Calcitriol affects hCG gene transcription in cultured human syncytiotrophoblasts
© Barrera et al; licensee BioMed Central Ltd. 2008
Received: 16 November 2007
Accepted: 22 January 2008
Published: 22 January 2008
In pregnancy, maternal serum concentrations of calcitriol significantly rise as a result of increased renal and placental contribution in order to assure calcium supply for the developing fetus. Considering that placenta is a site for vitamin D activation, and the versatility and potency of calcitriol, it is feasible that this hormone participates in fetal/placental development and physiology. In the present work we studied calcitriol actions upon human chorionic gonadotropin (hCG) secretion and expression in cultured trophoblasts, as well as vitamin D receptor (VDR) and CYP27B1 immunolocalization in placental villi.
Quantification of hCG in culture media was performed by immunoassay. Expression studies were carried out by real time PCR. Analysis of CYP27B1 and VDR localization in placental slides were performed by immunohistochemistry. Statistical significance was established by one way ANOVA using Tukey test for comparisons.
Calcitriol regulated hCG in a time-dependent manner: at 6 h the secosteroid stimulated hCG, whereas longer incubations (24 h) showed opposite effects. Interestingly, calcitriol stimulatory effects on hCG were accompanied by an increase in intracellular cAMP content and were abolished by pre-incubation of the cells with a selective protein kinase A inhibitor. Immunohistochemical techniques showed differential VDR localization in the syncytiotrophoblast layer or in the vascular smooth muscle cells depending on the epitope to which the antibodies were raised (specific for the carboxy- or amino-terminal regions, respectively). CYP27B1 was immunolocalized in the syncytiotrophoblast layer of placental villi.
The presence and location of the vitamin D activating enzyme CYP27B1 as well as the specific receptor for vitamin D were shown in placental sections. The latter, together with findings demonstrating specific effects of calcitriol acting through the VDR and the cAMP/PKA signaling pathway upon hCG expression and secretion, indicate that there is a functional vitamin D endocrine system in the placenta, and recognize calcitriol as an autocrine regulator of hCG.
Vitamin D is metabolized to the steroid hormone 1,25-dihydroxyvitamin D3 or calcitriol, which regulates calcium homeostasis, modulates the immune response, and promotes cellular differentiation, among other actions. Calcitriol, the most active vitamin D metabolite, exerts its biological effects by binding to the vitamin D receptor (VDR), which is a ligand-activated transcription factor that recognizes cognate vitamin D response elements (VDREs) in target genes, and can also elicit rapid responses mediated by membrane receptors . Placenta is a source and target of calcitriol . In a similar manner to the renal process, placental production of calcitriol is catalyzed by the mitochondrial CYP27B1 . In early reproductive events, calcitriol has shown to evoke specific biological effects such as regulation of the decidualization and implantation processes [4, 5]. In addition, calcitriol regulates placental lactogen expression as well as progesterone and estradiol secretion in cultured human syncytiotrophoblasts [6, 7]. Regarding other molecules that are regulated by calcitriol in the placenta, Evans et al showed that calcitriol acts in an autocrine/paracrine fashion to regulate both acquired and innate immune responses, decreasing synthesis of cytokines such as granulocyte-macrophage colony stimulating factor 2, tumor necrosis factor, and interleukin 6, but increasing expression of mRNA for the cathelicidin antimicrobial peptide . Since human chorionic gonadotropin (hCG) is a pivotal hormone for pregnancy maintenance, the aim of the present work was to broaden the knowledge of calcitriol actions in the placenta, focusing in the study of its effects upon hCG expression and secretion in cultured human syncytiotrophoblasts. The data presented herein display a functional vitamin D endocrine system present in human placenta and suggest its involvement in regulating placental physiology.
