| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
Division of Research,3 Department of Obstetrics, Gynecology and Women's Health, University of Louisville, Health Sciences Center, Louisville, Kentucky 40292
Laboratory of Radiation Safety,4 National Institute for Longevity Sciences, National Center for Geriatrics and Gerontology, Aichi 474-8522 Japan
ABSTRACT
We tested the hypothesis that hCG can upregulate human trophoblast indoleamine 2, 3-dioxygenase (INDO), which catalyzes the breakdown of tryptophan in villous circulation. The results revealed that it can. Treatment of human trophoblasts with hCG resulted in a time and dose dependent increase in INDO mRNA and protein levels and its enzyme activity. The hCG effect was hormone specific and required the dimer conformation of hCG. The hCG effect required its receptors and was mediated by a cAMP dependent, but protein kinase A independent, mitogen-activated protein kinase 3/1 (MAPK3/1) signaling mechanism. In summary, the present data demonstrate a novel hCG effect on human placental INDO, which probably plays a key role at maternal fetal interface in preventing fetal rejection.
human chorionic gonadotropin, mechanisms of hormone action, placenta, pregnancy, trophoblast
Indoleamine 2, 3-dioxygenase (INDO), an intracellular heme-containing enzyme, breaks tryptophan down to N-formylkynurenine [1]. This breakdown prevents the activation of lymphocytes. This prevention spares the fetoplacental tissues from attack by activated lymphocytes, leading to pregnancy loss [2]. Thus, lymphocyte inactivation at the maternal fetal interface could be one of the key mechanisms in the prevention of fetal rejection. Syncytiotrophoblasts, which are at the maternal-fetal interface, contain INDO [3–5]. Any regulatory agent that upregulates the INDO expression in synctiotrophoblasts could prevent fetal rejection, and conversely, any agent that downregulates INDO expression could promote fetal rejection [2]. The paradigm shift in hCG actions led to the recognition that it is a multifunctional molecule which targets many cells and tissues in the fetoplacental unit [6, 7]. This targeting results in several consequences, all of which promote the establishment and maintenance of pregnancy [6–10]. Since syncytiotrophoblasts contain both INDO and functional LHCG receptors (LHCGR) and the known pregnancy-promoting hCG actions, we hypothesized that hCG upregulates INDO expression in syncytiotrophoblasts. The studies to test this hypothesis demonstrate that hCG indeed can upregulate INDO, which suggests a potential mechanism by which hCG may prevent fetal rejection by promoting tryptophan degradation at the fetal-maternal interface.
The following were purchased from the indicated commercial sources: human interferon-
(IFNG), monoclonal antibodies against ß-tubulin, MAPK3/1 (also commonly known as extracellular signal-regulated kinase 1/2, ERK1/2) and phosphorylated MAPK3/1, polyclonal antibodies against cytokeratin, 
1-antichymotrypsin (ACT), chorionic somatomammotropin hormone 1 (CSH1, also known as placental lactogen), vimentin, normal rabbit and goat serum, Percoll, deoxyribonuclease (DNase) I, type III collagenase, Hank solution, Dulbecco modified Eagle medium (DMEM), HEPES, bovine fetal serum, antibiotic-antimycotic (penicillin-amphotericin B-streptomycin) solution, protease inhibitors (phenylmethylsulfonyl fluoride, EDTA tetrasodium salt, aprotinin, leupeptin, and pepstatin), guanidinium thiocyanate, chloroform, 8-bromo-cAMP, L-tryptophan, ascorbate, methylene blue, catalase and Ehrlich reagent (2 g 
-dimethylaminobenzaldehyde in 20% acetic acid) from Sigma-Aldrich (St. Louis, MO); Immobilon-P membranes from Millipore (Billerica, MA); enhanced chemiluminescence Western blotting detection kits from GE Healthcare (Piscataway, NJ); peroxidase-antiperoxidase immunostaining kits from Vector Laboratories (Burlingame, CA); oligo dT primer, AMV reverse transcriptase, deoxynucleotide mix, and Taq DNA polymerase from Promega Corp. (Madison, WI); 123 bp DNA ladder and cDNA cycle kit from Invitrogen (Carlsbad, CA); [-32P]dCTP (3000 Ci/mmol) from NEN Life Science Products (Boston, MA); H-89, PD98059, and Bisindolylmaleimide (Bis) from Calbiochem-Novabiochem Corp. (San Diego, CA). The following items were obtained as gifts: highly purified hCG (CR-127; 14,900 IU/mg), polyclonal antibody to hCG (lot 2), human luteinizing hormone (LH, AFP-0264B), hCG- (CR-125), hCG-ß (CR-129), follicle stimulating hormone (FSH, AFP-87929B) and thyroid stimulating hormone (TSH, AFP-4314C) from NIDDK's National Hormone & Pituitary Program and Dr. A. F. Parlow (Torrance, CA); and 8-Cl-cAMP from the Drug Synthesis and Chemistry Branch at the National Cancer Institute (Bethesda, MD). PCR primers for INDO and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and LHCGR antisense and sense oligodeoxynucleotides (ODN) were synthesized using a Pharmacia (Piscataway, NJ) LKB Gene Assembler Special automated DNA synthesizer using a standard phosphoramidite chemical procedure and desalted on NAP10 columns in our laboratory. The monoclonal antibody against human INDO (anti-INDO-Ab, IgG1 subclass) has been described previously [11].
