0013-7227/91/1291-0022$03.00/0 Endocrinology Copyright © 1991 by The Endocrine Society
Vol. 129, No. 1 Printed in U.S.A.
Transforming Growth Factor /?1 Inhibits Placental Differentiation and Human Chorionic Gonadotropin and Human Placental Lactogen Secretion DONALD W. MORRISH, DAMYANTI BHARDWAJ, AND MARIA TERESA PARAS University of Alberta, Edmonton, Alberta, Canada T6G 2S2
ABSTRACT. Previously, no inhibitors of placental differentiation have been described. In this study, we determined the effect of transforming growth factor /31 (TGF/31) on cytotrophoblast differentiation. Monolayer cultures of pure cytotrophoblasts were exposed to 0.001-10 ng/ml TGF/31 with and without the presence of 10 ng/ml epidermal growth factor (EGF), an inducer of placental differentiation. Over 7 days of culture, in 11 separate experiments, phase contrast microscopy demonstrated marked inhibition of EGF-induced syncytial formation by TGF/31. Basal human (h)CG and h-placental lactogen (PL) release were reduced compared to control by fractions of 0.75 (TGF /31/control) and 0.54, respectively. EGF alone induced
fractional (EGF/control) increases in hCG and hPL release of 2.46 and 2.68, respectively. However, this stimulation was significantly inhibited by 10 ng/ml TGF/31. Dose-response studies showed that maximal TGF/31 inhibition of EGF-stimulated hormone secretion occurred at 0.1 ng/ml or more TGF/31. Partial differentiation (syncytium formation) occurred despite the presence of TGF/31, suggesting a portion of cytotrophoblasts were committed to differentiation at the time of culture. We conclude that TGF/31 acts as a major inhibitor of trophoblast differentiation and concomitant peptide hormone secretion. (Endocrinology 129: 22-26, 1991)
P
LACENTAL differentiation occurs by the fusion of cytotrophoblasts to form a syncytium (1, 2). Few regulators of this process are known. Previously, we have shown that epidermal growth factor (EGF) (3), colonystimulating factor (CSF)-l and granulocyte-macrophage colony-stimulating factor (GM-CSF) (4) all induce differentiation, with concomitant increased secretion of the syncytial hormones human (h) CG and h-placental lactogen (PL). However, no inhibitors of differentiation have been previously described. A possible candidate is transforming growth factor /31 (TGF/31). TGF/31 is a multifunctional growth factor present in many cells and has both positive and negative effects on cell proliferation and differentiation (5, 6). At least five different molecular species are recognized, but most biological studies have used TGF/31. It induces differentiation of keratinocytes (7-9), ovarian thecal cells (10, 11) and bronchial epithelium (12), but can also inhibit differentiation in other cells, including myoblasts (13, 14), adipocytes (15), and early hematopoietic progenitor cells (16-18). In cartilage, TGF/31 can act both as an inducer (19, 20) and an inhibitor (20) of differentiation. Thus, because TGF/31 has potential action as an inhibitor of
differentiation, we wished to determine its effect on trophoblast differentiation.
Materials and Methods Term human placentae were obtained from normal deliveries and monolayer cultures of cytotrophoblasts prepared as previously described (3). In brief, pieces of placenta were subjected to 8 sequential 10-min treatments with 0.25% trypsin-10 U/ml DNAse I in Ringer-bicarbonate buffer. Cells were collected after the 5th-8th treatments, pooled, centrifuged at 300 X g, and resuspended in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal calf serum, 50 U/ml penicillin, 50 Mg/ml streptomycin, and 0.2raMglutamine. 1 X 106 cells per well were plated in 24-well plates and allowed to attach for 2 h, after which the medium was changed to DMEM without serum for the remainder of the experiment. Each test condition was conducted in triplicate or quadruplicate (for the doseresponse curve). Medium was changed daily and frozen at -20 C for subsequent hCG and hPL assay. After each medium change, plates received 10 ng/ml EGF, 0.001-10 ng/ml TGF/31, combinations of EGF and TGF/31, or carrier (final concentration 0.001% BSA or 0.04 mM HCl-0.001% BSA). Cells were photographed in situ using a Leitz (Wild Leitz Canada Ltd, Ottawa, Canada) phase contrast microscope. Purity of cytotrophoblast cell preparations was determined as previously described (3) and demonstrated over 98% cytotrophoblast cells. hCG and hPL RIA were performed as previously described (3, 21). Each datum point in Figs. 2 and 3 was the mean of determinations of samples from 11 separate placental experi-
Received December 7, 1990. Address all correspondence and requests for reprints to: Dr. D. W. Morrish, 362 Clinical Wing, Heritage Medical Research Centre, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada T6G 2S2.
