Understanding the mechanisms of endometrial carcinogenesis following neonatal xenoestrogen exposure continues to be of interest to regulatory agencies, industry, and the population at large due to the widespread presence of naturally-occurring and manufactured xenoestrogens found in the environment, food, and consumer products. Identification of
molecular signatures and a mode of action could be used to develop new or aid existing toxicity screening to protect human health. However, despite this model having been extensively studied for over 25 years, a clear mechanism remains to be determined. Newbold et al. (2007) first proposed that altered uterine gene expression functioned to promote endometrial
carcinogenesis in this model long after the timing of exposure. Subsequent studies primarily performed by the Newbold and Williams laboratories aimed to understand specific pathways altered by neonatal xenoestrogen that contribute to adverse reproductive effects later in life (Newbold et al., 2007; Jefferson et al., 2011; Jefferson et al., 2013). Permanent aberrant
expression of the oncofetal protein SIX1 following neonatal xenoestrogen exposure appeared to be a likely candidate for mediating carcinogenesis in this model and formed the basis for the research presented in this dissertation.
In the research presented here, I investigated both the correlation and causation of SIX1 in development of endometrial carcinoma following neonatal xenoestrogen exposure. I showed that SIX1 is a biomarker for neonatal xenoestrogen exposure and disease development, can be used to identify aberrant cell populations in exposed uteri, and is also present in human
endometrial cancer. However, I also showed that SIX1 is neither necessary for DES-induced endometrial cancer, nor sufficient to induce cancer by 6 months of age in the absence of
neonatal xenoestrogen exposure. Instead, I found that SIX1 expression induced by neonatal xenoestrogen exposure appears to act as a cellular differentiation factor and may regulate a population of endometrial progenitor cells that are susceptible to transformation.
Although other specific genes and gene networks, such as lactoferrin, P63, and TGFβ
(Newbold et al., 2007; Couse et al., 2001), have been proposed as potential mechanisms for endometrial carcinogenesis in this exposure model, I now think it is unlikely that a single gene is responsible for carcinogenesis and that instead, this model relies on several mechanisms that act synergistically to produce adverse effects. In this exposure model, cancer appears to arise due to a complex combination of the timing of exposure during cell patterning and
differentiation, a plethora of transient and permanent ER-mediated downstream effects that occur during initiation by neonatal exposure and subsequent promotion by later endogenous estrogens, and epigenetic modifications (Jefferson et al., 2011; Jefferson et al., 2013). Together, these factors (and likely others that remain unknown) highlight that xenoestrogen exposure and disease development in this model does not fit simply into the current way most toxicants are defined, with a straightforward mode of action and toxic/nontoxic criterion. Therefore, rather than continue with narrow studies targeting a single gene or pathway, and trying to identify a single mode of action pathway, I think the field should expand its focus to understand xenoestrogen-induced molecular changes on a global level. I also recommend that particular attention should be paid to global epigenetic modifications that result from exposure, as they likely play a prominent role in disease development, and might be the best way to identify a collection of molecular signatures of exposure and disease that could be used to 1) extrapolate animal data to humans, 2) connect exposure induced epigenetic modifications with adverse effects, and 3) be extended to establish toxicity screens during sensitive windows in development.
Investigating Epigenetic Modifications to Identify Molecular Signatures for Toxicity Studies
Previously, chromatin immunoprecipitation quantitative polymerase chain reaction (ChIP-PCR) experiments demonstrated that neonatal DES exposure results in permanent alterations in Six1 gene locus-specific epigenetic marks, including a significantly increased association with several activating histone marks (H3K4me3, H3K9ac, and H4K5ac) (Jefferson et al., 2013). The data presented here supported a link between exposure, locus-specific epigenetic modifications, and disease development, by showing that SIX1 is involved in
aberrant DES-induced endometrial epithelial cell differentiation. However, these data show only a fraction of the story. Future epigenetic regulation studies, such as histone ChIP-seq, to
explore in-depth alterations in epigenetic marks paired with gene expression studies (RNA-seq) in the neonatal xenoestrogen exposure model will provide a much fuller understanding of the molecular mechanisms of hormonal carcinogenesis. Recent advances in experimental techniques including cheaper and faster sequencing and ChIP protocols using smaller tissue samples now make global investigation of altered pathways feasible.
