hCG Reference Service - hCG, phantom hCG, hCG Reference Service, trophoblast

Gestational Trophoblastic Diseases:

1. Pathophysiology of Hyperglycosylated hCG

Laurence A. Cole, Donghai Dai , Stephen A. Butler , Kimberly K. Leslie and Ernest I. Kohorn

Department of Obstetrics and Gynecology, University of New Mexico, Albuquerque NM 87131

Yale Trophoblast Center, Obstetrics and Gynecology, Yale University, New Haven, CT 06520

Abbreviated title: Hyperglycosylated hCG in cancer invasion

Corresponding Author:

Laurence A. Cole PhD, Department of Obstetrics and Gynecology. MSC10 5580 1 University of New Mexico, Albuquerque, NM 87131-0001

ABSTRACT

OBJECTIVE

Hyperglycosylated hCG (hCG-H), is a glycosylation variant of hCG produced by cytotrophoblast cells at implantation of pregnancy and in choriocarcinoma. We investigated the biological function of hCG-H in invasion in vitro and in vivo and the use of hCG-H antibodies in blocking tumorigenesis and cancer growth in vivo.

METHODS AND RESULTS

hCG-H accounts for 43% to 100% of total hCG immunoreactivity in the culture fluid of choriocarcinoma cell lines, and 100% in primary cultures of pregnancy cytotrophoblast cells. We investigated the action of hCG and hCG-H on isolated cytotrophoblast cell primary cultures and on 3 different lines of choriocarcinoma cells cultured on Matrigel basement membrane inserts (culture models for assessing tumor invasion). The addition of hCG-H to medium significantly promoted invasion of membranes with both pregnancy and cancer cell line sources, while regular hCG had no significant effect.

JEG-3 human choriocarcinoma cells were transplanted subcutaneously into athymic nude mice. Tumors rapidly formed. B152, mouse monoclonal antibody against hCG-H, and non-specific mouse IgG (control) were administered twice weekly once tumors were clearly visible. While a correlation between time and growth was observed with the control group (r²=0.97), no correlation was observed with the B152-treated mice (r²=0.15). B152 blocked tumor growth (t test, IgG vs. B152, P=0.003). In a second experiment, antibody B152 or IgG were administered to mice at the time of choriocarcinoma transplantation. B152 significantly inhibited tumorigenesis (t test P=0.0071).

CONCLUSIONS

hCG-H is a critical promoter in human cytotrophoblast and human choriocarcinoma cell invasion in vivo and in vitro, promoting tumor growth and invasion through an autocrine mechanism. hCG-H is a signal for choriocarcinoma cell invasion, making it a biological tumor marker. Antibodies against hCG-H block tumor formation and growth. Human or humanized antibodies against hCG-H may be useful in treating and managing choriocarcinoma and other gestational trophoblastic malignancies.

INTRODUCTION

Measurement of human chorionic gonadotropin (hCG) is the basis of all pregnancy tests. hCG is produced by trophoblast cells of the placenta in pregnancy. It is also produced by trophoblast cells in gestational trophoblastic diseases (hydatidiform mole, gestational trophoblastic neoplasm (GTN) and choriocarcinoma. hCG is a glycoprotein composed of 2 dissimilar subunits, a- and b-subunit, coded by separate genes on separate chromosomes, held together by charge interactions. hCG a-subunit is composed of 92 amino acids and contains 2 N-linked oligosaccharides. hCG ß-subunit is composed of 145 amino acids and contains 2 N-linked and 4 O-linked oligosaccharides. The 8 oligosaccharide side chains comprise approximately 30% of the molecular weight of hCG, making it an exceptionally highly glycosylated glycoprotein [1-7].

hCG is a heterogeneous molecule. Peptide variants are detectable in serum and urine samples during pregnancy and gestational trophoblastic diseases [1]. Oligosaccharide variants originate from differences in the availability of sugars, variations in cellular metabolism, and the differential expression of different glycosyltransferases, the sugar-adding enzymes in cells [7-9]. It has long been recognized that the hCG molecule produced in choriocarcinoma is a larger molecule than that produced through most of normal pregnancy [10-12]. Structural studies have indicated the presence of a molecule with larger oligosaccharide side chains [3,4,13]. In 1987, a significant and consistent difference was demonstrated between the 4 O-linked oligosaccharides on hCG from choriocarcinoma patients and from those with normal first trimester pregnancy [4,5]. This observation was confirmed a year later by Amano et al [6].

