Expression of HOX gene products in normal and abnormal trophoblastic tissue

Available online at www.sciencedirect.com R

Gynecologic Oncology 90 (2003) 512–518

www.elsevier.com/locate/ygyno

Expression of HOX gene products in normal and abnormal trophoblastic tissue Lawrence S. Amesse,a,b,* Robert Moulton,a Yue Mei Zhang,a and Teresa Pfaff-Amessea a

Department of Obstetrics and Gynecology, Wright State University School of Medicine, Dayton, OH 45409, USA Division of Reproductive Endocrinology, Wright State University School of Medicine, Dayton, OH 45409, USA

b

Received 18 December 2002

Abstract Objective. The expression pattern of three homeobox genes products, HOX A11, HOX B6, and HOX C6, was examined in normal human placental tissue and abnormal trophoblastic tissue derived from complete hydatidiform moles and choriocarcinoma tumors. We sought to determine whether expression of these gene products during different states of trophoblastic differentiation and proliferation is constant or demonstrates variation. Variation in expression of these respective homeobox genes may provide insight into predicting which molar tissues are likely to develop into choriocarcinoma tumors. Methods. Tissue sections from a total of 12 samples were studied. Among these, six full-term human placentas, three complete hydatidiform moles, and three choriocarcinoma tumors were examined for expression of the homeobox HOX A11, HOX B6, and HOX C6 gene products, using immunohistochemistry staining methods. Results. Expression of HOX homeobox gene products, HOX A11, HOX B6, and HOX C6, was detected in full-term human placenta and tissue from complete hydatiform moles. Abnormal trophoblasts from complete moles demonstrated an immunoreactivity expression pattern comparable to that of normal trophoblasts from term pregnancies. However, definitive expression of these respective homeobox genes was not identified in tissue obtained from choriocarcinoma tumors. Conclusion. Variation in expression of HOX homeobox gene products, HOX A11, HOX B6, and HOX C6, was established in trophoblast tissue obtained from full-term human placentas, complete hydatiform moles, and choriocarcinoma tumors. This finding indicates that normal full-term trophoblasts and abnormal molar trophoblasts may share similar fundamental regulatory control mechanisms. The absence of definitive expression of these HOX gene products in trophoblastic cells derived from choriocarcinoma tumors indicates that while HOX A11, HOX B6, and HOX C6 genes may be involved in maintenance of some trophoblastic cell states, they may be either downregulated or have alterations in their expression in trophoblasts from choriocarcinoma tumors. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Complete hydatiform mole; Choriocarcinoma; Homeobox genes; Placenta

Introduction Homeobox genes are a family of regulatory genes that encode nuclear proteins, which in turn act as transcription factors. These homeoprotein transcription factors possess the capability of autoregulation and can also regulate other

* Corresponding author. Wright State University School of Medicine, Department of Obstetrics and Gynecology, 3800 CHE, Miami Valley Hospital, 128 Apple Street, Dayton, OH 45409, USA. Fax: ⫹1-937-2227255. E-mail address: [email protected] (L.S. Amesse).

homeobox genes [1,2]. Their mechanism of regulation is at the level of transcription and homeoproteins can either activate or inactivate downstream genes. Although initially described as control genes regulating pattern formation in developing embryos, homeobox genes are now known to be expressed in specific adult tissues, placenta, and cancers, in addition to playing a critical role in the regulation of cellular differentiation at many levels [3]. Two classes of homeobox genes are expressed in the mammalian placenta—HOX and non-HOX or divergent homeobox genes [2– 8]. Several members of the non-HOX homeobox gene family have been isolated from human placenta libraries and studied in detail

0090-8258/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0090-8258(03)00357-3

