C83
Molecular and Cellular Endocrinology, 14 (1990) C83-C86 Elsevier Scientific Publishers Ireland, Ltd.
MOLCEL
02427
At the Cutting
Edge
Estrogen receptor variants in human breast cancer Leigh C. Murphy Department of Biochemistry and Molecular Biology, University of Manitoba, Winnipeg. Manitoba R3E 0 W3, Canada
Key words: Estrogen
receptor
variant;
Breast cancer;
human
Human breast cancers which respond to endocrine therapy are classified as hormone-dependent. This phenotype is characterized by the presence of estrogen (ER) and progesterone receptors (PgR) in the tumor. However, only 70-80s of receptor-positive tumors respond to endocrine therapy, and 20-30% of receptor-positive tumors are resistant. Moreover, of those tumors which originally responded to endocrine therapy a substantial proportion eventually develop resistance. One possible explanation for this resistance is the development of a receptor-negative tumor cell population from an originally heterogenous population. Interestingly, however, many of the tumors which have developed resistance to endocrine therapy remain receptor positive (Rosen et al., 1977; Taylor et al., 1982). What are the mechanisms of this form of resistance? One possibility is the expression of variant or mutant estrogen receptors in resistant breast tumors. Depending on the type of variant/mutant ER, it is possible that the variant receptor may have (a) altered responses to estrogens and antiestrogens,
Address for correspondence: Leigh C. Murphy, Department of Biochemistry and Molecular Biology, University of Manitoba, 770 Bannatyne Ave., Winnipeg, Manitoba R3E 0W3, Canada. Supported by the National Cancer Institute (Canada) and the Medical Research Council (Canada). L.C.M. is an NC1 (Canada) Scientist.
0303-7207/90/$03.50
0 1990 Elsevier Scientific
Publishers
Ireland,
(b) altered
stability
of the ligand-receptor
com-
plex, (c) ligand-independent activity or (d) no activity of its own, but the ability to interfere with normal receptor activity by competing for ligand, DNA binding sequences, transcription factors, etc. Each one of these possible activities, or a combination of them, could be envisaged to lead to antihormone resistance or ligand-independent activity, and thus to an apparently hormone-independent phenotype. However, before any one of these hypotheses can be tested the existence of variant or mutant ER has to be addressed. The presence of abnormal ER proteins in some human breast cancer biopsy samples has been suggested from subcellular distribution and nuclear translocation experiments (MacFarlane et al., 1980; Leake et al., 1981). Moreover, studies using immunohistochemical staining techniques have added support to the presence of ER proteins which display functional abnormalities such as inability to bind to the nucleus when charged with ligand, or receptor proteins which interact with the nucleus in the absence of the ligand (Raam et al., 1988). Although these data are suggestive, they do not prove conclusively that mutant or variant estrogen receptors exist; they may also be explained by extra-receptor abnormalities, of the other factors necessary for normal estrogen receptor function. Such extra-receptor abnormalities may clearly be a possible mechanism for estrogen receptor-positive, antiestrogenresistant/ hormone-independent human breast Ltd
C84
cancer. On the other hand at least four separate groups of investigators, using different recombinant DNA techniques, have now reported variant forms of ER mRNA molecules in human breast cancer biopsy samples. The coding region of the normal estrogen receptor mRNA can be divided into six domains on structural and functional criteria (Kumar et al., 1987). These domains are labelled A, B, C, D, E, F from the 5’- to the 3’-end of the coding region. Garcia et al. (1988) using radiolabelled anti-sense riboprobes corresponding to various regions of the coding region of the normal ER mRNA and a solution hybridization/ RNase protection assay, demonstrated the presence of a small mutation in the B region of ER mRNA in some human breast cancer biopsy samples. The presence of this mutation was significantly correlated with low levels of estrogen binding, even though the A/B region of the ER is not the ligand-binding domain; on the other hand, depending on the estrogen-responsive promoter used, it does show transcriptional activation activity (Kumar et al., 1987). Garcia et al. (1988) hypothesized that this point mutation in the ER mRNA may result in receptors with altered characteristics, which