In 1997 we demonstrated that the difference in the 4 O-linked oligosaccharides is the principal difference between choriocarcinoma and pregnancy hCG. While first trimester normal pregnancy urine hCG contained primarily trisaccharide O-linked oligosaccharides, and 12.3 to 19% of a larger hexasaccharide side chains (mean = 15.6%), choriocarcinoma urine hCG contained 48 to 100% (mean = 74.2%) hexasaccharide side chains [7]. As such, larger O-linked oligosaccharides characterize choriocarcinoma hCG. The hCG produced in choriocarcinoma with primarily larger sugar side chains is called hyperglycosylated hCG (hCG-H). hCG-H, and the clinical test were licensed to Nichols Institute Diagnostics Inc. [14,15]. A specific monoclonal antibody (antibody B152) was generated against hCG-H [16], and manual and automated commercial immunoassays have been established detecting only hCG-H [14, 15]. Using these assays, serum hCG-H has been shown to be an outstanding marker for differentiating active choriocarcinoma plus gestational trophoblastic malignancies, needing chemotherapy from pre-malignant cases (quiescent gestational trophoblastic disease) [17]. hCG-H is as an absolute tumor marker (100% sensitivity and specificity demonstrated) in discriminating malignant and pre-malignant disease [18].

In 1998 O’Connor et al. used the B152-based assay to show that hCG-H is not only the principal form of hCG made in GTN and choriocarcinoma, but the principal form of hCG made during the initiation of pregnancy, at the time of and in the 2 weeks following implantation [19]. This finding has now been confirmed by these and other investigators [14, 15, 17, 19-22]. A similarity has long been suggested in the invasion processes of implantation and choriocarcinoma [20, 23, 24].

Root trophoblast cells, or cytotrophoblasts, are phenotypically invasive cells. These are the principal cells in choriocarcinoma tumors and in blastocysts at the time of implantation [20, 23, 24]. Cytotrophoblast cells produce hCG-H, and differentiated syncytiotrophoblast cells produce regular hCG [14, 20].

The primary function of hCG in pregnancy is to maintain progesterone production by corpus luteal cells [25]. hCG-H is only associated with invasive events, pregnancy implantation and GTN or choriocarcinoma. It is produced by invasive cytotrophoblasts rather than syncytiotrophoblasts. The choriocarcinoma hCG immunoreactivity in the conditioned medium of JAR choriocarcinoma cells is hCG-H [15]. Studies by Lei et al. [26], show that JAR cells are invasive in Matrigel basement membrane inserts (tumor invasion models) in vitro, and are rapidly tumorigenic when transplanted into athymic nude mice in vivo. Lei et al. [26] treated JAR cells with hCG a-subunit antisense cDNA. This blocked production of hCG-H, and also blocked Matrigel membrane insert invasion in vitro and tumorigenesis in athymic nude mice in vivo. This data indicated that hCG-H has a separate role to hCG, a critical function in trophoblast cell invasion.

This article examines the biological activities of hCG-H and its specific action in growth and invasion by trophoblast cells. It also investigates the use of monoclonal antibody against hCG-H to inhibit cancer cell growth in vivo, or its potential use in treatment of GTN or choriocarcinoma.

MATERIALS AND METHODS

Patient Samples

Serum and urine samples from early pregnancy and gestational trophoblastic diseases were obtained at Yale University (pregnancy and gestational trophoblastic disease urine) under an Internal Review Board (IRB) approved protocol, from 1997 to 1999, and at University of New Mexico under a separate IRB protocol (pregnancy and gestational trophoblastic disease serum), from 2002-2004. All serum samples were collected within one hour of phlebotomy and frozen at -80ºC, and thawed for immunoassays. Serum was tested for total hCG and hCG-H.

