Kathryn Horwitz - US grants

There is increasing clinical interest in progestins for two reasons: one, the use of progesterone receptor (PR) measurements to mark hormone-dependent breast, endometrial, ovarian and prostatic cancers; and two, the use of progestins and antiprogestins for contraception and for endocrine therapies of cancer. We have developed a stoichiometric nuclear exchange assay for PR and have used it to show have progesterone can regulate the levels of its own receptors freed of the interfering effects of estrogens. We have also shown that a widely used synthetic experimental progestin, R5020, has chronic suppressive effects on PR levels. This, if extrapolated to the clinical setting, suggests that PR in patients taking synthetic progestins may be incorrectly assigned. These studies have been possible because of the availability of permanent human breast cancer cell lines, particularly one called T47D, that synthesizes enormous levels of PR without requiring estrogen induction. Our aim in this application is to study in detail, the biology, metabolism and receptor mechanisms of progesterone, synthetic progestins, and a new contraceptive antiprogestin RU38 486. We plan to use T47D cells (estrogen independent PR) MCF-7 cells (estrogen-dependent PR) and BT-20 cells (PR negative): 1) To develop three biological responses to mark progestin action. These are inhibition of cell growth, induction of insulin receptors, and synthesis of secreted proteins. 2) We will use gas chromatography and mass spectrometry to study cell mediated progestin metabolism and identify the ultimate receptor-bound compound responsible for biological activity. 3) We plan to study the molecular biology of PR and their interaction with progestins: the regulation of PR levels by DNA methylation, histone acetylation and hormone resistance; holoreceptor and nuclear receptor structures by chromatography and electrophoresis; PR activation, translocation nuclear matrix PR binding sites by immunofluorescence; endo and exonuclease activity. 4) We propose to study the mechanisms of antiprogestin action using labeled and unlabeled RU38 486.

Progesterone is a steroid hormone of fundamental importance in reproduction. To generate probes for studying the mechanisms of progesterone action the progesterone receptor has been purified from T47D, a human cell line which provides a particularly rich source of the receptor. Three IgG monoclonal antibodies have been raised against the progesterone receptor proteins. Though only purified B- receptors (120,000 daltons) were the antigens, 1 of the 3 antibodies cross-reacts with A-receptors (94,000 daltons). In the continuation of this research immunochemistry, photoaffinity labeling and autoradiography will be applied to study how phosphorylation of the progesterone receptor proteins affects their action. The first goal is, using whole cells, to assess the effect of progestins, and intracellular A- and B-receptor localization on the phosphorylation state of the receptors, to analyze phospho-amino acids and peptides, and to study phosphorylation of receptor-associated proteins that do not bind hormone. The second goal is to use reconstituted, cell-free systems to characterize receptor kinases, particularly a putative nuclear kinase, and to assess receptor autophosphorylation capacity. The third goal is to determine whether the phosphorylation state of the receptor proteins varies with relevant physiological stimuli, including known modualtors of progesterone receptors such as estradiol. In sum, the overall goal is to understand how this covalent modification of the progesterone receptor proteins affects their ability to bind hormone, to acquire tight nuclear binding capacity, and to be down-regulated in nuclei. The results of this research should further our understanding of the mechanism of action of this important steroid hormone.

