- Email: [email protected]
Another Tie that Binds the MTA Family to Breast Cancer
Cell 142
Another Tie that Binds the MTA Family to Breast Cancer
In this issue of Cell, Fujita et al. (2003) demonstrate that MTA3 is an estrogen-dependent component of the NuRD complex and identify the Snail gene as its direct target. ER signaling upregulates MTA3 levels to negatively modulate Snail-mediated repression of E-cadherin. These findings may explain how ER status controls epithelial-to-mesenchymal transition in human breast tumors. Estrogen plays an essential role in normal breast development and in breast cancer development and progression. Its biosynthesis is regulated by aromatase, the upregulation of which in breast cancer stimulates growth in autocrine and paracrine manners. The principal target of estrogen, the estrogen receptor (ER), is found in 40%– 70% of breast tumors at diagnosis, together with a profile of ER-regulated genes. ER-positive tumors generally are responsive to anti-hormonal therapy, although a significant proportion of patients with this class of tumor do not respond to this therapy. In addition, most who do respond initially eventually develop hormone-independent tumors characterized by an aggressive clinical course and increased metastasis. Little is known about the molecular mechanisms by which ER-negative tumors become aggressive and metastatic. Although localized breast cancer can be cured by surgery, breast cancer has a high mortality rate due primarily to frequent metastasis while the primary tumor is undetected. Metastasis requires, among other steps, alterations in signaling pathways and target gene products, increased epithelial-to-mesenchymal transition (EMT), and aberrant expression and function of cell adhesion components. For example, loss of the cell adhesion molecule E-cadherin leads to epithelial dedifferentiation and increased metastasis. This loss is largely due to repression of the E-cadherin gene by the transcription factor Snail, the master regulator of EMT. Snail and aromatase levels can be inversely correlated in cancer tissue, and since Snail also downregulates aromatase expression (Chen et al., 2001), Snail may influence the levels of circulating estrogen as well as the development of hormone-independent, aggressive breast tumors. Metastasis-associated genes (MTAs) comprise a novel gene family with a growing number of members. Currently, there are three known genes encoding six isoforms (MTA1, MTA1s, MTA-ZG29p, MTA2, MTA3, MTA3L) (Wang and Kumar, 2003; Fujita et al., 2003). MTA1 was identified as a differentially expressed gene in rat metastatic tumors (Toh et al., 1994). Later studies identified other family members, including MTA3 (Simpson, et al., 2001). Since MTA proteins do not seem to possess enzymatic activity, the mechanism(s) of their function remains a mystery. The discovery that MTA1 is part of a nuclear remodeling and deacetylation complex (NuRD) suggested that the major function of these proteins might be to form a repressive chromatin state (Xue et al., 1998).
The recent finding that MTA1 interacts with CAK, a component of the TFIIH regulatory complex, suggests that MTA1 may also act as a signal transducer to mediate crosstalk between corepressor complexes and the general transcription machinery (Talukder et al., 2003). Unexpectedly, MTA1 was also identified as a target of growth factor signal transduction, and a role for MTA1 in hormonal independence was suggested (Mazumdar et al., 2001). This study found downregulation of ligandinduced ER transcriptional activity via recruitment of HDACs to the ER. Another study showed that a naturally occurring variant of MTA1, MTA1s, is overexpressed in ER-negative tumors (Kumar et al., 2002). MTA1s not only inhibits nuclear signaling by sequestering ER in the cytoplasm, but it also enhances ER cytoplasmic signaling and, thus, promotes tumorigenesis. Together, these studies suggest a complex role for MTA1 and MTA1s in blocking ER functions. The report by Fujita et al. (2003 [this issue of Cell]) offers yet another connection between MTAs and the ER, identifying MTA3 as an ER-regulated component of the Mi2/NuRD complex and showing that its expression is downregulated in ER-negative breast tumors. Given the established role of MTA family members in the NuRD complex, it was perhaps expected that MTA3 also would function in this repression complex. A pioneering aspect of this study, however, is the finding that MTA3 is an ERregulated gene and directly targets Snail. Modulation of E-cadherin expression via MTA3- and ER-signalingmediated changes in Snail expression has important phenotypic implications, and the authors provide correlative data for the existence of this ER-MTA3-Snail pathway in human breast tumor samples. Despite the important implications, the biochemical studies of Fujita et al. have yet to undergo functional assessment in a physiologically relevant model and, therefore, must be interpreted with caution. Identifying MTA3 as an ER-regulated gene supports the idea that MTA members