| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Departments of1 Obstetrics and Gynecology,2 Biostatistics and Documentation, and3 Central Blood Transfusion and Immunology, Medical University Innsbruck, Innsbruck, Austria;4 Institutes of Laboratory Medicine in Innsbruck and Wörgl, Tirol, Austria
aaddress correspondence to this author at: Department of Obstetrics and Gynecology, Medical University Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria; fax 43-512-504-23112, e-mail martin.widschwendter{at}uibk.ac.at
In normal human pregnancy, the uterus and its arterial system, including the decidua and the adjacent third of the myometrium, are invaded by cytotrophoblasts (1)(2), which initiate conversion of the decidual vascular system from a high-pressure/low-flow system to a low-pressure/high-flow system that meets the needs of the fetus and placenta (3). The trophoblastic invasion in humans occurs between weeks 8 and 18 of pregnancy (4). The invasion process into the uterus shows many similarities to the invasion of malignant cells during metastasis in that both types of cells have to pass through a basal membrane (5). One of the most important diseases of pregnant women, preeclampsia, is known to be associated with a failure of complete trophoblastic invasion in the first third of the myometrium (6).
Epigenetic alterations, including changes in DNA methylation status, are among the most common molecular alterations in human neoplasia (7). DNA methylation changes are also involved in mammalian development, starting with a wave of demethylation during cleavage, followed by genome-wide de novo methylation after implantation (8). Recently, Ohgane et al.(9) reported that the differentiation of a trophoblast lineage is associated with DNA methylation and demethylation. In many cases, aberrant methylation of the CpG island genes has been correlated with a loss of gene expression, and it has been proposed that DNA methylation provides an alternative to gene deletion or mutation for the loss of gene function (7). Moreover, it is now widely known that methylated DNA can be detected in various body fluids, including serum and plasma, and that the methylation status of some genes can be used for early detection of, or even risk assessment for, various types of human neoplasia (10).
Cell-free fetal and maternal DNA can be detected in the maternal bloodstream (11) and is reported to be robust and easy to obtain (12). Cell-free fetal DNA in maternal plasma seems to be of importance in noninvasive prenatal diagnosis (13) and in the diagnosis of some pregnancy-associated diseases, such as preeclampsia (14). Recently, Poon et al.(15) reported the first use of differential DNA methylation in maternal plasma to detect fetal DNA. The cellular origin of the increase in total plasma DNA is unclear at present, but Chan et al. (16) speculated that fetal DNA may be released into the plasma by trophoblasts, whereas maternal DNA may be released into the circulation by the decidua.
In this proof-of-principle study we addressed the question of whether the methylation pattern in the serum of pregnant women early in pregnancy of genes known to be involved in the invasion process shows similarities to the methylation pattern in patients with invasive cancers. We also asked whether such an invasion-specific methylation pattern in serum shows differences between women with normal pregnancies and women developing preeclampsia—a disease known to be accompanied by a disturbed invasion process in the first trimester (6). We chose a panel of three genes (CDH1, TIMP-3, and PTGS-2) known to play key roles in the invasion process of tumor cells (17)(18)(19), which are often regulated by promoter hypermethylation (20)(21)(22), or in the invasion process of trophoblast cells (2)(23)(24)(25). We also chose one gene (BLT1) involved in the regulation of immune response and regulated by promoter hypermethylation (26). In addition, we selected two genes (APC and RASSF1A) known to be involved in pathways counteracting metastasis that are reported to be methylated in several human neoplasias (27)(28). Recently, we were able to demonstrate that DNA sequences of these six genes are highly methylated in serum of patients with advanced breast cancer and show less methylation in primary breast cancer patients (29).
We analyzed the methylation status of the above-mentioned six genes early in gestation in 32 serum samples from healthy pregnant women with normal pregnancy outcomes and in serum samples from 17 healthy pregnant women who were later diagnosed with either severe preeclampsia (diastolic blood pressure >110 mmHg and 3+ proteinuria), eclampsia, or HELLP syndrome [gestational age at time of sampling, 10–15 weeks (median, 12 weeks) and 6–17 weeks (median, 12 weeks), respectively]. The serum samples were obtained during normal blood drawing for screening during the early gestational weeks. The median patient ages were 32 years (range, 19.7–41.3 years) for pregnant women with normal pregnancy outcomes and 28.5 years (range, 8.6–42.4 years) for women who later developed severe preeclampsia, eclampsia, or HELLP syndrome. Clinical data for all included patients were analyzed anonymously, and persons performing methylation analyses were totally blinded to the clinical data.