Culture Procedures

Purified cytotrophoblast cells from term pregnancies were kindly provided by Harvey Kliman at Yale University in Dulbecco’s High Glucose medium with 10% fetal calf serum (DHG-10%). Cytotrophoblast cell were purified by Percoll density centrifugation from trypsin dispersed term pregnancy villous trophoblast tissue using the methods used by Harvey Kliman previously [27, 28].

Spent culture fluid was tested for total hCG and hCG-H.

Culture medium was collected from 70-80% confluent flasks of BeWo, JAR, and JEG-3 choriocarcinoma cell line. Cells were cultured to confluency in RPMI-1640 medium with 10% fetal calf serum (RPMI-10%). Spent culture fluid was tested for total hCG and hCG-H.

Immunoassays

Total hCG (all forms of hCG, including hCG-H, hCG and their free ß-subunits) were measured using the DPC Inc. (Los Angeles CA) Immulite hCG assay on the Immulite automated immunoassay platform. This assay is calibrated in mIU/ml against the 3rd International Standard. Values were converted to ng/ml using the previously published conversion factor [1, 15]. This test has been shown to equally recognize, on a molar basis, regular hCG, hCG-H, and their free ß-subunits [15]. hCG-H was measured using the Nichols Institute Diagnostics Inc. (San Clemente CA), hCG-H test on the Nichols Advantage automated immunoassay platform (results in ng/ml).

Matrigel invasion assays

JEG-3 choriocarcinoma cells were harvested with trypsin and EDTA. Cytotrophoblast cells of JEG-3 choriocarcinoma cells were plated onto Matrigel membranes and control inserts, 5000 cell per membrane and per control insert (Biocoat Matrigel invasion membranes, BD Biosciences, Bedford, MA 01730). Cells were cultured at 37ºC for 24 hours in DHG-10% culture fluid containing no additives (controls), and with 10 (cytotrophoblast) and 100 (JEG-3) ng/ml regular hCG (hCG batch CR127) or hCG-H (hCG-H batch C7 [7]), each condition on triplicate membranes. These concentrations, 10 ng/ml and 100 ng/ml are both approximately 4 times that normally produced by cytotrophoblast and JEG-3 cultures, 2.3 ng/ml per 5000 cells and 22.5 ng/ml per 5000 cells, respectively. Matrigel membranes were processed and percentage invasion calculated as suggested by the manufacturer in package inserts. Briefly, membranes are rehydrated in DHG-10% in the incubator for 2 hours before use. Membranes and control inserts are then plated (25,000 cells in 0.5 medium per plate). Plates are cultured for 24 hours, and membranes removed from the inserts using a scalpel. Membranes are transferred to a slide using Cytoseal mounting medium (Stephens Scientific Inc., Riverdale NJ), exposing the under surface or the invaded cells. Cells are stained with DIF-Quick Stain (IMEB Inc., Chicago IL) to mark nuclei. Invaded cells are counted at 5 marked places, and the count averaged. Cell penetration or invasion of membranes is directly compared to that of correspondingly cultured control inserts and the percentage invasion is calculated using the formula provided by the manufacturer.

Nude mouse studies

Transplantation of JEG-3 choriocarcinoma cells into nude mouse was completed at the University of New Mexico. All procedures were approved by the University of New Mexico Health Sciences Center Animal Care and Use Committee. Six to eight week old athymic BALB/c, nu/nu nude mice were purchased from Charles River Laboratories (Wilmington MA), and hosted in the University of New Mexico Health Sciences Center Animal Care facility. JEG-3 cells were grown to 70% confluence in DHG-10% and harvested with trypsin and EDTA. Approximately 10 million cells were injected subcutaneously into each of the athymic mice. Mouse monoclonal antibody B152 was reconstituted with sterile PBS at 1mg/ml concentration, and 0.3ml was given through intra-peritoneal injection. Normal mouse IgG was used as control. In the first study, to test the effect of antibody B152 on established tumors, it was given 2 weeks after subcutaneous transplantation, once tumor were widely established , and continued twice a week for up to 2 weeks, or until the largest tumor reached 2cm, the maximal tumor size set by the animal use protocol. In the second study to test the action of monoclonal antibody B152 on the tumor development, B152 was given at time of transplantation with the dose described above. The tumor cross-section area was measured with calipers before every treatment according to the formula: length × width × 3.14 ¸ 4.

