Immunobiology of complete molar pregnancy and gestational trophoblastic tumor

Summary The unique curability of gestational trophoblastic tumors may in part be attributable to a host immunologic response. The occurrence of rapidly progressive and fatal choriocarcinoma may be favored by histocompatibility between patients and their partners. However, histocompatibility is not a prerequisite for the development and persistence of gestational choriocarcinoma. The expression of HLA by choriocarcinomateins in culture is enhanced following incubation with gamma-interferon and this may be of both biologic and clinical significance. Complete molar pregnancy is a complete allograft because all molar chromosomes are of paternal origin. Patients with complete mole are sensitized to paternal HLA antigen which is expressed in molar tissue. Other polymorphic antigen systems including trophoblast-leukocyte common antigens and placental-type alkaline phosphatase are also expressed in molar tissue. We have studied the immunopathology of the molar implantation site to investigate possible humoral and cellular immune responses. The Relationships among normal placenta, complete mole and choriocarcinoma are not clearly understood. The sattern of expression of oncofetal antigens in these three gestational tissues may be used to assess trophoblastic differentiation. In studies to date, molar trophoblast has the same pattern of expression of oncofetal antigens as normal placental trophoblast. We will review recent advances in our understanding of the immunobiology of gestational trophoblastic disease and suggest new directions for further research.Aearing Laboratory, New England Trophoblastic Disease Center, Division of Gynecologic Oncology, Brigham and Women's Hospital, Department of Obstetrics and Gynecology, Harward Medical SchoolSupported by NIH grant CA42738.     

Allelic Characterization of IGF2 and H19 Gene Polymorphisms in Molar Tissues

Background: To investigate the characteristics of allelic distribution of IGF2 and H19 gene polymorphisms in molar tissues compared to normal placentas. Materials and Methods: Forty-nine specimens of molar tissues as well as 100 control normal placental tissues, delivered on the same days, were collected. Polymerase chain reaction (PCR) with restriction fragment length polymorphism (RFLP) on 2% agarose gel electrophoresis was conducted to determine the allelic distribution. The ApaI polymorphism within exon 9 of IGF2 and the RsaI polymorphism within exon 5 of H19 were employed to identify the allelic distribution of the IGF2 and H19 genes, respectively. Then the data for these genes in the molar and normal placenta tissues were compared. Results: The allelic distribution of IGF2 genes found in molar tissue were 21 (42.9%) aa (undigested), 10 (20.4%) ab (heterozygous) and 18 (36.7%) bb (digested), while in normal placenta tissue the values were 22 (22%) aa, 51 (51%) ab, and 27 (27%) bb. The allelic distribution of H19 in molar tissues was 8 (16.2%) aa (undigested), 8 (16.3%) ab (heterozygous) and 33 (67.4%) bb (digested) and in normal placental tissue was 16 (16%) aa, 36 (36%) ab and 48 (48%) bb in normal placenta tissue. These results were significantly different with P values of 0.001 and 0.037 for the allelic distribution of IGF2 and H19, respectively. Conclusions: Molar tissues showed significant differences of allelic distribution of IGF2 and H19 from normal placenta tissues.     

Immunohistochemical localization of transforming growth factor-α and epithelial growth factor receptor in human fetal developing skin, psoriasis and restrictive dermopathy

Keratinocytes release a number of cytokines interacting with other intra- and subepidermal cells during the initiation and the perpetuation of skin inflammatory reactions. Cultured human keratinocytes overexpressing the transforming growth factor alpha (TGF-alpha) assumed a spindled morphology and displayed increased locomotion. Moreover, the receptor for TGF-alpha, the epithelial growth factor receptor (EGFR), is important for autocrine growth, promotion of cell survival, and regulation of cell migration. The expression of TGF-alpha and EGFR has not been widely studied in human developing skin and their roles in geno-dermatosis are not known. In this study, we investigated the expression of TGF-alpha and EGFR by immunohistochemistry in human developing skin at different gestational ages (14 th week, 20 th week, and 34 th week), in six patients with psoriasis, and, for the first time, in an infant affected with restrictive dermopathy, a very rare lethal genodermatosis, characterized by abnormal skin growth and differentiation with thin, tightly adherent skin. TGF-alpha and EGFR were expressed in the basal layer at the 14 th week and in all epidermal layers at the 20 th and 34 th week of gestation. In psoriasis, TGF-alpha was overexpressed in all layers of epidermis, while EGFR was expressed in the basal and first suprabasal layers. In restrictive dermopathy, we observed no expression of both TGF-alpha and EGFR at the level of the skin. The other organs showed comparable patterns to those of an age-matched infant. In conclusion, TGF-alpha and EGFR interact strictly to promote skin development during the intrauterine life. An interactive autocrine growth cycle between TGF-alpha and EGFR is present in psoriasis. A skin-localized alteration of the expression of TGF-alpha and EGFR may be at the basis of restrictive dermopathy. The delay of growth and differentiation of the skin in restrictive dermopathy may be related to the absent expression of TGF-alpha, which is probably due to a down regulation of EGFR by an abnormal autocrine mechanism.     