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

[7–9]. However, the role of HOX homeobox genes in trophoblastic differentiation and placental development has not been as well investigated [3,10]. Recent studies describing the presence of HOX genes in human trophoblast tissue suggest a role for HOX homeobox genes in trophoblast differentiation [11]. Indeed, they represent potential candidate genes for regulating trophoblast differentiation. They may also possess oncogenic properties associated with aberrant trophoblast cells as seen in gestational trophoblastic diseases. The biological behavior of trophoblast cells shares many striking similarities with the behavior of malignant cells. Indeed, trophoblast cells rapidly proliferate and infiltrate locally into maternal uterine tissue and endovasculature. Occasionally, trophoblasts can metastasize to distant sites, such as the lung, brain, liver, kidney, and bowel. In the normal or maintenance state, these processes are tightly regulated and self-limited, often ending with the culmination of pregnancy. However, when trophoblast invasion becomes unregulated, severe pathological conditions can result. Excessive trophoblast invasion can lead to deficient decidualization, resulting in such pregnancy-related conditions as placenta accreta, placenta increta, and placenta percreta. Moreover, aggressive pathological conditions such as gestational trophoblastic diseases can also result [12]. The putative role of homeobox HOX genes in oncogenesis is an emerging area of investigation. In various types of human leukemias, distinctive patterns of HOX gene expression have been described, while other HOX genes have expressed leukemia-associated translocations [13,14]. A number of recent reports have detected in solid tumors various alterations in HOX gene expression. As a brief overview, altered HOX gene expression has been reported in human primary carcinomas, including kidney and colon, as well as in metastatic colorectal cancer to the liver [15,16]. Clonal populations isolated from a metastatic human malignant melanoma cell line have also demonstrated aberrant HOX gene expression. These clonal populations have been associated with specific integrin and ICAM profiles [17]. Intriguingly, inactivation of multiple HOX genes, rather than misexpression, has been correlated with progression of the human primary tumor, small cell lung cancer [18]. Finally, expression of some HOX genes has assisted in the histological subclassification of some kidney cancers and Wilms’ tumors [19 –22]. Expression of the HOX A11 gene has recently been reported to be downregulated as human cytotrophoblastic cells differentiate into syncytiotrophoblasts [11]. Homeobox HOX B6 and HOX C6 genes have also been described in human placental tissue [3,11]. Whether these or other HOX genes affect oncogenic transformation of trophoblastic cells associated with gestational trophoblastic disease states is presently unknown. Gestational trophoblastic diseases (GTD) encompass a heterogeneous group of pregnancy-associated disorders that reflects a continuum of trophoblastic abnormalities. Each disorder along the contin-

513

uum represents a different derangement in trophoblast proliferation. Some members of GTD include complete and partial hydatiform moles, invasive mole, placental site trophoblastic tumor, and choriocarcinoma [23]. While their role in the malignant transformation of trophoblastic cells has not been elucidated, expression and misexpression of homeobox genes HOX A11, HOX B6, and HOX C6 have been reported in some human malignancies. Indeed, in vitro expression of HOX A11 in human chronic myelogenous leukemia, expression of HOX B6 in myeloid leukemia and misexpression at various stages of colonic cancer, and the demonstration of HOX C6 gene products in human breast cancer suggests that these genes function at some capacity during oncogenesis [24 –26]. Although expressed in trophoblastic tissue, there have been no reports of these three homeobox genes in association with GTDs, and this potential relationship is intriguing. Determining expression patterns of HOX A11, HOX B6, and HOX C6 gene products at various stages of trophoblastic regulation, differentiation, and proliferation as well as in neoplastic transformation, may provide insight into the function of these HOX genes in normal and aberrant trophoblastic tissue states, especially GTDs. Further, if variation in their gene expression exists, it may be possible to predict prognostic differences in determining which moles progress to choriocarcinoma. In this study, we sought to determine the expression pattern of three HOX homeobox gene products, HOX A11, HOX B6, and HOX C6, in various human trophoblast tissues that reflected normal and abnormal stages in trophoblast differentiation and proliferation, including full-term placentas and GTD tissue, complete hydatidiform moles, and choriocarcinomas.

Material and methods Tissue collection This study was conducted at the Miami Valley Hospital after institutional review board approval of tissue collection and preparation and informed consent was obtained. Placental tissue samples were collected from women with uncomplicated, singleton, third-trimester pregnancies, either after vaginal or elective cesarean section deliveries (n ⫽ 6). Samples from women with complete hydatidiform moles were obtained during evacuation procedures and one hysterectomy specimen (n ⫽ 3). Tissue samples from patients with choriocarcinoma tumors were collected from surgical specimens (n ⫽ 3). Immunohistochemistry All tissue was routinely fixed in 10% buffered formalin (pH 7.2), paraffin embedded, sectioned at a thickness of 4 ␮m, and mounted on poly-L-lysine coated glass slides. Sections were dried in an oven overnight at 37 °C and depar-