22
TGF/31 INHIBITS PLACENTAL DIFFERENTIATION
23
merits. The value for an individual placenta was the mean of determinations of samples from triplicate wells. The doseresponse curve (see Fig. 4) was performed using quadruplicate wells for each dose. Each datum point in Fig. 4 was the mean ± SEM of the average hCG or hPL concentration (expressed as a ratio of control) over days 2-7 of culture. Integrated hormone secretion was calculated by adding the value of each daily hormone determination, since these represent a cumulative 24h concentration. hCG and hPL secretion were used as indicators of trophoblast differentiation. Trypsin was obtained from ICN Biomedicals Canada Ltd. (Montreal, Quebec, Canada); DNAse I, DMEM, EDTA, penicillin-streptomycin, glutamine, and mouse EGF from Sigma Chemical Co. (St. Louis, MO); 24-well plates from Corning Glass Co. (Corning, NY); BSA from Miles Laboratories. (Elkhart, IN). TGF/31 extracted from human platelets was obtained from R & D Systems Inc. (Minneapolis, MN). Due to the wide variations occurring in hormone secretion among placentas, the data were normalized by using ratios (fractional or fold) stimulation or inhibition compared to controls. Because the data were not normally distributed, nonparametric tests of significance (Mann-Whitney, Wilcoxon) were employed.
Results Experiments using 11 separate placentae were performed. Phase contrast microscopy of cultured cells showed that control cells displayed slight spontaneous syncytia formation over 7 days (Fig. ID). During this period, there was a concomitant gradual rise and then decline in hCG and hPL concentrations (see legends for Figs. 2 and 3), as we have previously noted in this culture system (3). In comparison, EGF induced formation of large areas of syncytium after 7 days in culture (Fig. 1A). This effect was largely but not completely inhibited by concomitant addition of TGF/31 (Fig. IB). TGF/31 alone largely inhibited syncytium formation up to 7 days in culture in all experiments, and a representative example is shown in Fig. 1C. Compared to EGF-treated cells, TGF/31 -treated cells showed little syncytial formation throughout the culture period, similar to controls. Controls (Fig. ID), EGF plus TGF/31 (Fig. IB), and TGF/31 alone (Fig. 1C) thus all demonstrated qualitatively similar morphological changes. Before and during formation of syncytium, TGF/31 significantly inhibited basal hCG (Fig. 2) and hPL (Fig. 3) secretion after a lag phase of 1 day for hCG and 3 days for hPL (P < 0.005, P < 0.025, respectively). TGF/31 reduced hCG and hPL secretion to a fraction of 0.54 and 0.75 of their controls, respectively (Figs. 2 and 3). This effect was noted in all 11 separate placental experiments. In this culture system, hCG and hPL levels gradually rise during this culture period, as we have described (3, 21). The reduction in hCG and hPL secretion by TGF/31 thus represented a blunting of this normal rise by the
FIG. 1. Phase-contrast photomicrograph of trophoblast cells in culture after 6 days exposure to (A) 10 ng/ml EGF; composite of two representative areas, (B) 10 ng/ml EGF plus 10 ng/ml TGF/31, (C) 10 ng/ ml TGF/31, and (D) control; composite of two representative areas. Large syncytial islands developed in EGF-treated cells (A), but this differentiated state was mostly inhibited by concomitant addition of TGF/31 (B). Plates receiving TGF/31 alone were similar (C). Controls (D) showed little syncytial formation and were qualitatively similar to EGF plus TGF/31 (B), and to TGF/31 alone (C). s, Syncytia; C, cytotrophoblast. All magnifications X125.