Induction of breast cancers depends on ovarian steroids. However, at diagnosis, 2/3 of tumors have switched, from growth control by steroid agonists, to control by peptide growth factors. Hypothesis: progesterone primes breast cancers for the proliferative effects of growth factors, which then down-regulate progesterone receptors (PR) creating a negative feedback loop. The remaining 1/3 of steroid hormone-dependent tumors respond well to antagonists but acquire resistance as they progress. Hypothesis: acquired resistance to antagonists is associated with inappropriate expression of their agonist effects. Aim 1. Resistance to Agonists. Is steroid resistance biochemically linked to upregulation of growth factor signaling? Using stable cell lines expressing A- or B-isoforms of PR, or mutant PR, we will study priming effects of progesterone and the antiprogestin RU486: on acquired sensitivity to proliferative effects of EGF; on upregulation of MAPK and STAT signaling; on MAPK phosphorylation of PR which targets them for ubiquitylation and downregulation; on transcriptional synergism at p21 and c-fos promoters. These studies will define mechanisms for cross-talk between PR and EGF in the cytoplasm and nucleus. Aim 2. Acquired Resistance to Anatagonists. Are the agonist effects of mixed antagonists upregulated by recruitment of regulatory proteins to antagonist-occupied PR? Three novel proteins have been isolated in an antagonist-biased yeast two hybrid screen. ORF number 93 has multiple NR boxes and at least 3 TPR domains. We will clone it, and ask whether it acts as a scaffold to assemble PR in a multi-protein complex with EGF signaling molecules; is a coregulator; or assembles PR into a repressive transcription complex. The mechanisms of two other novel proteins will be assessed. These studies will describe the function of antagonist-occupied receptors as governed by three new receptor- interacting proteins. Aim 3. Acquired Resistance to Antagonists: Breast Cancers. Does the transcriptional coactivator to corepressor ratio determine outcome to tamoxifen? Tumors from tamoxifen responsive or resistant patients will be measured for transcript expression of the coactivators L7/SPA and SRC-1, and the corepressors N-CoR and SMRT, in relation to treatment response. We will show that mechanisms of SMRT repression are novel. Other studies address EGF signaling molecules and ORF number 93. These studies will document the relationships between coregulators and tamoxifen resistance, and demonstrate that antagonist-occupied steroid receptors recruit factors inappropriately. Overall, these studies will explain hormone-resistance mechanisms in breast cancers.

Steroid receptors belong to a superfamily of proteins that, in addition to the classic steroid hormones, also bind vitamin D3, retinoic acid and thyroid hormone. When occupied by hormone, receptors bind to their cognate DNA response elements and act as enhancers to regulate gene transcription. All steroid receptors studied to date are phosphoproteins but the function of phosphorylation is unknown. We have purified human progesterone receptors (PR) and have made monoclonal antibodies to them. There are two human PRs: B-receptors of 120 kDa and A-receptors of 94 kDa. By immunoaffinity purification, gel electrophoresis, immunoblotting, autoradiography, chemical cleavage, phosphotryptic peptide mapping, and reverse-phase high pressure liquid chromatography, we have shown that in cultured breast cancer cells, PRs are phosphorylated at multiple sites when unoccupied by hormone. This represents basal phosphorylation. A second, hormone-dependent phosphorylation occurs within 5 min of progestin treatment of cells, and increases the specific activity of [32P]orthophosphate labeling 5-10 fold. PR phosphorylation is on serine residues located predominately in the amino-terminal half of the receptors. In this application we plan to precisely map the serine residues involved in PR phosphorylation and determine the function of phosphorylation: Aim 1 is to sequence the amino acids of enzymatic phosphopeptide fragments eluted from HPLC and place the peptides in context on the full-length human PR protein. Aim 2 is to generate a series of cDNA mutants of the A- and B- receptors. First, to delete three phosphorylated clusters in the amino- terminus by mutagenesis of full-length hPR cDNAs encoding B- and A- receptors, or to delete the three major functional domains and test the mutant receptors for basal and hormone-dependent phosphorylation. Second, to use oligonucleotide-directed site specific mutagenesis to conservatively mutate key serine residues to alanine, and test the mutant receptors for phosphorylation and ability to trans-activate progestin responsive reporter genes. Third, to test the mutant receptors on simple and complex promoters to evaluate the function of phosphorylation. A number of iterative cycles of mutagenesis should converge on the residues whose mutation abrogates or modifies receptor function. In Aim 3, DNA binding receptor mutants will be constructed to evaluate the DNA binding requirement for PR phosphorylation. These studies will delineate the function of PR phosphorylation, define the specific roles of A- and B-receptors in transactivation, and serve as a model for the role of phosphorylation of other steroid receptors and transcription factors.

