Day 1 | Day 2 | Day 3

Day 1 of the 2009 Jensen Symposium on Nuclear Receptors featured opening remarks from the co-organizer, Dr. Sohaib Khan, followed by the keynote presentations (Bert O'Malley and Myles Brown) and talks by Craig Jordan, Donald McDonnell and Geoff Greene.

Opening Remarks

Sohaib Khan opened the symposium with the following remarks:

"I am delighted in welcoming this gathering to Cincinnati, which includes the founding members of the nuclear receptor field. From its inception in the late 1950s, when Elwood Jensen identified the estrogen-binding protein estrophilin (or ER), this field has evolved through a series of stunningly impressive scientific discoveries to a stage where it has morphed into a superfamily and assumed a leadership role in defining complex cellular functions, especially transcription.

Of particular mention is the discovery of coactivators, about which we will hear in the Keynote session. Needless to say the induction of coactivators and corepressors in the transcriptional machinery has tremendously expanded our understanding of this complex process. Moreover, a large number of research findings have had multitude of translational uses ranging from antiestrogen therapy for breast cancer to a possible chemical-based exercise program for the couch potato.

It's little wonder the drug companies are spending many billions of dollars to develop medicines for cancer and several metabolic disorders that involve nuclear receptors. The achievements of the nuclear receptor field have been recognized with Lasker Awards, the Presidential Medal of Honor and elections to the National Academy of Sciences. However, the scientific discoveries of these magnitudes certainly deserve recognition in Stockholm, which I strongly believe will come to pass. I am almost certain that this will happen and I predict that three amongst this gathering will share the ultimate honor.

I am happy to report that more than 300 delegates are attending this symposium and for those not able to attend, the NURSA website will publish a summary of the meeting. I am sure that we will hear fabulous science during this symposium for which I am so grateful to all the speakers and poster presenters. And with that, let us move on to the keynote session."

Talk summaries

The SRC/p160 family of coregulators is the prototypical coregulator family and includes SRC-1, SRC-2 and SRC-3. Bert O'Malley (Baylor College of Medicine) opened the meeting with a discussion of the roles of members of this family in cancer, with a focus on SRC-3, a powerful growth regulator and oncogene, and a coregulator that O’Malley casts in the role of a multitasking "master molecule" within the cell. Some of its activities, such as those in transcription, are historically well characterized, and he went on to review the work his lab has done in demonstrating additional roles in splicing (with CAPER), regulation of mRNA translation and enhancing mitochondrial energy production.

In his talk, O'Malley elegantly described a novel function of SRC-3 in motility that was attributable to a novel isoform of the protein. [Editor's note: A Drosophila protein related to SRC-3, taiman, has been previously implicated in invasive cell behavior]. He pointed out that up to 10-15% of cellular SRC-3 exists in the outer region of cells, and proceeded to show work characterizing a 136kDa splice variant of SRC-3 lacking exon 4 (SRC-3Δ4) that lacks a nuclear localization signal, and that predominates in the outer region of the cell. Using studies based on siRNA-specific knockdown of this isoform, he showed that it promotes epidermal growth factor (EGF)-mediated cancer cell migration in a focal adhesion kinase (FAK)-dependent manner. EGF, through EGF-receptor (EGFR) exerts a strong growth and motility stimulus on cells, and p21-associated kinase-1 (PAK1) is known to be associated with EGFR function. O'Malley showed that PAK1 phosphorylation of the cytoplasmic SRC-3Δ4 isoform increases its affinity for EGFR and FAK, thereby potentiating signaling along this axis, and ultimately cellular migration. In summary, the effects of SRC-3 on the cancer cell can be partitioned along isoform-specific lines, such that full length SRC-3 potentiates proliferation via nuclear, transcriptional effects, and the Δ4 isoform potentiates migration through cytoplasmic effects. O'Malley stressed the many different reports of SRC-3 post-translational modification - at least 45 reported to date - giving a high combinatorial potential for evolution of different functions with a single protein.

Myles Brown’s (Dana Farber Cancer Institute) first story focused on exploring the EGF-estrogen receptor-α (ERα)-17β-estradiol (E2) cistrome in breast cancer cells. It is known that ERα can be activated either in an E2-dependent manner, or hormone-independently through growth factor pathways leading to phosphorylation of the receptor. Looking first on the microarray level, Brown compared genes regulated by EGF with those regulated by E2 in MCF-7 cells, and showed the existence of quite distinct sets of genes, with a small subset of EGF-regulated genes, such as LIF, that were also Fulvestrant-sensitive, indicating a co-dependence on ERα and EGF. Extending this to breast tumors, he showed that the E2-induced ERα cistrome defines the luminal breast cancer subtype, and that ERα and EGF-regulated genes are upregulated both in MCF-7 cells and in ERBB2-amplified tumors, emphasizing their clinical importance. Extending these observations in microarrays to the cistrome, he showed that EGF pathway activation increases the repertoire of ERα binding sites. Proceeding to motif analysis, he asked whether motifs found at these binding sites were different across the genome depending on whether ER was brought to DNA by E2 or by EGF. He showed differential distribution of E2 response elements (EREs), forkhead motifs and AP-1 sites, depending on the initial stimulus. The presence of AP-1 sites and the absence of EREs were associated more with EGF stimulation, confirming previous observations, whereas forkhead motifs were preferentially associated with E2 stimulation. Again, relating this to breast cancer, he showed that the EGF-dependent ER cistrome identifies genes overexpressed in ERBB2-amplified breast tumors. 