In addition to investigating the unchallenged neonatal xenoestrogen exposure model (lacking subsequent exogenous estrogen exposure), manipulation of the promotion phase (2nd- hit) may help tease out cancer mechanisms. Several studies looking at altered gene expression in this model show that it is difficult to separate changes that are associated with carcinogenesis and other unrelated gene expression changes (eg. those that may contribute to other
phenotypes or have no effect at all) (Newbold et al., 2007; Huang et al., 2005; Jefferson et al., 2011). These studies predominantly focus on gene expression changes in control versus exposed animals at early time points (PND5, puberty). Several studies have also indicated that estradiol working through ERα mediates carcinogenesis in this model (Newbold et al., 1990; Leavitt et al., 1981; Couse et al., 2001; Lubahn et al., 1993). Future gene expression studies may elucidate the combination of changes that cause xenoestrogen induced carcinogenesis.
These include 1) comparing control and exposed mice at PND5 and the cancer endpoint (18 months of age) to reveal permanently altered pathways that could connect exposure with disease development; 2) comparing neontally xenoestrogen-exposed intact and ovariectomized mice +/- estradiol challenge, and 3) comparing neonatally-xenoestrogen exposed wildtype and conditional ERα knockout mice may reveal estrogen regulated pathways that are necessary for cancer promotion.
Using the DES-exposed Six1 Conditional Knockout Mouse as a Model for Normal and Abnormal Endometrial Epithelial Differentiation
Previous studies show that anterior-posterior patterning and cellular differentiation of the female reproductive tract is driven by homeobox transcription factors that coincide with specific female reproductive tract regions (Ma, 2009; Kobayashi and Behringer, 2003; Jefferson et al., 2012). Neonatal xenoestrogen exposure posteriorizes expression of these regulatory networks so factors normally controlling squamous epithelial differentiation in the vagina (eg. HOXA13, P63, SIX1) become expressed in the uterus where they facilitate basal cell differentiation (phenotype described in Chapter 3, evidence for role of SIX1 in Chapter 4). Several studies, including data presented in Chapter 3, show that the basal cell marker P63 is essential for vaginal epithelial differentiation and can be used as a marker of uterine basal cell differentiation (Kurita et al., 2004; Kurita, 2011). P63 and SIX1 have similar expression patterns in both normal vaginal epithelium and abnormal xenoestrogen-induced uterine epithelial differentiation
(Chapter 3). The vaginal and cervical epithelium in control P63 null mice differentiates into uterine epithelium (Kurita et al., 2004). However, neonatally DES-exposed P63 null mice still develop basal cell and squamous metaplasia within the uterine body and horns (Kurita et al., 2004). In contrast, the vaginal and cervical epithelium in control Six1 conditional knockout mice remains morphologically normal, but neonatally DES-exposed Six1 conditional knockout mice lack characteristic basal cell and squamous metaplasia within the uterine horns (Chapter 4). Interestingly, DES-induced uterine SIX1 transcript and protein expression precedes P63
(unpublished data). These data highlight the complexity of region-specific differentiation patterns, but also suggest that there is a common upstream regulator of P63 and SIX1 that controls location and exposure specific expression patterns and subsequent cell differentiation. Future studies using the DES-exposed Six1 conditional knockout model may contribute to understanding the molecular mechanisms of normal and abnormal female reproductive tract epithelial differentiation. To understand how DES-induced SIX1 alters endometrial epithelial differentiation, I propose that future gene expression studies should be performed comparing uteri from control and DES-exposed SIX1+/+ and Six1d/d mice. These comparisons may reveal
altered regulatory networks that control uterine-specific basal cell differentiation. These uterine regulatory networks can be further compared to those present in normal and P63 null (uterine- like) vagina. Together, these data may reveal region-specific control of vaginal and uterine-like epithelium within the vagina and uterine and vaginal-like epithelium within the uterus.