may in turn alter the half-life of the protein, its solubilization properties, or its ability to regulate gene expression. Further studies identified this abnormality as two point mutations, one of which was silent (G to C at nucleotide 261) and the other a C to T transition at nucleotide 257, which changes alanine to valine at amino acid 86 (Garcia et al., 1989). This variant ER was also found in uterine tissue from women without breast cancer and is thus considered a genetic polymorphism rather than a tumor-specific mutation. Interestingly, women who are heterozygous for this B-variant gene have a higher proportion of spontaneous abortions than those who are homozygous for the wild-type ER gene (Lehrer et al., 1990). Although this mutation is not associated with neoplasia, its association with low levels of estrogen binding suggests it might in turn be associated with a low response to hormone therapy and a higher incidence of antiestrogen resistance. A PuuII restriction fragment length polymorphism (RFLP) in 0.6 and 1.6 kb fragments of the human ER gene (Castagnoli et al., 1987) has also
been described. Hill et al. (1989) have shown that this RFLP is probably located in the sequences encoding the DNA-binding (C) or hormone-binding (E) domains of the ER, and absence of the 0.6 kb fragment is frequently associated with absence or low levels of ER expression (Hill et al., 1989). The molecular basis for the observation is unknown; the authors have speculated that perhaps the polymorphism may affect the splicing of the ER primary transcript, though they found no structural polymorphism in the ER mRNA by Northern blotting analysis. In contrast Murphy and Dotzlaw (1989) have demonstrated the presence of altered sized ER mRNA in some breast cancer biopsy samples, though the expression of these variant sized ER mRNAs was found to be unrelated to the PvuII RFLP in the ER gene. Since human breast cancer cell lines were found to be uniformly homozygous for one or other of the 0.6 or 1.6 kb fragments (Hill et al., 1989), the possibility that contaminating normal cells (to a different extent in different biopsy samples) may confound interpretation of Southern blot analysis of biopsy DNA cannot be excluded. Although this RFLP does not appear to be associated with neoplasia, its association with low or no ER expression may again be indirectly correlated with the success of hormone therapy, with prognosis and/or with a higher incidence of antiestrogen resistance. In this laboratory, we have approached the question of mutant/variant ER by looking for the presence of altered sized ER-like mRNA in human breast cancer biopsy samples on Northern blots (Murphy and Dotzlaw, 1989). Using this method, we have found smaller molecular weight (4.0, 3.8 and 2.5 kb) ER-like mRNA molecules in some human breast cancer biopsy samples, which are found only in the presence of normal transcripts. In some tumors, however, the abundance of these smaller transcripts is the same or greater than the normal sized transcript. If these variant ER-like mRNAs are translated into a stable protein, it is possible that they may mimic or interfere with normal ER function. Hybridization studies have shown that the smaller ER-like mRNAs are missing a substantial portion of the 3’-end of the normal ER mRNA, including those sequences encoding the hormone-
CSS
binding domain of the normal ER. Cloning and sequencing of the variant ER-like mRNAs from a human breast cancer biopsy confirmed the identity of these variant ER-like mRNAs, with the normal ER mRNA sequence at the 5’-end and entirely unique 3’-ends. Multiple variants were detected, with the sequence divergence of these variants from the normal receptor often occurring after the two zinc finger motifs. One of the variants contained an open reading frame which would encode a novel protein, identical with the normal ER protein up to and including the two zinc finger motifs. Thereafter the amino acid sequence differs from the normal ER with a sequence unique to the variant molecule (Mushy and Dotzlaw, unpublished). The ER gene in tumors expressing the variant sized ER-like mRNAs does not contain any major deletions or alterations (Murphy and Dotzlaw, 1989) so that it is possible that these variant sized ER-like mRNAs are differently processed forms of the primary ER transcript. Barrett-Lee et al. (1987), using Northern blot analysis, have also detected the presence of an approximately 3.7 kb ER-like transcript which, although present at low levels in MCF-7 human