play an important role in the ER pathway. Although earlier studies identified MTA1 and MTA2 as components of NuRD, Fujita et al.’s results are exciting and provocative. The failure to detect MTA1 and MTA2 in the MTA3-containing NuRD complex raises the possibility (adeptly proposed by the authors) that different NuRD complexes with distinct subunits of MTA family members exist. Whether this is true and whether they differ only in MTA family members or in other distinct subunits as well needs to be determined. What would be the normal functions of different NuRDs with distinct MTA-member subunits? Fujita et al. conclusively establish that loss of ER results in decreased MTA3 expression, thus activating Snail expression and, inferentially, EMT. The authors suggest that this pathway may have a role in the metastasis phenotype observed in ER-negative tumors. These data, however, also raise the new hypothesis that this pathway may be involved in ER-positive MTA1 deregulated tumors. Since MTA1 and MTA1s are potent repressors of nuclear ER functions, deregulation of these proteins may also downregulate MTA3 expression via ER repression, which can then lead to enhanced Snail expression and EMT. In addition, Snail-mediated repres-
Previews 143
Figure 1. Regulatory Interplay between MTA Family Members and ER Pathways Estrogen binding to ER leads to increased expression of MTA3, which in turn represses the expression of Snail, allowing upregulation of E-cadherin expression and cell adhesiveness. The loss of ER and the blockage of ER’s nuclear functions by MTA1 and MTA1s could interrupt ER regulation of MTA3 expression, permitting elevated levels of Snail to suppress E-cadherin expression. These biochemical alterations in the levels of Snail and E-cadherin promote loss of epithelial cell adhesion while enhancing EMT and invasiveness. Repression of aromatase expression by Snail can also contribute toward the development of ER-negative phenotypes. Black solid lines, activation pathways; red solid lines, inhibitory pathways; broken lines, postulated pathways yet to be validated; MTA-BPs, MTA binding proteins; CoA, ER coactivators.
sion of the aromatase gene will inhibit ER regulation of MTA3 by limiting the levels of estrogen (Figure 1). Future studies of the mechanisms by which ER regulates MTA3 expression and of the role of other MTAs or MTA binding proteins (Talukder et al., 2003) in regulating MTA3 are warranted. How does ER regulate MTA3 expression? Do ER coactivators or MICoA, a recently discovered MTA-interacting ER coactivator (Mishra et al., 2003), play a role? Does ER regulate MTA3 expression via the classical pathway or a nonclassical pathway, since there are three ER element half-sites (TGACCT) in the MTA3 promoter? Fujita et al. suggest that ER-mediated regulation of MTA3 depends on new protein synthesis. Understanding the regulatory elements in the MTA3 promoter and the regulatory proteins involved is crucial to exploring the proposed ER-MTA3 signaling pathway and gaining further insight into the connection between ER and MTAs. It also is likely that MTA3 modulates transactivation functions of ER and regulates other essential proteins besides E-cadherin. The identification of specific binding partners for MTA3 may shed light on other critical regulators within this new regulatory pathway. Another important question regarding MTA3 regulation concerns the role that selective estrogen-receptor modulators (SERMs) play. Despite being a partial agonist in endometrial Ishikawa cells, tamoxifen (an antagonist in MCF-7 breast cancer cells) reduced the expression of MTA3 in endometrial cells (Fujita et al., 2003). This finding raises the possibility that ER corepressor proteins also play a role in the regulation of MTA3. Do new-generation SERMs regulate the expression of MTA3? Additionally, ER may help in regulation of MTA3, raising
the question of whether MTA3 regulation is specific to ER␣ or ER. It will be important to know if MTA3 expression is deregulated in other hormone-dependent/hormone-resistant cancers, and if restoring MTA3 expression can reduce the metastatic potential of cancer cells. All the known MTA family members are widely expressed in normal tissues, so understanding MTA activities in normal development is critical. It is equally important to understand how MTA family members regulate NuRD complex activity. Since MTAs contain several protein-protein interaction motifs, identifying potential binding proteins will help elucidate the function of MTA genes. Identification of potential targets of each MTA family member is another worthy goal. MTA members are overexpressed in metastatic tumors, are part of NuRD complexes, and are regulated by growth factors/oncogenes. They modulate ER functions and may participate in EMT. These data suggest that MTA proteins play an important and fundamental role in pathological processes. The report by Fujita et al. should provide impetus for future studies directed at increasing our understanding of the functions and regulation of MTA family members.