Investigating SIX1 as an Exposure and Disease Biomarker in Women Prenatally Exposed to DES
As described previously, the reproductive cancer most directly correlated with prenatal DES exposure is vaginal clear-cell adenocarcinoma among women exposed during early gestation (Troisi et al., 2007). Thus far, no increased risk of endometrial carcinoma has been reported in association with prenatal DES exposure (IARC, 2012). However, postmenopausal follow up studies have not yet been completed as prenatally DES-exposed women currently range in age from ~44-78. In the mouse model of hormonal carcinogenesis, endometrial
carcinoma development was most prominent in late life (12-18 months of age) suggesting that a similarly altered risk of endometrial cancer development in DES-exposed women might not become apparent until postmenopausal studies have been completed. In the data presented here, I show that although SIX1 does not play a direct role in carcinogenesis, it may mediate a putative cancer progenitor cell population, and is also an excellent biomarker for neonatal xenoestrogen exposure, abnormal differentiation, and neoplasia. I also showed that SIX1 is
present in neoplastic cells within human endometrial cancer lesions but not in normal cancer- adjacent tissue, or normal endometrial tissue from patients without cancer. It is unknown if the endometrial cancer biopsies analyzed in Chapter 2 came from women who were prenatally exposed to DES. However, these data support further investigation for the use of SIX1 as an exposure and disease biomarker. It would be particularly informative to investigate uterine SIX1 expression in cohorts of women with known prenatal DES exposure such as the NIEHS Sister Study or the Nurse’s Study (D'Aloisio et al., 2010; Mahalingaiah et al., 2014).
Investigating the Windows of Susceptibility
There are many published studies demonstrating that different windows of exposure to estrogenic chemicals have distinct outcomes in terms of reproductive tract development. Most of the studies described here, including the original research presented in Chapters 2-4, investigate adverse effects resulting from early life xenoestrogen exposure (prenatal exposure or exposure up to postnatal day 5). However, the window of susceptibility in the neonatal xenoestrogen exposure model, and when it closes, is not well understood. For instance, gland development defines another period in female reproductive tract differentiation that could be particularly susceptible to disruption by xenoestrogens. Gland development occurs after the classic PND1-5 xenoestrogen exposure, beginning around PND7, and is not complete until PND15 (Cooke et al., 2013). Most studies compare prenatal exposures to early neonatal exposures, (Newbold and McLachlan, 1982; Leavitt et al., 1981), but comparisons of effects resulting from xenoestrogen exposure during different postnatal windows has not been
completed. One study in Eker rats showed that DES exposure during PND3-5 and PND10-12, but not following exposure during PND17-19, resulted in increased incidences of leiomyoma, indicating that the window of susceptibility closed by this time (Cook et al., 2007). The window of susceptibility for DES-induced endometrial carcinoma in this mouse model may continue up until the completion of gland development on PND15. Future studies could investigate the various exposure windows. Investigating other exposure windows in this animal model could be
useful for understanding if SIX1 is a neonatal-specific exposure biomarker or if it could be extended to exposures during other time points.
Future studies investigating endometrial cancer and uterine SIX1 expression in cohorts of women exposed to DES during other sensitive developmental periods may be useful for understanding sensitive windows in humans. Cohorts could include women exposed to DES during pregnancy, women prescribed DES as a hormone replacement therapy during
menopause (IARC, 1987; IARC, 2012), and adolescent girls exposed to DES to decrease their adult height (Venn et al., 2004).
Taken together, these studies will help us understand the molecular mechanisms of adverse effects caused by xenoestrogen exposures.
REFERENCES
AAP. (1998). Soy Protein-based Formulas: Recommendations for Use in Infant Feeding. Pediatrics, 101, 148-153.
Armat, M., Oghabi Bakhshaiesh, T., Sabzichi, M., Shanehbandi, D., Sharifi, S., Molavi, O., Mohammadian, J., Saeid Hejazi, M. and Samadi, N. (2016). The role of Six1 signaling in paclitaxel-dependent apoptosis in MCF-7 cell line. Bosn. J. Basic Med. Sci., 16, 28-34.
Behbakht, K., Qamar, L., Aldridge, C.S., Coletta, R.D., Davidson, S.A., Thorburn, A. and Ford, H.L. (2007). Six1 overexpression in ovarian carcinoma causes resistance to TRAIL-mediated apoptosis and is associated with poor survival. Cancer Res., 67, 3036-3042.
Bernbaum, J.C., Umbach, D.M., Ragan, N.B., Ballard, J.L., Archer, J.I., Schmidt-Davis, H. and Rogan, W.J. (2008). Pilot studies of estrogen-related physical findings in infants. Environ. Health Perspect., 116, 416-420.
Bhatia, J. and Greer, F. (2008). Use of soy protein-based formulas in infant feeding. Pediatrics,
121, 1062-1068.