breast cancer cells, was more abundant in some human breast cancer biopsy samples. These authors, however, found no abnormalities of the ER-mRNA size in tumors from a group of 14 patients who were ER positive but did not respond to antiestrogen therapy. Furthermore, Rio et al. (1987) have examined 114 human breast tumors but failed to detect any abnormally sized transcripts. Fuqua et al. (1990) have approached the question of mutant ER in human breast cancer in yet another way. They have taken RNA from ERnegative/PgR-positive tumors, as well as from OR-positive/PgR-positive tumors, and have made single-stranded cDNA by reverse transcription primed with oligonucleotides specific to exons 4 and 6, sequences encoding part of the hormone binding domain of the normal ER mRNA. The DNA was then amplified by polymerase chain reaction and sequenced. They have detected the presence of a variant ER mRNA which is deleted entirely of the exon 5 sequences of the normal ER gene. This transcript was a minor component of ER-positive/PgR-positive tumors but was the pre-
dominant species in the ER-negative/PgR-positive tumors. Interestingly, in a yeast expression system, the protein expressed by this variant ERlike sequence was found to have lo-158 of the ~~u~~-activating activity of the normal ER, although the truns-activation activity was ligand-independent. This group has also detected other variant forms of ER-like mRNAs in human breast cancer biopsy samples (Fuqua and McGuire, personal communication). T-47D,,, a human breast cancer cell line which is resistant to antiestrogens (Horwitz et al., 1982), has been found to have little or no binding of radiolabelled estradiol. However, ER mRNA of the normal size (appr~~mately 6.5 kb) is detectable on Northern blot analysis (Berkenstam et al., 1989). Furthermore, low amounts of an immunologically reactive, normal sized ER (68 kDa) were also detected in these cells. The authors thus speculated that the estrogen and antiestrogen resistance of these cells may be due to a defective ER protein. Grahame et al. (1990) have now reported the isolation of mutant ER cDNAs from the antiestrogen-resistant T-47D,, cells. Some of these cDNAs have been sequenced and the abnormalities found include a frame-soft mutation which would be predicted to encode an ERlike molecule truncated just beyond the second zinc finger of the DNA binding domain of the normal ER; another that would encode an ER-like molecule truncated near the end of exon 5, which encodes part of the hormone binding domain; and an in-frame deletion of the nuclear-localization signal in exon 4 to the end of exon 5. As for the biopsy findings of Murphy and Dotzlaw (1989), there appear to be multiple variant ER mRNAs present in a single human breast cancer cell line. However, in contrast to human breast cancer biopsy samples (Murphy and Dotzlaw, 1989), variant sized mRNA species are not detectable in the T-47D,, cells by Northern blot analysis (Grahame et al., 1990). In summary, apart from the existence of genetic variants of the ER gene which exist normally unassociated with tumorigenesis, the evidence presented in this review strongly supports the presence of variant/mutant forms of ER-like mRNA$ in some human breast cancers. Though the functional capabilities of most of the proteins encoded
C86
by the mutant ER mRNAs are presently unknown, these data should be available in the near future. There is no direct evidence, however, that these mRNAs are translated into stable proteins in vivo. Although one study has found a correlation of a variant form of the ER with a high incidence of spontaneous abortion, the relationship of any of the other variant ER-like mRNAs to antihormone resistance and progression from hormone dependence to hormone independence is presently a matter for speculation. References Barrett-Lee, P.J., ‘D-avers, M.T., McClelland, R.A., Luqmani, Y. and Coombes, R.C. (1987) Cancer Res. 47, 6653-6659. Berkenstam, A.. Glaumann, H., Martin, M., Gustafsson, J.-A. and Norstedt, G. (1989) Mol. Endacrinol. 3, 22-28. Castagnoli, A., Maestri, I. and Del Sanno, L. (1987) Nucleic Acids Res. 15, 866. Fuyua. S.A.W., McDonnell, D.P., Nawaz, Z., O’Malley, B.W. and McGuire, W.L. (1990) Proceedings of the 72nd Annual Meeting of the Endocrine Society, Atlanta, Abstract No. 1317, p. 354. Garcia, T., Lehrer, S., Bloomer, W.D. and Schachter, B. (1988) Mol. Endocrinol. 2, 785-791. Garcia, T., Sanchez, M., Cox, J.L., Shaw, P.A., Ross, J.B.A., Lehrer, S. and Schachter. B. (1989) Nucleic Acids Res. 17, 8364.