Statistical analysis

All data was analyzed in a Microsoft Excel 2003 (Microsoft Inc., Redmond WA). Mean, and standard deviation (SD) were determined and Student’s t-test used to compare Matrigel invasion results and tumor cross sectional area values.

RESULTS

The occurrence of hCG-H as a component of total hCG in isolated cytotrophoblast cell and choriocarcinoma cell line culture fluids was investigated. Table 1 shows that hCG-H accounts for 43% to 100% of hCG immunoactivity in 3 independent choriocarcinoma cell lines. Cytotrophoblast cells purified from term placenta and placed into primary culture were also examined. hCG-H accounted for 100% of the total hCG produced.

Cytotrophoblasts and JEG-3 choriocarcinoma cells each produced only hCG-H, and not other forms of hCG. The ability of these cells to invade Matrigel basement membranes was examined during 24 hours culture. Table 2 shows that 48 ± 11% of choriocarcinoma cells with no additive, and 40 ± 10% of pregnancy cytotrophoblasts with no additive, penetrated the membranes. The addition of 4-fold excess hCG-H (over that normally produced by cells) to the cultures, significantly enhanced penetration. This was 88 ± 6% for JEG-3 choriocarcinoma and 66 ± 13% for pregnancy cytotrophoblast. A significant difference is observed upon enhancement with hCG-H (P=0.005 and P=0.05, respectively). By contrast, identical concentrations of hCG, instead of hCG-H, slightly reduced invasion, 38 ± 3% and 34 ± 95, respectively. This, however, was not statistically different from non-treated controls.

Human choriocarcinoma cells rapidly formed tumors when transplanted into athymic nude mice [26]. The action of hCG-H in vivo using athymic nude mice subcutaneously transplanted with JEG-3 choriocarcinoma cells was investigated (Figure 1). Subcutaneous tumors were clearly visible in mice after 2 weeks. At this time, mouse blood contained 1818 ± 1842 ng/ml (± SD) hCG-H. Mice were then either treated twice weekly with intraperitoneal injections of B152 anti-hCG-H or with a similar concentration of non-specific IgG (controls). In the time that followed, tumors rapidly doubled in size in the control group receiving no B152. In contrast, tumors ranged from just -18% to +7% of the size at the start of the study, or changed minimally in those receiving monoclonal antibody B152. A significant difference was observed by t test between animals receiving and not receiving B152 at all growth points (2.5, 3 and 3.5 weeks) P=0.003. While a clear growth trend was observed in the control group (r²=0.97), none was observed in those receiving antibody B152 (r²=0.15).

In a further experiment initial tumor formation was investigated in athymic nude mice newly transplanted with JEG-3 cells (Figure 2). Mice were treated starting immediately after transplantation with twice weekly intraperitoneal injections of B152 anti-hCG-H or with a similar concentration of non-specific IgG (controls). In the control group, tumor first appeared at 2 weeks and grew rapidly thereafter. In contrast, much smaller tumors, approximately one quarter to one eight of the size of the control group, formed in animals receiving monoclonal antibody B152. A significant difference was observed by t test between animals receiving and not receiving B152 at 2, 3 and 4 weeks, P=0.0071, 0.0031 and 0.012, respectively.