Rakesh Kumar Department of Molecular and Cellular Oncology The University of Texas M.D. Anderson Cancer Center Houston, Texas 77030 Selected Reading Chen, S., Zhou, D., Yang, C., Okubo, T., Kinoshita, Y., Yu, B., Kao, Y.C., and Itoh, T. (2001). J. Steroid Biochem. Mol. Biol. 79, 35–40. Fujita, N., Jaye, D.L., Kajita, M., Geierman, C., Moreno, C.S., and Wade, P.A. (2003). Cell 113, this issue, 207–219. Kumar, R., Wang, R., Mazumdar, A., Talukder, A.H., Mandal, M., Yang, Z., Bagheri-Yarmand, R., Sahin, A., Hortobagyi, G., Adam, L., et al. (2002). Nature 418, 654–657. Mazumdar, A., Wang, R., Mishra, S.K., Adam, L., Yarmand, R.B., Mandal, M., Vadlamudi, R., and Kumar, R. (2001). Nat. Cell Biol. 3, 30–37. Mishra, S.K., Mazumdar, M., Wang, R., Li, F., Yu, W., Vadlamudi, R.K., Jordan, V.C., Santen, R.J., and Kumar, R. (2003). J. Biol. Chem., in press. Published online March 14, 2003. 10.1074/jbc. M301968200. Simpson, A., Uitto, J., Rodeck, U., and Mahony, M.G. (2001). Gene 273, 29–39. Talukder, A.H., Mishra, S.K., Mandal, M., Balasenthil, S., Mehta, S., Sahin, A.A., Barnes, C.J., and Kumar, R. (2003). J. Biol. Chem. 278, 11676–11685. Toh, Y., Pencil, S.D., and Nicolson, G.L. (1994). J. Biol. Chem. 269, 22958–22963. Wang, R., and Kumar, R. (2003). In The Identities of Membrane Steroid Receptors (Boston: Kluwer Academic Publishers), pp. 119–124. Xue, Y., Wong, J., Moreno, G.T., Young, M.K., Cote, J., and Wang, W. (1998). Mol. Cell 2, 851–861.
Another Tie that Binds the MTA Family to Breast Cancer
In this issue of Cell, Fujita et al. (2003) demonstrate that MTA3 is an estrogen-dependent component of the NuRD complex and identify the Snail gene as its direct target. ER signaling upregulates MTA3 levels to negatively modulate Snail-mediated repression of E-cadherin. These findings may explain how ER status controls epithelial-to-mesenchymal transition in human breast tumors. Estrogen plays an essential role in normal breast development and in breast cancer development and progression. Its biosynthesis is regulated by aromatase, the upregulation of which in breast cancer stimulates growth in autocrine and paracrine manners. The principal target of estrogen, the estrogen receptor (ER), is found in 40%– 70% of breast tumors at diagnosis, together with a profile of ER-regulated genes. ER-positive tumors generally are responsive to anti-hormonal therapy, although a significant proportion of patients with this class of tumor do not respond to this therapy. In addition, most who do respond initially eventually develop hormone-independent tumors characterized by an aggressive clinical course and increased metastasis. Little is known about the molecular mechanisms by which ER-negative tumors become aggressive and metastatic. Although localized breast cancer can be cured by surgery, breast cancer has a high mortality rate due primarily to frequent metastasis while the primary tumor is undetected. Metastasis requires, among other steps, alterations in signaling pathways and target gene products, increased epithelial-to-mesenchymal transition (EMT), and aberrant expression and function of cell adhesion components. For example, loss of the cell adhesion molecule E-cadherin leads to epithelial dedifferentiation and increased metastasis. This loss is largely due to repression of the E-cadherin gene by the transcription factor Snail, the master regulator of EMT. Snail and aromatase levels can be inversely correlated in cancer tissue, and since Snail also downregulates aromatase expression (Chen et al., 2001), Snail may influence the levels of circulating estrogen as well as the development of hormone-independent, aggressive breast tumors. Metastasis-associated genes (MTAs) comprise a novel gene family with a growing number of members. Currently, there are three known genes encoding six isoforms (MTA1, MTA1s, MTA-ZG29p, MTA2, MTA3, MTA3L) (Wang and Kumar, 2003; Fujita et al., 2003). MTA1 was identified as a differentially expressed gene in rat metastatic tumors (Toh et al., 1994). Later studies identified other family members, including MTA3 (Simpson, et al., 2001). Since MTA proteins do not seem to possess enzymatic activity, the mechanism(s) of their function remains a mystery. The discovery that MTA1 is part of a nuclear remodeling and deacetylation complex (NuRD) suggested that the major function of these proteins might be to form a repressive chromatin state (Xue et al., 1998).
