- Email: [email protected]
In vitro production of pregnancy-associated plasma protein A by human decidua and trophoblast
In vitro production of pregnancy-associated plasma protein A by human decidua and trophoblast P. Bischof, S. DuBerg, M. T. Sizonenko, A.M. Schindler, F. Beguin, W. L. Herrmann, and P. C. Sizonenko Geneva, Switzerland Immunohistochemical techniques and direct measurements of pregnancy-associated plasma protein A (PAPP-A) have demonstrated the presence of PAPP-A in trophoblast and decidua. The purpose of the present study was to investigate the possibility that these tissues are capable of producing PAPP-A in vitro. Trophoblast and decidua were obtained from term deliveries and from legal surgical terminations of pregnancy (7 to 12 weeks). In addition to trophoblast and decidua, myometrium was also obtained during two hysterectomies in the first trimester of pregnancy. Tissues were incubated in medium 199 at 37" C under an oxygen/carbon dioxide atmosphere. Media containing either pregnancy-associated serum or non-pregnancy-associated serum were changed after 8 hours of incubation in medium 199 alone. In addition to PAPP-A, human placental lactogen (hPL) and prolactin (Prl) were measured in homogenates and media by radioimmunoassays in order to confirm the viability of the cultured tissues. Addition of pregnancy-associated serum to the media induced a significant release of PAPP-A from trophoblast and decidua when compared to that in control cultures. Non-pregnancy-associated serum had no effect. Myometrium did not release any measurable PAPP-A into the medium even in the presence of pregnancy-associated serum. Cycloheximide added to pregnancy-associated serum significantly inhibited the release of PAPP-A from trophoblast and decidua. These last tissues, irrespective of the culture condition, released significantly more PAPP-A as well as hPL and Prl than was initially present in the tissue. These data demonstrate that PAPP-A is released in vitro by trophoblast and decidua (but not by myometrium) and that this release can be magnified by a factor present only in pregnancy-associated serum. The release of PAPP-A, hPL, and Prl is considered as a de novo production since concentration of these proteins are higher in media and tissues after incubation compared to concentrations initially present in the tissue before culture and since cycloheximide significantly inhibits the release of PAPP-A, Prl, and hPL from the cultured tissues. (AM. J. 0BSTET. GYNECOL. 148:13, 1984.)
Pregnancy-associated plasma protein A (PAPP-A) was discovered by Lin and co-workers in 1974 1 in the plasma of pregnant women. Subsequent purification and characterization 2 • '1 demonstrated that PAPP-A is a macromolecular glycoprotein (molecular weight of 800,000) of dimeric structure, each monomer being composed of two apparently identical subunits with a molecular weight of 200,000. 2 • '1 PAPP-A appears in increasing concentrations as pregnancy progresses, 4 - 6 almost exclusively in the maternal circulation. Although its biologic role is still unknown, in vitro experiments suggest an immunomodulatory function. 7 - 9 By the use of immunohistochemical techniques,
From the Division of Clinical Biology of Growth and Reproduction, the Department of Obstetrics and Gynecology, and the Centre de Cytologie et de Depistage du Cancer, University of Geneva. Supported by Grants No. 3.931.0.80 and No. 3.909.0.82 from the Swiss National Fund for Scientific Research. Received for publication April 7, 1983. Accepted August 1, 1983. Reprint requests: Dr. P. Bischof, Laboratoire d'hormonologie, Maternite, 1211 Geneva 4, Switzerland.
PAPP-A could be localized on the syncytiotrophoblast and in decidual cells. 10 - 12 Direct measurements of PAPP-A in different fetomaternal tissues 1'1• 14 also revealed its presence in trophoblast and decidua as well as in endometrium.I" Immunohistochemical localization and direct measurements of PAPP-A thus suggest a possible decidual as well as a trophoblastic origin for PAPP-A. The present study was undertaken to determine whether these tissues are capable of producing PAPP-A in vitro. Material and methods
Placentas and membranes were collected from 10 uncomplicated term pregnancies terminated by spontaneous labor and delivery or cesarean section. Trophoblast and decidua were obtained from I 4 patients undergoing surgical termination of pregnancy between weeks 7 and 12 of gestation. Finally, myometrium, decidua, and trophoblast were also obtained from two patients undergoing hysterectomy in the first trimester of pregnancy for reasons other than malignancy. Term placentas were cut midway between the basal and chorionic plates, and the tissue was cut in small pieces. Term decidua was peeled from the fetal mem13
14 Bischof et al.
January 1, 1984 Am. ]. Obstet. Gynecol.
PAPP-A l~g/g)
mean
I TROPHOBLAST I
:t
PAPP-A
SEM
l~g/g)
100,-------:;.,.';==:::;------" early
•••• ••
'•
pre 9 nancy·······r·
- - - 199
•••••}""
b
200
T = Trophoblast
- - - - 199. NPs ···········199•PS
160
0 =Decidua M = Myometrium
so
·i/
term pregnancy
/
I
,-.....-.......l······ .
1\
•
120
~~·· O~~~~~~=:=-~t=-~~====;===~T~~~ I DECIDUA I 100 early pregnancy
_I110)
PAPP-A (~g/g)
{10)
40
T
term pregnancy
0
l . .) . . .
\
so
0
\/- /
I .1••••• .J"" •••••••• J••••••
~ I
••
T
16
24
T
36
J(4)
48
0
8
16
24
141
(10)
36
0
12
24
36
M 24
48
36
48
hours of incubation
',J.,
~-L-~
01.)