Distribution of tumor-associated antigen cea and cross-reacting NCA in fetal organs

The organ distribution of the tumor-associated car-cinoembryonic antigen (CEA), and that of the normal tissue component NCA (non-specific cross-reacting antigen) have been investigated in the fetus. Organ extracts from five fetuses between 14 and 21 weeks of age were analysed by radioimmunoassay using specific antisera. CEA was detected in large amounts (800-1,650 ng/g) in fetal colon and in barely detectable amounts in lung and placental tissue. This differed from NCA, which could be detected in almost all organ extracts analysed. The highest concentration of NCA was measured in fetal colon and the content increased with the gestational age of the fetus. High amounts of NCA were also found in the liver, spleen and placenta! tissue. The gel elution profiles of CEA and NCA from an amniotic fluid pool and a pool of colonic extracts were also determined. CEA eluted similarly to the marker ¹²⁵I-CEA purified from liver metastasis of colonic carcinoma. The NCA-reactive material was found in three distinct peaks.     

Metastatic trophoblastic disease after an initial diagnosis of partial hydatidiform mole: genotyping and chromosome in situ hybridization analysis

BACKGROUND: Hydatidiform mole (HM) is classified into partial (PHM) and complete (CHM) subtypes according to histopathologic and genetic criteria. Traditionally, it is believed that PHM carries a better prognosis and rarely develops metastasis. However, making a distinction between PHM and CHM using histologic criteria alone may be difficult. METHODS: The authors used fluorescent microsatellite genotyping following laser-capture microdissection and chromosome in situ hybridization (CISH) to perform a genetic analysis of six patients with histologically diagnosed PHM who subsequently developed metastatic gestational trophoblastic neoplasia. RESULTS: Patients ranged in age from 25 years to 44 years (mean, 33.2 years). The gestational age of the molar pregnancies varied from 6 weeks to 20 weeks. All six patients had pulmonary metastases, with additional liver metastasis in two patients. Among the six patients with histologically diagnosed PHM, it was found that four patients had a diploid karyotype and no maternal alleles; thus, their neoplasms actually were CHM. Maternal genome was detected in the remaining two patients consistent with a biparental origin, and these patients had a triploid karyotype. CISH findings in all patients correlated with the genotyping findings. Triploid HM had maternally derived alleles, whereas diploid HMs were purely androgenetic. CONCLUSIONS: In the current study, which may be the largest series of genetically analyzed metastatic PHMs to date, the difficulty of histologic distinction between PHM and CHM was confirmed. Molecular analysis may help to refine the classification of HM. Although the current findings support the belief that most aggressive trophoblastic diseases are derived from CHM, a small number of PHMs do progress to metastatic disease. Thus, the current study reaffirmed that all patients with HM should be followed closely irrespective of histologic subclassification.     

Epidermal growth factor: receptor and ligand expression in human breast cancer

The epidermal growth factor (EGF) gene is expressed by most human breast cancer cell lines as well as 83% of human breast cancers in vivo. Furthermore, EGF mRNA is detectable in normal human breast tissue. These data suggest that EGF may have a functional role in both normal and neoplastic human breast tissue. Expression of EGF was generally highest in steroid receptor positive human breast tumor biopsies and cell lines. EGF expression was increased by progestins in T-47D and ZR 75 human breast cancer cells. Furthermore, progestins specifically increased the level of TGF-alpha and EGF-receptor mRNA in T-47D cells. Under these same conditions progestins inhibit growth of the cells. Regulation of expression of EGF, TGF-alpha and the EGF-receptor is unlikely to be directly related to the mechanism of progestin induced growth inhibition in T47-D cells. T-47D-5 cells are more sensitive than T-47D cells to progestin and antiestrogen induced growth inhibition. T-47D-5 cells do not express EGF and contain very low levels of TGF-alpha mRNA. The higher level of EGF and TGF-alpha expression in T-47D cells may be one mechanism by which these cells decrease their sensitivity to growth inhibition by progestins and antiestrogens.     