514

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

affinized by immersion into xylene for three changes at 15–20 min per change. Rehydration was performed by immersion through descending dilutions of ethanol followed by three rinses in PBS. Endogenous peroxidase activity was quenched using 3% hydrogen peroxide in methanol. Sections were immersed into 0.5% Triton for 30 min followed by three rinses in PBS. Blocking of nonspecific binding was achieved by applying normal goat serum diluted in PBS to each section followed by incubation in a moist chamber for 60 min at room temperature. Excess serum was removed by blotting the area directly surrounding the tissue sections. Immunohistochemistry was performed using the avidin– biotin complex immunoperoxidase method (Vectastain® Elite ABC Kit; Vector Laboratories, Burlingame, CA). Polyclonal rabbit antibodies for HOX A11, HOX B6, and HOX C6 gene products were generated against peptide sequences specific to the various epitopes of the respective target gene products and were purchased from BABCo (Berkley, CA). All primary antibodies were applied at a dilution of 1:250 in 0.5% bovine serum albumin in PBS to each tissue section, except for the negative control slides. The sections were incubated in a moist chamber for 60 min followed by three PBS rinses. Secondary antibody, using diluted biotinylated goat anti-rabbit IgG antibody (1:200), was applied to all sections, which were incubated in a moist chamber for 45 min, followed by three rinses in PBS. Next, Vectastain® Elite Reagent was applied to all sections. After a 45-min incubation at room temperature, the sections were rinsed three times in PBS. For color visualization of the primary antigen–antibody complex, peroxidase substrate solution, (DAB stain kit, Vector Labs, Burlingame, CA) was applied to all sections followed by incubation in a dark, moist chamber for 5 min. After rinsing in tap water, a light hematoxylin counterstain was applied and the sections were dehydrated through graded concentrations of ethanol and cleared through three changes of xylene. Mounting media and coverslips were applied and the sections were examined by light microscopy. Sections from normal full-term human placenta with known positive immunoreactivity for HOX A11, HOX B6, and HOX C6 were used as the positive control tissue [11]. Sections from each sample of human placenta, complete hydatiform moles, and choriocarcinoma tissue were used as negative controls with the primary antibody replaced with Tris-buffered saline. Assessment of immunostaining Using an Olympus microscope, quantitative and qualitative evaluations of immunohistochemical staining were conducted on all tissue samples by two experienced pathologists (YMZ and TPA). Quantitative evaluation was performed by counting between 200 and 300 target cells from five separate, distinct areas of each tissue sample and determining the percentage of HOX A11, HOX B6, and HOX C6 antigen positive cells. Only areas representing

Table 1 Clinical pathologic features of placenta tissue Patient age (y)

Gestational age (weeks)

Mode of delivery

Diagnosis

28 25 20 19 19 28

38 40 37 40 37 37

C/S Vaginal C/S C/S C/S Vaginal

3rd TM placenta 3rd TM placenta 3rd TM placenta 3rd TM placenta 3rd TM placenta Early 3rd TM placenta

Note. C/S, cesarean section; TM, trimester.

viable, morphologically characteristic tissue were examined, while necrotic, hemorrhagic, and artifactual areas were entirely avoided. The following schema was applied: NR or (⫺) ⫽ no evidence of immunoreactivity or negative staining; (⫹/⫺) ⫽ positive immunoreactivity in ⬍1% of the total cell number; ⫹ ⫽ positive immunoreactivity in 1%– 10% of the total cell number; ⫹⫹ ⫽ positive immunoreactivity in 10%–50% of the total cell number; ⫹⫹⫹ ⫽ positive immunoreactivity in 50%–90% of the total cell number; and ⫹⫹⫹⫹ ⫽ positive immunoreactivity in ⬎90% of the total cell number. Various staining parameters were examined for the qualitative evaluation. The character and quality of positive immunoreactivity in cytotrophoblasts, intermediate cytotrophoblasts, and syncytiotrophoblasts were categorized as follows: focal or diffuse staining; mild, moderate, or strong immunoreactivity; and coarse or faint granularity. The specific location of positive immunoreactivity in the target cells was noted as follows: predominately cytoplasmic, thick, uniform nuclear membrane staining, nucleolar staining, and/or endovascular staining.