Culture media, fetal bovine serum (FBS) and Trizol were from Invitrogen (NY, USA). TaqMan Master reaction, TaqMan probes and the transcriptor RT system were from Roche (Roche Applied Science, IN, USA), calcitriol (1α,25-dihydroxycholecalciferol) was kindly donated from Hoffmann-La Roche Ltd (Basel, Switzerland). 3-Isobutyl-1-methylxanthine (IBMX), 8-Bromo cAMP (8-Br-cAMP), H-89 and the enzymes used for cell cultures were from Sigma-Aldrich (MO, USA). Immunoassay for hCG was from Immunometrics Ltd, (London, UK). CYP27B1 antibody (sheep anti-murine 25-hydroxyvitamin D-1α-hydroxylase) was from The Binding Site (Birmingham, UK). The VDR antibodies (rabbit polyclonal anti-VDR N-20 sc-1009 and anti-VDR C-20 sc-1008), as well as the secondary antibodies rabbit anti-sheep-horseradish peroxidase, and mouse anti-rabbit IgG-HRP were purchased from Santa Cruz Biotechnology (CA, USA). DAB (3,3'-diaminobenzidine tetrahydrochloride) was from Zymed Laboratories Inc. (CA, USA).
This study was approved by the Institutional Human Ethical Committee (Hospital de Gineco-Obstetricia "Luis Castelazo Ayala", IMSS, México), and written informed consents forms were obtained from each placental donor. Term placentae (37–42 weeks of gestation) were acquired from uncomplicated pregnancies.
Fresh placental tissue from 5 term placentas was embedded in paraffin after fixation in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. Serial sections (7 μm) were obtained according to standard procedures. Slides were treated with methanol-hydrogen peroxide in order to block the endogenous peroxidase activity. Normal rabbit serum and 1% BSA were used as blockers to decrease non-specific signal. Slides were then incubated with primary antibodies (anti-CYP27B1, anti-VDR-N and anti-VDR-C) during 45 minutes at room temperature, followed by further washing and incubation with secondary antibodies for another 45 minutes. Staining was developed using DAB substrate and the chromogen was contrasted with Mayer's hematoxylin. Immunolabeling specificity was tested by omitting the primary antibody.
Trophoblast cell culture
Trophoblasts were cultured as previously described [3, 9, 10]. Briefly: Villous cytotrophoblasts were obtained by enzymatic dispersion and cells were separated on density Percoll gradients. Trophoblasts were plated at a density of 8 × 105 cells/mL in supplemented medium [(DMEM) 100 U/ml penicillin, 100 mg/ml streptomycin, 0.25 mg/ml Fungizone], containing 20% heat-inactivated FBS. Incubations were performed in humidified 5% CO2-95% air at 37°C. The morphological aspects of cells were examined daily, secreted hCG was measured by immunoassay (EIA) following manufacturer instructions and results were normalized against total protein content. Protein was determined by the method of Bradford .
Calcitriol effects on hCG secretion
Two days-cultured trophoblasts were incubated in the presence of different concentrations of calcitriol or ethanol as vehicle, in serum-free DMEM-F12 during 6 h or 24 h. Additional experiments were performed incubating the cells with a selective protein kinase A inhibitor (H-89). Incubations were stopped by media collection, cell lysis with RIPA buffer (9.1 mM dibasic sodium phosphate, 1.7 mM monobasic sodium phosphate, 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS, pH 7.4) was used for protein determination and hCG was quantified in culture media.
Calcitriol effects on hCG expression
For expression studies 3 × 106 cells were plated in 25 cm2 cell culture flasks and subjected to the same treatments as stated above. Total RNA was extracted using Trizol and 1 μg was reverse transcribed using the transcriptor RT system. Real-time PCR was carried out using the LightCycler 2.0 from Roche (Roche Diagnostics, Mannheim, Germany), according to the following protocol: activation of Taq DNA polymerase and DNA denaturation at 95°C for 10 min, proceeded by 45 amplification cycles consisting of 10 s at 95°C, 30 s at 60°C, and 1 s at 72°C. The primer pair was targeted to the β subunit of the hCG mRNA and the sequences were as follows: GCTCACCCCAGCATCCTAT and CAGCAGCAACAGCAGCAG. The house keeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was also amplified as an internal control, using the primers: AGCCACATCGCTGAGACAC and GCCCAATACGACCAAATCC. The sizes of the resulting amplicons were 131 bp and 66 bp, and the probes utilized were # 79 and # 60 (Roche human universal probe library), for hCG and GAPDH, respectively. The expression of CYP24A1 used as a control for calcitriol effects was evaluated using the following sense and anti-sense primers: CATCATGGCCATCAAAACAA and GCAGCTCGACTGGAGTGAC and probe # 88 from Roche human universal probe library.