Placental Localization by Placental Scanning and Soft-Tissue Radiography

A review has been made of 105 patients investigated during the past 18 months in whom there has been a strong clinical possibility of placenta praevia. The techniques used have been soft tissue radiography and placental scanning. Some 75 placental scans and 72 radiographic examinations have been reviewed. Technical details are discussed with special reference to placental scanning techniques and their problems. Our experience suggests that placental scanning is a highly accurate means of localizing the placenta. Placental scanning should be used as the first means of diagnosis of ante-partum haemorrhage from 26 to 34 weeks' gestation. After 34 weeks' gestation soft tissue radiography should, in general, be the first means of diagnosis. This will often accurately localize the placenta and other useful information can be obtained, especially in the primigravida.     

Methylation of the adenomatous polyposis coli (APC) gene in human placenta and hypermethylation in choriocarcinoma cells

Methylation of the human APC gene promoter is associated with several different types of cancers and has also been documented in some pre-cancerous tissues. We have examined the methylation of APC gene promoters in human placenta and choriocarcinoma cells. This revealed a general hypomethylation of the APC-1b promoter and a pattern with monoallelic methylation of the APC-1a promoter in full term placental tissue. However, there was no evidence of a parent-of-origin effect, suggesting random post zygotic origin of methylation. Increased methylation of this promoter was observed in all choriocarcinoma-derived trophoblast cell lines, suggesting a trophoblastic origin of placental APC methylation and implicating APC hypermethylation in the development of this group of gestational tumours. Our demonstration of placental methylation of the APC-1a promoter represents the first observation of monoallelic methylation of this gene in early development, and provides further support for a role of canonical Wnt signalling in placental trophoblast invasiveness. This also implicates tumour suppressor gene silencing as an integral part of normal human placental development.     

Macrophage Colony-stimulating Factor Induces the Growth and Differentiation of Normal Pregnancy Human Cytotrophoblast Cells and Hydatidiform Moles but Does Not Induce the Growth and Differentiation of Choriocarcinoma Cells

In the present study, we examined whether or not macrophage colony-stimulating factor (M-CSF; CSF-1) is involved in the growth and differentiation of human chorionic, hydatidiform mole and choriocarcinoma cells. M-CSF promotes the growth of early gestation chorionic cells, hydatidiform mole cells, and a human term placenta cell line (tPA30-1). However, the growth of choriocarcinoma cells, BeWo, Jar, Jeg-3, and NUC-1, was not influenced at all by M-CSF. M-CSF promoted the secretion of human chorionic gonadotropin (hCG) and human placental lactogen (hPL), which are secreted from differentiated trophoblast, from early gestation chorionic cells and from hydatidiform mole cells. However, the secretion of hCG and hPL from choriocarcinoma cells was not affected by M-CSF. When M-CSF localization was examined by immunohistochemical staining, M-CSF was detected in chorionic and hydatidiform mole cells, but was absent in choriocarcinoma cells. These results suggest that the growth and differentiation of normal chorionic and hydatidiform mole cells are M-CSF-dependent, while the growth and differentiation of choriocarcinoma cells are not.     

Altered p16 and Bcl-2 Expression Reflects Pathologic Development in Hydatidiform Moles and Choriocarcinoma

Abnormal trophoblast differentiation is the main cause of gestational trophoblast diseases in the case of hydatidiform moles and choriocarcinomas. Here we investigated the expression patterns of two gene products, p16 and Bcl-2, implicated in cell cycle regulation and apoptosis, respectively, using immunohistochemistry during normal placenta differentiation, hydatidiform moles (partial, complete and invasive) and post-molar choriocarcinomas. The p16 protein shows a gradual expression in cytotrophoblast of normal villous, from a p16 weak proliferative phenotype to a p16 strong invasive phenotype reaching a maximum around 17 weeks of gestation. The expression pattern in cytotrophoblast was similar in moles in contrast to the villous mesenchyme of invasive moles where p16 was strongly expressed. Bcl-2 expression was syncytiotrophoblast specific in normal placenta and moles and increased gradually during normal differentiation. The results explain the persistence of normal and molar villous fragments during their development and their dramatic invasion in the uterine arteries in case of invasive moles. In choriocarcinomas the weak Bcl-2 expression is associated with weak p16 expression indicating a great apoptotic and proliferative potentials. The results suggest that strong p16 expression in the villous mesenchyme may be responsible in part of the morbidity of the moles, and the key of cancer progression in the choriocarcinomas would be a fast cell-cycle turnover.     