Grahame, M.L., Krett, N.L., Miller, L.A., Leslie, K.K., Gorden, D.F., Wood, W.M., Wei, L.L. and Horwitz, K.B. (1990) Cancer Res. (in press). Hill, S.M., Fuqua, S.A.W., Chamness, G.C., Greene, G.L. and McGuire, W.L. (1989) Cancer Res. 49, 145-148. Horwitz, K.B., Mockus, M.B. and Lessey. B.A. (1982) Cell 28, 633-642. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J.R. and Chambon, P. (1987) Cell 51, 941-951. Leake, R.E., Lain& L., Calman, K.C., MacBeth, F.R., Crawford, D. and Smith, D.C. (1981) Br. J. Cancer 43, SQ-66. Lehrer, S., Sanchez, M., Song, H.K., Dalton, J., Levine, E., Savoretti, P., Tbung, S.N. and Schachter, B. (1990) Lancet 335, 622-624. MacFarlane, J.K., Fleiszer, D. and Fazekas, A.G. (1980) Cancer 452998-3003. Murphy, L.C. and Dotzlaw, H. (1989) Mof. Endocrinol. 3, 687-693. Raam, S., Robert, N., Pappas, CA. and Tamura, H. (1988) J. Natl. Cancer Inst. 80, 756-761. Rio, MC, Bellocq, J.P., Gairard, B., Rasmussen, U.B., Krust. A., Koehl, C., Calderoli, H., Schiff, V., Renaud, R. and Chambon, P. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 9243-9241. Rosen, P.P., Memedez-Botet, C.J., Urban, J.A., Fracchia, A. and Schwartz, M.K. (1977) Cancer 39, 2194-2200. Taylor, R.E., Powles, T.J., Humphreys, J., Bettelheim, R.. Dowsett, M., Casey, A.J., Neville, A.M. and Coombes, R.C. (1982) Br. J. Cancer 45, 80-85.
Molecular and Cellular Endocrinology, 14 (1990) C83-C86 Elsevier Scientific Publishers Ireland, Ltd.
MOLCEL
02427
At the Cutting
Edge
Estrogen receptor variants in human breast cancer Leigh C. Murphy Department of Biochemistry and Molecular Biology, University of Manitoba, Winnipeg. Manitoba R3E 0 W3, Canada
Key words: Estrogen
receptor
variant;
Breast cancer;
human
Human breast cancers which respond to endocrine therapy are classified as hormone-dependent. This phenotype is characterized by the presence of estrogen (ER) and progesterone receptors (PgR) in the tumor. However, only 70-80s of receptor-positive tumors respond to endocrine therapy, and 20-30% of receptor-positive tumors are resistant. Moreover, of those tumors which originally responded to endocrine therapy a substantial proportion eventually develop resistance. One possible explanation for this resistance is the development of a receptor-negative tumor cell population from an originally heterogenous population. Interestingly, however, many of the tumors which have developed resistance to endocrine therapy remain receptor positive (Rosen et al., 1977; Taylor et al., 1982). What are the mechanisms of this form of resistance? One possibility is the expression of variant or mutant estrogen receptors in resistant breast tumors. Depending on the type of variant/mutant ER, it is possible that the variant receptor may have (a) altered responses to estrogens and antiestrogens,
Address for correspondence: Leigh C. Murphy, Department of Biochemistry and Molecular Biology, University of Manitoba, 770 Bannatyne Ave., Winnipeg, Manitoba R3E 0W3, Canada. Supported by the National Cancer Institute (Canada) and the Medical Research Council (Canada). L.C.M. is an NC1 (Canada) Scientist.
