Current Research Areas

Prevention of Anti-cancer Therapy-induced Premature Ovarian Failure
This project is directed at elucidation of the biochemical, molecular and genetic events underlying inappropriate destruction of the ovaries in young girls and women treated for cancer. In 1997, we published a study using mice as a model (Nature Medicine 1997 3: 1228-1232) showing that the programmed cell death pathway of apoptosis is the mechanism responsible for loss of female germ cells (oocytes) following exposure of females to anti-cancer therapies. Moreover, several effector molecules were identified as key participants in the oocyte cell death program. These data have subsequently been used to develop gene therapy-based (Molecular Endocrinology 1999 13: 841-850) and small molecule therapy-based (Nature Medicine 2000 6: 1109-1114; Nature Medicine 2002 8: 901-902) approaches for protecting oocytes from both radiation and chemical therapy-induced death in vivo. The latter work also has provided the framework for a utilization patent application.

Approaches to Delay or Eliminate Menopause
This project is directed at clarifying the events underlying normal (developmental) oocyte loss and ovarian aging, with the goal of developing strategies to slow the time to natural menopause as a means to prevent the onset of post-menopausal health problems in women. We published a paper in 1999 (Nature Genetics 1999 21: 200-203) providing proof-of-concept that, at least in female mice, prolongation of functional ovarian life span into advanced chronological age can be achieved by repressing the normal programmed death of oocytes during postnatal life. In addition to the biological value of this animal for future study of how sustained ovarian function influences the aging body, we anticipate that this unique model will enable the identification of other ovarian-derived factors beneficial for women's health. Such factors could then be used to improve the efficacy and safety of steroid-based hormone replacement therapies currently offered to post-menopausal women.

Mechanisms Underlying Environmental Toxicant-induced Early Menopause
This project was developed to determine if chemicals present at high abundance in the environment and in tobacco smoke (polycyclic aromatic hydrocarbons or PAHs) activate an orphan transcription factor (termed the aryl hydrocarbon receptor or AhR) in oocytes, leading to de novo expression of genes which trigger oocyte death. Consistent with this hypothesis, we have observed that the AhR is abundantly expressed in oocytes, and that Ahr gene knockout in female mice leads to increased oocyte survival (Endocrinology 2000 141: 450-453). Moreover, in our most recent studies (Nature Genetics 2001 28: 355-360), we have shown that the chemically-activated AhR induces transcription of the pro-apoptotic bax gene in oocytes, and that expression of both the Ahr and bax genes are functionally required for these chemicals to cause oocyte depletion and ovarian failure. Using a human ovarian xenograft model, we have also shown in this same study that an induction of bax gene expression and apoptosis occurs in human primordial and primary oocytes following PAH exposure in vivo. These findings provide insight into why women who smoke are more likely to undergo menopause earlier than non-smokers are. Finally, we have followed these experiments with a similar investigation of the impact of prenatal exposure to these chemicals on fetal ovaries (Endocrinology 2002 143: 615-620). These studies showed that a brief exposure to PAHs during pregnancy is sufficient to kill over 80% of the developing female germ cells, and that this death pathway involves the AhR and Bax. Based on these findings, we believe that the reduced fecundity reported for women born to mothers who smoked during pregnancy is due to PAH-driven death of the fetal ovarian germ line.

Genetics of Developmental (Prenatal) Oocyte Death
Over two-thirds of the human female oocyte population is deleted for as-yet-unknown reasons by apoptosis prior to birth. This project was initiated to determine the contribution of specific gene products to programmed death of oocytes in the developing fetal ovaries resulting from cytokine insufficiency versus meiotic defects (the two presumed principal initiators of oocyte loss). As a first step, we developed and validated a fetal ovarian culture model to explore the regulation of prenatal oocyte death under controlled conditions ex vivo (Endocrinology 1999 140: 941-949 and 140: 2696-2703). We then utilized several gene knockout mouse lines to begin to build a molecular framework for the cell death program activated in fetal oocytes (for example, see Genes & Development 1998 12: 1304-1314; Endocrinology 2000 141: 450-453; Nature Medicine 2000 6: 1109-1114). Moreover, we have found that oocyte loss resulting from cytokine insufficiency versus meiotic defects is executed via two genetically distinct programs (Cell Death Differentiation 2001 8: 614-620). Such studies should impact on many aspects of female reproduction, including our understanding of Turner syndrome in human females since it is believed that gonadal dysgenesis in the XO genetic background results from deletion of increased numbers of meiotically-defective germ cells carrying unpaired chromosomes.