DESCRIPTION (adapted from the applicant's abstract): The proposed studies address molecular mechanisms that may explain tissue-specific and gene- specific regulation by progestins. Aim 1 is to assess the functional differences between A- and B-receptors in breast and endometrial cells. Aim 2 is to investigate BUS (the B-upstream segment) as a third transcriptional activation domain unique to B-receptors. Finally, Aim 3 is to determine the mechanisms of TAF-3 action in the BUS segment.

DESCRIPTION: (Adapted from the applicant's abstract) There are two progesterone receptors (PR): 933 aa B-receptors, and 769 aa A-receptors with a 164 aa N-terminal truncation. The B:A ratio differs among tissues; transcription by B vs. A differs in ligand-, promoter- and cell-specific ways; B are transactivators when A are inhibitors. The PR N-termini containing the homologous DNA binding domain (DBD) but lacking the hormone binding domain (HBD), are independent functional domains. They are constitutive transactivators that recapitulate the functional differences of the full-length receptors. Hypothesis: structural differences at their N-termini explain functional differences of the two PR. Aim 1. Structure of A- (NT-A) and B-receptor (NT-B) N-termini. The functionally autonomous domains, NT-A and NT-B, plus full-length A- and B- receptors, will be purified to homogeneity. Biochemical function will be documented by in vitro transcription, DNA binding affinity and cooperativity will be quantified; self-association properties will be analyzed by sedimentation equilibrium; size and shape defined by sedimentation velocity. The structure of purified N-termini will be compared by limited proteolysis and spectroscopic assays. Conformational changes due to DNA binding and the HBD will be mapped. These studies will define structural differences at the N-termini of the two PR. AIM 2. Isolation and characterization of proteins that mediate the unique functional properties of the two PR N-termini. We postulate that structural differences at the N-termini lead to recruitment of unique transcriptional co-regulators to each PR isoform. NT-A and NT-B will be used as "bait" in yeast one-and two- hybrid assays and in conventional chromatographic protein isolation and sequencing that interact with both PR N-termini, or ones that interact only with one, and confer unique properties to each PR isoform. AIM 3. Mutational and transcriptional analyses of structural subdomains in NT-A vs. NT-B. Preliminary data show that NT-A is organized into 7 protease inaccessible cassettes separated by 6 solvent-exposed loops and that its activation function is a stable subunit. These domains, and newly defined structural and functional domains in NT-B, will be mutated to assess their role in constitutive transcription by NT-A vs. NT-B, and ligand-regulated transcription by full-length PR. The proposed studies address the presently unknown functions of the two PR N-termini.

DESCRIPTION (provided by applicant): Breast cancer is a disease of women due to the actions of estrogens (E) and progesterone (P). While oophorectomy is protective, hormone replacement therapy with E+P increases breast cancer risk compared to E-only. This implicates P, either alone or with E, in tumor promotion. Though many studies address the actions of E or P alone, we contend that one must understand how these two hormones work together. E and P activate estrogen (ER) or progesterone (PR) receptors. Two PR isoforms, PR-A and PR-B, are equal in the normal breast but dysregulated in cancers. Tamoxifen treated patients whose tumors are PR-A rich relapse faster than patients with PR-B rich tumors. Importantly, PR-A and PR-B differentially influence E-dependent tumor growth in either the presence or absence of P. We postulate that PR-A and PR-B in the absence or presence of P, differentially influence effects of E and ER on tumors and that PR-A are especially harmful. AIM 1. Mechanisms of ligand dependent (LD) and ligand independent (LI) transcription by PR.A vs. PR-B. In the presence of P, PR-A and PR-B exert different effects on gene transcription with PR-B more powerful. In the absence of P, PR-A and PR-B are also different with PR-A more powerful. We also show that in addition to Progesterone Response Elements (PRE), LD and LI PR-regulated promoters contain co-Response elements (coRE). This aim dissects the role of PRE and coRE in differential gene regulation by PR-A and PR-B, and addresses mechanisms of LI gene regulation. AIM 2. PR isoforms and in vivo LI and LD effects on E regulated tumor biology and therapeutics. Our preliminary data, using ER+, E-dependent xenografts in ovariectomized non-P supplemented mice, show that presence of PR-A suppresses tumor growth. This aim asks: Why are PR-A+ tumors smaller than PR-B+ ones? How do PR influence tumor responses to antiestrogens? Why do PR block apoptosis by Taxanes? AIM 3. The role of ER and PR and their hormones in breast cancer cell metastasis and growth. Metastasis of ER+, PR+ breast cancers is the major killer of women yet conventional "wisdom" holds that such tumors do not metastasize. Our models show metastasis of ER+, PR+ tumors. This aim develops new models of E-dependent ER+, PR+ metastatic breast cancer and defines the role of steroid receptors and hormones in this process. In sum, we will develop new models to test the hypothesis that PR, even in the absence of P, modify ER+ tumor cell behavior, and that PR-A increase tumor aggressiveness and are harmful.

