|Home | About | Journals | Submit | Contact Us | Français|
Lung cancer has long been thought of as a cancer that mainly affects men, but over the past several decades, because of the high increase in tobacco use by women, there has been a corresponding dramatic increase in lung cancer among women. Since 1998, lung cancer deaths in women have surpassed those caused by breast cancer in the United States. Annual lung cancer deaths among women in the US also currently surpass those caused by breast, ovarian, and cervical cancers combined. Women are more likely than men to be diagnosed with adenocarcinoma and small-cell carcinoma of the lung compared to squamous cell carcinoma, and never smokers diagnosed with lung cancer are almost three times more likely to be female than male. These observations in the population, coupled to the findings that both estrogen receptors and aromatase, the enzyme that synthesizes 17β-estradiol, are expressed by lung tumors, suggest a role for female steroid hormones in control of lung cancer growth. Pre-clinical data and clinical data are increasingly emerging to support this concept, and to suggest that a local production of estrogen and expression of ERs occurs in lung tumors that rise in men as well as women. An additional protein that recognizes 17β-estradiol with high affinity, GPR30, is also expressed in lung tumors at high levels and may be responsible for some of the proliferation signals induced by estrogen.
Although there are conflicting data on whether or not tobacco exposure is significantly more carcinogenic to women as the same dose is to men, a number of observations about sex differences in lung cancer are well-accepted. The proportional occurrence of histologic lung carcinoma subtypes differs significantly (p<0.0001) between men and women (1,2). In men, squamous cell carcinoma is the most common sub-type. In women, adenocarcinoma is the most common histologic subtype (representing over 44% of all lung tumors), while squamous cell carcinoma makes up only about 21% (1,2). The extent of DNA damage elicited by tobacco exposure after correcting for dose has also been found in several studies to be higher in the lungs and tumors of women compared to men (3,4), and this has been attributed to a decreased DNA repair capacity in women (5). Induction of cytochrome P450 enzymes that activate tobacco carcinogens by β-estradiol has also been documented (3-5). In a recent study, Fu et al. found significantly more women were diagnosed with lung cancer at an age younger than 50, when they are presumably largely pre-menopausal, than were men younger than 50 (6). Never smokers diagnosed with lung cancer are much more likely to be female than male (7), and in general, women diagnosed with lung cancer have smoked less than men diagnosed with the disease. In a recent analysis of several large prospective cohorts, Wakelee et al. found that the underlying risk for lung cancer in women who never smoked was over 2 times that of men who never smoked (8). These observations suggest some preference for female-associated pathways in lung cancer development.
Although it is well-established that on average women with advanced lung cancer live longer than men, recent work suggests this effect is confined to older, post-menopausal women. Pre-menopausal women in general present with more advanced lung cancer with less differentiation, a sign of more aggressive biology. Women over the age of 60 showed a distinct survival advantage over both men and younger women; the difference compared to younger women could be due to higher levels of circulating estrogen in the under 50 population (9). Men in contrast did not show an age effect in survival.
Hormone replacement therapy (HRT) has been examined in relation to lung cancer survival. Ganti et al. have reported a significant association between both a younger median age at lung cancer diagnosis and a shorter median survival time in women who used HRT around the time of diagnosis compared to those who did not (10). The Women's Health Initiative also recently reported a strong adverse effect on survival after a lung cancer diagnosis in women who took hormone replacement therapy containing both β-estradiol and a progesterone (11). In the WHI randomizd trial, more than 16,000 post-menopausal women received either placebo or daily HRT (containing estrogen plus medroxyprogesterone acetate) for over five years. Smoking history and age, two factors that influence lung cancer diagnosis and survival, were balanced between the two study arms. There was a trend toward more lung cancer diagnoses in the HRT group compared to placebo, which did not reach statistical significance. However, the HRT group experienced a significantly greater likelihood of dying from lung cancer (46% mortality in the HRT arm versus 27% in the placebo arm). These observations strongly suggest that HRT provides a tumor growth advantage in lung cancer, as it does in breast cancer.
