydrodenaso activity in lluman endonetrium. Prostaglandins 19: 115-120 (lY80). 16. Schatz, F. and Gurpide, E. Effects of estrddiol 0)) prostaglandin P&evels in primary monolayer cultures of epithelial cell tram human proliferative endometrium. Endocrinoi 113: 1274-127Y (lYti3).
17. Narkiewicz, L., Schatz, G., Barg, P. and Gurpide, E.
Prostaglandin F2,outyut by human endometrium under superfusion auJ orga0 culture conditions. .J Steroid Biochem 22: 231-235 (19dSj.
JULY 1987 VOL. 36 NO. 1
33
CONTRACEPTION
18. Schatz, F., tlarkieuicz,L., Barg, P., and Gurpide, E. -In vitro inhibition with antiestrogens of estradiol effects on prostaglandin F20cproduction by human endometrium and endometrial rpithelidl cells. Endocri;lol118: 408-412 (1386). 19. Gurpide, E., Markiewicz, L., Schatz, F. and Hirata, F. Lipocortin output oy numan endometrium -in vitro. J Clin Endocr Metab 53: 162-166 (lY36). 23. i
JULY 1987 VOL. 36 NO. 1
CONTRACEPTION
30. Kees, X.C.P., Anderson, A.B.d., Deters, L.M. and Turnbull, A.C. Endometrial and myometrial prostaglandin release during the menstrual cycle in relation to menstrual tiloodloss;. J Ciin End Metab 58: 813-818 (1984). Of 31. Downing, I., Hutchon, D.J.K. and Poyser, N.L. ,,ptdKz [ 3H]-ardchidonic acid by human endometrium. Difference between normal and menorrhagic tissue. Prostaglandins 26: 55-67 (1933).
32. Demers, L.li.,Kees, i1.C.P.and Turllbull,A.C. Aracnidouic acid metabolism by the nonpregnant uterus. Prostagland Leukotriene Med 14: 175-180 (1934). 33. Granstrijm,E., Swahn, X-L. and Lundstrom, V. The pos;;iblcroles of prostaglandins and related compounds in endometrial uleeding. Acta Obstet Gynaecol Stand Suppl 113: 31-99 (lYU3). 34. Myatt, L., Bray, M.A., Gordon, D. and Morley, J. Macrophages on intrauterine contraceptive devices produce 2rostaglandias. :Jattirc 257: 227-228 (1975). 35. Kelly, 1I.U.an.lAbel, m.H. Coyger dnd zinc inhibit tne metabolism of prostaglandin by the human uterus. Biol Reprod 28: 883-88Y (1983). 36. tia&enfeldt,%. Studies of tilemode of action of the copper-T device. Acta Endocrin (Cph) Suppl 169: l-37 (lY72). 37. Salaverry, G., Mendes, C., Ziyper, J. and rledel,?I. Copper determination and localization in different morphoiogical components of human endometrium during tileiuenstrualcycle in copper intrauterine contraceptive device wearers. Am J Obstet Gynecol 115: 103-158 (lY73). 38. Peplov, P.V. and Hurst, P.R. Prostailandin synthesis by rdt uterine homogenate in the presence of copper ions, and the effects of 21: indomethacin and salicyclic acid. Prosta~laad Leukotriene lleil 15-22 (1986). 3r. Christaens, G.C.M.L., Siama, J.J. and Haspels, A.A. Haemostasis in menstrual endometrium in the presence of an intrauterine device. Br J Obstet Gynaecol 88: 32j-835 (lY81). 40. Sileppard,B.L. and Bonnar, J. The effects of intrauterine contraceptive devices on the ultrastructure of endometrium in relation to bleeding complications. Am J Obstet Gynecol 146: 829-534,(1983).
JULY
1987 VOL. 36 NO. 1
35