Journal Facts

Journal
BoneKEy Knowledge Environment
ISSN
1533-4368
Volume
Year
2002

Article Facts

Title
Nuclear receptor comodulators do matter!
DOI
10.1138/2002042
Authors

Online Date
05/23/2002

Article Links

BoneKEy-Osteovision | Commentary

Nuclear receptor comodulators do matter!


DOI:10.1138/2002042

Commentary on: Shang Y, Brown M. Molecular determinants for the tissue specificity of SERMs. Science. 2002 Mar 29;295(5564):2465-8.

Background

The role of estrogen in female reproductive functions is well described. Tissue targets for estrogen include the ovaries, uterus, vagina and breast. Estrogen's biological activity is much more pleiotropic, however, extending well beyond the reproductive tract to a number of tissues not only in females, but in males as well. Non-reproductive tissue targets include the immune system, the cardiovascular system, the central nervous system and the skeleton (). Indeed, it may be rare to find a mammalian tissue that does not express small numbers of receptors for estrogen (ER), extending dramatically the number of tissues potentially sensitive to this sex steroid. The biological range of action of estrogen has been highlighted most recently in male and female ERα and ERβ knockout mice ().

Mechanism of estrogen action

The genomic mechanism of action of estrogen in general terms is reasonably well understood (). The ligand enters the cell by diffusion and forms a complex with nuclear ER. Complex formation prompts the dissociation of inhibitory proteins from the receptor, homodimer formation and finally interaction with unique DNA sequences or estrogen response elements (EREs). Recent studies by numerous investigators indicate that ER functions to a large extent as a nucleation sites for the recruitment of comodulatory complexes that served as enzymatic mediators of the transcriptional regulatory process. The recruitment of these comodulatory complexes is likely cell-specific and may be gene promoter-specific. Comodulator functions include prompting changes in chromatin structure that are essential to the initiation of transcription as well as facilitating direct contact with the general transcriptional apparatus for regulating transcriptional output. These actions can be either positive or negative. In the former case, comodulators such as SRC-1 and GRIP both retain intrinsic histone acetyltransferase (HAT) activity essential to chromatin modification and also recruit additional HATs (). Corepressors such as NCoR and SMRT act in turn to recruit histone deacetyltransferases that function to limit chromatin accessibility (). Additional enzymatic modifications include phosphorylation and more recently methylation (). The importance of several of these comodulators has been highlighted recently in mice in gene knockout experiments ().

Estrogen via its receptor also functions to modify ongoing transcription through indirect mechanisms. One mode of action involves the association of receptor-ligand complexes with transcription factors bound directly to active promoters, thus enhancing or in many cases repressing existing levels of activity (). Since ER associates with an extensive number of basal as well as regulatory transcription factors, this mechanism of estrogen action has the potential both to modify more broadly the expression patterns of target tissues and to provoke unique and perhaps tissue-selective profiles of activity. In a second mode, the ability of estrogen to regulate patterns of gene expression may also be enhanced though the ligand's apparent capacity to trigger at the plasma membrane activation of several signaling cascades that converge on the MAPKs (). It is unclear currently how such a mechanism might work, although studies are ongoing to more precisely define the molecular basis for these effects and to assess their relevance in vivo. A more thorough understanding of this activity is critical, since most of the steroid hormones appear capable of exploiting this type of signaling mechanism.

Molecular and biologic actions of the SERMs

As indicated above, the ability of ER to recruit either a comodulator or a corepressor is hypothesized to represent a central determinant of ER's capacity to activate or repress transcription. This ability lies at the heart of selective estrogen receptor modulator (SERM) action. SERMs such as tamoxifen (T) and raloxifene (R) represent ER ligands that induce unique conformations within ER to produce unique tissue-selective actions. The ability of different ligands to induce unique ER conformations is not in dispute. X-Ray crystallographic studies reveal that the structure of ER is indeed different when bound to estrogen, T or R (). That unique ligands induce recruitment of different comodulators is also clear in vitro. Both biochemical and molecular biological studies have demonstrated that interaction between ER and specific comodulators is dramatically influenced by ER ligands such as estrogen, T, and R and that these interactions appear to have transcriptional consequences (). Perhaps more importantly, the above studies show that both promoter context and cellular background exert profound influence on the transcriptional capabilities of ER ligands. Accordingly, while estrogen and R function in both breast and endometrial cells as an agonist and antagonist, respectively, T displays both agonist (endometrial cells) and antagonist (breast cells) properties (). These differential activities in cultured cells reflect those observed in both the breast and uterus in vivo, although a direct cause-and-effect relationship has not been clearly established. Missing from these studies is a direct demonstration using key endogenous genes as reporters that 1) estrogen agonist activity is associated with recruitment of coactivators, 2) ER antagonist activity is associated with corresponding corepressor recruitment, and 3) cell-selective partial agonist activity such as that seen in the uterus with T involves unique coactivator participation.