In contrast, some reports suggest that HRT use prior to lung cancer diagnosis could actually protect women from developing lung cancer, especially if they smoked (12). An inverse relationship was observed between HRT use and NSCLC risk in postmenopausal women with lung tumors that were ER-positive, but not ER-negative (13). These data suggest that there are different effects on the balance between cell differentiation and cell proliferation in response to estrogen in normal or preneoplastic lung epithelium compared to malignant epithelium. Because lung tumors are also known to produce aromatase (see below), it is possible that in normal lung, exogenous hormone use reduced local estrogen production by negative feedback regulation of aromatase.
Steroid hormone pathways have been effectively targeted in breast and prostate cancer, where hormone-dependent growth is well understood. However, steroid hormone receptors such as the ER are found in many tissues outside the reproductive tract, such as the heart and blood vessels (14). Estrogens are also known to have biological effects in non-reproductive tumors, such as renal cancer (15). Steroid receptors are able to signal independently of steroid ligands; the best-characterized effect in this regard is through activation of steroid receptors by phosphorylation (16). Thus, steroid hormone receptors could be biologically active through both steroid-induced signalling and steroid-independent signalling. As discussed below, estrogen receptor (ER) signaling pathways that are proliferative have been found in a number of studies in non-small cell lung cancer (NSCLC). Progesterone receptor (PR) may be active as a differentiation-inducing pathway in lung cancer as well, which could potentially protect against cancer progression (17).
Estrogen receptors (ERs) are members of the nuclear steroid receptor superfamily. Two forms of the ER have been identified, ERα and ERβ, that are the products of two separate genes. Both ERα and ERβ contain functional domains involved in their nuclear signaling including a sequence called AF-1 near the amino terminus that acts as a ligand-independent transcriptional activation domain; a DNA binding domain located centrally in the protein; and a domain in the carboxy-terminus that contains the ligand binding domain. Also contained in the carboxy-terminus is another transcriptional activation function, termed AF-2, that carries out ligand-dependent activation of the receptor (18). ERα has been known for decades, while a second high affinity estrogen receptor, ERβ, was discovered by two different laboratories simultaneouly in the mid-1990's (19,20). Although the affinity for estrogen of the ERβ protein is similar to that of ERα, the AF-1 domain of ERβ is trucated compared to ERα and also is less active at inducing transcriptional activation (21). The two ERs display different tissue distributions, with ERα highest in breast, ovarian, and endometrial tissues and ERβ highest in ovaries and lung (19,22). When bound by an agonist such as 17β-estradiol, a conformational change in the AF-2 domain is induced, creating an interaction surface for ER coactivator proteins (23,24), which causes chromatin unwinding and production of an active gene transcription complex. Activation of an ERE by 17β-etradiol and other selective estrogen response modifiers (SERMs) has been found in lung tumor cells (25,26).
Genetic manipulation has been used to demonstrate which tissues outside the reproductive tract contain biologically functional ERs. A transgenic mouse in which an ERE-luciferase reporter construct was expressed in 26 different tissues was produced, and estrogen-responsive tissues were identified in ovarectomized female mice treated with 17β-estradiol or vehicle. The lungs displayed 15-fold induction of luciferase reporter gene activity that was 17β-estradiol dependent, a greater effect than that observed in known hormone-responsive tissues such as bone, uterus and mammary gland (27). In another ERE-luciferase reporter gene mouse model, Lemmen et al. showed that 17β-estradiol also significantly induced luciferase activity in the lungs of male and female mice (28), suggesting the lungs of both sexes are estrogen responsive. Targeted inactivation of ERβ, but not ERα was shown to result in lung abnormalities. The number of alveoli is significantly decreased in lungs from female ERβ−/− mice at 3 months of age compared to WT controls, and ERβ−/− females also display reduced surfactant (29). Both male and female ERβ-deficient mice exhibit significant lung dysfunction (abnormal extracellular matrix deposition and alveolar collapse) by 5 months, causing systemic hypoxia (30). The observation that this phenotype occurs in both adult female and male mice suggests either that effects of lung development through ERβ occur in a ligand-independent manner, or that estrogens act via a local autocrine mechanim in the lungs of both sexes.