Calcitriol effects on cAMP accumulation
Cells were incubated in the presence of calcitriol or its vehicle in DMEM-F12 supplemented with IBMX (0.05 mM). Incubations were terminated after 10 minutes by media collection and homogenization of the cells in RIPA buffer. Samples were boiled during 5 min for phosphodiesterases inactivation and intracellular cAMP was measured by specific radioimmunoassay (RIA) as previously described . Results were normalized against total protein content and expressed as fmol cAMP/mg protein.
Data are presented as the mean ± S.D. Statistical significance among groups was established by one way ANOVA using Tukey test. A P value ≤ 0.05 was considered statistically significant.
Calcitriol effects on hCG in cultured trophoblasts
Serum concentrations of biologically active hCG depend on the rate of synthesis of its specific β subunit; whereas at the cellular level multiple factors modulate hCG production by interacting with specific membrane receptors on placental trophoblasts. The most studied factors that modulate hCG are the gonadotropin releasing hormone (GnRH), hCG itself and other molecules that activate cAMP-dependent signal transduction pathways [16–18]. Indeed, both hCGα and hCGβ genes are highly transcriptionally induced by cAMP [19–21]. In this study we showed that calcitriol is an additional factor that modulates hCG in human trophoblasts. In fact, calcitriol regulated hCG in a time-dependent manner, stimulating or inhibiting hormone secretion and expression. The stimulatory effect after 6 hours was probably due, as demonstrated in this study, to the rapid calcitriol-dependent increase in intracellular cAMP. This assumption was further supported by results in the presence of H-89, a selective inhibitor of PKA. Indeed, blocking the cAMP/PKA signal transduction cascade impaired the ability of calcitriol to elicit transcriptional induction of hCGβ gene, as well as hCG secretion into the culture media. Rapid cAMP generation induced by calcitriol has been previously reported in other cell types [15, 22], and may be the result of its interaction with membrane-VDR or other surface proteins. In addition, since it has been demonstrated that calcium ion channels are involved in GnRH dependent-hCG secretion , calcitriol could also release stored hCG through promoting a rapid calcium entry into the cell. Further studies are needed in order to clarify this matter.
The concentration of hCG was also measured after 12, 24 and 48 h of calcitriol treatment, but the results reported in the present study were only those that differed significantly when compared with the vehicle alone. The stimulatory effects observed at 6 h were no further evident after 12 or 24 h, and when cells were incubated in the presence of calcitriol during 2 consecutive days, the effects were rather inhibitory. Inhibition was evident at the mRNA level after 24 hours treatment, preceding the observed response in hCG protein. These data are probably more likely to be reflective of the true biological situation. Indeed, our results that calcitriol inhibited hCG were in line with previous data from this and other laboratories where low serum calcitriol and high serum hCG levels were found in preeclampsia [24–26], that conjointly with the fact that [3H]25-hydroxyvitamin D bioconversion into [3H]1,25-dihydroxyvitamin D was significantly reduced in preeclamptic placentas , may suggest a direct regulatory effect of calcitriol on hCG production. Regarding the inhibitory effects of calcitriol on hCG, it is likely that a secondary metabolic C23/C24 calcitriol oxidation pathway might play a role, since the resulting trihydroxylated metabolite is considered biologically inactive . Alternatively, since calcitriol has been shown to stimulate progesterone secretion  and in turn this hormone inhibits hCG secretion , this mechanism could additionally participate in calcitriol long term inhibitory effects in placenta. In any case, the demonstration in this study of genomic mediated effects of calcitriol on hCG suggested the presence of VDR dependent regulatory regions on hCG promoters. Indeed we found five putative VDR/RXR heterodimer binding sites in the hCGβ-5 gene promoter , which probably may be acting as calcitriol dependent-transcriptional regulatory regions. Nevertheless, the sole presence of the VDREs in the hCGβ-5 promoter is not sufficient to indicate transcriptional function; therefore, functional evaluation of the putative VDREs deserves further investigation.