The recent finding that MTA1 interacts with CAK, a component of the TFIIH regulatory complex, suggests that MTA1 may also act as a signal transducer to mediate crosstalk between corepressor complexes and the general transcription machinery (Talukder et al., 2003). Unexpectedly, MTA1 was also identified as a target of growth factor signal transduction, and a role for MTA1 in hormonal independence was suggested (Mazumdar et al., 2001). This study found downregulation of ligandinduced ER transcriptional activity via recruitment of HDACs to the ER. Another study showed that a naturally occurring variant of MTA1, MTA1s, is overexpressed in ER-negative tumors (Kumar et al., 2002). MTA1s not only inhibits nuclear signaling by sequestering ER in the cytoplasm, but it also enhances ER cytoplasmic signaling and, thus, promotes tumorigenesis. Together, these studies suggest a complex role for MTA1 and MTA1s in blocking ER functions. The report by Fujita et al. (2003 [this issue of Cell]) offers yet another connection between MTAs and the ER, identifying MTA3 as an ER-regulated component of the Mi2/NuRD complex and showing that its expression is downregulated in ER-negative breast tumors. Given the established role of MTA family members in the NuRD complex, it was perhaps expected that MTA3 also would function in this repression complex. A pioneering aspect of this study, however, is the finding that MTA3 is an ERregulated gene and directly targets Snail. Modulation of E-cadherin expression via MTA3- and ER-signalingmediated changes in Snail expression has important phenotypic implications, and the authors provide correlative data for the existence of this ER-MTA3-Snail pathway in human breast tumor samples. Despite the important implications, the biochemical studies of Fujita et al. have yet to undergo functional assessment in a physiologically relevant model and, therefore, must be interpreted with caution. Identifying MTA3 as an ER-regulated gene supports the idea that MTA members play an important role in the ER pathway. Although earlier studies identified MTA1 and MTA2 as components of NuRD, Fujita et al.’s results are exciting and provocative. The failure to detect MTA1 and MTA2 in the MTA3-containing NuRD complex raises the possibility (adeptly proposed by the authors) that different NuRD complexes with distinct subunits of MTA family members exist. Whether this is true and whether they differ only in MTA family members or in other distinct subunits as well needs to be determined. What would be the normal functions of different NuRDs with distinct MTA-member subunits? Fujita et al. conclusively establish that loss of ER results in decreased MTA3 expression, thus activating Snail expression and, inferentially, EMT. The authors suggest that this pathway may have a role in the metastasis phenotype observed in ER-negative tumors. These data, however, also raise the new hypothesis that this pathway may be involved in ER-positive MTA1 deregulated tumors. Since MTA1 and MTA1s are potent repressors of nuclear ER functions, deregulation of these proteins may also downregulate MTA3 expression via ER repression, which can then lead to enhanced Snail expression and EMT. In addition, Snail-mediated repres-
Previews 143
Figure 1. Regulatory Interplay between MTA Family Members and ER Pathways Estrogen binding to ER leads to increased expression of MTA3, which in turn represses the expression of Snail, allowing upregulation of E-cadherin expression and cell adhesiveness. The loss of ER and the blockage of ER’s nuclear functions by MTA1 and MTA1s could interrupt ER regulation of MTA3 expression, permitting elevated levels of Snail to suppress E-cadherin expression. These biochemical alterations in the levels of Snail and E-cadherin promote loss of epithelial cell adhesion while enhancing EMT and invasiveness. Repression of aromatase expression by Snail can also contribute toward the development of ER-negative phenotypes. Black solid lines, activation pathways; red solid lines, inhibitory pathways; broken lines, postulated pathways yet to be validated; MTA-BPs, MTA binding proteins; CoA, ER coactivators.