There are inconsistent reports of the presence of ERs in lung tumors, although early studies probed for the presence of the classical ERα protein only. With the development of antibodies that can differentiate between ERα and ERβ, it is now apparent that ERβ is expressed in the majority of human NSCLC cell lines, and is also present in primary tissues of human non-small cell lung carcinomas (NSCLCs) from both men and women (25, 31-34). The role of ERα in lung cancer is less clear. Antibodies that are used clinically to detect ERα in breast tumors show little or no reactivity in lung tumors. Using an antibody that recognizes the ERα carboxyterminus, staining was mainly found in the cytoplasm and cell membrane in immunohistochemical studies (25). Based on immunoblot analysis, the proteins detected by this antibody consist mainly of smaller, variant proteins, which was confirmed by the presence of many alternatively spliced mRNAs, with little full-length mRNA (25). This non-nuclear ERα pool presumably is comprised of variant isoforms that lacks the amino-terminus because the proteins are not detected by antibodies that recognize the ERα amino-terminus.
In contrast, ERβ was found localized to both the nucleus and the cytoplasm, and was comprised of mainly full-length protein in addition to some variants (25). In a number of studies, nuclear localization of ERβ was observed in lung tumors; between 45-69% of lung cancer cases were positive (31-34), while ERα was rarely detected. There is little consensus on whether ERβ expression is a factor in survival. Some studies have suggested a protective effect of nuclear expression (32-34), which may only be significant in men (33). These results are opposite of what has been demonstrated for ER status and prognosis of breast cancer patients (35,36). It is possible the presence of nuclear ERβ confers a hormone dependence for growth, rather than dependence on other more aggressive oncogenic pathways, leading to comparatively better survival. However, lung tumors with ERβ expression are still lethal. The specificity of some ERβ antibodies used in early studies has been disputed, and early studies also did not consider the cytoplasmic ER compartment, which is involved in non-genomic signaling.
The classical model for 17β-estradiol regulation of gene expression predicts that transcription of estrogen-responsive genes is increased when ligand-bound ERs translocate to DNA and bind EREs in the promoters of these genes. We have transfected NSCLC cells expressing endogenous ERs with an ERE-tk-luciferase reporter gene. Luciferase reporter gene activity, which serves as a measure of ER-mediated transcription, was quantified after 24 hours. Physiologic levels of 17β-estradiol resulted in a 2-fold activation of gene reporter activity, which could be blocked by the ER down-regulator fulvestrant (25,26). The genomic activation in NSCLC cell lines could be induced by ligands that are specific for ERβ, while ligands specific for ERα showed no activity (26). In addition, only an ERβ ligand could increase NSCLC tumor growth in a xenograft model (26).
Estrogen is also known to rapidly activate non-genomic signaling pathways in other systems, leading to increases in second messengers such as cAMP, calcium, and activation of PI3K and MAPK. This rapid signaling leads to changes in protein structure/function and gene transcription and is necesary for full responses to estrogen. For example, inhibition of PI3K or MAPK with pharmacologic agents prevents the 17β-estradiol-dependent proliferation of endothelial cells (14). Whether the estrogen binding protein at the plasma membrane is identical to ERα or ERβ is in dispute, and alternative estrogen receptors such as the G-protein coupled receptor GPR30 have been detected (37). Pharmacological inhibitors for GPR30 have been made, which will aid in understanding the relative importance of this protein (37). It is possible that both a classical ER and GPR30 are involved in lung cancer responses to estrogens.
A number of studies have found that 17β-estradiol initiates membrane signaling in NSCLC cells, as evidenced by rapid phosphorylation of p42/p44 MAPK and/or Akt (26, 38, 39, 40). Plasma membrane fractions from NSCLC cells contain 17β-estradiol binding activity, and isolated caveolae from NSCLC cells contain proteins that react with ER-specific antibodies (38). These data support the hypothesis that at least a part of the ER pool is present at the plasma membrane and triggers membrane signaling in NSCLC cells.