0303-7207/90/$03.50
0 1990 Elsevier Scientific
Publishers
Ireland,
(b) altered
stability
of the ligand-receptor
com-
plex, (c) ligand-independent activity or (d) no activity of its own, but the ability to interfere with normal receptor activity by competing for ligand, DNA binding sequences, transcription factors, etc. Each one of these possible activities, or a combination of them, could be envisaged to lead to antihormone resistance or ligand-independent activity, and thus to an apparently hormone-independent phenotype. However, before any one of these hypotheses can be tested the existence of variant or mutant ER has to be addressed. The presence of abnormal ER proteins in some human breast cancer biopsy samples has been suggested from subcellular distribution and nuclear translocation experiments (MacFarlane et al., 1980; Leake et al., 1981). Moreover, studies using immunohistochemical staining techniques have added support to the presence of ER proteins which display functional abnormalities such as inability to bind to the nucleus when charged with ligand, or receptor proteins which interact with the nucleus in the absence of the ligand (Raam et al., 1988). Although these data are suggestive, they do not prove conclusively that mutant or variant estrogen receptors exist; they may also be explained by extra-receptor abnormalities, of the other factors necessary for normal estrogen receptor function. Such extra-receptor abnormalities may clearly be a possible mechanism for estrogen receptor-positive, antiestrogenresistant/ hormone-independent human breast Ltd
C84
cancer. On the other hand at least four separate groups of investigators, using different recombinant DNA techniques, have now reported variant forms of ER mRNA molecules in human breast cancer biopsy samples. The coding region of the normal estrogen receptor mRNA can be divided into six domains on structural and functional criteria (Kumar et al., 1987). These domains are labelled A, B, C, D, E, F from the 5’- to the 3’-end of the coding region. Garcia et al. (1988) using radiolabelled anti-sense riboprobes corresponding to various regions of the coding region of the normal ER mRNA and a solution hybridization/ RNase protection assay, demonstrated the presence of a small mutation in the B region of ER mRNA in some human breast cancer biopsy samples. The presence of this mutation was significantly correlated with low levels of estrogen binding, even though the A/B region of the ER is not the ligand-binding domain; on the other hand, depending on the estrogen-responsive promoter used, it does show transcriptional activation activity (Kumar et al., 1987). Garcia et al. (1988) hypothesized that this point mutation in the ER mRNA may result in receptors with altered characteristics, which may in turn alter the half-life of the protein, its solubilization properties, or its ability to regulate gene expression. Further studies identified this abnormality as two point mutations, one of which was silent (G to C at nucleotide 261) and the other a C to T transition at nucleotide 257, which changes alanine to valine at amino acid 86 (Garcia et al., 1989). This variant ER was also found in uterine tissue from women without breast cancer and is thus considered a genetic polymorphism rather than a tumor-specific mutation. Interestingly, women who are heterozygous for this B-variant gene have a higher proportion of spontaneous abortions than those who are homozygous for the wild-type ER gene (Lehrer et al., 1990). Although this mutation is not associated with neoplasia, its association with low levels of estrogen binding suggests it might in turn be associated with a low response to hormone therapy and a higher incidence of antiestrogen resistance. A PuuII restriction fragment length polymorphism (RFLP) in 0.6 and 1.6 kb fragments of the human ER gene (Castagnoli et al., 1987) has also