In non pregnant women the physiological concentration of calcitriol fluctuates between 40–100 pM. In the present study the calcitriol doses tested were: 100 pM, 1 nM and 10 nM. The lowest concentration (100 pM) is within the physiological range of circulating calcitriol levels in mexican pregnant women (127 pM and 151 pM) as observed previously [24, 29]. The other doses tested were supraphysiological, nevertheless, calcitriol effects upon hCG were evident starting with the lowest concentration.
Placenta is considered not only as a source but also as a target of calcitriol . In order to get insights on calcitriol paracrine/autocrine effects in placenta, we investigated the immunolocalization of VDR and CYP27B1 in placental chorionic villi. In accordance with previous reports , CYP27B1 protein was located in the syncytiotrophoblast layer, corroborating that the endocrine phenotype of trophoblasts cells is responsible for vitamin D activation in placenta. To answer where the locally produced calcitriol acts in the placenta, we looked for VDR protein in placental sections. To our knowledge, this is the first report to show immunoreactive VDR in different locations in the placental villi, since VDR expression has been mainly addressed at the mRNA level in placenta [2, 31]. The antibodies showed the presence of VDR in the endocrine placental cells and VSMC, suggesting that calcitriol could be involved in regulating hormonal production and vascular remodeling through the VDR. The latter assumption derives from previous studies demonstrating that calcitriol acts in the vasculature promoting VSMC growth and migration [32, 33]. Interestingly, the C-terminus antibody intensely stained the syncytiotrophoblast layer and faintly stained the surrounding cells of placental vessels, whereas the N-terminus antibody detected a strong signal in the endothelial and VSMC. These observations may indicate different epitopes recognized by the antibodies depending on the topological position of the VDR. An interesting challenge would be to define specific VDR responses in different placental structures.
In summary, the present study broadens the knowledge of placental vitamin D endocrine system by demonstrating the physiological effects of calcitriol on an important biochemical placental function marker such as hCG. In addition, this is the first report to show immunoreactive VDR in different locations in human placental villi, and opens the field to address important research questions on the role of calcitriol in specific placental structures.
D.B. is a Ph.D. student from the posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México (UNAM) under a fellowship from the Consejo Nacional de Ciencia y Tecnología (CONACyT, México). This work was supported by a grant (45937/A1) from CONACyT, México. We acknowledge with special thanks to Hoffmann-La Roche Ltd for calcitriol donation and to Hospital de Gineco-Obstetricia "Luis Castelazo Ayala", IMSS, México, for placental donation.
- Dusso AS, Brown AJ, Slatopolsky E: Vitamin D. Am J Physiol Renal Physiol. 2005, 289: F8-28. 10.1152/ajprenal.00336.2004.View ArticlePubMedGoogle Scholar
- Avila E, Diaz L, Halhali A, Larrea F: Regulation of 25-hydroxyvitamin D3 1alpha-hydroxylase, 1,25-dihydroxyvitamin D3 24-hydroxylase and vitamin D receptor gene expression by 8-bromo cyclic AMP in cultured human syncytiotrophoblast cells. J Steroid Biochem Mol Biol. 2004, 89–90: 115-119. 10.1016/j.jsbmb.2004.03.090.View ArticlePubMedGoogle Scholar