Commentary

Yongfeng Shang and Myles Brown address these issues directly in their recent paper entitled “Molecular determinants for the tissue specificity of SERMs” () Briefly, the authors establish using MCF-7 breast cancer and Ishikawa endometrial cell lines that estrogen acts in both cell types to induce two classes of endogenous genes: those that contain estrogen response elements or EREs in their promoters and involve direct ER binding to DNA (ERE-dependent) and those that do not contain EREs and involve interaction of ER with other transcription factors that are bound to DNA (ERE-independent). By contrast, raloxifene has no activity on either of these gene classes in either cell type and, as we will see, acts as an active antagonist. Tamoxifen is selective, however, both at the level of the cell and at the level of the promoter. Accordingly, tamoxifen functions as an antagonist in MCF-7 cells and as an agonist in Ishikawa cells. This agonist activity is, nevertheless, restricted to the c-myc and IGF-1 promoters, promoters not regulated by the ER through the presence of an ERE. These results establish excellent cellular models to explore ER-dependent comodulator recruitment to endogenous gene promoters at the molecular level.

The authors proceed to explore the activities of these ligands using a particularly useful technique termed Chromatin ImmunoPrecipitation Assay or ChIP. In this assay, intact cells are subjected to specific treatments known to modify transcription. Following this, the cells are treated with reagents that induce covalent protein-protein and protein-DNA cross-links. Chromatin is then sheared to short lengths (300 to 500 bp) and the fragments subjected to immunoprecipitation using antibodies to receptors, comodulators, or other chromatin or DNA-binding proteins of interest. Finally, gene promoter-selective primers are employed in a PCR reaction to determine the presence of a specific gene in the immunoprecipitate. Importantly, successful amplification provides strong evidence that the protein used to immunoprecipitate the chromatin fragment was associated directly with that gene promoter in the intact cell. The authors elegantly show that regardless of the cell type or the gene class examined, estrogen initiates recruitment of the comodulator SRC-1, GRIP, AIB1, CBP, and acetylated histone. All of these proteins are known to participate in transcriptional induction. As expected, estrogen is unable to recruit corepressors such as NCoR or SMRT, or the histone deacetylases HDAC2 or HDAC4. R functions in reverse. Accordingly, R induces recruitment of NCoR, SMRT, HDAC2 and HDAC4 but none of the coactivators and does so irrespective of the gene promoter class. This result indicates that R promotes active repression rather than passive antagonism that would occur in the presence of estrogen.

Tamoxifen is very different, however. While tamoxifen promotes the recruitment of corepressors to both classes of gene promoter classes in MCF-7 cells, in Ichikawa cells this compound appears to recruit corepressors to the two ERE-dependent genes but coactivators to the two ERE-independent genes c-myc and IGF-1. These latter two gene targets are the very ones that were induced by tamoxifen in the Ishikawa cell line. Interestingly, these data suggest that tamoxifen is not merely a partial agonist in the uterine cell but rather both an agonist and an antagonist depending upon the gene target. Thus, the ability of a compound to display agonist or antagonist properties is not cell- or tissue-specific, but rather gene promoter-specific. One is tempted to speculate that perhaps only ERE-independent gene promoters such as those for c-Myc or IGF-1 can be induced through tamoxifen partial agonist activity. This is not likely to be the case, however, as tamoxifen exhibits partial agonist activity on the complement 3 promoter, a promoter that is regulated through an ERE-based mechanism. Thus, future studies will almost certainly focus on the use of a more extensive array of estrogen-regulated gene targets.