T
80
*"'-"'*-=-=-=;~~,-=;Jiill41
0
'l
a
80
48
Fig. 2. Release of PAPP-A in culture media expressed in micrograms per gram of tissue ex planted (tissues obtained from two hysterectomies during pregnancy). a: In medium 199 alone; b: in medium 199 supplemented by pregnancy-associated serum.
hours of incubation
Fig. 1. Release of PAPP-A in culture media expressed in micrograms per gram of tissue explanted. 199: Medium 199 alone; NPS: non-pregnancy-associated serum (pool of normal female serum); PS: pregnancy serum (pool of third-trimester pregnancy serum).
branes according to the technique of Riddick and Kusmik. 16 Decidua and trophoblast from early pregnancy aspirates or products of hysterectomy were separated by hand under sterile conditions. Routine histologic sections were taken from all tissues for evaluation of possible contamination by other tissues. All material was collectd under sterile conditions and treated as quickly as possible after delivery or operation (within 20 minutes). Tissues were minced and washed in phosphatebuffered saline (PBS; 0.05M phosphate and 0.2M sodium chloride, pH 7.4) containing 2.5% penicillin and 2.5% streptomycin, and approximately 300 mg of tissue was incubated separately in Ehrlenmeyer flasks containing 2 ml of medium 199 (Difco, Detroit, Michigan). Incubation was carried out in a shaking water bath at 37° C with a 5% carbon dioxide and 95% oxygen atmosphere. Medium 199 was changed at 2, 4, 6, and 8 hours. At this point, the effects of media containing pooled serum from nonpregnant women (pool of normal serum) or pooled serum from pregnant women (pool of third-trimester) (both at concentrations of 10% in medium 199) were com pared to the effects of medium 199 in control incubations. Media were changed at 16, 24, 36, and 48 hours after the beginning of incubation. In another set of experiments the effects of
(mean±SEM)
Tissue and medium concentrations
PAPP-A
CJ
before and
~
after culture
(~g/g) .----------------~
term pregnancy 200
early pregnancy
Fig. 3. Mean concentrations of PAPP-A in the trophoblast before culture and total concentration of PAPP-A released in the medium and left in the trophoblast after culture. 199, PS, and NPS as in Fig. 1. * = p < 0.05; ** = p < 0.01; *** = p < 0.001.
In vitro production of PAPP-A
Volume 148 Number 1
(mean± SEM)
Tissue and medium concentrations before and ~ after culture
CJ
HPL.---------------------~ (J,Jg/g)
200
(mean:t:SEM)
15
Tissue and medium concentrations
CJ
before and
~
after culture
PAPP-A r--------------------------------. (J.Jg/g)
term pregnancy
200
100
NPS
early
200.-----------------------------~
early pregnancy
200
100
199 Fig. 4. Mean concentrations of hPL in trophoblast before cul-
ture and total concentration of hPL released in the medium and left in the trophoblast after culture. 199, PS, and NPS as in Fig. 1. *** = p < 0.001. cycloheximide (100 p,g/ml of medium) were evaluated under the same conditions. All media were stored at -20° C until assayed. Samples of all tissues (approximately 300 mg) taken before incubation and stored at -20° C served as controls for the tissues taken after 4 8 hours of culture. These tissues were homogenized in 2 ml of medium 199 with a polytron (Kinematica, Lucerne, Switzerland). Human placental lactogen (hPL) and prolactin (PrJ), used as viability indices, were measured in media and homogenates and determined with commercially available radioimmunoassay kits (CEA-CIS, Medipro, Teufen, Switzerland). PAPP-A was measured in the same samples with a solid-phase radioimmunoassay as described previously.n Results were expressed in nanograms per gram or in micrograms per gram of fresh tissue as weighed before incubation. In the media
PS
NPS
Fig. 5. Mean concentrations of PAPP-A in decidua before culture and total concentration of PAPP-A released in the medium and left in the decidua after culture. 199, PS, and NPS as in Fig. 1. *=p<0.05; **=p
to which human serum had been added, the concentration ofhPL, PrJ, and PAPP-A present in the medium before incubation was subtracted from the concentration of these proteins in the media after incubation in order to evaluate exclusively the contribution of these tissues in culture. Statistical evaluation was performed with Student's t test. Results
According to histologic examination, the tissues prepared for incubation were not contaminated with other tissues. Kinetics of PAPP-A release. PAPP-A concentrations in the medium decreased to a low level during the first 8 hours of incubation, reaching a minimum at 8 hours. This decrease was observed in trophoblast and decidua
16
Bischof et al. Am.
J.
January I, 1984 Obstet. Gynecol.
Table I. Effects of cycloheximide on total PAPP-A, hPL, and Prl release from early pregnancy trophoblast and decidua cultures (mean ± SEM, n = 4) During and after 48 hr of culture
1
Protein
Before culture
Medium 199
Decidua
PAPP-A (J.~-g/gm) Prl (ng/gm)
10.1 :!:: 2.7 432 :!:: 210
28.6 :!:: 8.2§ 2,351 :±: 550~
22.8 :!:: 2.3 1,092 :±: 416
Trophoblast
PAPP-A (J.~-g/gm) hPL (J.~-glgm)
14.7 :!:: 4.4 7.7 :!:: 0.9
41.3:!:: 10.3~ 91.2 :!:: 30.7
23.2 :!:: 4.4 38.8 :!:: 7.7
Tissue
Medium 199 plus C*
Total release was calculated by adding the concentration of proteins in the media at 2, 4, 6, 8, 16, 36, and 48 hours, expressed in micrograms per gram of tissue, plus the concentration of the same proteins found in tissues after culture. *Medium 199 plus cycloheximide, 100 J.l,glml tMedium 199 plus pregnancy-associated serum (10% v/v) :j:Medium 199 plus pregnancy-associated serum plus cycloheximide. §p < 0.05 as compared to value "before culture." lip < 0.005 as compared to value with 109 plus PS. ~p < 0.001 as compared to value "before culture." #p < 0.05 as compared to value with 199 plus PS. **p < 0.0 I as compared to value with 199 plus PS.