fractional amounts given in Figs. 2 and 3. As previously described (3, 21), EGF significantly increased hCG and hPL secretion 2.68- and 2.46-fold after day 1 and day 4, respectively (Figs. 2 and 3; P < 0.001, P < 0.005, respectively). The later stimulation during the culture period of hPL secretion by EGF compared to that of hCG has been noted previously in our system (3, 21). EGF-induced hormone secretion was significantly inhibited by 10 ng/ ml TGF/31 (P < 0.001 for integrated secretion on days 2-7 for hCG; P < 0.01 for integrated secretion on days 3-7 for hPL) and not different from control. TGF/31 reduced the EGF-induced increases by as much as 0.79 (hCG) and 0.64 (hPL) times the control value. TGF/31 thus prevented EGF from inducing increases in hormone secretion and maintained the levels similar to those of controls (which gradually rise during the culture period, as noted above). Dose-response studies demonstrated that significant inhibition of EGF-stimulated hCG and hPL secretion (P < 0.005, P < 0.025, respectively) occurred with 0.1 ng/ml or more TGF/31 (Fig. 4). Despite the presence of TGF/31, a small degree of differentiation continued to occur as noted in Fig. 1. Control plates also demonstrated some syncytia development by 6-7 days in
TGF01 INHIBITS PLACENTAL DIFFERENTIATION
24 5T
O • 4-- A
OEGF/Control •TGF01/Control A(EGF + TGF/31 )/Control
4T hCG
ESS hPL
o
o
Endo •1991 Vol 129 • N o l
(J
'O 2--
T
T
T
^
A
1-
8 2+
§
AH
0.001 0.01 0.1 1 10 TGF/3 Dose (ng/ml)
Doy of Culture FIG. 2. Effects of EGF and TGF/31 on hCG secretion. The fractional changes are the mean ± SEM of 11 separate placental experiments. Ten nanograms per ml EGF induced up to a 2.68 fractional increase in hCG secretion (P < 0.001). Ten nanograms per ml TGF/31 alone inhibited hCG secretion after 1 day in culture, reducing basal secretion fractionally to 0.54 of control (P < 0.005). Ten nanograms per ml TGF/31 added concomitantly with 10 ng/ml EGF significantly inhibited EGF-induced hCG secretion (P < 0.001) to levels similar to that of control values. hCG concentrations for control cells were 31.2 ± 27.0, 22.4 ± 13.5, 35.8 ± 12.7, 54.3 ± 18.3, 49.4 ± 14.4, 36.1 ± 12.8, and 36.3 ± 15.1 ng/ml (mean ± SEM) on culture days 1-7, respectively. 5T 4--
O
OEGF/Control
•
•TGF/31/Control
A — A ( E G F + TGF/31 )/Control 3--
1
2-
1"
0
T /
.6'
• =
Vol. 129, No. 1 Printed in U.S.A.
Transforming Growth Factor /?1 Inhibits Placental Differentiation and Human Chorionic Gonadotropin and Human Placental Lactogen Secretion DONALD W. MORRISH, DAMYANTI BHARDWAJ, AND MARIA TERESA PARAS University of Alberta, Edmonton, Alberta, Canada T6G 2S2
ABSTRACT. Previously, no inhibitors of placental differentiation have been described. In this study, we determined the effect of transforming growth factor /31 (TGF/31) on cytotrophoblast differentiation. Monolayer cultures of pure cytotrophoblasts were exposed to 0.001-10 ng/ml TGF/31 with and without the presence of 10 ng/ml epidermal growth factor (EGF), an inducer of placental differentiation. Over 7 days of culture, in 11 separate experiments, phase contrast microscopy demonstrated marked inhibition of EGF-induced syncytial formation by TGF/31. Basal human (h)CG and h-placental lactogen (PL) release were reduced compared to control by fractions of 0.75 (TGF /31/control) and 0.54, respectively. EGF alone induced
fractional (EGF/control) increases in hCG and hPL release of 2.46 and 2.68, respectively. However, this stimulation was significantly inhibited by 10 ng/ml TGF/31. Dose-response studies showed that maximal TGF/31 inhibition of EGF-stimulated hormone secretion occurred at 0.1 ng/ml or more TGF/31. Partial differentiation (syncytium formation) occurred despite the presence of TGF/31, suggesting a portion of cytotrophoblasts were committed to differentiation at the time of culture. We conclude that TGF/31 acts as a major inhibitor of trophoblast differentiation and concomitant peptide hormone secretion. (Endocrinology 129: 22-26, 1991)
P
LACENTAL differentiation occurs by the fusion of cytotrophoblasts to form a syncytium (1, 2). Few regulators of this process are known. Previously, we have shown that epidermal growth factor (EGF) (3), colonystimulating factor (CSF)-l and granulocyte-macrophage colony-stimulating factor (GM-CSF) (4) all induce differentiation, with concomitant increased secretion of the syncytial hormones human (h) CG and h-placental lactogen (PL). However, no inhibitors of differentiation have been previously described. A possible candidate is transforming growth factor /31 (TGF/31). TGF/31 is a multifunctional growth factor present in many cells and has both positive and negative effects on cell proliferation and differentiation (5, 6). At least five different molecular species are recognized, but most biological studies have used TGF/31. It induces differentiation of keratinocytes (7-9), ovarian thecal cells (10, 11) and bronchial epithelium (12), but can also inhibit differentiation in other cells, including myoblasts (13, 14), adipocytes (15), and early hematopoietic progenitor cells (16-18). In cartilage, TGF/31 can act both as an inducer (19, 20) and an inhibitor (20) of differentiation. Thus, because TGF/31 has potential action as an inhibitor of
differentiation, we wished to determine its effect on trophoblast differentiation.