Non-genomic ER signaling may also interact with other membrane growth factor pathways, such as the epidermal growth factor receptor (EGFR/HER-1). The EGFR/HER family of tyrosine kinase receptors are known to be involved in NSCLC growth, protection from apoptosis, and angiogenesis. An interaction between the ER and EGFR has been demonstrated, in which estrogen can rapidly activate the EGFR in lung cancer cell lines through release of EGFR ligands at the cell surface (39). The combination of fulvestrant and gefitinib, an EGFR tyrosine kinase inhibitor, in NSCLC can maximally inhibit cell proliferation and induce apoptosis in vitro and in vivo (39). The EGFR tyrosine kinase inhibitor erlotinib also gave superior anti-tumor activity in NSCLC tumor xenografts when used in combination with fulvestrant (41). Membrane ERs were also found to be co-localized with EGFR in lung tumors (38). Ligand-independent ER non-genomic signaling may also occur, via EGFR downstream mediators that directly phosphorylate ER at specific serine residues (16). These serines were phosphorylated in over 80% of ER positive lung tumors examined (41). It is possible EGFR and ER pathways act as alternate signaling mechanisms in NSCLC, because EGFR protein expression was down-regulated in response to estrogen and up-regulated in response to fulvestrant in NSCLC cell lines. This suggests that the EGFR pathway is activated when estrogen is depleted (39). Conversely, ERβ protein expression was down-regulated in response to EGF and up-regulated in response to gefitinib (39). This provides a rationale for dual targeting of these pathways.
A phase I clinical trial using drugs that target EGFR and ER was performed to assess the toxicity of combined treatment of gefitinib with fulvestrant in 22 post-menopausal women (42). These two targeted agents in combination were found to be safe and to have anti-tumor activity in female patients with stage IIIB/V NSCLC. Presence of a high percentage of cells within the tumors with nuclear ERβ immunostaining was correlated with improved patient survival. Phase II trials examining the combination of erlotinib with fulvestrant are also underway (43). Combination therapy may increase the duration of response in patients whose tumors harbour an EGFR mutation as well as an improved response rate in patients whose tumors are EGFR wild type.
Estrogens are known to increase mitochondrial function, and this effect has been hypothesized to lead to greater longevity in women compared to men. ER has been found in the mitochondria of breast cancer cells (44). Mechanistic studies examining effects of 17β-estradiol on the mitochondria has recently shown that nuclear transcription of Nuclear Respiratory Factor-1 (NRF-1) is increased by 17β-estradiol in breast cancer and lung cancer cells (45), that ERβ changes mitochondrial sensitivity to oxidative stress (46), and that mitochondrial DNA EREs respond to estrogen to regulate gene expression in mitochondria (47-49). NRF-1 encodes a transcription factor needed to promote mitochondrial DNA transcription and replication (50), and NRF-1 mRNA expression is estrogen responsive in both breast and lung cancer (45). The NRF-1 target gene, TFAM as well as mitochondrial genes regulated by TFAM were also positivly regulated by 17β-estradiol treatment. Stimulation of mitochondrial function/biogenesis through nuclear genomic regulation of the NRF-1 promoter by ERs may be a key function of estrogen.
ERβ has been shown to localize to mitochondria in many cell types such as neurons, skeletal muscle cells, and human lens cells, pointing to a general role for this ER sub-type in regulating function of mitochondria (51-54). Hippocampal HT-22 cells with engineered loss of ERβ are refractory to cell death induced by mitochondrial stressors and showed reduced superoxide production after oxidative insult (46). This suggests that ERβ promotes mitochondrial vulnerability, suggesting that an increase in estrogen signaling through ERβ in the lung would promote lung cancer development or progression by reducing cellular susceptibility to oxidative insults. Conversely, it is predicted that lung cancer cells would become more susceptible to stress- inducing agents under conditions of estrogen deficiency. Mitochondrial localization of ERβ in NSCLC cell lines has been reported (45), and studies are required to determine whether ERβ is expressed in the mitochondria of primary lung cancer cells and what role is has in oxidative processes.
Several preclinical mouse models have been used to study the role of estrogen in lung cancer. In mice in which lung adenocarcinomas were induced by K-ras activation and p53 deletion, 17β-estradiol promoted tumor progression. Administration of β-estradiol at physiologic levels doubled the number of tumors observed in whole lung mounts and this difference was similar in both males and ovariectomized females (55). In lung cancer cell lines in culture, 17β-estradiol significantly increased cell growth in vitro and in tumor xenografts, increased gene expression of several ERE-responsive genes, and increased VEGF secretion. Extent of p44/p42 mitogen activated protein kinase (MAPK) was also increased (16,38,39, 56,57). These proliferative and survival responses to 17β-estradiol may be responsible for the poorer clinical outcomes observed in NSCLC patients who have high endogenous or exogenous estrogen levels. The ER antagonist fulvestrant significantly inhibited the 17β-estradiol induced growth of NSCLC cell lines in culture and in tumor xenografts (39), providing strong rationale for the evaluation anti-estrogens in lung cancer therapy.