Blevins, M.A., Towers, C.G., Patrick, A.N., Zhao, R. and Ford, H.L. (2015). The SIX1-EYA transcriptional complex as a therapeutic target in cancer. Expert Opin. Ther. Targets, 19, 213- 225.
Cao, Y., Calafat, A.M., Doerge, D.R., Umbach, D.M., Bernbaum, J.C., Twaddle, N.C., Ye, X. and Rogan, W.J. (2009). Isoflavones in urine, saliva, and blood of infants: data from a pilot study on the estrogenic activity of soy formula. J. Expo. Sci. Environ. Epidemiol., 19, 223-234. Christensen, K.L., Patrick, A.N., McCoy, E.L. and Ford, H.L. (2008). The six family of homeobox genes in development and cancer. Adv. Cancer Res., 101, 93-126.
Cieply, B., Farris, J., Denvir, J., Ford, H.L. and Frisch, S.M. (2013). Epithelial-mesenchymal transition and tumor suppression are controlled by a reciprocal feedback loop between ZEB1 and Grainyhead-like-2. Cancer Res., 73, 6299-6309.
Coletta, R.D., Christensen, K., Reichenberger, K.J., Lamb, J., Micomonaco, D., Huang, L., Wolf, D.M., Muller-Tidow, C., Golub, T.R., Kawakami, K. and Ford, H.L. (2004). The Six1
homeoprotein stimulates tumorigenesis by reactivation of cyclin A1. Proc. Natl. Acad. Sci. U. S. A., 101, 6478-6483.
Coletta, R.D., Christensen, K.L., Micalizzi, D.S., Jedlicka, P., Varella-Garcia, M. and Ford, H.L. (2008). Six1 overexpression in mammary cells induces genomic instability and is sufficient for malignant transformation. Cancer Res., 68, 2204-2213.
Cook, J.D., Davis, B.J., Goewey, J.A., Berry, T.D. and Walker, C.L. (2007). Identification of a sensitive period for developmental programming that increases risk for uterine leiomyoma in Eker rats. Reprod. Sci., 14, 121-136.
Cooke, P.S., Spencer, T.E., Bartol, F.F. and Hayashi, K. (2013). Uterine glands: development, function and experimental model systems. Mol. Hum. Reprod., 19, 547-558.
Couse, J.F., Davis, V.L., Hanson, R.B., Jefferson, W.N., McLachlan, J.A., Bullock, B.C.,
Newbold, R.R. and Korach, K.S. (1997). Accelerated onset of uterine tumors in transgenic mice with aberrant expression of the estrogen receptor after neonatal exposure to diethylstilbestrol. Mol. Carcinog., 19, 236-242.
Couse, J.F., Dixon, D., Yates, M., Moore, A.B., Ma, L., Maas, R. and Korach, K.S. (2001). Estrogen receptor-alpha knockout mice exhibit resistance to the developmental effects of neonatal diethylstilbestrol exposure on the female reproductive tract. Dev. Biol., 238, 224-238. Couse, J.F. and Korach, K.S. (1999). Estrogen receptor null mice: what have we learned and where will they lead us? Endocr. Rev., 20, 358-417.
Couse, J.F. and Korach, K.S. (2004). Estrogen receptor-alpha mediates the detrimental effects of neonatal diethylstilbestrol (DES) exposure in the murine reproductive tract. Toxicology, 205,
55-63.
Creasman, W.T. (2015). Endometrial Cancer Staging.
http://emedicine.medscape.com/article/2006791-overview.
D'Aloisio, A.A., Baird, D.D., DeRoo, L.A. and Sandler, D.P. (2010). Association of intrauterine and early-life exposures with diagnosis of uterine leiomyomata by 35 years of age in the Sister Study. Environ. Health Perspect., 118, 375-381.
Davis, V.L., Newbold, R.R., Couse, J.F., Rea, S.L., Gallagher, K.M., Hamilton, K.J., Goulding, E.H., Jefferson, W., Eddy, E.M., Bullock, B.C. and Korach, K.S. (2012). Expression of a dominant negative estrogen receptor alpha variant in transgenic mice accelerates uterine cancer induced by the potent estrogen diethylstilbestrol. Reprod. Toxicol., 34, 512-521. Dieckmann, W.J., Davis, M.E., Rynkiewicz, L.M. and Pottinger, R.E. (1953). Does the administration of diethylstilbestrol during pregnancy have therapeutic value? Am. J. Obstet. Gynecol., 66, 1062-1081.