Genomic DNA from serum samples was isolated by use of the High Pure Viral Nucleic Acid Kit (Roche Diagnostics) according to the manufacturer’s protocol with some modifications for multiple loading of the DNA extraction columns to obtain a sufficient amount of DNA (29). Sodium bisulfite conversion of genomic DNA and the MethyLight assay for methylation analysis were performed as described previously (30)(31). Primers and probes for the analyzed genes were also as published previously (29). Using the MethyLight assay, we obtained percentages of fully methylated reference values, and in this study we deemed a percentage of fully methylated reference value >0 as positive for methylation.
We detected various degrees of DNA methylation of the six loci in sera from pregnant women early in pregnancy who showed normal as well as pathologic pregnancy outcomes (Table 1
). We found statistically significant differences between the methylation status of APC in sera of women who later developed severe preeclampsia, eclampsia, or HELLP syndrome (n = 17) vs healthy pregnant women (n = 32; P = 0.041) and an inverse trend (P = 0.067) for RASSF1A methylation in sera of these two groups of pregnant women (Table 1
).
|
We also compared the methylation profiles of the pregnant women in early pregnancy who had normal as well as pathologic pregnancy outcomes with the methylation status of sera from 10 healthy controls, pretreatment sera from 26 patients with primary breast cancer, and sera from 10 patients with metastasized breast cancer. The sera of these patients had been analyzed previously for another reason in a recently published study (29). All 10 control patients underwent core biopsies of the breast and were confirmed to have benign disease of the breast (age range, 20.5–71.5 years; median, 42.4 years). Within the group of primary breast cancer patients (age range, 36.1–83.9 years; median, 58 years), 2, 18, and 6 patients had pT1, pT2, and pT3 cancers, respectively, and 15, 10, and 1 patients had lymph node-negative, -positive, and unknown disease, respectively. The 10 patients with advanced breast cancer were diagnosed with metastases in the bone, lung, brain, or liver (age range, 49.3–68.7 years; median, 53.6 years).
We first addressed the question whether sera from the two groups of pregnant women had a methylation status different from that of healthy controls. We found strong statistically significant differences between the methylation status of PTGS2 and BLT1 in sera from pregnant women (n = 59) vs healthy controls (n = 10; P <0.001 for both), a statistically significant difference for RASSF1A methylation (P = 0.038), a trend (P = 0.051) for CDH1 methylation status, and no significant differences in the TIMP-3 and APC methylation status (Table 1
). Looking at the methylation status of these six gene loci in sera from healthy controls vs pretreatment sera from primary breast cancer patients we found no statistically significant differences (data not shown). Otherwise, looking at various methylation changes in healthy controls compared with those in patients with advanced breast cancer revealed statistically significant differences for CDH1, PTGS2, APC, and RASSF1A methylation (P = 0.013, 0.003, 0.001, and 0.005, respectively), a trend for BLT1 methylation (P = 0.087), and no significant differences for TIMP-3 methylation (Table 1
).
With this proof-of-principle study we show for the first time that methylation changes in TIMP-3, CDH1, PTGS2, BLT1, APC, and RASSF1A can be detected in pregnant women. From our point of view, the most important finding is the similarity between pregnant women and metastasized breast cancer patients in the methylation changes in genes that are known to be involved in metastasis and tumor cell invasion (17)(18)(19)(27)(28) or even in the invasion process of trophoblast cells (2)(23)(24)(25) and in the regulation of the immune response (26). As sera of those cancer patients without evidence of metastasis at the time of diagnosis lacked the methylation changes found in advanced breast cancer and pregnancy (Table 1
), we speculate that the observed methylation pattern reflects DNA release from invasive cells, specifically trophoblast cells and tumor cells.
In summary, a statistically significant difference in methylation of APC was seen in sera of healthy pregnant women and women who later developed severe preeclampsia, eclampsia, or HELLP syndrome, perhaps offering a possible tool for early detection of this severe disease in pregnancy. We also describe for the first time in a phenomenologic way that methylation changes in sera of women in early pregnancy are similar to those in sera of patients with advanced breast cancer. Further studies are needed to clarify the importance of DNA methylation in regulating the invasion process of cells in general and of trophoblast cells in particular.
Acknowledgments
We thank Lisl Perkmann and Inge Gaugg for technical assistance. Research funding was received from the Austrian "Fonds zur Förderung der wissenschaftlichen Forschung" (Grants P15995-B05 and P16159-B05) and "Jubiläumsfonds der Österreichischen Nationalbank" (Grants 9856 and 7585). We also thank the laboratory of Dr. Richard Rohrer at Innsbruck for serum sample collection from women in early pregnancy.
Footnotes
1 H.M. Müller and L. Ivarsson contributed equally to this work 
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