Receptor conformation induced by individual ligands determines the ability of ER to recruit either coactivators (E) or corepressors (R). Nevertheless, the capacity of T to act on the same gene target as both an agonist in Ishikawa cells and as an antagonist in MCF-7 cells suggests an addition determinant as well - the level of comodulators present in the two cell. Indeed, a search of comodulator concentration revealed a substantial reduction in the level of expression of SRC-1 in MCF-7 cells as compared to Ishikawa cells. This finding led the authors to hypothesize that low levels of expression of SRC-1 might account for the inability of tamoxifen to induce c-Myc or IGF1 expression in MCF-7 cells and that high levels of SRC-1 may account for the agonist activity in Ishikawa cells. To test this, the authors enhanced SRC-1 levels in MCF-7 cells through forced expression and suppressed SRC-1 levels in Ishikawa cells using short interfering RNA molecules and then assess the activity of tamoxifen on c-Myc and IGF-1 expression. Remarkably, overexpression of SRC-1 in MCF-7 cells converted tamoxifen from an antagonist to an inducer of c-Myc and IGF-1 expression whereas reduction of SRC-1 expression in Ishikawa cells suppressed the compound's partial agonist activity.

The essential role of SRC-1 in the agonist activity of tamoxifen was further strengthened by the observation that reduction in SRC-1 expression in Ishikawa cells almost completely blocks tamoxifen-induced uterine cell growth. Surprisingly, the agonist activities of estrogen were largely unperturbed by increasing or decreasing SRC-1 activity in MCF-7 cells or Ishikawa cells, respectively. This unexpected finding suggests that specific coactivator requirements are different for estrogen and tamoxifen activation. Regardless, these studies support the idea that while receptor conformation is a primary determinant of comodulator recruitment, the levels of key comodulators play a significant role in the ability of a given ligand to function differentially in unrelated cells or tissues.

Are comodulators relevant in bone?

Estrogens have long been known to play an important role in adult bone remodeling, a process that involves coupled bone resorption and bone formation at selected sites throughout the skeleton (). Accordingly, loss of ovarian function following menopause leads to a substantial increase in bone turnover and a critical imbalance between bone formation and resorption. This imbalance leads to a progressive loss of trabecular bone mass, osteoporosis, and eventually bone fracture. Importantly, this loss of bone can be prevented clinically in postmenopausal women through estrogen replacement therapy (). The imbalance is due at least in part to an increase in the production, activity and survival of the osteoclast, the major cellular entity responsible for bone resorption in vivo (). Thus, estrogens are believed to exert their bone-sparing effects on the skeleton largely through antiresorptive mechanisms that involve regulation of the osteoclast.

A number of mechanisms have been described to account for postmenopausal bone loss. At their core is the ability of estrogen to suppress the expression of a variety of inflammatory cytokines including IL-1, TNFα, and IL-6 and other regulatory molecules such as M-CSF, osteoprotegerin and Rank ligand that function to modulate osteoclast formation, activity and survival (). An overarching theme is emerging wherein estrogens also regulate the cell populations that produce these molecules. Clearly, both B and T cell populations rise following estrogen loss as do additional cell types with a similar capacity to modulate osteoclast formation, activity and survival (). The physiologic impact of upregulating numbers of cells that produce regulatory cytokines is likely to be far greater than that achieved by simply altering cellular cytokine expression. Finally, estrogens also exert direct effects on osteoclast precursors and on the osteoclast itself. Accordingly, loss of estrogen leads to an upregulation of CFU-GM, one of the earliest established osteoclast precursors (). Estrogen depletion also leads to an elevation in the number of more proximal precursors, and an increase in osteoclast numbers, activity and survival time (). Studies also suggest that estrogens can suppress osteoclast formation directly in vitro through a mechanism that retards the signaling pathways essential to Rank ligand- induced osteoclast differentiation (). While the myriad of findings attest to the amazing complexity of estrogen's actions in bone cells, the common theme that emerges is one of suppression or inhibition.

Studies in animal models as well as in the clinic suggest that tamoxifen and raloxifene both function in the skeleton predominantly as estrogen mimetics (). These studies therefore suggest a common cellular mechanism of action. Since each of the compounds binds directly to ER and utilizes this mediator to regulate gene expression, one must conclude that the interaction of the ER with a common comodulator lies at the heart of this regulatory mechanism. Thus, the answer to the above question “Are comodulators relevant to bone?” is almost certainly “YES”, although that must now be demonstrated directly. The mimetic nature of tamoxifen and raloxifene, coupled with insights that are emerging from discoveries such as those found in the current paper, should facilitate a more complete understanding of the molecular mechanism of action of estrogen in bone in the coming years.

Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.