Dixon, D., Alison, R., Bach, U., Colman, K., Foley, G.L., Harleman, J.H., Haworth, R., Herbert, R., Heuser, A., Long, G., Mirsky, M., Regan, K., Van Esch, E., Westwood, F.R., Vidal, J. and Yoshida, M. (2014). Nonproliferative and proliferative lesions of the rat and mouse female reproductive system. J. Toxicol. Pathol., 27, 1s-107s.
Dodds, E.C., Goldberg, L., Lawson, W. and R., R. (1938). Oestrogenic activity of certain synthetic compounds. Nature, 141, 247-248.
Doerge, D.R., Twaddle, N.C., Banks, E.P., Jefferson, W.N. and Newbold, R.R. (2002).
Pharmacokinetic analysis in serum of genistein administered subcutaneously to neonatal mice. Cancer Lett., 184, 21-27.
Dontu, G., El-Ashry, D. and Wicha, M.S. (2004). Breast cancer, stem/progenitor cells and the estrogen receptor. Trends Endocrinol. Metab., 15, 193-197.
Dunlap, K.A., Filant, J., Hayashi, K., Rucker, E.B., 3rd, Song, G., Deng, J.M., Behringer, R.R., DeMayo, F.J., Lydon, J., Jeong, J.W. and Spencer, T.E. (2011). Postnatal deletion of Wnt7a inhibits uterine gland morphogenesis and compromises adult fertility in mice. Biol. Reprod., 85,
El-Hashash, A.H., Al Alam, D., Turcatel, G., Rogers, O., Li, X., Bellusci, S. and Warburton, D. (2011). Six1 transcription factor is critical for coordination of epithelial, mesenchymal and vascular morphogenesis in the mammalian lung. Dev. Biol., 353, 242-258.
FDA. (1972). Selected Item from the FDA Drug Bulletin—November 1971: Diethylstilbestrol Contraindicated in Pregnancy. Calif. Med., 116, 85-86.
Forsyth, B.W., McCarthy, P.L. and Leventhal, J.M. (1985). Problems of early infancy, formula changes, and mothers' beliefs about their infants. J. Pediatr., 106, 1012-1017.
Franke, A.A., Custer, L.J. and Tanaka, Y. (1998). Isoflavones in human breast milk and other biological fluids. Am. J. Clin. Nutr., 68, 1466s-1473s.
Franke, A.A., Halm, B.M., Custer, L.J., Tatsumura, Y. and Hebshi, S. (2006). Isoflavones in breastfed infants after mothers consume soy. Am. J. Clin. Nutr., 84, 406-413.
Giusti, R.M., Iwamoto, K. and Hatch, E.E. (1995). Diethylstilbestrol revisited: a review of the long-term health effects. Ann. Intern. Med., 122, 778-788.
Grifone, R., Demignon, J., Houbron, C., Souil, E., Niro, C., Seller, M.J., Hamard, G. and Maire, P. (2005). Six1 and Six4 homeoproteins are required for Pax3 and Mrf expression during myogenesis in the mouse embryo. Development, 132, 2235-2249.
Hanahan, D. and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell, 144,
646-674.
Harlid, S., Adgent, M., Jefferson, W.N., Panduri, V., Umbach, D.M., Xu, Z., Stallings, V.A., Williams, C.J., Rogan, W.J. and Taylor, J.A. (2016). Soy Formula and Epigenetic Modifications: Analysis of Vaginal Epithelial Cells from Infant Girls in the IFED Study. Environ. Health
Perspect., 125, 447-452.
Hayashi, K., Yoshioka, S., Reardon, S.N., Rucker, E.B., 3rd, Spencer, T.E., DeMayo, F.J., Lydon, J.P. and MacLean, J.A., 2nd. (2011). WNTs in the neonatal mouse uterus: potential regulation of endometrial gland development. Biol. Reprod., 84, 308-319.
Herbst, A.L. (1981). Diethylstilbestrol and other sex hormones during pregnancy. Obstet. Gynecol., 58, 35S-40S.
Herbst, A.L. and Anderson, D. (2015). Diethylstilbestrol (DES) Pregnancy Treatment: A