Results Sample classification The clinical–pathological features of the placental tissue included in this study are summarized in Table 1. The six examined placentas were obtained from women with uncomplicated, singleton pregnancies. The age of the women ranged from 19 to 28 years with an average age of 23 years. The gestational age of the placentas ranged from 37 to 40 weeks with an average gestation of 38 weeks. Four placentas were retrieved from elective cesarean section deliveries and two from normal vaginal deliveries. All were histologically third-trimester placentas without evidence of inflammatory infiltrate, chorioamnionitis, or other significant pathologic changes. In Table 2 the clinical–pathological characteristics of the complete hydatidiform molar tissue are summarized. A total of three different molar specimens were examined. The age of the patients ranged from 23 to 51 years with an average

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

515

Table 2 Clinical–pathologic features of complete hydadiform moles Patient age (y)

Preoperative Dx

Postop Dx and pathology

Ploidy status

23 32 51

Missed abortion Molar pregnancy Endometrial carcinoma/sarcoma

CHM, noninvasive; persistent CHM, noninvasive CHM, noninvasive

Diploid Diploid Diploid

Note. Postop, postoperative; Dx, diagnosis; CHM, complete hydatidiform mole.

age of 35 years. The gestational age of the molar pregnancies ranged from 10 to 15 weeks. The preoperative diagnosis in one woman was a missed abortion. In another patient, the presumptive diagnosis was endometrial carcinoma or uterine leiomyosarcoma. In only one case was the preoperative diagnosis a molar pregnancy. Pathologic diagnosis revealed complete hydatidiform mole for all three cases. In one case, a subsequent diagnosis of persistent mole was rendered based on persistent serum hCG levels and tissue derived from a second evacuation. Nuclear DNA index cell cycle analysis revealed diploid populations in all three complete moles. Parental origin studies were not routinely performed on these specimens. Table 3 summarizes the clinical–pathological features of the tissue from choriocarcinoma tumors. The age of the patients ranged from 31 to 35 years with an average age of 33 years. The tissue was obtained from two females and one male. The preoperative diagnosis was a hemothorax with lung mass extending into the chest wall in one patient whose ␤-hCG level was markedly elevated at ⬎500,000 mIU/ml. In another patient, a testicular mass was identified. The preoperative diagnosis was gestational trophoblastic disease in one case. Two cases were pure choriocarcinomas while the testicular mass was diagnosed as a mixed choriocarcinoma and embryonal tumor. The choriocarcinoma diagnosed in the 31-year-old female revealed deep myometrial and vascular invasion.

fusely positive immunoreactivity for all three gene products (Fig. 1B–D). The same expression pattern was identified in cytotrophoblast, intermediate trophoblast, and syncytiotrophoblast cells from third-trimester placental tissue. However, between 50% and 90% (or ⫹⫹⫹) of the targeted trophoblast cells displayed moderate, diffusely positive immunoreactive staining for all three homeobox gene products (Fig. 2B–D). In both placenta and complete molar tissue, the positive immunostaining was located predominately in the cytoplasm with coarse granularity evident. Occasional positive, thick, and uniformly circumferential nuclear membrane staining was identified. Some nucleoli demonstrated prominent, strongly positive immunoreactivity. When present, endovascularly located trophoblasts also demonstrated strongly positive staining (Fig. 1C). In the complete hydatidiform molar tissue, positive immunoreactivity staining for all three HOX homeobox gene products, HOX A11, HOX B6, and HOX C6, was more widespread and intense when compared to trophoblasts in the third-trimester placenta tissue. A definitive expression pattern for HOX A11, HOX B6, and HOX C6 gene products was not observed immunohistochemically in the human choriocarcinoma tumor specimens (Fig. 3B–D). Trophoblasts from this tissue demonstrated negative immunoreactivity for these proteins. Stromal tissue from all specimens also demonstrated negative immunoreactivity for HOX A11, HOX B6, and HOX C6 gene products.