been described. Hill et al. (1989) have shown that this RFLP is probably located in the sequences encoding the DNA-binding (C) or hormone-binding (E) domains of the ER, and absence of the 0.6 kb fragment is frequently associated with absence or low levels of ER expression (Hill et al., 1989). The molecular basis for the observation is unknown; the authors have speculated that perhaps the polymorphism may affect the splicing of the ER primary transcript, though they found no structural polymorphism in the ER mRNA by Northern blotting analysis. In contrast Murphy and Dotzlaw (1989) have demonstrated the presence of altered sized ER mRNA in some breast cancer biopsy samples, though the expression of these variant sized ER mRNAs was found to be unrelated to the PvuII RFLP in the ER gene. Since human breast cancer cell lines were found to be uniformly homozygous for one or other of the 0.6 or 1.6 kb fragments (Hill et al., 1989), the possibility that contaminating normal cells (to a different extent in different biopsy samples) may confound interpretation of Southern blot analysis of biopsy DNA cannot be excluded. Although this RFLP does not appear to be associated with neoplasia, its association with low or no ER expression may again be indirectly correlated with the success of hormone therapy, with prognosis and/or with a higher incidence of antiestrogen resistance. In this laboratory, we have approached the question of mutant/variant ER by looking for the presence of altered sized ER-like mRNA in human breast cancer biopsy samples on Northern blots (Murphy and Dotzlaw, 1989). Using this method, we have found smaller molecular weight (4.0, 3.8 and 2.5 kb) ER-like mRNA molecules in some human breast cancer biopsy samples, which are found only in the presence of normal transcripts. In some tumors, however, the abundance of these smaller transcripts is the same or greater than the normal sized transcript. If these variant ER-like mRNAs are translated into a stable protein, it is possible that they may mimic or interfere with normal ER function. Hybridization studies have shown that the smaller ER-like mRNAs are missing a substantial portion of the 3’-end of the normal ER mRNA, including those sequences encoding the hormone-
CSS
binding domain of the normal ER. Cloning and sequencing of the variant ER-like mRNAs from a human breast cancer biopsy confirmed the identity of these variant ER-like mRNAs, with the normal ER mRNA sequence at the 5’-end and entirely unique 3’-ends. Multiple variants were detected, with the sequence divergence of these variants from the normal receptor often occurring after the two zinc finger motifs. One of the variants contained an open reading frame which would encode a novel protein, identical with the normal ER protein up to and including the two zinc finger motifs. Thereafter the amino acid sequence differs from the normal ER with a sequence unique to the variant molecule (Mushy and Dotzlaw, unpublished). The ER gene in tumors expressing the variant sized ER-like mRNAs does not contain any major deletions or alterations (Murphy and Dotzlaw, 1989) so that it is possible that these variant sized ER-like mRNAs are differently processed forms of the primary ER transcript. Barrett-Lee et al. (1987), using Northern blot analysis, have also detected the presence of an approximately 3.7 kb ER-like transcript which, although present at low levels in MCF-7 human breast cancer cells, was more abundant in some human breast cancer biopsy samples. These authors, however, found no abnormalities of the ER-mRNA size in tumors from a group of 14 patients who were ER positive but did not respond to antiestrogen therapy. Furthermore, Rio et al. (1987) have examined 114 human breast tumors but failed to detect any abnormally sized transcripts. Fuqua et al. (1990) have approached the question of mutant ER in human breast cancer in yet another way. They have taken RNA from ERnegative/PgR-positive tumors, as well as from OR-positive/PgR-positive tumors, and have made single-stranded cDNA by reverse transcription primed with oligonucleotides specific to exons 4 and 6, sequences encoding part of the hormone binding domain of the normal ER mRNA. The DNA was then amplified by polymerase chain reaction and sequenced. They have detected the presence of a variant ER mRNA which is deleted entirely of the exon 5 sequences of the normal ER gene. This transcript was a minor component of ER-positive/PgR-positive tumors but was the pre-
