Perspective

Crossroads of Estrogen Receptor and NF-κB Signaling

See allHide authors and affiliations

Science's STKE  14 Jun 2005:
Vol. 2005, Issue 288, pp. pe27
DOI: 10.1126/stke.2882005pe27

Abstract

Cellular homeostasis in higher organisms is maintained by balancing cell growth, differentiation, and death. Two important systems that transmit extracellular signals into the machinery of the cell nucleus are the signaling pathways that activate nuclear factor κB (NF-κΒ) and estrogen receptor (ER). These two transcription factors induce expression of genes that control cell fates, including proliferation and cell death (apoptosis). However, ER has anti-inflammatory effects, whereas activated NF-κB initiates and maintains cellular inflammatory responses. Recent investigations elucidated a nonclassical and nongenomic effect of ER: inhibition of NF-κB activation and the inflammatory response. In breast cancer, antiestrogen therapy might cause reactivation of NF-κB, potentially rerouting a proliferative signal to breast cancer cells and contributing to hormone resistance. Thus, ER ligands that selectively block NF-κB activation could provide specific potential therapy for hormone-resistant ER-positive breast cancers.

In a recent report, Chadwick et al. (1) presented an innovative concept regarding the mode of action of the ER when it interacts with selective ligands. These authors demonstrated that the nonsteroidal compound WAY-169916 blocked proinflammatory signals mediated by the transcription factor NF-κB. Inhibition of NF-κB signaling required binding of the selective ligand WAY-169916 to either of the two forms of ER (ERα or ERβ) and, apparently, this pathway-selective anti-inflammatory effect was achieved through a nonclassical mode of action by the receptor. In animal models, systemic administration of WAY-169916 reversed disease states caused by an abnormal inflammatory response. Furthermore, WAY-169916 failed to induce familiar effects of estrogen, mediated by its classical ligand-dependent function in promoting gene expression. Thus, WAY-169916 appears to be an anti-inflammatory agent that requires ER for its action and blocks proinflammatory signaling by NF-κB.

What brought these authors to examine the effects of an ER ligand on inflammation? In clinical medicine, states of estrogen excess and in particular pregnancy ameliorate symptoms of inflammatory diseases such as rheumatoid arthritis, inflammatory bowel disease (Crohn’s disease and ulcerative colitis), and multiple sclerosis. These effects are attributed to the increased concentrations of estrogen produced during pregnancy. Furthermore, molecular approaches have documented antagonistic cross-talk between the NF-κΒ and ER pathways by demonstrating the ability of estrogen-activated ER to quell NF-κΒ signals. The mechanisms can only be inferred but may include direct protein-protein interaction, inhibition of binding of NF-κB to DNA, or unbalanced sharing of transcriptional coactivators. Chadwick et al. (1) exploited cross-talk between ER and NF-κB pathways by developing an ER ligand that selectively inhibited NF-κB and inflammation, without inducing classical estrogen effects.

The ovarian hormone estrogen controls cell proliferation and differentiation in reproductive organs such as the uterus, pituitary gland, mammary gland, and ovary. However, estrogen has other effects in humans, among which are effects on the skeletal, cardiovascular, and nervous systems that influence bone density, concentrations of blood lipids, and cognitive function. The NF-κB transcription factor is activated by a multitude of stimuli, including cytokines, growth factors, viral and bacterial infections, and various mediators of cell stress (2, 3). In health, NF-κB signaling is required for the normal inflammatory response caused by immune activation. NF-κB is linked to disease states by way of overactivity, usually a consequence of aberrant stimulation by otherwise normal signals.

In the canonical view of ER signaling, estrogen and related sex steroids bind to the ER to promote formation of a receptor homodimer, which is released from cytoplasmic chaperones to enter the nucleus and to transactivate responsive target genes (Fig. 1). This classical signaling by estrogen and ER may be designated "genomic" signaling. The system is activated by estrogen to perform normal functions in female reproduction, or it can mediate abnormal proliferation of mammary and endometrial cells in breast and uterine cancer. In the NF-κB signaling pathway, pleiotropic extracellular factors interact with cell surface receptors and cause the release of active NF-κB (a family of at least five distinct subunits, which form homodimers or heterodimers when activated) from phosphorylated cytoplasmic inhibitory protein κB (IκB) (Fig. 1B). IκB-kinase (IKK) and, perhaps, other protein kinases catalyze phosphorylation of IκB (4). Like ER, activated NF-κB is translocated to the nucleus, where it binds to the promoter regions of a cohort of target genes and ultimately influences cell proliferation and evasion of cell death (apoptosis).

Fig. 1.