Immunohistochemistry analysis In the present study, mononuclear cytotrophoblast cells, intermediate trophoblasts, and multinucleated syncytiotrophoblast cells obtained from complete molar tissue demonstrated an expression pattern for all three homeobox, HOX A11, HOX B6, and HOX C6, gene products. In all examined complete hydatidiform molar samples, more than 90% (or ⫹⫹⫹⫹) of the target cells demonstrated strong, dif-

Discussion Expression of homeobox genes HOX B6, HOX C6, and, more recently, HOX A11 has been reported in human placenta tissue [3,11,27]. This study demonstrates for the first time the differential expression pattern of three HOX homeobox gene products in trophoblast tissue obtained from

Table 3 Clinical pathologic features of choriocarcinomas (y)

Gender

Preoperative Dx

Postop Dx and pathology

33 35 31

Female Male Female

Hemothorax, lung mass w/chest wall ext. testicular mass GTD

Choriocarcinoma Mixed choriocarcinoma and embryonal carcinoma Choriocarcinoma; Myometrial and vascular invasion

Note. Ext, extension; GTD, gestational trophoblastic disease.

516

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

Fig. 1. Complete hydatidiform molar tissue. (A) Hydropic villi contain foci of projecting areas of trophoblastic proliferation alternating with attenuated trophoblasts (H&E). (B) Strongly positive immunoreactive staining is revealed for HOX A11. Inset demonstrates, in detail, trophoblastic proliferation and immunostaining. (C) Expression of HOX B6 gene products in hydropic villi where the trophoblasts are predominately attenuated. Inset shows strongly positive immunostaining of intermediate trophoblasts surrounding a vascular space where the endothelial cells have been replaced by trophoblasts. (D) HOX C6 immunostaining of hydropic villi is demonstrated. Inset shows detail of the strong immunoreactivity of trophoblastic proliferation.

normal term placentas and two distinct forms of GTD. The results from this study are consistent with other reports in which homeobox genes appear to regulate extraembryonic tissue or placental development [7–9,27,28]. Because placenta development is nonsegmental, the hypothesis that homeobox genes may serve in more diversified roles is strengthened. Remarkably, each of the three genes under investigation has also been implicated in various human malignancies [24 –26]. This suggests a potential role for these genes not only in trophoblast regulation, but also in neoplastic transformation.

Fig. 2. Full-term placenta demonstrating mature chorionic villi, syncytrophoblasts, intermediate trophoblasts, and inconspicuous cytotrophoblasts. Syncytial knots are evident. (A) H&E. Moderately positive immunoreactive staining for HOX A11, HOX B6, and HOX C6 gene products is demonstrated in B, C, and D, respectively.

Fig. 3. Choriocarcinoma characterized by clusters of cytotrophoblasts separated by sheets of syncytiotrophoblasts. (A) H& E. Negative immunoreactive staining is demonstrated for HOX A11, HOX B6, and HOX C6 gene products in B, C, and D, respectively (high power).

The expression pattern of HOX A11, HOX B6, and HOX C6 gene products in complete hydatidiform molar tissue was characterized by ⬎90% of the target cells demonstrating strong, positive, diffuse immunoreactivity. The location of the staining was predominately cytoplasmic with intermittent immunostaining of nuclear membranes and nucleoli also evident. The strong expression pattern displayed by cytotrophoblasts, intermediate trophoblasts, and syncytiotrophoblasts found in complete hydatidiform molar tissue was in direct contrast to the expression pattern in malignant trophoblasts from choriocarcinoma tissue. Target cells from that tissue were notable for the absence of any demonstrable positive immunoreactivity. The expression pattern of HOX A11, HOX B6, and HOX C6 gene products in third-trimester placental tissue was similar to that observed in the molar tissue. Between 50% and 90% of cytotrophoblastic, intermediate trophoblasts, and syncytiotrophoblastic cells demonstrated moderately positive, diffuse immunoreactivity with a location similar to that observed in complete hydatidiform molar trophoblasts. This qualitative level of expression in trophoblast cells from full-term placentas most likely reflects the usual or baseline expression patterns for these genes during normally regulated placental growth. The similarity in expression patterns of homeobox HOX A11, HOX B6, and HOX C6 gene products in normal and abnormal trophoblasts from third-trimester placenta and molar tissues, respectively, suggests that many of the developmental control mechanisms found in maintaining the normal, mature placenta are also present in complete hydatidiform moles, but may either be inactivated or downregulated in choriocarcinoma. Term placenta trophoblasts are morphologically similar to those in early to midgestation, but differ predominately in volume and number as well as by some functions. With the exclusion of intermediate trophoblast subpopulations, most have undergone terminal differentiation by the second tri-