Table II. Net PAPP-A release during 48 hours of culture (mean ± SEM)
Trophoblast Early pregnancy Term pregnancy Decidua Early pregnancy Term pregnancy
Medium 199 alone
Medium 199 plus NPS
Medium 199 plus PS
31.2:!:: 10.0 (14) 21.3 :!:: 5.9:j: (10)
64.5 :!:: 12.6* (4) 22.8 :!:: 5.0 (4)
311.8 :!:: 78.1 t (14) 88.9 ± 13.3 (10)
27.1 :±: 4.9 (14) 50.7 ± 12.U (10)
28.6 :!:: 8.5* (4) 76.6 :!:: 32.9 (4)
78.1 :!:: 18.6t (14) 127.3 :!:: 40.9 (10)
PAPP-A release was calculated as follows: Sum of amounts present in the media (expressed in micrograms per gram of tissue) at 2, 4, 6, 8, 16, 36, and 48 hours plus concentration in the tissue after incubation minus concentration in the tissue before incubation. NPS: Non-pregnancy-associated serum. PS: pregnancy-associated serum. Numbers in parentheses are numbers of experiments. *p < 0.05. tp < 0.005. :j:p < 0.03.
cultures from early or term pregnancies (Fig. 1). From 8 hours (time at which the different media were added) and up to 48 hours, the concentration of PAPP-A in the media containing pregnancy-associated serum increased significantly (p < 0.01) for trophoblast and decidua obtained from both early and term pregnancies. No increase was observed with tissues incubated in medium 199 or in media containing non-pregnancyassociated serum. No significant difference could be detected between tissues incubated in medium 199 and those incubated in medium 199 supplemented with non-pregnancy-associated serum. The release of PAPP-A in the different culture media for early decidua as compared to term decidua was not significantly different. In contrast, PAPP-A
concentrations released in medium 199 supplemented with pregnancy-associated serum were significantly higher (p < 0.01) for early pregnancy trophoblast as compared to term pregnancy trophoblast (Fig. 1). In cultures with tissues obtained from two hysterectomies during the first trimester of pregnancy, the kinetics of PAPP-A release from decidua and trophoblast were essentially the same as previously described (Fig. 2). Myometrium cultured under the same conditions did not release measurable quantities of PAPP-A (Fig. 2). PAPP-A, hPL, and Prl. The amount of PAPP-A, hPL, and Prl present in the tissues before incubation was compared to the total amount released into the medium (expressed per gram of cultured tissue) plus
In vitro production of PAPP-A
Volume 148 Number I
(mean±SEM)
Prol. During and after 48 hr of culture Medium 199 plus PSt
Medium 199 plus C:f:
255.5 ± 40.2 1,400 ± 304
54.8 ± 15.0/1 1,029 ± 164
141.2 ± 47.0 229.0 ± 43.3
17
Tissue and medium concentrations
Cl
before and
~after
culture
r-------------------------------------, (IJg/g) term pregnancy
40
30.4 ± 6.3# 61.6 ± I 2.4**
the amount present in the tissue at the end of incubation. Irrespective of the nature of the medium in which trophoblast had been incubated, the total amount of PAPP-A released into the medium and found in tissues at the end of the culture was always significantly higher than the amount present in the tissues before culture (Fig. 3). The same observation also applies to hPL (Fig. 4). The total amount of PAPP-A in decidual tissue and media after culture was significantly higher than the amount initially present in the tissue (Fig. 5). This observation is true irrespective of the composition of the medium in which decidua was incubated. The same tendency was observed with Prl; however, the differences were not statistically significant (Fig. 6). The amounts (per gram of tissue) of PAPP-A, hPL, and Prl found in tissues and media after culture in medium 199 supplemented with cycloheximide were not significantly lower than those measured when tissues were incubated in medium 199 alone (Table I). PAPP-A and hPL concentrations after culture of trophoblast in media containing pregnancy-associated serum and cycloheximide were significantly lower (p < 0.05) than concentrations after culture in media containing pregnancy-associated serum alone (Table I). PAPP-A concentrations after culture of decidua in media containing pregnancy-associated serum and cycloheximide were significantly lower (p < 0.005) than concentrations after culture in media containing pregnancy-associated serum alone (Table I). The "net release" (in micrograms per gram) of PAPP-A was calculated by adding the amounts of PAPP-A present in the media to the concentration in the tissue after incubation minus the concentration in the tissues before culture. This net release was found to be significantly different from zero in all cases (Table II).