Role of the Sphingomyelin Pathway in Ovarian Corpus Luteum (CL) Regression
Low levels of progesterone secretion caused by luteal insufficiency through the first 7-8 weeks of gestation in women often results in loss of the pregnancy. Cytokines, once thought to be ancillary factors to the process of corpus luteum regression, are proving to be active players. This project was designed to determine the role of cytokines in CL development, function and regression, and elucidate their mechanisms of action. Our working hypothesis is that cytokines stimulate sphingomyelin hydrolysis in luteal cells, resulting in the production of the pro-apoptotic second messenger, ceramide. The elevated levels of ceramide inhibit both basal and gonadotropin-stimulated progesterone production by the CL, and induce death of luteal cells leading to CL regression. An understanding of the mechanisms that initiate and coordinate luteal regression is crucial for developing therapies to combat infertility resulting from luteal insufficiency.

Elucidating the Mechanism of Action of Progestin Therapy in Endometrial Cancer
Endometrial cancer is primarily attributed to unopposed estrogen exposure in the absence of the cell-differentiating effects of progesterone. Endometrial hyperplasia and some endometrial cancers can be successfully treated with progesterone therapy. However, the mechanism(s) by which progesterone elicits its response is not clearly established. This project was designed to determine how progesterone alters uterine cell function, and why progesterone therapy is effective in some, but not all, all patients with early stage endometrial cancer. With this information, we can begin to design and develop novel strategies to identify those sub-populations of women who can be successfully treated by progestin therapies rather than complete hysterectomy. Such an approach is especially important for reproductive-age women diagnosed with this type of cancer who wish to retain their fertility.

Identification of Novel Tumor Suppressor Proteins in Gynecologic Cancers
In collaboration with Dr. Lawrence Zukerberg (MGH Department of Pathology), we have identified a protein that is lost with high frequency during the early stages of malignancy in the human endometrium. This protein is a negative regulator of the cell cycle via its ability to enhance the inhibitory phosphorylation of the Cdk2 protein (a key modulator of the G1-S phase transition in mitosis). Consequently, loss of function would theoretically result in uncontrolled cellular proliferation as a first step towards cancer. We have just completed studies showing that this cell cycle regulatory protein is regulated by hormones in human endometrial epithelial cells. In addition, we have generated a gene mutant mouse, which is deficient in this cell cycle regulatory protein. The mutant mice show evidence of atypical hyperplasia, often considered the precursor to endometrial cancer. These results are in perfect agreement with the results of pathology specimen screening assays of human uterine cancer biopsies. We observed a consistent loss of this protein in almost all cancer biopsies and, most importantly, intermittent loss of the protein in those biopsies with pre-cancerous atypical hyperplasia. These studies are currently under review for publication. Upon acceptance more specific information on the cell cycle regulatory protein can be provided.

The Bi-Directional Molecular Dialog of Early Pregnancy
For pregnancy to be first established, and then maintained, the embryo must communicate with the mother. It does so by generating molecular and cellular signals which modify the uterus, converting it from an indifferent, cyclic environment to one that is nurturing. In response to the embryo, the mother in turn produces a variety of factors that facilitates formation of the placenta, the embryonic lifeline to the outside world. An understanding of this sophisticated and complex bi-directional signaling is paramount since it is estimated that 30-60 percent of mammalian conceptions are routinely lost during early pregnancy due to faulty maternal-fetal interactions. Such studies may also shed new light on underlying causes of preeclampsia. It is hypothesized that the pathology of preeclampsia may stem from dysfunctional events that occur earlier in pregnancy as the embryo establishes itself within the uterus.

Identification of Novel Modulators in the Pathogenesis of Endometriosis
Endometriosis is a polygenic disease with complex multifactorial etiologies affecting reproductive-aged women. Although this disorder is commonly treated in clinical practice, the mechanisms by which ectopic endometrium is disseminated and proliferates, is not completely understood. There are a number of different factors, which have been implicated in either the genesis or the propagation of endometriosis. They include but are not limited to prostaglandins, cytokines, growth factors, chemokines, cell adhesion molecules and steroid hormones. It is, however, difficult to investigate the functional role of these factors without adequate in vivo model systems. We are currently developing mouse models which have been manipulated to ‘mimic’ the human disease in order that we may study the effect of specific factors on the growth of ectopic endometrium. The mouse model was chosen to incorporate the power of mouse genetics. Using mutant mouse strains which are devoid or over express one or more of the factors described above we should be able to delineate their cause an effect relationships with the pathogenesis of the disease. Using this strategy in tandem with gene and protein analysis we hope to gain a better understanding of the underlying mechanisms of endometriosis. This information will also serve to develop more effective alternative treatment modalities.