Recent reports demonstrate that lung cancer cells can produce their own estrogen (58). The aromatase enzyme, a cytochrome P450 protein, catalyzes the conversion of the androgens androstenedione and testosterone to estrone and estradiol, respectively, and is expressed in the lung (59,60). Aromatase protein was detected in NSCLC cell lines and primary tumor tissue and was shown to be functional, based on detection of β-estradiol release over time (61). A large decrease in size of lung tumor xenografts treated with anastrozole was also observed (61). Aromatase inhibitor therapy in lung cancer is further supported by Coombes and colleagues, who reported a decreased incidence of primary lung cancer in breast cancer patients treated long-term with the potent aromatase inhibitor exemestane after tamoxifen therapy (4 cases) compared with continued tamoxifen treatment (12 cases) (62). In addition, a recent report showed that aromatase was a protective marker of survival in women over age 65 with early stage lung cancer (63). This finding suggests the effect of local estrogen production is highest in patients where circulating estrogens are low because the ovaries have ceased functioning. Whether aromatase could be important in the survival of men with lung cancer is not known.
Given the preclinical data supporting the ability of estrogen ligands to increase gene expression and stimulate the growth of NSCLC cells, there is a strong rationale to evaluate anti-tumor activity of estrogen down-modulators in lung cancer. The available strategies for targeting estrogen signaling clincially include antagonists of ER function such as tamoxifen, down-regulation of ER function through agents such as fulvestrant and reduction of estrogen levels through aromatase inhibitors, such as the reversible non-steroidal agents letrozole and anastrozole and the irreversible steroidal inactivator exemestane (64,65). Tamoxifen has a partial agonist effects in certain tissues, such as endometrium, and is known to act as an agonist for some ERβ functions, such as stimulating AP-1-mediated transcription (66). Tamoxifen has been shown by our group to increase lung tumor xenograft growth (unpublished observations) and is not an appropriate choice of therapy for NSCLC. Additionally, results from the Tamoxifen Breast Cancer Prevention trial as part of the National Surgical Adjuvant Breast and Bowel Project did not show any decreased risk of lung cancer (67). Seventeen tumors of the lung, trachea, and bronchus were reported among the placebo group and 20 in the women who had received tamoxifen therapy. Although not statistically significant, these results suggest that tamoxifen could have some agonistic effects in the lung. Findings that lung tumors express high amounts of the enzyme aromatase and contain measureable intra-tumoral 17β-estradiol and its metabolites (68), as well as the recent observation of increased lung cancer deaths in women on HRT therapy (11), point to aromatase inhibitors as a viable potential therapeutic option. In addition, a recent case report described a post-menopausal woman with lung cancer whose tumor enlarged during a course of HRT (69). It is possible that reducing estrogenic signaling will have activity on its own, and also increase the efficacy of chemotherapy and other targeted therapies.
Understanding the biology and signaling of estrogen in lung cancer will be of benefit to both women and men with this disease. Lung tumors from both male and female patients express ERs and aromatase, and cell lines derived from both sexes respond to estrogens, anti-estrogens, and aromatase inhibitors. Therefore, therapeutic treatments that down-modulate estrogen signalling could benefit all patients, not solely women. Data are strongest for estrogen levels playing an important role in survival of women with lung cancer, and the effect of estrogen on men with lung cancer is less well studied. The relative contribution of androgen and progesterone is largely unstudied. Further understanding of the role of steroid hormones in lung cancer will provide rationales for future targeting of these pathways for therapy throughout the course of disease and possibly for lung cancer prevention. Additional understanding of the role of non-nuclear versus nuclear receptors and how to most effectively target the signaling from each cellular compartment that contributes to lung cancer progression will be important for designing effective treatments.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
From The University of Pittsburgh P50 CA090440 Specialized Program of Research Excellence (SPORE) in Lung Cancer
Jill M. Siegfried, Department of Pharmacology and Chemical Biology, University of Pittsburgh, Pittsburgh, PA.
Pamela A. Hershberger, Department of Pharmacology and Chemical Biology, University of Pittsburgh, Pittsburgh, PA.
Laura P. Stabile, Department of Pharmacology and Chemical Biology, University of Pittsburgh, Pittsburgh, PA.