ER and NF-κB signaling pathways. (Left) Fundamentals of ER signaling. This elementary scheme illustrates the genomic actions of the ER, which entails the expression of estrogen-responsive genes. The ovarian hormone estrogen (E2) interacts with its cytoplasmic receptor (ER). On estrogen binding, monomeric ER forms a dimer in a process chaperoned by heat shock protein 90. ER bound to E2 is released from the cytoplasm, and the ligand-activated factor traffics to the nucleus where it binds to ER response elements (ERE) in the promoter region. In the presence of specific coactivators, ER initiates transcription of responsive genes, which are necessary for cell proliferation and differentiation of cells in reproductive organs of the female. (Right) Activation of NF-κB. NF-κB is a transcription factor that exists as homo- or heterodimers of Rel (reticuloendotheliosis) family proteins. One group, the processed members of the family, includes RelA (p65), RelB, and c-Rel. The second group contains the unprocessed members p105 (precursor to p50, NF-κB1) and p100 (precursor to p52, NF-κB2), which are cleaved to generate p50 and p52 in the cytoplasm. All these proteins have the common Rel homology domain (RHD), which contains DNA binding, nuclear localization, transactivation, and the IκB-binding domains. The p65-p50 heterodimer is the most commonly detected and most abundant form of NF-κB in different cell types.

NF-κB is located in the cytoplasm in most cells (with the exception of B cells) in an inactive state sequestered with IκB protein. Pleiotropic extracellular factors (PEFs), including ligand triggering of cell surface receptors, initiate phosphorylation cascades that lead to activation of IKK. Activated IKK phosphorylates IκB, marking it for degradation by proteasomes, thereby releasing and allowing translocation of the active NF-κB dimer into the nucleus. The activated NF-κΒ then binds to its NF-κΒ response element (NRE) in the promoter region of responsive genes and aids their expression. NF-κΒ is a multifunctional transcription factor and modulates the expression of genes that influence cell cycle progression, regulated cell death (apoptosis), inflammatory reactions, immune response, metastasis, stress-related genes, and integrated viral genes.

The cell proliferative actions of both estrogen and active NF-κΒ are mediated by increased expression of the cell cycle regulatory protein cyclin D1. Cyclin D1 forms complexes with cyclin-dependent kinases 4 and 6 (Cdk4 and Cdk6), and the holoenzyme phosphorylates the retinoblastoma protein (Rb), which causes the release of the transcription factor E2F-1. Free E2F-1 then augments expression of specific genes responsible for driving S phase and cell cycle progression (Fig. 2). Thus regulation of cyclin D1 is a point in the "crossroads" at which estrogen and NF-κB signaling merge, and in this case, both promote cell cycle progression (Fig. 2). In the immune system, NF-κB activates inflammatory cells and propels inflammation. In female reproductive organs, estrogen and its receptor act on target genes to cause mixed proliferative and differentiation effects. Estrogen signaling through ER also inhibits NF-κB activation, in a manner that appears to be independent of gene transcription, a so-called "nongenomic" effect of estrogen and ER signaling (Fig. 2). Therefore, the crossroads of these two pathways, exemplified by the inhibition of NF-κB activation, is potentially important in cells that simultaneously carry signaling traffic over both routes.

Fig. 2.

Crossroads of ER and NF-κΒ signaling. Estrogen interacts with ER (Fig. 1), which initiates a sequence of events leading to modulation of the expression of responsive genes (blue arrows). Estrogen can also bind to ER and exert nongenomic effects, which inhibit NF-κΒ activation and the cellular response to inflammation (red arrows). The outcomes of these two diverging actions of estrogen are tissue specific. In reproductive tissues and cells, proliferation predominates and in inflammatory cells, NF-κΒ is inhibited and proliferation retarded. WAY-169916 is a nonsteroidal ER ligand that can selectively inhibit NF-κΒ in inflammatory cells, by using the first intersection of these two pathways. NF-κΒ is activated in steps (Fig. 1) and transits into the nucleus, where it binds to NRE in the promoter regions of responsive genes. In the G1 phase of the cell cycle, the NF-κΒ and ER pathways that control cell proliferation intersect a second time by controlling the expression of cyclin D1. Cyclin D1 activates the cyclin-dependent kinases 4 and 6 (Cdk4 and Cdk6), which phosphorylate the retinoblastoma protein (Rb) and release the transcription factor E2F-1 from the RB-E2F-1 complex. E2F-1 drives the expression of genes responsible for cell cycle progression and cell proliferation. In immune cells (lymphocytes, neutrophils, and macrophages), NF-κΒ also stimulates the production of cytokines and inhibits apoptosis of immune effector cells, thereby amplifying the process of inflammation. Estrogen activates both the proliferative and anti-inflammatory pathways of ER, whereas the selective action of WAY results in only the anti-inflammatory effect.