dominant species in the ER-negative/PgR-positive tumors. Interestingly, in a yeast expression system, the protein expressed by this variant ERlike sequence was found to have lo-158 of the ~~u~~-activating activity of the normal ER, although the truns-activation activity was ligand-independent. This group has also detected other variant forms of ER-like mRNAs in human breast cancer biopsy samples (Fuqua and McGuire, personal communication). T-47D,,, a human breast cancer cell line which is resistant to antiestrogens (Horwitz et al., 1982), has been found to have little or no binding of radiolabelled estradiol. However, ER mRNA of the normal size (appr~~mately 6.5 kb) is detectable on Northern blot analysis (Berkenstam et al., 1989). Furthermore, low amounts of an immunologically reactive, normal sized ER (68 kDa) were also detected in these cells. The authors thus speculated that the estrogen and antiestrogen resistance of these cells may be due to a defective ER protein. Grahame et al. (1990) have now reported the isolation of mutant ER cDNAs from the antiestrogen-resistant T-47D,, cells. Some of these cDNAs have been sequenced and the abnormalities found include a frame-soft mutation which would be predicted to encode an ERlike molecule truncated just beyond the second zinc finger of the DNA binding domain of the normal ER; another that would encode an ER-like molecule truncated near the end of exon 5, which encodes part of the hormone binding domain; and an in-frame deletion of the nuclear-localization signal in exon 4 to the end of exon 5. As for the biopsy findings of Murphy and Dotzlaw (1989), there appear to be multiple variant ER mRNAs present in a single human breast cancer cell line. However, in contrast to human breast cancer biopsy samples (Murphy and Dotzlaw, 1989), variant sized mRNA species are not detectable in the T-47D,, cells by Northern blot analysis (Grahame et al., 1990). In summary, apart from the existence of genetic variants of the ER gene which exist normally unassociated with tumorigenesis, the evidence presented in this review strongly supports the presence of variant/mutant forms of ER-like mRNA$ in some human breast cancers. Though the functional capabilities of most of the proteins encoded
C86
by the mutant ER mRNAs are presently unknown, these data should be available in the near future. There is no direct evidence, however, that these mRNAs are translated into stable proteins in vivo. Although one study has found a correlation of a variant form of the ER with a high incidence of spontaneous abortion, the relationship of any of the other variant ER-like mRNAs to antihormone resistance and progression from hormone dependence to hormone independence is presently a matter for speculation. References Barrett-Lee, P.J., ‘D-avers, M.T., McClelland, R.A., Luqmani, Y. and Coombes, R.C. (1987) Cancer Res. 47, 6653-6659. Berkenstam, A.. Glaumann, H., Martin, M., Gustafsson, J.-A. and Norstedt, G. (1989) Mol. Endacrinol. 3, 22-28. Castagnoli, A., Maestri, I. and Del Sanno, L. (1987) Nucleic Acids Res. 15, 866. Fuyua. S.A.W., McDonnell, D.P., Nawaz, Z., O’Malley, B.W. and McGuire, W.L. (1990) Proceedings of the 72nd Annual Meeting of the Endocrine Society, Atlanta, Abstract No. 1317, p. 354. Garcia, T., Lehrer, S., Bloomer, W.D. and Schachter, B. (1988) Mol. Endocrinol. 2, 785-791. Garcia, T., Sanchez, M., Cox, J.L., Shaw, P.A., Ross, J.B.A., Lehrer, S. and Schachter. B. (1989) Nucleic Acids Res. 17, 8364.
Grahame, M.L., Krett, N.L., Miller, L.A., Leslie, K.K., Gorden, D.F., Wood, W.M., Wei, L.L. and Horwitz, K.B. (1990) Cancer Res. (in press). Hill, S.M., Fuqua, S.A.W., Chamness, G.C., Greene, G.L. and McGuire, W.L. (1989) Cancer Res. 49, 145-148. Horwitz, K.B., Mockus, M.B. and Lessey. B.A. (1982) Cell 28, 633-642. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J.R. and Chambon, P. (1987) Cell 51, 941-951. Leake, R.E., Lain& L., Calman, K.C., MacBeth, F.R., Crawford, D. and Smith, D.C. (1981) Br. J. Cancer 43, SQ-66. Lehrer, S., Sanchez, M., Song, H.K., Dalton, J., Levine, E., Savoretti, P., Tbung, S.N. and Schachter, B. (1990) Lancet 335, 622-624. MacFarlane, J.K., Fleiszer, D. and Fazekas, A.G. (1980) Cancer 452998-3003. Murphy, L.C. and Dotzlaw, H. (1989) Mof. Endocrinol. 3, 687-693. Raam, S., Robert, N., Pappas, CA. and Tamura, H. (1988) J. Natl. Cancer Inst. 80, 756-761. Rio, MC, Bellocq, J.P., Gairard, B., Rasmussen, U.B., Krust. A., Koehl, C., Calderoli, H., Schiff, V., Renaud, R. and Chambon, P. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 9243-9241. Rosen, P.P., Memedez-Botet, C.J., Urban, J.A., Fracchia, A. and Schwartz, M.K. (1977) Cancer 39, 2194-2200. Taylor, R.E., Powles, T.J., Humphreys, J., Bettelheim, R.. Dowsett, M., Casey, A.J., Neville, A.M. and Coombes, R.C. (1982) Br. J. Cancer 45, 80-85.