- Diaz L, Sanchez I, Avila E, Halhali A, Vilchis F, Larrea F: Identification of a 25-hydroxyvitamin D3 1alpha-hydroxylase gene transcription product in cultures of human syncytiotrophoblast cells. J Clin Endocrinol Metab. 2000, 85: 2543-2549. 10.1210/jc.85.7.2543.PubMedGoogle Scholar
- Halhali A, Acker GM, Garabedian M: 1,25-Dihydroxyvitamin D3 induces in vivo the decidualization of rat endometrial cells. J Reprod Fertil. 1991, 91: 59-64.View ArticlePubMedGoogle Scholar
- Du H, Daftary GS, Lalwani SI, Taylor HS: Direct regulation of HOXA10 by 1,25-(OH)2D3 in human myelomonocytic cells and human endometrial stromal cells. Mol Endocrinol. 2005, 19: 2222-2233. 10.1210/me.2004-0336.View ArticlePubMedGoogle Scholar
- Barrera D, Avila E, Hernandez G, Halhali A, Biruete B, Larrea F, Diaz L: Estradiol and progesterone synthesis in human placenta is stimulated by calcitriol. J Steroid Biochem Mol Biol. 2007, 103: 529-532. 10.1016/j.jsbmb.2006.12.097.View ArticlePubMedGoogle Scholar
- Stephanou A, Ross R, Handwerger S: Regulation of human placental lactogen expression by 1,25-dihydroxyvitamin D3. Endocrinology. 1994, 135: 2651-2656. 10.1210/en.135.6.2651.PubMedGoogle Scholar
- Evans KN, Nguyen L, Chan J, Innes BA, Bulmer JN, Kilby MD, Hewison M: Effects of 25-hydroxyvitamin D3 and 1,25-dihydroxyvitamin D3 on cytokine production by human decidual cells. Biol Reprod. 2006, 75: 816-822. 10.1095/biolreprod.106.054056.View ArticlePubMedGoogle Scholar
- Diaz L, Arranz C, Avila E, Halhali A, Vilchis F, Larrea F: Expression and activity of 25-hydroxyvitamin D-1 alpha-hydroxylase are restricted in cultures of human syncytiotrophoblast cells from preeclamptic pregnancies. J Clin Endocrinol Metab. 2002, 87: 3876-3882. 10.1210/jc.87.8.3876.PubMedGoogle Scholar
- Kliman HJ, Nestler JE, Sermasi E, Sanger JM, Strauss JF: Purification, characterization, and in vitro differentiation of cytotrophoblasts from human term placentae. Endocrinology. 1986, 118: 1567-1582.View ArticlePubMedGoogle Scholar
- Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976, 72: 248-254. 10.1016/0003-2697(76)90527-3.View ArticlePubMedGoogle Scholar
- Avila E, Diaz L, Barrera D, Halhali A, Mendez I, Gonzalez L, Zuegel U, Steinmeyer A, Larrea F: Regulation of Vitamin D hydroxylases gene expression by 1,25-dihydroxyvitamin D3 and cyclic AMP in cultured human syncytiotrophoblasts. J Steroid Biochem Mol Biol. 2007, 103: 90-96. 10.1016/j.jsbmb.2006.07.010.View ArticlePubMedGoogle Scholar
- Ulloa-Aguirre A, August AM, Golos TG, Kao LC, Sakuragi N, Kliman HJ, Strauss JF: 8-Bromo-adenosine 3',5'-monophosphate regulates expression of chorionic gonadotropin and fibronectin in human cytotrophoblasts. J Clin Endocrinol Metab. 1987, 64: 1002-1009.View ArticlePubMedGoogle Scholar
- Jameson JL, Lindell CM: Isolation and characterization of the human chorionic gonadotropin beta subunit (CG beta) gene cluster: regulation of transcriptionally active CG beta gene by cyclic AMP. Mol Cell Biol. 1988, 8: 5100-5107.PubMed CentralView ArticlePubMedGoogle Scholar
- Santillan GE, Boland RL: Studies suggesting the participation of protein kinase A in 1, 25(OH)2-vitamin D3-dependent protein phosphorylation in cardiac muscle. J Mol Cell Cardiol. 1998, 30: 225-233. 10.1006/jmcc.1997.0577.View ArticlePubMedGoogle Scholar
- Petraglia F, Santuz M, Florio P, Simoncini T, Luisi S, Plaino L, Genazzani AR, Genazzani AD, Volpe A: Paracrine regulation of human placenta: control of hormonogenesis. J Reprod Immunol. 1998, 39: 221-233. 10.1016/S0165-0378(98)00024-2.View ArticlePubMedGoogle Scholar
- Licht P, Cao H, Lei ZM, Rao CV, Merz WE: Novel self-regulation of human chorionic gonadotropin biosynthesis in term pregnancy human placenta. Endocrinology. 1993, 133: 3014-3025. 10.1210/en.133.6.3014.PubMedGoogle Scholar