Comment PAPP-A is present in large quantities in the maternal circulation during pregnancy, and immunohistochemical studies 10 - 12 as well as direct measurements of PAPP-A in tissues 13 - 15 point to the trophoblast and to the decidua as potential sources of PAPP-A. The re-
199
PS
NPS
early pregnancy 40
20
199
PS
NPS
Fig. 6. Mean concentrations of Prl in decidua before culture and total concentration of Prl released in the medium and left in the decidua after culture. 199, PS, and NPS as in Fig. 1. * = p < 0.05.
suits presented here are in agreement with these suggestions and demonstrate an in vitro release of PAPP-A which is specific for decidua and trophoblast since myometrium did not release measurable quantities of PAPP-A. The fact that these tissues were alive throughout the culture period is demonstrated by the fact that they were capable of releasing other proteins, such as hPL for trophoblast and PrJ for decidua. That the described release of PAPP-A from decidua and trophoblast is not simply a leaching out from the cells is demonstrated by the fact that the concentration of PAPP-A in the medium decreased during the first 8 hours of culture and increased significantly thereafter. Furthermore, greater amounts of PAPP-A were released during incubation com pared to the amounts initially present before incubation. This strongly suggests that PAPP-A is being synthesized de novo by the decidua as well as by the trophoblast during culture. The same has been observed with hPL and Prl. The effects of cycloheximide also lead to the same conclusion. The rather large variations in PAPP-A, hPL, and
18
Bischof et al. Am.
particularly Prl "production" from one experiment to another may reflect the variability seen in the maternal circulation6 or possibly the heterogeneous population of cells present in the trophoblast as well as in the decidua. The release of PAPP-A from both tissues is statistically significant even in the absence of human serum, but the presence of pregnancy-associated serum significantly enhances the release of PAPP-A from both tissues. This effect is specific for PAPP-A and for pregnancy-associated serum since normal human serum does not exert the same effect on PAPP-A and since no further increase in hPL and Prl could be observed. Thus, pregnancy-associated serum contains a factor, not present in nonpregnant subjects, which specifically stimulates the in vitro release of PAPP-A from decidua and trophoblast, and the nature of this factor is still unknown. Although the in vitro situation cannot be strictly compared to the in vivo one, it is tempting to speculate that the lower PAPP-A production by term trophoblast as compared to early pregnancy trophoblast reflects the in vivo situation. This would imply that in late pregnancy PAPP-A is essentially of decidual origin and would explain why circulating PAPP-A levels increase in the maternal circulation at a time when placental growth levels off. 4 - 6 PAPP-A is not pregnancy specific since measurable quantities have been detected in nonpregnant women. 14 • 17 The source of PAPP-A in nonpregnant individuals is unknown, but the present results raise the question of whether PAPP-A can also be produced by nondecidualized endometrium in particular, since the endometrial PAPP-A concentration is rather high and varies with phase of the menstrual cycle. 16 It is interesting to note that trophoblast and decidua are anatomically in close contact but are genetically of different origins (trophoblast being of fetal origin and decidua of maternal origin). Both of these tissues are capable of in vitro production of a protein, PAPP-A, which has been shown to exert immunomodulatory effects.7-9 The presence of PAPP-A in these different tissues raises the question of the possible local action of PAPP-A at the fetomaternal border, where this protein could contribute to mechanisms that negate the rejection of the fetoplacental unit as an allograft. We wish to thank Mrs. C. Gruffat, M. Mendez, and Mr. G. Rus for technical assistance. REFERENCES l. Lin, T. M., Halbert, S. P., Kiefer, D., and Spellacy, W. N.:
Three pregnancy associated human plasma proteins:
2. 3.
4. 5.
6.
7. 8.
9.
10. 11. 12.
13.
14.
15.
16. 17.
January 1, 1984 Gynecol.
J. Obstet.
purification, monospecific antisera and immunological identification, Int. Arch. Allergy 47:35, 1974. Bischof, P.: Purification and characterization of pregnancy-associated plasma protein-A (PAPP-A), Arch. Gynecol. 227:315, 1979. Sutcliffe, R. G., Kukulska-Langlands, B. M., Coggins, ]. R., Hunter,]. B., and Gore, G. H.: Studies on human pregnancy-associated plasma protein-A. Purification by affinity chromatography and structural comparison with a2-macroglobulin, Biochem.]. 191:799, 1980. Lin, T. M., Halbert, S. P., and Spellacy, W. N.: Measurements of pregnancy-associated plasma proteins during human gestation, J. Clin. Invest. 54:576, 1974. Folkersen,J., Grudzinskas,J. G., Hindersson, P., Teisner, B., and Westergaard, J. G.: Pregnancy-associated plasma protein A: Circulating levels during normal pregnancy, AM. J. 0BSTET. GYNECOL. 139:910, 1981. Bischof, P., DuBerg, S., Herrmann, W. L., and Sizonenko, P. C.: Amniotic fluid and circulating levels of pregnancyassociated plasma protein-A (PAPP-A) throughout pregnancy. Comparison with other foeto-placental products, Br. J. Obstet. Gynaecol. 89:358, 1982. Bischof, P.: Pregnancy-associated plasma protein-A (PAPP-A): an inhibitor of the complement system, Placenta 2:29, 1981. Bischof, P., Lauber, K., de Wurstemberger, B., and Girard, J. P.: Inhibition oflymphocyte transformation by pregnancy-associated plasma protein-A (PAPP-A),]. Clin. Lab. Immunol. 7:61, 1982. Bischof, P., DuBerg, S., and Schindler, A. M.: Is pregnancy-associated plasma protein-A (PAPP-A) an immunomodulator during pregnancy? Placenta (Suppl.) 4:93, 1982. Lin, T. M., and Halbert, S. P.: Placental localization of human pregnancy-associated plasma proteins, Science 193:1249, 1976. Wahlstrom, T., Teisner, B., and Folkersen, J.: Tissue localization of pregnancy-associated plasma protein-A (PAPP-A) in normal placenta, Placenta 2:253, 1981. Schindler, A. M., Bordignon, P., and Bischof, P.: Immunohistochemical localization of pregnancy-associated plasma protein-A in decidua and trophoblast. Comparison with hCG and fibrin, Placenta, 1983. In press. Smith, R., Bischof, P., Hughes, G., and Klopper, A.: Studies on pregnancy-associated plasma protein-A in the third trimester of pregnancy, Br. J. Obstet. Gynaecol. 86:882, 1979. DuBerg, S., Bischof, P., Schindler, A. M., Beguin, F., Herrmann, W. L., and Sizonenko, P. C.: Tissue and plasma concentrations of pregnancy-associated plasma protein-A (PAPP-A). Comparison with other foeto-placental products, Br. J. Obstet. Gynaecol. 89:352, 1982. Bischof, P., DuBerg, S., Schindler, A.M., Obradovich, D., Wei!, A., Faigaux, R., Herrmann, W. L., and Sizonenko, P. C.: Endometrial and plasma concentrations of pregnancy-associated plasma protein-A (PAPP-A), Br. J. Obstet. Gynaecol. 89:701, 1982. Riddick, D. H., and Kusmik, W. F.: Decidua: A possible source of amniotic fluid prolactin, AM.]. 0BSTET. GvNECOL. 127:187, 1977. Bischof, P., Haenggeli, L., Sizonenko, M. T., Herrmann, W. L., and Sizonenko, P. C.: A radioimmunoassay for the measurement of pregnancy-associated plasma protein-A (PAPP-A) in humans, Bioi. Reprod. 24:1076, 1981.