sion of the aromatase gene will inhibit ER regulation of MTA3 by limiting the levels of estrogen (Figure 1). Future studies of the mechanisms by which ER regulates MTA3 expression and of the role of other MTAs or MTA binding proteins (Talukder et al., 2003) in regulating MTA3 are warranted. How does ER regulate MTA3 expression? Do ER coactivators or MICoA, a recently discovered MTA-interacting ER coactivator (Mishra et al., 2003), play a role? Does ER regulate MTA3 expression via the classical pathway or a nonclassical pathway, since there are three ER element half-sites (TGACCT) in the MTA3 promoter? Fujita et al. suggest that ER-mediated regulation of MTA3 depends on new protein synthesis. Understanding the regulatory elements in the MTA3 promoter and the regulatory proteins involved is crucial to exploring the proposed ER-MTA3 signaling pathway and gaining further insight into the connection between ER and MTAs. It also is likely that MTA3 modulates transactivation functions of ER and regulates other essential proteins besides E-cadherin. The identification of specific binding partners for MTA3 may shed light on other critical regulators within this new regulatory pathway. Another important question regarding MTA3 regulation concerns the role that selective estrogen-receptor modulators (SERMs) play. Despite being a partial agonist in endometrial Ishikawa cells, tamoxifen (an antagonist in MCF-7 breast cancer cells) reduced the expression of MTA3 in endometrial cells (Fujita et al., 2003). This finding raises the possibility that ER corepressor proteins also play a role in the regulation of MTA3. Do new-generation SERMs regulate the expression of MTA3? Additionally, ER may help in regulation of MTA3, raising
the question of whether MTA3 regulation is specific to ER␣ or ER. It will be important to know if MTA3 expression is deregulated in other hormone-dependent/hormone-resistant cancers, and if restoring MTA3 expression can reduce the metastatic potential of cancer cells. All the known MTA family members are widely expressed in normal tissues, so understanding MTA activities in normal development is critical. It is equally important to understand how MTA family members regulate NuRD complex activity. Since MTAs contain several protein-protein interaction motifs, identifying potential binding proteins will help elucidate the function of MTA genes. Identification of potential targets of each MTA family member is another worthy goal. MTA members are overexpressed in metastatic tumors, are part of NuRD complexes, and are regulated by growth factors/oncogenes. They modulate ER functions and may participate in EMT. These data suggest that MTA proteins play an important and fundamental role in pathological processes. The report by Fujita et al. should provide impetus for future studies directed at increasing our understanding of the functions and regulation of MTA family members.
Rakesh Kumar Department of Molecular and Cellular Oncology The University of Texas M.D. Anderson Cancer Center Houston, Texas 77030 Selected Reading Chen, S., Zhou, D., Yang, C., Okubo, T., Kinoshita, Y., Yu, B., Kao, Y.C., and Itoh, T. (2001). J. Steroid Biochem. Mol. Biol. 79, 35–40. Fujita, N., Jaye, D.L., Kajita, M., Geierman, C., Moreno, C.S., and Wade, P.A. (2003). Cell 113, this issue, 207–219. Kumar, R., Wang, R., Mazumdar, A., Talukder, A.H., Mandal, M., Yang, Z., Bagheri-Yarmand, R., Sahin, A., Hortobagyi, G., Adam, L., et al. (2002). Nature 418, 654–657. Mazumdar, A., Wang, R., Mishra, S.K., Adam, L., Yarmand, R.B., Mandal, M., Vadlamudi, R., and Kumar, R. (2001). Nat. Cell Biol. 3, 30–37. Mishra, S.K., Mazumdar, M., Wang, R., Li, F., Yu, W., Vadlamudi, R.K., Jordan, V.C., Santen, R.J., and Kumar, R. (2003). J. Biol. Chem., in press. Published online March 14, 2003. 10.1074/jbc. M301968200. Simpson, A., Uitto, J., Rodeck, U., and Mahony, M.G. (2001). Gene 273, 29–39. Talukder, A.H., Mishra, S.K., Mandal, M., Balasenthil, S., Mehta, S., Sahin, A.A., Barnes, C.J., and Kumar, R. (2003). J. Biol. Chem. 278, 11676–11685. Toh, Y., Pencil, S.D., and Nicolson, G.L. (1994). J. Biol. Chem. 269, 22958–22963. Wang, R., and Kumar, R. (2003). In The Identities of Membrane Steroid Receptors (Boston: Kluwer Academic Publishers), pp. 119–124. Xue, Y., Wong, J., Moreno, G.T., Young, M.K., Cote, J., and Wang, W. (1998). Mol. Cell 2, 851–861.