- Milsted A, Cox RP, Nilson JH: Cyclic AMP regulates transcription of the genes encoding human chorionic gonadotropin with different kinetics. DNA. 1987, 6: 213-219.View ArticlePubMedGoogle Scholar
- Ringler GE, Kao LC, Miller WL, Strauss JF: Effects of 8-bromo-cAMP on expression of endocrine functions by cultured human trophoblast cells. Regulation of specific mRNAs. Mol Cell Endocrinol. 1989, 61: 13-21. 10.1016/0303-7207(89)90185-8.View ArticlePubMedGoogle Scholar
- Jameson JL, Hollenberg AN: Regulation of chorionic gonadotropin gene expression. Endocr Rev. 1993, 14: 203-221. 10.1210/er.14.2.203.PubMedGoogle Scholar
- Johnson W, Albanese C, Handwerger S, Williams T, Pestell RG, Jameson JL: Regulation of the human chorionic gonadotropin alpha- and beta-subunit promoters by AP-2. J Biol Chem. 1997, 272: 15405-15412. 10.1074/jbc.272.24.15405.View ArticlePubMedGoogle Scholar
- Fleet JC: Rapid, membrane-initiated actions of 1,25 dihydroxyvitamin D: what are they and what do they mean?. J Nutr. 2004, 134: 3215-3218.PubMedGoogle Scholar
- Sharma SC, Rao AJ: Effect of calcium ion channel antagonists on chorionic gonadotropin secretion. Biochem Mol Biol Int. 1997, 43: 1101-1106.PubMedGoogle Scholar
- Halhali A, Tovar AR, Torres N, Bourges H, Garabedian M, Larrea F: Preeclampsia is associated with low circulating levels of insulin-like growth factor I and 1,25-dihydroxyvitamin D in maternal and umbilical cord compartments. J Clin Endocrinol Metab. 2000, 85: 1828-1833. 10.1210/jc.85.5.1828.PubMedGoogle Scholar
- August P, Marcaccio B, Gertner JM, Druzin ML, Resnick LM, Laragh JH: Abnormal 1,25-dihydroxyvitamin D metabolism in preeclampsia. Am J Obstet Gynecol. 1992, 166: 1295-1299.View ArticlePubMedGoogle Scholar
- Basirat Z, Barat S, Hajiahmadi M: Serum beta human chorionic gonadotropin levels and preeclampsia. Saudi Med J. 2006, 27: 1001-1004.PubMedGoogle Scholar
- Yamamoto T, Matsumoto K, Kurachi H, Okamoto Y, Nishio Y, Sakata M, Tasaka K, Murata Y: Progesterone inhibits transcriptional activation of human chorionic gonadotropin-alpha gene through protein kinase A pathway in trophoblast cells. Mol Cell Endocrinol. 2001, 182: 215-224. 10.1016/S0303-7207(01)00580-9.View ArticlePubMedGoogle Scholar
- Genomatix. [http://www.genomatix.de/]
- Halhali A, Villa AR, Madrazo E, Soria MC, Mercado E, Diaz L, Avila E, Garabedian M, Larrea F: Longitudinal changes in maternal serum 1,25-dihydroxyvitamin D and insulin like growth factor I levels in pregnant women who developed preeclampsia: comparison with normotensive pregnant women. J Steroid Biochem Mol Biol. 2004, 89–90: 553-556. 10.1016/j.jsbmb.2004.03.069.View ArticlePubMedGoogle Scholar
- Zehnder D, Bland R, Williams MC, McNinch RW, Howie AJ, Stewart PM, Hewison M: Extrarenal expression of 25-hydroxyvitamin d(3)-1 alpha-hydroxylase. J Clin Endocrinol Metab. 2001, 86: 888-894. 10.1210/jc.86.2.888.PubMedGoogle Scholar
- Zehnder D, Evans KN, Kilby MD, Bulmer JN, Innes BA, Stewart PM, Hewison M: The ontogeny of 25-hydroxyvitamin D(3) 1alpha-hydroxylase expression in human placenta and decidua. Am J Pathol. 2002, 161: 105-114.PubMed CentralView ArticlePubMedGoogle Scholar
- Rebsamen MC, Sun J, Norman AW, Liao JK: 1alpha,25-dihydroxyvitamin D3 induces vascular smooth muscle cell migration via activation of phosphatidylinositol 3-kinase. Circ Res. 2002, 91: 17-24. 10.1161/01.RES.0000025269.60668.0F.View ArticlePubMedGoogle Scholar
- Mitsuhashi T, Morris RC, Ives HE: 1,25-dihydroxyvitamin D3 modulates growth of vascular smooth muscle cells. J Clin Invest. 1991, 87: 1889-1895.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.