Pregnancy-associated plasma protein A (PAPP-A) was discovered by Lin and co-workers in 1974 1 in the plasma of pregnant women. Subsequent purification and characterization 2 • '1 demonstrated that PAPP-A is a macromolecular glycoprotein (molecular weight of 800,000) of dimeric structure, each monomer being composed of two apparently identical subunits with a molecular weight of 200,000. 2 • '1 PAPP-A appears in increasing concentrations as pregnancy progresses, 4 - 6 almost exclusively in the maternal circulation. Although its biologic role is still unknown, in vitro experiments suggest an immunomodulatory function. 7 - 9 By the use of immunohistochemical techniques,
From the Division of Clinical Biology of Growth and Reproduction, the Department of Obstetrics and Gynecology, and the Centre de Cytologie et de Depistage du Cancer, University of Geneva. Supported by Grants No. 3.931.0.80 and No. 3.909.0.82 from the Swiss National Fund for Scientific Research. Received for publication April 7, 1983. Accepted August 1, 1983. Reprint requests: Dr. P. Bischof, Laboratoire d'hormonologie, Maternite, 1211 Geneva 4, Switzerland.
PAPP-A could be localized on the syncytiotrophoblast and in decidual cells. 10 - 12 Direct measurements of PAPP-A in different fetomaternal tissues 1'1• 14 also revealed its presence in trophoblast and decidua as well as in endometrium.I" Immunohistochemical localization and direct measurements of PAPP-A thus suggest a possible decidual as well as a trophoblastic origin for PAPP-A. The present study was undertaken to determine whether these tissues are capable of producing PAPP-A in vitro. Material and methods
Placentas and membranes were collected from 10 uncomplicated term pregnancies terminated by spontaneous labor and delivery or cesarean section. Trophoblast and decidua were obtained from I 4 patients undergoing surgical termination of pregnancy between weeks 7 and 12 of gestation. Finally, myometrium, decidua, and trophoblast were also obtained from two patients undergoing hysterectomy in the first trimester of pregnancy for reasons other than malignancy. Term placentas were cut midway between the basal and chorionic plates, and the tissue was cut in small pieces. Term decidua was peeled from the fetal mem13
14 Bischof et al.
January 1, 1984 Am. ]. Obstet. Gynecol.
PAPP-A l~g/g)
mean
I TROPHOBLAST I
:t
PAPP-A
SEM
l~g/g)
100,-------:;.,.';==:::;------" early
•••• ••
'•
pre 9 nancy·······r·
- - - 199
•••••}""
b
200
T = Trophoblast
- - - - 199. NPs ···········199•PS
160
0 =Decidua M = Myometrium
so
·i/
term pregnancy
/
I
,-.....-.......l······ .
1\
•
120
~~·· O~~~~~~=:=-~t=-~~====;===~T~~~ I DECIDUA I 100 early pregnancy
_I110)
PAPP-A (~g/g)
{10)
40
T
term pregnancy
0
l . .) . . .
\
so
0
\/- /
I .1••••• .J"" •••••••• J••••••
~ I
••
T
16
24
T
36
J(4)
48
0
8
16
24
141
(10)
36
0
12
24
36
M 24
48
36
48
hours of incubation
',J.,
~-L-~
01.)
T
80
*"'-"'*-=-=-=;~~,-=;Jiill41
0
'l
a
80
48
Fig. 2. Release of PAPP-A in culture media expressed in micrograms per gram of tissue ex planted (tissues obtained from two hysterectomies during pregnancy). a: In medium 199 alone; b: in medium 199 supplemented by pregnancy-associated serum.
hours of incubation
Fig. 1. Release of PAPP-A in culture media expressed in micrograms per gram of tissue explanted. 199: Medium 199 alone; NPS: non-pregnancy-associated serum (pool of normal female serum); PS: pregnancy serum (pool of third-trimester pregnancy serum).