DISCUSSION

The study examined the occurrence and biological properties of hCG-H, and use of monoclonal antibodies to block hCG-H action. Previous studies using cell lines and primary cultures to show that hCG-H accounts for a high proportion of hCG in choriocarcinoma and all of the hCG immunoreactivity produced by cytotrophoblast cells, were confirmed here [14-19]. It is inferred that hCG-H is produced in invasive conditions; that is by the cytotrophoblast cells that invade the decidua at implantation of pregnancy [20], those involved in mid-trimester implantation [23], and by those involved in choriocarcinoma invasion. Malignant choriocarcinoma cell are primarily invasive cytotrophoblast cells [29 - 31]. In some cases, a varying proportion of syncytiotrophoblast cells are present [29 - 31]. This may explain why with primary cultures of placental cytotrophoblast cells (no syncytiotrophoblast present, 28) we find 100% of hCG immunoreactivity due to hCG-H, and with JEG-3 choriocarcinoma cells, 100% of hCG immunoreactivity due to hCG-H. However, with JAR and BeWo choriocarcinoma cells, or other circumstances where syncytiotrophoblast cells are also present, lower proportions of total hCG immunoreactivity are due to hCG-H. These findings confirm the studies of Kovalevskaya et al [20] showing that hCG-H is produced by cytotrophoblast cells.

Matrigel invasion model studies clearly demonstrate that hCG-H, but not regular hCG, promotes growth and invasion of membranes by pregnancy cytotrophoblast cells and choriocarcinoma cells in vitro. Furthermore, results show that mouse monoclonal antibody against hCG-H blocks growth and tumorigenesis of human choriocarcinoma cells transplanted into athymic nude mice, in vivo.

The biological function of hCG is to promote progesterone production at the corpus luteal LH/hCG receptor. hCG-H is not effective in this biological action. Choriocarcinoma hCG-H has one fifth of the steroid-promoting activity of normal pregnancy hCG [2]. It is inferred that hCG-H is a separate molecule to regular hCG. Regular hCG is optimal at the LH/hCG receptor [1, 9], while hCG-H is not. As shown here, hCG-H promotes cytotrophoblast and choriocarcinoma cell growth and invasion, while hCG does not. This is a unique endocrine situation. There are 2 genes coding for the a-subunit and ß-subunit of the hCG/hCG-H common polypeptides [7]. hCG and hCG-H are produced by separate types of cells [20] and have separate functions. hCG, is an endocrine, produced by syncytiotrophoblast cells and acting on he LH/hCG receptor on corpus luteal cells. hCG-H, by contrast, is an autocrine agent since it is secreted by choriocarcinoma cells, circulates in the blood, and is acting on the same cell as produce it (as shown by the culture models). It is a growth and invasion-promoting agent. It is not know if it acts through an LH/hCG receptor or a separate receptor. This needs to be the subject of future investigations.

The studies presented here clearly show a role for hCG-H in promoting both growth and invasion by pregnancy cytotrophoblast and choriocarcinoma cells in vivo and in vitro. These observations support and expand upon the studies by Lei et al. [26], showing that the hCG-related molecules produced by JAR choriocarcinoma cells, or hCG-H are critical to cell invasion in Matrigel membrane inserts in vitro, and to cell tumorigenesis in athymic nude mice in vivo.

A mouse monoclonal antibody against hCG-H effectively inhibited human choriocarcinoma tumor formation and tumor growth and development in athymic nude mice in vivo. Human monoclonal antibodies with the same specificity, or appropriately modified mouse antibodies (humanized antibodies), or antagonists of hCG-H could be clinically useful as a treatment for choriocarcinoma and GTN. They may be also be clinically useful in the prevention of persistence of hydatidiform mole or of recurring trophoblastic diseases. Alternatively, a segment or modified preparation of hCG-H may be useful as a vaccine in promoting endogenous production of human hCG-H antibodies to prevent persistence or recurrence of hydatidiform mole or choriocarcinoma.