Approximately 27 term pregnancy and 3 early pregnancy placentas were used. Term pregnancy placentas were obtained from uncomplicated cesarean section deliveries performed at the University of Louisville affiliated hospitals and early pregnancy placentas were obtained from legal terminations performed during 8–12 weeks at a local clinic. All the placentas were transported immediately to the laboratory on ice. Their use was approved by our institutional Human Studies Committee.
Isolation of Human Placental Trophoblasts
Kliman and associates' procedure was modified to isolate cytotrophoblasts [12, 13]. Briefly, placental villi were thoroughly washed and then incubated for 30 min at 37°C in a shaking water bath with type III collagenase and DNase I in Hanks solution containing 0.1 mM CaCl2 and MgSO4. The digestion was repeated three times. The supernatants containing dispersed cells were filtered through a nylon mesh and centrifuged for 20 min at 1200 x g in a 10%–70% Percoll gradient. The middle layer (density 1.048–1.062) containing cytotrophoblasts was removed and washed once. The viability of cytotrophoblasts, as determined by trypan blue exclusion, was greater than 90%. The purity, as determined by immunostaining for the markers of cytotrophoblasts (cytokeratin), syncytiotrophoblasts (hCG and CSH1), fibroblasts (vimentin), and macrophages (ACT) was greater than 95%. Thus, less than 5% of the cells were immunostained positively for vimentin, hCG, and CSH1. The ACT immunopositive cells were not detected, indicating that macrophage contamination was minimal. The cytotrophoblast yield per placenta was about 1 x 106 cells for early pregnancy and 1 x 108 cells for term pregnancy.
Cytotrophoblasts were cultured for 48 h at 37°C in 12-well micro-plates in DMEM, 25 mM HEPES, 0.5% antibiotic-antimycotic mixture and 10% bovine fetal serum in a humidified 5% CO2 and 95% air incubator. As previously demonstrated, the differentiation of cytotrophoblasts into syncytiotrophoblasts was nearly completed by 48 h [13, 14]. The syncytiotrophoblasts were then subjected to various treatments, as indicated in figure legends.
In some experiments, the treatments included the addition of 21-mer phosphorothioate antisense (5'-GCCGAGAACCGCTGCTTCATG-3') or sense (5'-CATGAAGCAGCGGTTCTCGGC-3') ODNs. The ODNs were targeted to the region surrounding the translation initiation codon, ATG, of the human LHCGR cDNA sequence. As we previously demonstrated, the antisense, but not sense ODN was effective in the suppression of LHCGR protein levels in human endometrial cells, adrenal cortical H295R cells, and isolated human trophoblasts [14–16].
Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
Total RNA was isolated by a single step guanidinium thiocyanate-chloroform extraction method [17, 18]. Three µg of total RNA were reverse transcribed into cDNA with oligo dT primer and AMV reverse transcriptase using the cDNA cycle kit. The resulting cDNA was co-amplified by 25 PCR cycles with Taq DNA polymerase, [32P] labeled dCTP and unlabeled dTTP, dATP, dGTP, INDO primers (forward – 5'-TCTGCCAAATCCACAGGAAA-3', reverse – 5'-GCAAGACCTTACGGACATCTCC-3'), and GAPDH primers (forward – 5'-TGGACTCCACGACGTACTCA-3', reverse – 5'-CTCTCTGCTCCTCCTGTTCG-3'). Each cycle consisted of 1 min denaturation at 94°C, 1 min annealing at 61°C, and 1 min extension at 72°C. The sizes of the amplified fragments, which were predicted to be 280 bp for INDO and 370 bp for GAPDH, were determined from comparison with a 123 bp DNA ladder run in an adjacent lane. PCR products were resolved by electrophoresis in 5% polyacrylamide gels and the bands identified by autoradiography. The bands were then excised and radioactivity was eluted and counted in a liquid scintillation counter.
This procedure was performed with an enhanced chemiluminescence detection system [19–21]. Briefly, the cells were homogenized in 50 mM Tris-HCl, pH 7.4 buffer containing protease inhibitors. Twenty-five µg protein aliquots were electrophoresed in SDS-PAGE gels and transferred to Immobilon-P membranes. INDO protein was detected by using 1:1200 dilution of 5 mg/ml of anti-INDO antibody. The detection of ß-tubulin, by using 1:1500 dilution of 3 mg/ml of its antibody, served as a loading control. The phosphorylated MAPK3/1, which represents an activated form of the enzyme, was detected by using 1:2000 dilution of 2 mg/ml of its monoclonal antibody. The total (phosphorylated and nonphosphorylated forms) MAPK3/1 was detected by using 1:2000 dilution of 2 mg/ml of anti-MAPK3/1 monoclonal antibody. The intensities of the bands were quantified by a Z-gel Scanning System (Zaxis Inc., Hudson, OH) and expressed as ratios with ß-tubulin.
Measurement of INDO Enzyme Activity
The enzyme activity was measured according to Takikawa et al [11, 22]. Briefly, 50 µl aliquots of cell extracts were mixed with 100 mM potassium phosphate buffer, pH 6.5, 40 mM ascorbate, 20 µM methylene blue, 200 µg/ml catalase, and 800 µM L-tryptophan and incubated for 30 min at 37°C. The reaction was stopped by the addition of 30% trichloroacetic acid, followed by centrifugation for 5 min at 10 000 x g. The supernatants containing L-tryptophan breakdown product, kynurenine, were mixed with Ehrlich reagent and absorbance at 480 nm was measured.
Each experiment on term pregnancy placentas was performed in triplicate and repeated three times on cells from different specimens. All three early pregnancy placentas were for the experiments shown in Figure 4. The data were analyzed by One-Way Analysis of Variance (ANOVA) followed by Turkey-Kramer multiple comparison test using an Instat Version 3.06 program (Graphpad Sofware, Inc., San Diego, CA). A P value less than 0.05 was considered statistically significant.
|
The hCG Effect on INDO mRNA and Protein Levels and Its Enzyme Activity
Syncytiotrophoblasts contain detectable steady state levels of INDO mRNA, which increased with hCG treatment. The first significant increase was seen at 6 h, followed by a gradual decline to base line levels by 48 h of incubation (Fig. 1A). The hCG effect was seen at 500 mIU/ml. Further increase in the hCG concentration resulted in either a modest or no further increase (Fig. 1B). The LH, but not other members of the glycoprotein hormone family, such as FSH and TSH, mimicked hCG (Fig. 1C). Isolated - and ß-subunits of hCG also failed to increase INDO mRNA levels (Fig. 1C). Western blotting and colorimetric assay revealed that hCG treatment also significantly increased INDO protein levels and its enzyme activity (Fig. 2).
|
|
IFNG Effect and Synergism with hCG
IFNG is a known inducer of INDO in other tissues [11, 23, 24]. Fig. 3 shows that it also can increase INDO mRNA levels in syncytiotrophoblasts. The increase was similar to that induced by hCG. A combination of IFNG and hCG, however, resulted in an additive increase in INDO mRNA levels (Fig. 3).
|
The hCG Effect on INDO in Early Pregnancy Placentas
Early pregnancy is more vulnerable to loss than late pregnancy. If hCG has an immunoprotective role by upregulating INDO, then its role would be relatively more important in early than in late pregnancy. Therefore, we determined the hCG effect on INDO protein levels in syncytiotrophoblasts isolated from early pregnancy placentas. Consistent with the results obtained from term placenta, Western blotting revealed that hCG treatment significantly increased INDO protein levels in 3 different placentas (Fig. 4).