branes according to the technique of Riddick and Kusmik. 16 Decidua and trophoblast from early pregnancy aspirates or products of hysterectomy were separated by hand under sterile conditions. Routine histologic sections were taken from all tissues for evaluation of possible contamination by other tissues. All material was collectd under sterile conditions and treated as quickly as possible after delivery or operation (within 20 minutes). Tissues were minced and washed in phosphatebuffered saline (PBS; 0.05M phosphate and 0.2M sodium chloride, pH 7.4) containing 2.5% penicillin and 2.5% streptomycin, and approximately 300 mg of tissue was incubated separately in Ehrlenmeyer flasks containing 2 ml of medium 199 (Difco, Detroit, Michigan). Incubation was carried out in a shaking water bath at 37° C with a 5% carbon dioxide and 95% oxygen atmosphere. Medium 199 was changed at 2, 4, 6, and 8 hours. At this point, the effects of media containing pooled serum from nonpregnant women (pool of normal serum) or pooled serum from pregnant women (pool of third-trimester) (both at concentrations of 10% in medium 199) were com pared to the effects of medium 199 in control incubations. Media were changed at 16, 24, 36, and 48 hours after the beginning of incubation. In another set of experiments the effects of
(mean±SEM)
Tissue and medium concentrations
PAPP-A
CJ
before and
~
after culture
(~g/g) .----------------~
term pregnancy 200
early pregnancy
Fig. 3. Mean concentrations of PAPP-A in the trophoblast before culture and total concentration of PAPP-A released in the medium and left in the trophoblast after culture. 199, PS, and NPS as in Fig. 1. * = p < 0.05; ** = p < 0.01; *** = p < 0.001.
In vitro production of PAPP-A
Volume 148 Number 1
(mean± SEM)
Tissue and medium concentrations before and ~ after culture
CJ
HPL.---------------------~ (J,Jg/g)
200
(mean:t:SEM)
15
Tissue and medium concentrations
CJ
before and
~
after culture
PAPP-A r--------------------------------. (J.Jg/g)
term pregnancy
200
100
NPS
early
200.-----------------------------~
early pregnancy
200
100
199 Fig. 4. Mean concentrations of hPL in trophoblast before cul-
ture and total concentration of hPL released in the medium and left in the trophoblast after culture. 199, PS, and NPS as in Fig. 1. *** = p < 0.001. cycloheximide (100 p,g/ml of medium) were evaluated under the same conditions. All media were stored at -20° C until assayed. Samples of all tissues (approximately 300 mg) taken before incubation and stored at -20° C served as controls for the tissues taken after 4 8 hours of culture. These tissues were homogenized in 2 ml of medium 199 with a polytron (Kinematica, Lucerne, Switzerland). Human placental lactogen (hPL) and prolactin (PrJ), used as viability indices, were measured in media and homogenates and determined with commercially available radioimmunoassay kits (CEA-CIS, Medipro, Teufen, Switzerland). PAPP-A was measured in the same samples with a solid-phase radioimmunoassay as described previously.n Results were expressed in nanograms per gram or in micrograms per gram of fresh tissue as weighed before incubation. In the media
PS
NPS
Fig. 5. Mean concentrations of PAPP-A in decidua before culture and total concentration of PAPP-A released in the medium and left in the decidua after culture. 199, PS, and NPS as in Fig. 1. *=p<0.05; **=p
to which human serum had been added, the concentration ofhPL, PrJ, and PAPP-A present in the medium before incubation was subtracted from the concentration of these proteins in the media after incubation in order to evaluate exclusively the contribution of these tissues in culture. Statistical evaluation was performed with Student's t test. Results
According to histologic examination, the tissues prepared for incubation were not contaminated with other tissues. Kinetics of PAPP-A release. PAPP-A concentrations in the medium decreased to a low level during the first 8 hours of incubation, reaching a minimum at 8 hours. This decrease was observed in trophoblast and decidua
16
Bischof et al. Am.
J.
January I, 1984 Obstet. Gynecol.
Table I. Effects of cycloheximide on total PAPP-A, hPL, and Prl release from early pregnancy trophoblast and decidua cultures (mean ± SEM, n = 4) During and after 48 hr of culture
1
Protein
Before culture
Medium 199
Decidua
PAPP-A (J.~-g/gm) Prl (ng/gm)
10.1 :!:: 2.7 432 :!:: 210
28.6 :!:: 8.2§ 2,351 :±: 550~
22.8 :!:: 2.3 1,092 :±: 416
Trophoblast
PAPP-A (J.~-g/gm) hPL (J.~-glgm)
14.7 :!:: 4.4 7.7 :!:: 0.9
41.3:!:: 10.3~ 91.2 :!:: 30.7
23.2 :!:: 4.4 38.8 :!:: 7.7
Tissue
Medium 199 plus C*
Total release was calculated by adding the concentration of proteins in the media at 2, 4, 6, 8, 16, 36, and 48 hours, expressed in micrograms per gram of tissue, plus the concentration of the same proteins found in tissues after culture. *Medium 199 plus cycloheximide, 100 J.l,glml tMedium 199 plus pregnancy-associated serum (10% v/v) :j:Medium 199 plus pregnancy-associated serum plus cycloheximide. §p < 0.05 as compared to value "before culture." lip < 0.005 as compared to value with 109 plus PS. ~p < 0.001 as compared to value "before culture." #p < 0.05 as compared to value with 199 plus PS. **p < 0.0 I as compared to value with 199 plus PS.