Materials and Methods Term human placentae were obtained from normal deliveries and monolayer cultures of cytotrophoblasts prepared as previously described (3). In brief, pieces of placenta were subjected to 8 sequential 10-min treatments with 0.25% trypsin-10 U/ml DNAse I in Ringer-bicarbonate buffer. Cells were collected after the 5th-8th treatments, pooled, centrifuged at 300 X g, and resuspended in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal calf serum, 50 U/ml penicillin, 50 Mg/ml streptomycin, and 0.2raMglutamine. 1 X 106 cells per well were plated in 24-well plates and allowed to attach for 2 h, after which the medium was changed to DMEM without serum for the remainder of the experiment. Each test condition was conducted in triplicate or quadruplicate (for the doseresponse curve). Medium was changed daily and frozen at -20 C for subsequent hCG and hPL assay. After each medium change, plates received 10 ng/ml EGF, 0.001-10 ng/ml TGF/31, combinations of EGF and TGF/31, or carrier (final concentration 0.001% BSA or 0.04 mM HCl-0.001% BSA). Cells were photographed in situ using a Leitz (Wild Leitz Canada Ltd, Ottawa, Canada) phase contrast microscope. Purity of cytotrophoblast cell preparations was determined as previously described (3) and demonstrated over 98% cytotrophoblast cells. hCG and hPL RIA were performed as previously described (3, 21). Each datum point in Figs. 2 and 3 was the mean of determinations of samples from 11 separate placental experi-
Received December 7, 1990. Address all correspondence and requests for reprints to: Dr. D. W. Morrish, 362 Clinical Wing, Heritage Medical Research Centre, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada T6G 2S2.
22
TGF/31 INHIBITS PLACENTAL DIFFERENTIATION
23
merits. The value for an individual placenta was the mean of determinations of samples from triplicate wells. The doseresponse curve (see Fig. 4) was performed using quadruplicate wells for each dose. Each datum point in Fig. 4 was the mean ± SEM of the average hCG or hPL concentration (expressed as a ratio of control) over days 2-7 of culture. Integrated hormone secretion was calculated by adding the value of each daily hormone determination, since these represent a cumulative 24h concentration. hCG and hPL secretion were used as indicators of trophoblast differentiation. Trypsin was obtained from ICN Biomedicals Canada Ltd. (Montreal, Quebec, Canada); DNAse I, DMEM, EDTA, penicillin-streptomycin, glutamine, and mouse EGF from Sigma Chemical Co. (St. Louis, MO); 24-well plates from Corning Glass Co. (Corning, NY); BSA from Miles Laboratories. (Elkhart, IN). TGF/31 extracted from human platelets was obtained from R & D Systems Inc. (Minneapolis, MN). Due to the wide variations occurring in hormone secretion among placentas, the data were normalized by using ratios (fractional or fold) stimulation or inhibition compared to controls. Because the data were not normally distributed, nonparametric tests of significance (Mann-Whitney, Wilcoxon) were employed.