Transforming growth factor alpha and epidermal growth factor levels in bladder cancer and their relationship to epidermal growth factor receptor.

We have examined levels of epidermal growth factor (EGF) and transforming growth factor alpha (TGF-alpha) in neoplastic and non-neoplastic bladder tissue using a standard radioimmunoassay technique. Tumour samples had much higher TGF-alpha levels compared with EGF and TGF-alpha levels in malignant tissue were significantly higher than in benign bladder samples. There was, in addition, a difference in mean EGF levels from ''normal'' bladder samples from non-tumour bearing areas of bladder in patients with bladder cancer compared with ''normal'' bladder tissue obtained at the time of organ retrieval surgery. Levels of EGF and TGF-alpha did not correlate with levels of EGF receptor (EGFR) as determined by a radioligand binding method but levels of TGF-alpha > 10 ng gm-1 of tumour tissue did correlate with EGFR positivity defined using immunohistochemistry. These data suggest that TGF-alpha is the likely ligand for EGFR in bladder tumours.     

The immunocytochemical location of two membrane-associated placental tissue proteins in human and cynomolgus monkey placentae

Using the PAP technique, the location of two new membrane-associated placental tissue proteins, MP1 and PP4 was studied in the placenta, its membranes, decidua and umbilical cord of human and cynomolgus monkeys. The results were the same throughout pregnancy. MP1 was located in the syncytiotrophoblast and trophoblastic cells of the reflected chorion. Other placental tissue components (cytotrophoblast, amnion, decidua, and umbilical cord) were negative. At present, MP1 appears to be specific for trophoblast. PP4 was located in the syncytiotrophoblast. Furthermore clear positive staining for PP4 was found in the villous cytotrophoblast, reflected and basal chorion, amnion, umbilical cord and decidua. In addition, PP4 was positive in some granulocyte-like blood cells in the intervillous space. Immunocytochemically, the most positive staining for both proteins was observed in the membrane of villous syncytiotrophoblastic cells. The findings in the placenta of cynomolgus monkeys were similar to those in women. The monkey could, thus, serve as a model for the investigation of these new membrane-associated placental tissue proteins.     

Modulation of benzo[a]pyrene-induced covalent DNA modifications in adult and fetal mouse tissues by gestation stage

In these studies, we investigated the influence of gestation age on the induction of covalent DNA modifications by benzo[a]pyrene (B[a]P). Timed-pregnant ICR mice were given a single treatment of B[a]P (80 mg/kg, p.o.) on different days of gestation, killed 24 h later and analyzed for the presence of B[a]P-induced DNA adducts using the P1 nuclease version of the 32P-postlabeling method. Our results showed that B[a]P bound to embryonic, placental, fetal and maternal DNA throughout gestation with gestation-stage dependency. Overall, B[a]P bound less to maternal DNA during organogenesis and placentation compared to other stages of gestation and to the non-pregnant stage. The ontogenesis of B[a]P-induced DNA adducts in fetal tissues exhibited organ specificity that had two different types of profiles. With advancing gestation age, one type (lung, carcass and placenta) exhibited a steady linear increase, and the other type [gastrointestinal tract (GIT) and skin] a biphasic increase. In the fetal and maternal organs, adduct levels peaked 2 days before parturition. Over the course of gestation, fetal adduct levels were 70-100% of adult levels in the skin, 7-12% in the GIT, 25-40% in the liver and 15-80% in the lung. The adduct levels in many fetal organs exhibited little relationship to placental adduct levels throughout gestation. Collectively, our results indicate that: (i) transplacental DNA damage induced by B[a]P is determined mainly by fetal competence in metabolic activation and/or detoxification of B[a]P; and (ii) events occurring during placentation and organogenesis inhibit B[a]P binding to maternal tissues.     