Table II. Net PAPP-A release during 48 hours of culture (mean ± SEM)
Trophoblast Early pregnancy Term pregnancy Decidua Early pregnancy Term pregnancy
Medium 199 alone
Medium 199 plus NPS
Medium 199 plus PS
31.2:!:: 10.0 (14) 21.3 :!:: 5.9:j: (10)
64.5 :!:: 12.6* (4) 22.8 :!:: 5.0 (4)
311.8 :!:: 78.1 t (14) 88.9 ± 13.3 (10)
27.1 :±: 4.9 (14) 50.7 ± 12.U (10)
28.6 :!:: 8.5* (4) 76.6 :!:: 32.9 (4)
78.1 :!:: 18.6t (14) 127.3 :!:: 40.9 (10)
PAPP-A release was calculated as follows: Sum of amounts present in the media (expressed in micrograms per gram of tissue) at 2, 4, 6, 8, 16, 36, and 48 hours plus concentration in the tissue after incubation minus concentration in the tissue before incubation. NPS: Non-pregnancy-associated serum. PS: pregnancy-associated serum. Numbers in parentheses are numbers of experiments. *p < 0.05. tp < 0.005. :j:p < 0.03.
cultures from early or term pregnancies (Fig. 1). From 8 hours (time at which the different media were added) and up to 48 hours, the concentration of PAPP-A in the media containing pregnancy-associated serum increased significantly (p < 0.01) for trophoblast and decidua obtained from both early and term pregnancies. No increase was observed with tissues incubated in medium 199 or in media containing non-pregnancyassociated serum. No significant difference could be detected between tissues incubated in medium 199 and those incubated in medium 199 supplemented with non-pregnancy-associated serum. The release of PAPP-A in the different culture media for early decidua as compared to term decidua was not significantly different. In contrast, PAPP-A
concentrations released in medium 199 supplemented with pregnancy-associated serum were significantly higher (p < 0.01) for early pregnancy trophoblast as compared to term pregnancy trophoblast (Fig. 1). In cultures with tissues obtained from two hysterectomies during the first trimester of pregnancy, the kinetics of PAPP-A release from decidua and trophoblast were essentially the same as previously described (Fig. 2). Myometrium cultured under the same conditions did not release measurable quantities of PAPP-A (Fig. 2). PAPP-A, hPL, and Prl. The amount of PAPP-A, hPL, and Prl present in the tissues before incubation was compared to the total amount released into the medium (expressed per gram of cultured tissue) plus
In vitro production of PAPP-A
Volume 148 Number I
(mean±SEM)
Prol. During and after 48 hr of culture Medium 199 plus PSt
Medium 199 plus C:f:
255.5 ± 40.2 1,400 ± 304
54.8 ± 15.0/1 1,029 ± 164
141.2 ± 47.0 229.0 ± 43.3
17
Tissue and medium concentrations
Cl
before and
~after
culture
r-------------------------------------, (IJg/g) term pregnancy
40
30.4 ± 6.3# 61.6 ± I 2.4**
the amount present in the tissue at the end of incubation. Irrespective of the nature of the medium in which trophoblast had been incubated, the total amount of PAPP-A released into the medium and found in tissues at the end of the culture was always significantly higher than the amount present in the tissues before culture (Fig. 3). The same observation also applies to hPL (Fig. 4). The total amount of PAPP-A in decidual tissue and media after culture was significantly higher than the amount initially present in the tissue (Fig. 5). This observation is true irrespective of the composition of the medium in which decidua was incubated. The same tendency was observed with Prl; however, the differences were not statistically significant (Fig. 6). The amounts (per gram of tissue) of PAPP-A, hPL, and Prl found in tissues and media after culture in medium 199 supplemented with cycloheximide were not significantly lower than those measured when tissues were incubated in medium 199 alone (Table I). PAPP-A and hPL concentrations after culture of trophoblast in media containing pregnancy-associated serum and cycloheximide were significantly lower (p < 0.05) than concentrations after culture in media containing pregnancy-associated serum alone (Table I). PAPP-A concentrations after culture of decidua in media containing pregnancy-associated serum and cycloheximide were significantly lower (p < 0.005) than concentrations after culture in media containing pregnancy-associated serum alone (Table I). The "net release" (in micrograms per gram) of PAPP-A was calculated by adding the amounts of PAPP-A present in the media to the concentration in the tissue after incubation minus the concentration in the tissues before culture. This net release was found to be significantly different from zero in all cases (Table II).
Comment PAPP-A is present in large quantities in the maternal circulation during pregnancy, and immunohistochemical studies 10 - 12 as well as direct measurements of PAPP-A in tissues 13 - 15 point to the trophoblast and to the decidua as potential sources of PAPP-A. The re-
199
PS
NPS
early pregnancy 40
20
199
PS
NPS
Fig. 6. Mean concentrations of Prl in decidua before culture and total concentration of Prl released in the medium and left in the decidua after culture. 199, PS, and NPS as in Fig. 1. * = p < 0.05.
suits presented here are in agreement with these suggestions and demonstrate an in vitro release of PAPP-A which is specific for decidua and trophoblast since myometrium did not release measurable quantities of PAPP-A. The fact that these tissues were alive throughout the culture period is demonstrated by the fact that they were capable of releasing other proteins, such as hPL for trophoblast and PrJ for decidua. That the described release of PAPP-A from decidua and trophoblast is not simply a leaching out from the cells is demonstrated by the fact that the concentration of PAPP-A in the medium decreased during the first 8 hours of culture and increased significantly thereafter. Furthermore, greater amounts of PAPP-A were released during incubation com pared to the amounts initially present before incubation. This strongly suggests that PAPP-A is being synthesized de novo by the decidua as well as by the trophoblast during culture. The same has been observed with hPL and Prl. The effects of cycloheximide also lead to the same conclusion. The rather large variations in PAPP-A, hPL, and
18
Bischof et al. Am.