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

mester. However, trophoblasts from normal gestations are fundamentally different from abnormal trophoblasts found in gestational trophoblastic diseases; such as complete hydatidiform moles and choriocarcinoma tissues used in this study. These fundamental differences include changes in morphologic features, variations in cellular regulation and differentiation, and genetic composition. Abnormal trophoblasts from complete moles are similar morphologically to trophoblasts found in the early developing placenta and implantation site. However, when compared to normal placental trophoblasts at gestational ages equivalent to those associated with hydatidiform moles, they differ by a number of features. Briefly, abnormal molar trophoblasts demonstrate cytological atypia, have an increased propensity for proliferation, show markedly elevated apoptosis levels, and reveal some distinguishing immunoreactive phenotypes [29,30,31]. Malignant trophoblasts found in choriocarcinoma reflect a highly primitive cell state. They morphologically recapitulate early previllous trophoblasts of the implanting blastocyst, and because of alterations in their regulatory machinery, invade in an aggressive and highly destructive manner. Choriocarcinoma tumors in this study failed to express the same HOX gene products that in complete molar tissues exhibited unequivocally strongly positive immunoreactivity and in term placentas, demonstrated moderately strong immunointensity. The coordinated expression of genes involved in trophoblast development, differentiation, and proliferation, in addition to alterations involving neoplastic transformation, may involve transcription factors including the HOX gene family. It is possible that HOX gene-derived homeoproteins orchestrate this process by either activating or inactivating downstream effector genes. In molar disorders, derangements in the normal regulatory mechanisms controlling trophoblast development result in pathologic proliferative diseases. Hydropic chorionic villi are surrounded by large sheets of hyperplastic trophoblasts, which secrete hCG. Clinically, molar diseases are often associated with the development of preeclampsia during the first trimester of pregnancy. Most molar disorders are selfcontained and respond well to reduction in tissue mass. However, in approximately 17% of all complete moles, invasive moles develop and in a smaller proportion, 2%, choriocarcinoma develops [32]. The regulatory mechanisms responsible for this critical transition have not been fully examined. Epidemiologic studies indicate that partial moles, characterized by triploidy, i.e., one maternal plus two paternal sets of chromosomes along with an abnormal fetus, rarely become invasive or metastatic [23,33]. This is in contrast to invasive moles, which nearly always arise from complete moles, occurring in approximately 16% of all cases [33]. Factors determining which complete mole will transform into a highly invasive disease or choriocarcinoma are unknown in part, because few of such molecular studies have been conducted and gestational trophoblastic diseases are relatively rare [34].

517

Modulation of HOX A11, HOX B6, and HOX C6 gene products may be important in the pathogenesis of gestational trophoblastic diseases. Because HOX genes are vital to normal development and play various roles in oncogenesis and tumor progression, it is likely that a parallel role also exists for HOX genes in trophoblasts [1,13–22,24 –26]. Understanding the regulation of these transcription factors may be crucial in determining the molecular basis of neoplastic transformation in trophoblast cells. A potential relationship between the differential expression pattern of HOX A11, HOX B6, and HOX C6 and the development of trophoblastic tumors may exist, and this relationship may be important in predicting which complete moles will undergo malignant transformation into choriocarcinoma. The absence of HOX A11, HOX B6, and HOX C6 gene product expression in malignant trophoblast cells from choriocarcinoma tumors may provide insight into the underlying mechanisms involved in the pathogenesis of gestational trophoblastic disease. HOX developmental control genes are usually expressed midgestation and are involved in pattern formation in the developing embryo [6]. The absence of definitive expression of developmental control genes, such as the HOX homeobox genes in this study, suggests that the development of choriocarcinoma may precede the expression of trophoblastic-specific HOX genes. It is also possible, and perhaps more intriguing, that these HOX genes function at the level of cellular maintenance in trophoblast differentiation and when downregulated, misexpressed, or inactivated, neoplastic transformation of trophoblastic cells results. An analogous situation exists in human small cell lung cancer. Tumor cells from this primitive and aggressive tumor originate from neural crest derivatives and tumor progression has been correlated with inactivation of multiple HOX genes [18]. Should this hypothesis prove correct, then examining molar tissue to determine the quantitative expression patterns of these HOX gene products may provide predictive indicators for which mole has a molecular or genetic predilection for malignant transformation. The specific stages in gestation when trophoblastic-specific HOX genes are expressed have not been established. Knowledge of their expression timeline would provide insight into various pregnancy-related disorders ranging from preeclampsia to choriocarcinoma. Additional studies directed at examining placental tissue at various stages of embryonic and fetal development will be necessary for definitive elucidation of placental-specific HOX genes function.