Results Experiments using 11 separate placentae were performed. Phase contrast microscopy of cultured cells showed that control cells displayed slight spontaneous syncytia formation over 7 days (Fig. ID). During this period, there was a concomitant gradual rise and then decline in hCG and hPL concentrations (see legends for Figs. 2 and 3), as we have previously noted in this culture system (3). In comparison, EGF induced formation of large areas of syncytium after 7 days in culture (Fig. 1A). This effect was largely but not completely inhibited by concomitant addition of TGF/31 (Fig. IB). TGF/31 alone largely inhibited syncytium formation up to 7 days in culture in all experiments, and a representative example is shown in Fig. 1C. Compared to EGF-treated cells, TGF/31 -treated cells showed little syncytial formation throughout the culture period, similar to controls. Controls (Fig. ID), EGF plus TGF/31 (Fig. IB), and TGF/31 alone (Fig. 1C) thus all demonstrated qualitatively similar morphological changes. Before and during formation of syncytium, TGF/31 significantly inhibited basal hCG (Fig. 2) and hPL (Fig. 3) secretion after a lag phase of 1 day for hCG and 3 days for hPL (P < 0.005, P < 0.025, respectively). TGF/31 reduced hCG and hPL secretion to a fraction of 0.54 and 0.75 of their controls, respectively (Figs. 2 and 3). This effect was noted in all 11 separate placental experiments. In this culture system, hCG and hPL levels gradually rise during this culture period, as we have described (3, 21). The reduction in hCG and hPL secretion by TGF/31 thus represented a blunting of this normal rise by the
FIG. 1. Phase-contrast photomicrograph of trophoblast cells in culture after 6 days exposure to (A) 10 ng/ml EGF; composite of two representative areas, (B) 10 ng/ml EGF plus 10 ng/ml TGF/31, (C) 10 ng/ ml TGF/31, and (D) control; composite of two representative areas. Large syncytial islands developed in EGF-treated cells (A), but this differentiated state was mostly inhibited by concomitant addition of TGF/31 (B). Plates receiving TGF/31 alone were similar (C). Controls (D) showed little syncytial formation and were qualitatively similar to EGF plus TGF/31 (B), and to TGF/31 alone (C). s, Syncytia; C, cytotrophoblast. All magnifications X125.
fractional amounts given in Figs. 2 and 3. As previously described (3, 21), EGF significantly increased hCG and hPL secretion 2.68- and 2.46-fold after day 1 and day 4, respectively (Figs. 2 and 3; P < 0.001, P < 0.005, respectively). The later stimulation during the culture period of hPL secretion by EGF compared to that of hCG has been noted previously in our system (3, 21). EGF-induced hormone secretion was significantly inhibited by 10 ng/ ml TGF/31 (P < 0.001 for integrated secretion on days 2-7 for hCG; P < 0.01 for integrated secretion on days 3-7 for hPL) and not different from control. TGF/31 reduced the EGF-induced increases by as much as 0.79 (hCG) and 0.64 (hPL) times the control value. TGF/31 thus prevented EGF from inducing increases in hormone secretion and maintained the levels similar to those of controls (which gradually rise during the culture period, as noted above). Dose-response studies demonstrated that significant inhibition of EGF-stimulated hCG and hPL secretion (P < 0.005, P < 0.025, respectively) occurred with 0.1 ng/ml or more TGF/31 (Fig. 4). Despite the presence of TGF/31, a small degree of differentiation continued to occur as noted in Fig. 1. Control plates also demonstrated some syncytia development by 6-7 days in
TGF01 INHIBITS PLACENTAL DIFFERENTIATION
24 5T
O • 4-- A
OEGF/Control •TGF01/Control A(EGF + TGF/31 )/Control
4T hCG
ESS hPL
o
o
Endo •1991 Vol 129 • N o l
(J
'O 2--
T
T
T
^
A
1-
8 2+
§
AH
0.001 0.01 0.1 1 10 TGF/3 Dose (ng/ml)
Doy of Culture FIG. 2. Effects of EGF and TGF/31 on hCG secretion. The fractional changes are the mean ± SEM of 11 separate placental experiments. Ten nanograms per ml EGF induced up to a 2.68 fractional increase in hCG secretion (P < 0.001). Ten nanograms per ml TGF/31 alone inhibited hCG secretion after 1 day in culture, reducing basal secretion fractionally to 0.54 of control (P < 0.005). Ten nanograms per ml TGF/31 added concomitantly with 10 ng/ml EGF significantly inhibited EGF-induced hCG secretion (P < 0.001) to levels similar to that of control values. hCG concentrations for control cells were 31.2 ± 27.0, 22.4 ± 13.5, 35.8 ± 12.7, 54.3 ± 18.3, 49.4 ± 14.4, 36.1 ± 12.8, and 36.3 ± 15.1 ng/ml (mean ± SEM) on culture days 1-7, respectively. 5T 4--
O
OEGF/Control
•
•TGF/31/Control
A — A ( E G F + TGF/31 )/Control 3--
1
2-
1"
0
T /
.6'
• =