Myeloid and lymphoid progenitor cells, mature lymphocytes, and neutrophils express ERα, ERβ, or both receptors (5). Furthermore, knockout mice devoid of expression of one or both receptor isoforms show problems with immune system development and even develop a mixed myeloproliferative and lymphoproliferative disorder (6). NF-κB activation is detected in inflammatory diseases that are associated with increased amounts of inflammatory cytokines (7, 8). Some of these cytokine’s genes are direct targets that are both expressed in response to NF-κB activation and can feed-forward to activate NF-κB by causing release of NF-κB from IκB. This can amplify the immune response and prolong inflammation, either in a physiologic response or in a disease state. Estrogen or a selective compound like WAY-169916 can act through the nonclassical and anti-inflammatory pathway between ER and NF-κB and can inhibit the inflammatory chain reaction.

Human breast cancer is another disease state where the NF-κB and ER pathways intersect. Sixty percent of human breast cancers express ER (the ER-positive cancers), and the predominant effect of ER and estrogen is to stimulate cancer cell growth. A probable consequence of the cross-talk between ER and NF-κB in breast cancer cells was observed in human breast tumor specimens (9). Active NF-κB detected by DNA binding was found in extracts of the majority of tumor samples that lacked ER expression (ER-negative tumors), whereas DNA-bound NF-κB was practically absent in ER-positive specimens. It is likely that ambient estrogens bound to ER and suppressed NF-κB activity in ER-positive human breast tumor samples. Because primary human breast tumors were taken from patients before treatment, the effect of hormone therapy on NF-κΒ status in specimens from treated patients has not been addressed.

Restoration of NF-κΒ signaling during antiestrogen treatment might send proliferative signals to mammary cancer cells. Pharmaceuticals are available that inhibit NF-κB, some of which are in clinical use (for example, bortezomib, a proteosome inhibitor that stabilizes IκB). Many other low-molecular-weight compounds under development target the very specific phosphorylation of IκB by IKK and the subsequent release of active NF-κB. These drugs may find their way into the treatment of both ER-negative breast cancer (with generally high constitutive activity of NF-κB) and perhaps hormone-resistant ER-positive breast cancer.

Currently available drugs that target estrogen signaling can be divided into the selective estrogen response modulators (SERMs), which act as partial agonists; completely antagonistic agents (fulvestrant), which cause rapid degradation of receptor; and the aromatase inhibitors, which inhibit estrogen synthesis. Neither the SERM raloxifene nor the antagonist fulvestrant fully inhibits NF-κB activation produced by interleukin 1. In patients with ER-positive breast cancer, the consequences of estrogen blockade are less certain. Resistance to SERMs is at least partly explained by their acquiring agonist activity with continued usage. Whether NF-κΒ is reactivated in parallel with the resistance to SERMs is potentially an important issue but remains an unanswered question. Furthermore, if NF-κB activity is modulated during antihormone treatment, will this be beneficial to or detrimental for patients with ER-positive breast cancer? Blocking the chronic repression of NF-κΒ activation by estrogen signaling, as well as its release by antiestrogen treatment, may play some role in resistance to endocrine therapy of breast cancer.

WAY-169916 requires the presence of one of the ER receptor isoforms for its inhibition of NF-κΒ. Both fulvestrant and raloxifene antagonized the effect of WAY-169916 in vitro, and fulvestrant blocked NF-κB inhibition in an animal model of inflammatory bowel disease (that is, it blocked the beneficial effect of WAY-169916). Aromatase inhibitors are often prescribed for postmenopausal women with ER-positive breast cancer. In this circumstance, ER is freed from its natural ligands and remains available for binding to a selective ligand like WAY-169916. This natural "back door" into NF-κΒ activation machinery provides a way to inhibit NF-κΒ by co-opting the ER in ER-positive breast cancer.

In cultured cell systems and animal models (911), NF-κB activation has been implicated in breast cancers, particularly those driven by two members of the Erb family of receptors: ErbB1 (epidermal growth factor receptor EGFR) or ErbB2 (HER2 or Neu). Increased expression of ErbB2 due to increased gene copy number (amplification) is found in 20% of human breast cancers, and at least one-half of these ErbB2-positive tumors are ER-positive. Furthermore, most patients with ER-positive cancers receive endocrine therapy with estrogen-receptor inhibitors The consequences of releasing NF-κB from inhibition by estrogen during endocrine attack on breast cancers are unknown and only speculative. Compounds such as WAY-169916 might be beneficial even for therapy of hormone-resistant breast cancers by decreasing the effects of NF-κB activation. Thus, the existence of a crossroads of estrogen and NF-κB signaling traffic could provide rationale for targeting NF-κB activation in both ER-negative and ER-positive breast cancers during endocrine treatment.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.

Stay Connected to Science Signaling

Navigate This Article