Human choriocarcinomas: placental growth factor-dependent preclinical tumor models

Choriocarcinomas are a rare form of cancer that develops in the uterus from tissue which would normally become the placenta. Choriocarcinomas are a trophoblastic gestational disease and have been studied largely to investigate conditions related to pregnancy such as preeclampsia. Choriocarcinomas are highly angiogenic and produce high levels of placental growth factor (PlGF) to promote the development of blood vessels. Upregulation of PlGF expression also occurs during the development of other human malignancies such as breast cancer and melanoma. Both tumor specimens and plasma samples have higher levels of PlGF than normal tissues. Hence, PlGF has emerged as a valid target for anti-angiogenic therapy. The cell lines BeWo, JAR and JEG-3, derived from human choriocarcinomas, were investigated in vitro and in vivo for suitability as PlGF-dependent models. BeWo, JAR and JEG-3 cells were characterized in culture and were implanted into immunodeficient mice to generate subcutaneous tumors. The PlGF and VEGF angiogenic profiles of the choriocarcinoma cells and tumors were investigated by ELISA and by immunohistochemical methods. Double immunofluorescence methods were applied to choriocarcinoma xenograft sections to characterize the cellular components of the blood vessels. sFLT01, a fusion protein that neutralizes PlGF, was assessed in cell culture experiments and xenograft studies. The novel results presented here validate the importance of human choriocarcinoma cell lines and xenografts in further exploring the role of PlGF in tumor angiogenesis, for evaluating PlGF as an anti-angiogenic target, and for developing therapies that may provide clinical benefit.     

Epidermal growth factor and transforming growth factor alpha characteristics of human oral carcinoma cell lines.

This study examined the expression of epidermal growth factor (EGF) cell-surface receptors, the response to exogenous ligand and the autocrine production of transforming growth factor alpha (TGF-alpha) in normal and carcinoma-derived human oral keratinocytes. One of eight malignant cell lines overexpressed EGF receptors, while the remainder expressed receptor numbers similar to normal cells. Exogenous EGF stimulated incorporation of tritiated thymidine in a dose-dependent manner. In keratinocytes expressing normal numbers of EGF receptors, the cellular response to exogenous EGF correlated positively with total EGF receptor number. SCC-derived keratinocytes produced more TGF-alpha than normal cells. There was no statistical correlation between the autocrine production of TGF-alpha, EGF cell-surface receptor expression and cellular response to exogenous EGF. While the growth-stimulatory effects of exogenous TGF-alpha were inhibited by the addition of a neutralising antibody, the presence of this antibody in conditioned medium failed to produce a similar decrease in growth. The results indicate that overexpression of EGF receptors is not an invariable characteristic of human oral squamous carcinoma-derived cell lines. Further, the contribution of TGF-alpha to the growth of normal and carcinoma-derived human oral keratinocytes in vitro may be less significant than previously documented.     

Transforming growth factor alpha in chemically transformed hamster oral keratinocytes

The cheek pouch of the Syrian hamster is an excellent tissue for the experimental induction of oral cancer by carcinogenic chemicals. Lysate prepared from a cell line (HCPC-1) derived from one of these hamster oral tumors greatly increased the growth of these oral tumor cells in vitro. We now show that the mitogenic substance, transforming growth factor alpha (TGF-alpha), is present in all of the chemically transformed hamster oral tumors examined (in vitro and in vivo). In no adult normal tissue of the Syrian hamster can we detect expression of TGF-alpha. TGF-alpha could be partly or wholly responsible for the mitogenic activity detected in the lysate of the chemically transformed hamster oral keratinocytes. Both normal and chemically transformed hamster oral keratinocytes express the receptor to epidermal growth factor. The consistent detection of TGF-alpha and epidermal growth factor receptor mRNAs in these hamster oral tumor cells suggests that an autocrine growth mechanism might be operative. This hamster cheek pouch oral cancer model can be used for the molecular analysis of how TGF-alpha and epidermal growth factor receptor might be involved in the malignant transformation of epithelial tissues.     

Transforming growth-factor-alpha and factor-beta(1) levels in plasma and effusions from patients with cancer or hematologic malignancies

Transforming growth factors-alpha and -beta1 are thought to play a role in carcinogenesis. Using a sandwich linked immunosorbent assay, we have measured TGF-alpha and -beta1 levels in malignant human plasma and effusions. TGF-alpha and -beta1 plasma levels were not significantly different among normal volunteers, patients with solid tumors, and patients with hematologic malignancies (TGF-alpha, p=0.225; TGF-beta1, p=0.354). Statistically significant differences were also not found in levels between malignant and non-malignant effusions (TGF-alpha, p=0.327; TGF-beta1, p=0.095). However, a trend for the majority of elevated TGF-alpha or TGF-beta1 levels to be in malignant effusions warrants further studies with larger numbers of samples.