Acknowledgments The authors acknowledge Drs. Daniel Hood from Miami Valley Hospital for his assistance and Neil Rote, Cleveland Metro Health Center, for his advice.

518

L.S. Amesse et al. / Gynecologic Oncology 90 (2003) 512–518

References [1] Xu F, Li H, Jin T. Cell type-specific autoregulation of the caudalrelated homeobox gene Cdx-2/3. J Biol Chem 1999;274:34310 – 6. [2] Arcioni L, Simeone A, Guazzi S, Zappavigna V, Boncinelli E, Mavilo F. The upstream region of the human homeobox gene HOX 3D is a target for regulation by retinoic acid and hox homeoproteins. EMBO J 1992;11:265–77. [3] Su BC, Strand D, McDonough P, McDonald JF. Temporal and constitutive expression of homeobox-2-gene (Hu-2), human heat shock gene (hsp-70), and oncogenes C-sis and N-myc in early trophoblast. Am J Obstet Gynecol 1988;159:1195–9. [4] McGinnis W, Levine MS, Hafen E, Kuroiwa A, Gehring WJ. A conserved DNA sequence in homeotic genes of the Drosphilia Antennapedia and bithorax complexes. Nature 1984;308:428 –33. [5] McGinnis W, Garber RL, Wirz J, Kuroiwa A, Gehring WJ. A homologous protein-coding sequence in Drosophila homeotic genes and its conservation in other metazoans. Cell 1984;37:403– 8. [6] Ruddle FH, Bartels JL, Bentley KL, Kappen C, Murtha MT, Pendleton JW. Evolution of HOX genes. Ann Rev Genet 1994;28:423– 42. [7] Quinn LM, Johnson BV, Nicholl J, Sutherland GR, Kalionis B. Isolation and identification of homeobox genes from the human placenta including a novel member of the Distal-less family, DLX4. Gene 1997;187:55– 61. [8] Quinn LM, Latham SE, Kalionis B. The homeobox gene MSX2 and MOX2 are candidates for regulating epithelial-mesenchymal cell interactions in the human placenta. Cell Biol 1995;21:550 – 4. [9] Quinn LM, Latham SE, Kalionis B. Homeobox gene HB24, a regulator of hematopoiesis, is a candidate for regulating differentiation of the extraembryonic trophoblast cell lineage. Reprod Fertil Dev 1997; 9:617–23. [10] Sharpe PT, Miller JR, Evans EP, Burtenshaw MD, Gaunt SJ. Isolation and expression of a new mouse homeobox gene. Development 1988; 102:397– 407. [11] Zhang YM, Xu B, Rote N, Peterson L, Amesse LS. Expression of homeobox gene transcripts in trophoblastic cells. Am J Obstet Gynecol 2002;187:24 –32. [12] Strickland S, Richards WG. Invasion of the trophoblasts. Cell 1992; 71:355–7. [13] Celetti A, Barbar P, Cillo C, Rotoli B, Boncinelli E, Magli MC. Characteristic patterns of HOX gene expression in different types of human leukemia. Int J Cancer 1993;53:237– 44. [14] Borrow J, Shearman AM, Stanton VP Jr, Becher R, Colins T, Williams AJ, Dube I, Katz F, Kwong YL, Morris C, Ohyashiki K, Toyama K, Rowley J, Houseman DE. The t(7;11)(p15;p15) translocation in acute myleoid leukaemia fuses the genes for nucleoporin NUP98 and class I homeoprotein HOXA9. Nat Genet 1996;12:159 – 167. [15] Cillo C, Barbar P, Freschi G, Bucciarelli G, Magli MC, Boncinelli E. HOX gene expression in normal and neoplastic human kidney. Int J Cancer 1992;51:892–7. [16] DeVita G, Barba P, Odartechenko N, Givel JC, Freschi G, Bucciarelli G, Magli, MC, Boncinelli, E, Cillo C. Expression of homeoboxcontaining genes in primary and metastatic colorectal cancer. Eur J Cancer 1993;29A:887–93. [17] Cillo C, Cantile M, Mortarini R, Barba P, Parmiani G, Anichini A. Differential patterns of HOX gene expression are associated with