hCG-H has been previously shown to be a reliable marker for differentiating active choriocarcinoma and GTN cases needing chemotherapy from pre-malignant cases (quiescent gestational trophoblastic disease) [17, 18, 32]. The use of hCG-H as an absolute tumor marker (100% sensitivity and specificity demonstrated) in discriminating malignant and pre-malignant disease, and in predicting hydatidiform moles requiring and not-requiring chemotherapy is considered in an accompanying publication [18]. Common tumor markers, such as PSA, CA125, and CEA are cell surface proteins or glycoproteins. While they signify malignancy, particularly advanced malignancy, they are byproducts of the malignancy and not the causes of the malignancy. Results presented here indicate that hCG-H is a specific promoter or stimulant of cytotrophoblast and choriocarcinoma invasion in vitro, and choriocarcinoma growth and development in vivo. Blockage of hCG-H with monoclonal antibodies obstructs choriocarcinoma tumorigenesis and growth and development in vivo. As such, hCG-H is not only a tumor marker of choriocarcinoma, but a unique marker, in that it is the signal for malignancy, or a biological marker. Can there be a better tumor marker than the signal that make the tumor progress? This is discussed in the accompanying publications [18].

REFERENCES

1. Cole LA. Immunoassay of human chorionic gonadotropin, its free subunits, and metabolites. Clin Chem 1997;43:2233-43.

2. Cole LA, Kardana A, Andrade-Gordon P, Gawinowicz MA, Morris JC, Bergert ER, O'Connor J, Birken S. The Heterogeneity of hCG: III. The occurrence, biological and immunological activities of nicked hCG. Endocrinology 1991;129:1559-67.

3. Cole LA, Hussa RO. The carbohydrate on human chorionic gonadotropin produced by cancer cells. Adv Exp Med Biol 1984;176:245-70.

4. Cole LA. O-Glycosylation of proteins in the normal and neoplastic trophoblast. Troph Res 1987;2:139-48.

5. Cole LA. The O-linked oligosaccharides are strikingly different on pregnancy and choriocarcinoma hCG. J Clin Endocrinol Metab 1987;65:811-13.

6. Amano J, Nishimura R, Mochizuki M, Kobata A. Comparative study of the mucin-type sugar chains of human chorionic gonadotropin present in the urine of patients with trophoblastic diseases and healthy pregnant women. J Biol Chem 1988;263:1157-65.

7. Elliott MM, Kardana A, Lustbader JW, Cole LA. Carbohydrate and peptide structure of the a- and b-subunits of human chorionic gonadotropin from normal and aberrant pregnancy and choriocarcinoma. Endocrine 1997;7:15-32.

8. Takamatsu S, Oguri S, Toba Minowa M, Yoshida A, Nakamura K, Takeuchi M, Kobata A.

Unusally High Expression of N-Acetylglucosaminyltransferase-IVa in Human Choriocarcinoma Cell Lines: A Possible Enzymatic Basis of the Formation of Abnormal Biantennary Sugar Chain

Cancer Res 1999;59:3949-3953.

9. Kobata A, Takeuchi M. Structure, pathology and function of the N-linked sugar chains of human chorionic gonadotropin. Biochim Biophys Acta. 1999;1455:315-26

10. Peters BP, Krzesicki RF, Hartle RJ, Perini F, Ruddon RW A kinetic comparison of the processing and secretion of the alpha beta dimer and the uncombined alpha and beta subunits of chorionic gonadotropin synthesized by human choriocarcinoma cells. J Biol Chem. 1984;259:15123-30.

11. Hussa RO. Immunologic and physical characterization of human chorionic gonadotropin and its subunits in cultures of human malignant trophoblast. J Clin Endocrinol Metab 1977;44:1154-62.

12. Mann K, Karl HJ. Molecular heterogeneity of human chorionic gonadotropin and its subunits in testicular cancer. Cancer 1983;52:654-60.

13. Imamura S, Armstrong GA, Birken S, Cole LA, Canfield RE. Detection of desialylated forms of human chorionic gonadotropin. Clin Chim Acta 1987;163:339-49.

14. Cole LA, Shahabi S, Oz UA, Bahado-Singh RO, Mahoney MJ. Hyperglycosylated human chorionic gonadotropin (invasive trophoblast antigen) immunoassay: A new basis for gestational Down syndrome screening. Clin Chem 1999;45:2109-19.