Receptor Requirement in the hCG Action
The ODN approach was used to investigate whether receptors are required for hCG to upregulate INDO in syncytiotrophoblasts. Treatment with antisense receptor phosphorothioate ODN, made from human receptor sequence, which inhibits receptor synthesis [14], prevented hCG from increasing INDO mRNA levels (Fig. 5). The sense receptor ODN, on the other hand, which does not affect receptor synthesis, failed to prevent increase in INDO mRNA levels by hCG treatment (Fig. 5). Neither antisense nor sense ODN had any effect on basal INDO mRNA levels (data not shown).
|
Signaling in the Action of hCG in Syncytiotrophoblasts
Cyclic AMP is a signaling molecule that mediates the hCG actions in placenta [13, 14, 25]. In order to determine whether it was involved in the hCG action to upregulate INDO, we first tested whether 8-bromo-cAMP, a stable and membrane permeable analog, could mimic hCG in increasing INDO mRNA levels. The results in Fig. 6A show that it does. 8-Cl-cAMP, a site-selective analog that inhibits type I PKA by increasing the RI regulatory subunit degradation [26], had no effect on its own nor could it influence the upregulatory actions of hCG or 8-bromo-cAMP (Fig. 6A). This finding suggested that cAMP dependent (but PKA independent) signaling could be involved in the hCG action. This suggestion was further tested by using PKA, PKC, and MAPK3/1 inhibitors, H-89, Bis, and PD98059, respectively. None of these inhibitors, except PD98059, had any effect on basal as well as on hCG stimulated increases in INDO mRNA levels. PD98059, on the other hand, was able to prevent the hCG action (Fig. 6B). The potential MAP kinase involvement in the hCG actions via the PKA independent pathway suggested that hCG treatment may increase phosphorylated MAPK3/1 and this increase should be prevented by PD98059, but not by H-89. Indeed, as shown in Figure 6C, hCG treatment increased phosphorylated MAPK3/1, which was prevented by PD98059 but not by H-89. Given the limitations in using pharmacologic kinase inhibitors, our data suggest that hCG used a cAMP dependent, but PKA independent, mechanism to activate MAPK3/1.
|
INDO is a cytosolic and heme-containing monomeric enzyme present in many tissues with high levels in human placenta [3–5]. This enzyme degrades tryptophan transported through cells, which creates a tryptophan deficient microenvironment [27, 28]. The degradation of tryptophan, an essential amino acid, is not harmful to fetal growth because tryptophanyl tRNA synthase, which is required for protein synthesis, has a lower Km for tryptophan than INDO [29, 30]. This allows protein synthesis to continue in the face of tryptophan degradation. The tryptophan depletion, on the other hand, arrests lymphocyte proliferation because they have a strict tryptophan requirement at the G1 phase of cell cycle and thus becoming more susceptible to apoptosis [31–33].
Although syncytiotrophoblasts are known to contain INDO [3–5], its regulation is virtually unknown. Since hCG has other autocrine and paracrine actions in placenta, such as regulation of cytotrophoblast differentiation, hCG synthesis, cAMP secretion, glycogen breakdown, steroid hormone and prostacyclin production, etc. [6], it is possible that hCG could also regulate INDO. Direct verification experiments revealed that hCG treatment can indeed increase INDO mRNA, protein, and enzyme activity. The hCG effect was time and dose dependent and hormone specific, in that the other members of the same glycoprotein hormone family that do not activate the same receptor had no effect on INDO. The requirement for dimer hormone conformation was suggested by the failure of and ß subunits of hCG to increase INDO.
It is likely that despite high saturating peripheral levels, hCG may still regulate INDO in vivo due to its continuous removal from the maternal-fetal interface by the circulating blood and pulsatile hCG secretion [34]. In addition, hCG secreted by villus trophoblasts is likely to act on vascular cells in the uterus and also on other hCG receptor bearing cells in fetoplacental environment. However, whether these other sites of hCG action involve INDO upregulation is not known.