particularly Prl "production" from one experiment to another may reflect the variability seen in the maternal circulation6 or possibly the heterogeneous population of cells present in the trophoblast as well as in the decidua. The release of PAPP-A from both tissues is statistically significant even in the absence of human serum, but the presence of pregnancy-associated serum significantly enhances the release of PAPP-A from both tissues. This effect is specific for PAPP-A and for pregnancy-associated serum since normal human serum does not exert the same effect on PAPP-A and since no further increase in hPL and Prl could be observed. Thus, pregnancy-associated serum contains a factor, not present in nonpregnant subjects, which specifically stimulates the in vitro release of PAPP-A from decidua and trophoblast, and the nature of this factor is still unknown. Although the in vitro situation cannot be strictly compared to the in vivo one, it is tempting to speculate that the lower PAPP-A production by term trophoblast as compared to early pregnancy trophoblast reflects the in vivo situation. This would imply that in late pregnancy PAPP-A is essentially of decidual origin and would explain why circulating PAPP-A levels increase in the maternal circulation at a time when placental growth levels off. 4 - 6 PAPP-A is not pregnancy specific since measurable quantities have been detected in nonpregnant women. 14 • 17 The source of PAPP-A in nonpregnant individuals is unknown, but the present results raise the question of whether PAPP-A can also be produced by nondecidualized endometrium in particular, since the endometrial PAPP-A concentration is rather high and varies with phase of the menstrual cycle. 16 It is interesting to note that trophoblast and decidua are anatomically in close contact but are genetically of different origins (trophoblast being of fetal origin and decidua of maternal origin). Both of these tissues are capable of in vitro production of a protein, PAPP-A, which has been shown to exert immunomodulatory effects.7-9 The presence of PAPP-A in these different tissues raises the question of the possible local action of PAPP-A at the fetomaternal border, where this protein could contribute to mechanisms that negate the rejection of the fetoplacental unit as an allograft. We wish to thank Mrs. C. Gruffat, M. Mendez, and Mr. G. Rus for technical assistance. REFERENCES l. Lin, T. M., Halbert, S. P., Kiefer, D., and Spellacy, W. N.:
Three pregnancy associated human plasma proteins:
2. 3.
4. 5.
6.
7. 8.
9.
10. 11. 12.
13.
14.
15.
16. 17.
January 1, 1984 Gynecol.
J. Obstet.
purification, monospecific antisera and immunological identification, Int. Arch. Allergy 47:35, 1974. Bischof, P.: Purification and characterization of pregnancy-associated plasma protein-A (PAPP-A), Arch. Gynecol. 227:315, 1979. Sutcliffe, R. G., Kukulska-Langlands, B. M., Coggins, ]. R., Hunter,]. B., and Gore, G. H.: Studies on human pregnancy-associated plasma protein-A. Purification by affinity chromatography and structural comparison with a2-macroglobulin, Biochem.]. 191:799, 1980. Lin, T. M., Halbert, S. P., and Spellacy, W. N.: Measurements of pregnancy-associated plasma proteins during human gestation, J. Clin. Invest. 54:576, 1974. Folkersen,J., Grudzinskas,J. G., Hindersson, P., Teisner, B., and Westergaard, J. G.: Pregnancy-associated plasma protein A: Circulating levels during normal pregnancy, AM. J. 0BSTET. GYNECOL. 139:910, 1981. Bischof, P., DuBerg, S., Herrmann, W. L., and Sizonenko, P. C.: Amniotic fluid and circulating levels of pregnancyassociated plasma protein-A (PAPP-A) throughout pregnancy. Comparison with other foeto-placental products, Br. J. Obstet. Gynaecol. 89:358, 1982. Bischof, P.: Pregnancy-associated plasma protein-A (PAPP-A): an inhibitor of the complement system, Placenta 2:29, 1981. Bischof, P., Lauber, K., de Wurstemberger, B., and Girard, J. P.: Inhibition oflymphocyte transformation by pregnancy-associated plasma protein-A (PAPP-A),]. Clin. Lab. Immunol. 7:61, 1982. Bischof, P., DuBerg, S., and Schindler, A. M.: Is pregnancy-associated plasma protein-A (PAPP-A) an immunomodulator during pregnancy? Placenta (Suppl.) 4:93, 1982. Lin, T. M., and Halbert, S. P.: Placental localization of human pregnancy-associated plasma proteins, Science 193:1249, 1976. Wahlstrom, T., Teisner, B., and Folkersen, J.: Tissue localization of pregnancy-associated plasma protein-A (PAPP-A) in normal placenta, Placenta 2:253, 1981. Schindler, A. M., Bordignon, P., and Bischof, P.: Immunohistochemical localization of pregnancy-associated plasma protein-A in decidua and trophoblast. Comparison with hCG and fibrin, Placenta, 1983. In press. Smith, R., Bischof, P., Hughes, G., and Klopper, A.: Studies on pregnancy-associated plasma protein-A in the third trimester of pregnancy, Br. J. Obstet. Gynaecol. 86:882, 1979. DuBerg, S., Bischof, P., Schindler, A. M., Beguin, F., Herrmann, W. L., and Sizonenko, P. C.: Tissue and plasma concentrations of pregnancy-associated plasma protein-A (PAPP-A). Comparison with other foeto-placental products, Br. J. Obstet. Gynaecol. 89:352, 1982. Bischof, P., DuBerg, S., Schindler, A.M., Obradovich, D., Wei!, A., Faigaux, R., Herrmann, W. L., and Sizonenko, P. C.: Endometrial and plasma concentrations of pregnancy-associated plasma protein-A (PAPP-A), Br. J. Obstet. Gynaecol. 89:701, 1982. Riddick, D. H., and Kusmik, W. F.: Decidua: A possible source of amniotic fluid prolactin, AM.]. 0BSTET. GvNECOL. 127:187, 1977. Bischof, P., Haenggeli, L., Sizonenko, M. T., Herrmann, W. L., and Sizonenko, P. C.: A radioimmunoassay for the measurement of pregnancy-associated plasma protein-A (PAPP-A) in humans, Bioi. Reprod. 24:1076, 1981.