15. Cole LA, Khanlian SA, Sutton JM, Davies S, Stephens N. hCG-H (Invasive Trophoblast Antigen, hCG-H) a Key Antigen for Early Pregnancy Detection. Clin Biochem, 2003;36:647-655

16. Birken S, Krichevsky A, O’Connor J, Schlatterer J, Cole LA, Kardana A, Canfield R. Development and characterization of antibodies to a nicked and hyperglycosylated form of hCG from a choriocarcinoma patient: generation of antibodies that differentiate between pregnancy hCG and choriocarcinoma hCG. Endocrine 1999;10:137-44.

17. Khanlian SA, Smith HO, Cole LA. Persistent Low Levels of hCG: A Pre-malignant gestational trophoblastic disease. Am J Obstet Gynecol, 188: 1254-1259, 2003.

18. Cole LA, Butler SA, Khanlian SA, Giddinbgs A, Seckl MJ, Kohorn EI. Gestational trophoblastic diseases: 2. Hyperglycosylated hCG as a reliable marker of active neoplasia. Gyn Oncol, (next paper, please update reference)

19. O’Connor JF, Ellish N, Kakuma T, Schlatterer J, Kovalevskaya G. Differential urinary gonadotrophin profiles in early pregnancy and early pregnancy loss. Prenat Diagn 1998;18:1232–40.

20. Kovalevskaya G, Genbacev O, Fisher SJ, Caceres E, O’Connor JF. Trophoblast origin of hCG isoforms: cytotrophoblasts are the primary source or choriocarcinoma-like hCG. Mol Cell Endocrinol 2002;94:147-55.

21. Butler SA, Khanlian SA, Cole LA. Detection of early pregnancy forms of human chorionic gonadotropin by home pregnancy test devices. Clin Chem 2001;47:2131-06.

22. Cole LA, Khanlian SA, Sutton JM, Davies S, Rayburn WF. Accuracy of home pregnancy tests at the time of missed menses. Am J Obstet Gynecol 2004;190:100-05.

23. Genbacev O. DiFederico E. McMaster M. Fisher SJ. Invasive cytotrophoblast apoptosis in pre-eclampsia. Human Reproduction. 1999;2:59-66.

24. Tarrade A. Goffin F. Munaut C. Lai-Kuen R. Tricottet V. Foidart JM. Vidaud M. Frankenne F. Evain-Brion D. Effect of matrigel on human extravillous trophoblasts differentiation: modulation of protease pattern gene expression. Biology of Reproduction. 2002;67:1628-37

25. Friedman MH, Lapham ME. A simple rapid procedure for the laboratory diagnosis of early pregnancies. Am J Obstet Gynecol 1931;21:405-410

26. Lei ZM, Taylor DD, Gercel-Taylor C, Rao CV. Human chorionic gonadotropin promotes tumorigenesis of choriocarcinoma JAR cells. Troph Res 1999;13:147-59.

27. Kliman H.J., Nestler, J.E., Sermasi, E., Sanger, J.M., Strauss, J.F. Purification, characterization, and in vivo differentiation of cytotrophoblasts from human term placentae. Endocrinology 118:1567-1582, 1986.

28. Kliman H.J., Feinman, M.A., Strauss, J.F. Differentiation of human cytotrophoblasts into synctiotrophoblasts in culture. Trop Res 2:407-421, 1987.27.

29. Lala PK, Graham CH. Mechanisms of trophoblast invasiveness and their control: the role of proteases and protease inhibitors. Cancer Metast Rev 1990; 9: 369-379.

30. Strickland S, Richards WG. Invasion of the Trophoblasts. Cell 1992; 71:355-357.

31. Paradinas FJ, Sebire NJ, Rees HC Pathology In: Gestational Trophoblastic Disease 2nd edition, eds. Hancock BW, Newlands ES, Berkowitz RS and Cole LA, Sheffield University Press, 175-181, 2003 (internet: http://www.isstd.org/gtd/index.html)

32. Cole LA, Khanlian SA Inappropriate management of women with persistent low hCG results. J Reprod Med 2004; 49: 423-432.

Table 1. Production of total hCG and hCG-H by choriocarcinoma cell lines and by term pregnancy isolated cytotrophoblast cells, after 24 hours culture in defined medium. The proportion of immunoreactivity due to hCG-H is calculated as hCG-H ¸ total hCG.