[18]

[19]

[20]

[21] [22] [23] [24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

specific integrin and ICAM profiles in clonal populations isolated from a single human melanoma metastatsis. Int J Cancer 1996;66: 692–7. Tiberio C, Barba P, Magli MC, Arvelo F, Le Chevalier T, Poupon MF, Cillo C. HOX gene expression in human small-cell lung cancers xenographed into nude mice. Int J Cancer 1994;58:608 –15. Redline RW, Hudock P, MacFee M, Patterson P. Expression of Abd-B-type homeobox genes in human tumors. Lab Invest 1994;71: 663–70. Clapp WL, Abrahamson DR. Regulation of kidney organogenesis: homeobox genes, growth factors and Wilms’ tumor. Curr Opin Nephrol Hypertens 1993;2:419 –29. Cillo C, Wilmore HP, Barba P, Brown KW. Expression and structure of HOX genes in Wilms’ tumour. Int J. Oncol 1995;7:11145–50. Sciavolino PJ, Abate-Shen C. Molecular biology of prostate development and prostate cancer. Ann Med 1998;30:357– 68. Kurman RJ. Pathology of trophoblast. Monogr Pathol 1991;33:195– 227. Fujino T, Suzuki A, Ito Y, Ohyashiki K, Hatano Y, Miura I, Nakamura T. Single-translocation and double-chimeric transcripts: detection of NUP98-HOXA9 in myeloid leukemias with HOXA11 or HOXA13 breaks of the chromosomal translocation t(7;11)(p15;p15). Blood 2002;99:1428 –33. Ohnishi K, Tobita T, Sinjo K, Takeshita A, Ohno R. Modulation of homeobox B6 and B9 genes expression in human leukemia cell lines during myelomonocytic differentiation. Leuk Lymphoma 1998;31: 599 – 608. Bodey B, Bodey B Jr, Siegel SE, Kaiser HE. Immunocytochemical detection of the Homebox B3, B4 and C6 gene products in breast carcinomas. Anticancer Res 2000;20:3281– 6. Oudejans C, Pannese M, Simeone A, Meijer C, Boncinelli E. The three most downstream genes of the Hox-3 cluster are expressed in human extraembryonic tissues including trophoblast of androgenetic origin. Development 1990;108:471–7. Guillemot F, Nagy A, Auerbach A, Rossant J, Joyner AL. Essential role of Mash-2 in extraembryonic development. Nature 1994;371: 333– 6. Montes M, Roberts D, Berkowitz RS, Genest DR. Prevalence and significance of implantation site trophoblastic atypia in hydatidiform moles and spontaneous abortions. Am J Clin Pathol 1996;105:411– 6. Qiao S, Nagasaka T, Harada T, Nakashima N. p53, Bax and Bcl-2 expression, and apoptosis in gestational trophoblast of complete hydatidiform mole. Placenta 1998;19:361–9. Shih IM, Kurman RJ. Ki-67 labeling index in the differential diagnosis of exaggerated placental site, placental site trophoblastic tumor, and choriocarcinoma: a double immunohistochemical staining technique using Ki-67 and Mel-CAM antibodies. Hum Pathol 1998;29: 27–33. Vider BZ, Zimber A, Hirsch D, Estlein D, Chastre E, Prevot S, Gespach C, Yaniv A, Gazit A. Human colorectal carcinogenesis is associated with deregulation of homeobox gene expression. Biochem Biophys Res Commun 1997;232:742– 8. Lurain JR, Brewer JI, Torok EE, Halpern B. Natural history of hydatidiform mole after primary evacuation. Am J Obstet Gynecol 1983;145:591–5. Roberts DJ, Mutter GL. Advances in the molecular biology of gestational trophoblastic disease. J Reprod Med 1994;39:201– 8.