DESCRIPTION (provided by applicant): This renewal continues research on estrogens (E) and progestins (P) in luminal, estrogen (ER) and progesterone (PR) receptor-positive human breast cancer. In the preceding cycle 27 papers were published. Relevant to this renewal, we: 1. Developed fluorescent mouse models of ER+PR+ tumors that metastasize to lymph nodes (LN). Compared to primary tumors, LN metastases are E resistant. 2. Demonstrated that ER+PR+ tumors contain a minor ER-PR- subpopulation that expresses cytokeratin 5 (CK5). These rare cells may be tumor-initiating and are expanded by P. 3. Initial ER+PR+ models of distant metastasis show hormonal influences on organ-specific engraftment. Metastatic breast cancer kills more than 40,000 American women each year and two-thirds of these tumors retain ER or PR. Despite clinical evidence that ER+PR+ tumors metastasize, the role of women's steroid hormones or their receptors on metastases is unknown, due to lack of models. Hypotheses: 1. E and P play critical roles on LN and distant metastasis of ER+PR+ disease. 2. In ER+PR+ disease, an ER-PR-CK5+ subpopulation with tumor-initiator properties is expanded by P and influences metastasis, dormancy and drug resistance. 3. Luminal ER+PR+ breast cancers exhibit receptor plasticity characterized by receptor loss, in a process driven by P. AIM 1. Metastasis Models of ER+PR+, hormone dependent breast cancer and role of estrogens. To develop models of ER+PR+ metastasis starting from solid tumors or circulating tumor cells, and test the hypothesis that E and P play a role in metastatic engraftment of ER+PR+ disease. AIM 2. Tumor-initiating cells in ER+PR+ breast cancer, tumor dormancy and drug resistance. To test the hypothesis that ER+PR+CK5- breast cancers harbor rare pre-existing ER-PR-CK5+ cells with tumor-initiating properties. ER-PR-CK5+ cells are expanded by P. We engineer models to study constitutive and P-regulated live ER-PR-CK5+ tumor cells and their role in recurrent disease. AIM 3. Towards a new biology for progesterone in luminal breast cancer. To explore a novel view of P in ER+PR+ breast cancers focused on P regulation of tumor-cell phenotype. We test the hypothesis that luminal breast cancers exposed t exhibit receptor plasticity associated with receptor loss. We study cell- biological and molecular mechanisms of this plasticity, develop methods to distinguish among putative ER-PR- cell subtypes, determine whether or not they represent a continuum of the same cells, and study CK5 regulation by P. In sum, the majority of breast cancer metastases are ER+PR+, so E and P must play critical roles in this process. E is the proliferative hormone. P is not. Rather, P partially extinguishes receptor expression. Receptor loss in a subset of ER+PR+ tumor cells is dangerous because these cells acquire tumor-initiating properties that secondarily promote tumor expansion. This has consequences on metastasis and disease recurrence. PUBLIC HEALTH RELEVANCE: The majority of breast cancers are luminal estrogen (ER) and progesterone (PR) receptor-positive, so estrogens (E) and progestins (P) must play critical roles in this disease subtype. The proliferative role of E is established. The role of P is unclear. We suggest that P influence the phenotype of luminal cancers and target the plasticity of this disease by partially extinguishing receptor expression. Receptor loss in a subset of cells is dangerous because these cells have tumor- initiating properties that secondarily promote tumor expansion and drug resistance. This has consequences for metastasis and tumor recurrence.