Source	Total hCG ng/ml	hCG-H ng/ml	Proportion hCG-H

BeWo choriocarcinoma cell line (30000 cells seeded)	1962	841	43%
JAR choriocarcinoma cell line (30000 cells seeded)	140	126	90%
JEG-3 choriocarcinoma cell line (30000 cells seeded)	150	165	100%

Pregnancy isolated cytotrophoblast cells (5000 cells seeded)	2.3	2.3	100%

Table 2. Action of hCG-H and regular hCG on cytotrophoblast invasion and choriocarcinoma cell invasion of Matrigel membranes. Isolated cytotrophoblast cells were prepared from term placenta. Cytotrophoblasts and JEG-3 choriocarcinoma cells were separately cultured 24 hours (5000 cells) on Matrigel basement membranes and control inserts in triplicate. Concentrations of hCG and hCG-H used to promote invasion were approximately 4 times that (ng/ml per 1000 cells) normally produced by the cells, 10 ng/ml for term pregnancy cytotrophoblast and 100ng/ml for JEG-3 choriocarcinoma cells. The underside of Matrigel basement membranes, containing penetrated or invaded cells, was stained and counted. Cell penetration was compared with that of control inserts. The percentage penetration or invasion was calculated using the formula described by the manufacturer.

	Mean penetration ± standard deviation
A) Pregnancy cytotrophoblasts
Control cultures	40 ± 10%
hCG-H, 10 ng/ml	66 ± 13% ^a
Regular hCG, 10 ng/ml	34 ± 9% ^b

B) JEG-3 choriocarcinoma cells
Control cultures	48 ± 11%
hCG-H, 100 ng/ml	88 ± 6% ^a
Regular hCG, 100 ng/ml	38 ± 3% ^b

^a A significant difference was observed by t test in % penetration between control cultures and those with added hCG-H in both cell sources, P=0.05 (cytotrophoblast cells) and P=0.005 (JEG-3 choriocarcinoma cells).

^b While no significant difference was observed in % penetration between control cultures and those with added regular hCG in either cell source, a significant difference was recorded by t test between those with added regular hCG and added hCG-H, P=0.025 (cytotrophoblast cells) and P=0.0002 (JEG-3 cells).

Figure 1. Effect of anti-hCG-H antibody B152 on tumor growth and progression. Athymic nude mice were subcutaneously transplanted with JEG-3 choriocarcinomas cells. After subcutaneous tumor was clearly visible (2 weeks), mice were either treated with intraperitoneal injections, twice each week, with non-specific IgG (controls, solid diamonds and solid line) or with B152 anti-hCG-H antibody (solid squares, dashed lines). Results were averages with 6 mice. In the control group, relative tumor size was 100%, 107±22%, 142±43% and 206±53%, and in those given B152 was, 100%, 82±11%, 92±11% and 108 ±11%, respectively, for weeks 2, 2.5, 3 and 3.5 following transplantation. By t test a significant difference was noted between all the changes at all time points (2.5, 3 and 3.5 weeks) with the B152-treated and the control mice (P=0.003). A correlation between time and growth was observed with the control group (r²=0.97), but no growth trend was observed with the B152-treated mice (r²=0.15).

Figure 2. Effect of anti-hCG-H antibody B152 on tumorigenesis. Athymic nude mice were transplanted subcutaneously with JEG-3 choriocarcinomas cells. At the time of transplantation, mice were treated with either intraperitoneal injections, twice each week, with non-specific IgG (controls, solid diamonds and solid line) or with B152 anti-hCG-H antibody (solid squares, dashed lines). Results are average results with 11 mice. In those given non-specific IgG, cross section size of tumor was 0, 0, 79±58, 121±68 and 149±98 mm², respectively. In those given B152 was 0, 0, 13±7.6, 27±15 and 43±22 mm², respectively, for weeks 0, 1, 2, 3 and 4 following transplantation. A significant difference was observed by t test at 2, 3 and 4 weeks, P=0.0071, 0.0031 and 0.012, respectively.