The IFNG, a Th1 type cytokine produced mainly by activated T-lymphocytes, macrophages, and natural killer cells [1, 24] is known to enhance expression of the INDO gene in a number of animal and human cell types [1, 24]. Our results showed that IFNG was also capable of increasing INDO in the placenta. The comparison of hCG versus IFNG regulation of placental INDO is misleading because hCG maintains pregnancy, whereas IFNG promotes pregnancy loss [35, 36] and yet both can upregulate INDO. The key to understanding this dilemma may rely on the findings that IFNG is not normally produced by placenta unless it is infected with viruses [37]. In such a case, IFNG induced INDO increase may help fight infection, as viral proliferation requires tryptophan, as do intracellular pathogens [38, 39]. The additive INDO induction by hCG with IFNG suggests that hCG may also help fight viral infections, as suggested [40, 41]. In murine pregnancy, uNK cells in mesometrial side of the uterus can produce IFNG, which contributes to midgestational decidual health [42]. However, whether this might be relevant to INDO upregulation is not known.
The actions of hCG require its receptors, as inhibition of their synthesis using an antisense ODN resulted in an abrogation of the hCG effect. This finding is consistent with all the previous evidence that nongonadal hCG actions are receptor mediated, just as hCG gonadal actions are [7, 14, 15].
The findings that 8-bromo-cAMP mimics hCG and PKA inhibitor could not block either the effects of hCG or 8-bromo-cAMP suggest that hCG uses a cAMP dependent, but PKA independent, signaling mechanism. Bypassing PKA to activate MAP kinase can occur through members of the Ras superfamily [43]. In this mechanism, cAMP binds to guanine nucleotide exchange factor and activates Rap1A, which then increases MAPK phosphorylation [43, 44].
The hCG regulation of INDO may begin from the time of implantation through the rest of the pregnancy until the onset of labor. Consistent with the implication that hCG might be acting through the upregulation of INDO, hCG treatment increased INDO protein in trophoblasts from three different early pregnancy placentas. How hCG helps at the time of implantation and during early pregnancy may differ from how it helps to maintain pregnancy. For example, hCG promotes trophoblast invasion [45–48]. For it to be successful, surrounding lymphocytes must be inhibited and hCG induced INDO may play an important role in this inhibition, as it does during the rest of the pregnancy, even though invasion is not as important in late pregnancy as in early pregnancy. However, lymphocytes must nevertheless be suppressed to maintain an intact maternal-fetal interface [31, 49]. Pregnancy complications, such as miscarriages, preterm labor, etc., are likely to occur when hCG regulation of INDO is defective. In fact, it has now been shown that these complications can be successfully treated with hCG [50].
In summary, the present results demonstrate that hCG upregulates INDO, which degrades tryptophan. This degradation removes the stimulus for lymphocyte activation at the maternal-fetal interface. Consistent with this possibility, our preliminary data showed that hCG treatment of an allogeneic lymphocyte and syncytiotrophoblast coculture resulted in an increase in lymphocyte death. These findings suggest that one of the mechanisms by which therapeutic hCG use prevents spontaneous as well as habitual miscarriages and prematurity may involve lymphocyte inactivation at maternal fetal interface.
FOOTNOTES
2Current address: University of Texas at Houston, Houston, TX 77030. 
Correspondence: 1C.V. Rao, Division of Research, Department of Obstetrics, Gynecology and Women's Health, 438 MDR Building, 511 South Floyd St., University of Louisville Health Sciences Center, Louisville, KY 40292. FAX: 502 852 0881; e-mail: cvrao001{at}louisville.edu
Received: 1 September 2006.
First decision: 9 October 2006.
Accepted: 18 December 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  H. Wan, M. A. Versnel, L. M. E. Leijten, C. G. van Helden-Meeuwsen, D. Fekkes, P. J. M. Leenen, N. A. Khan, R. Benner, and R. C. M. Kiekens Chorionic gonadotropin induces dendritic cells to express a tolerogenic phenotype J. Leukoc. Biol., April 1, 2008; 83(4): 894 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-L. Meng, O. M. Rennert, and W.-Y. Chan Human Chorionic Gonadotropin Induces Neuronal Differentiation of PC12 Cells through Activation of Stably Expressed Lutropin/Choriogonadotropin Receptor Endocrinology, December 1, 2007; 148(12): 5865 - 5873. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