Brown next discussed the development of a predictive model for defining transcription factor binding sites, with a specific emphasis on the androgen receptor (AR) in prostate cancer cells. He showed an interesting pattern of nucleosome positioning around enhancer elements: after four hours of treatment with dihydrotestosterone (DHT), compared to the absence of ligand, he demonstrated the dismissal of a single nucleosome containing the AR binding site, and the appearance of two sharply and well-defined nucleosomes just upstream and downstream of the AR binding site. Brown went on to describe how his group monitored the appearance of paired positioned nucleosomes at 0h and 4h and 16h of androgen treatment, and reported that AR binding sites were highly enriched after 4h of treatment relative to 0h. In contrast, when 16h androgen treatment was compared to 4h, motifs representing other transcription factors were enriched, emphasizing the relative importance of secondary factors such NKX3.1 in regulating genes at this time point.

Craig Jordan (Georgetown University) described an intriguing novel biology of E2 action that has emerged from cellular models developed and characterized by his group over the past 15 years that make elegant connections between cellular biology and the clinic. They provide a context in which to shed light on observations in the clinic that contradict the paradigm of E2 being a positive, survival stimulus for the growth of breast cancer. Jordan described how Sir Alexander Haddow made seminal observations in the clinic over 60 years ago that led to high dose of synthetic estrogens becoming the standard of care in post-menopausal women with breast cancer. With the development of antiestrogens, treatment with Tamoxifen gradually became the norm in treatment of breast cancer, but over the years it became apparent that the consequence of long term antihormonal therapy was the acquisition of drug resistance. Jordan sketched a multiphase model for hormone resistance in ER positive tumors: in Phase I resistance, E2 and SERM treatment stimulate tumor growth. This is followed after about five years by a second phase, in which E2 inhibits breast cancer growth.

Jordan posed the question: could these clinical observations be recapitulated in models in the laboratory, and could these models be used to tease out a molecular basis for the cell-killing functions of E2? He described how his group originally studied the development of antiestrogen resistance in vitro and established an antiestrogen resistant variant of MCF-7 cells (MCF-7:5C) after long term culture in E2-free medium. When these MCF-7:5C cells were exposed to physiological levels of E2, they died after becoming apoptotic, both in culture and when transplanted into athymic animals. Jordan showed that the molecular events that triggered the E2-induced apoptosis in MCF-7:5C cells involved induction of caspases, specifically CASP4 as an initial event. The group went on to test whether apoptosis in MCF-7:5C cells could be enhanced through blockade of a survival oncogene, focusing on SRC, and the SRC inhibitor PP2. To their surprise, they found that inhibition of SRC actually protects against E2-induced apoptosis in MCF-7:5C cells, and induces a state in which both E2 and SERMs stimulate growth, a situation effectively resembling Phase I resistance. Jordan’s models will likely yield additional novel perspectives in years to come on the remarkable context specificity of E2 in regulating the growth of breast cancer cells.

Donald McDonnell (Duke University) described a new player in androgen receptor biology, HOX13B, which appears to play multiple contextual mechanistic roles in determining the direction of gene regulation by AR. McDonnell’s group carried out a phage display screen for proteins differentially interacting with the androgen receptor in the presence of multiple AR ligands, showing that the interaction profile of a specific ligand could predict biological activity, specifically regulation of known AR target genes in LNCaP cells. In order to prioritize the list of AR-interacting proteins for the next phase of the study, they ranked them according to a number of variables, including likely links to prostate cancer and AR biology, and among the top-ranked proteins in this list was HOXB13, a member of the homeobox family of genes. The homeobox family of genes is involved in the specification of differentiation along the anterior-posterior axis in the developing embryo, and HOXB13 has been shown to be required for normal prostate development. McDonnell showed a series of experiments which demonstrated that siRNA knockdown of HOXB13 dramatically alters AR transcriptional responses in LNCaP cells, and which defined different clusters of R1881-induced genes, some for which HOXB13 played no role, others in which it played a repressive role (so called Class II genes, including PSA and FASN), and others in which it played an activatory role (e.g.ORM1). GO term analysis showed that the genes activated by HOXB13 were involved in proliferation, and that genes repressed by HOXB13 were involved in differentiation, an observation which resolved the distinct context-dependent roles that androgen can play in both stimulating cellular proliferation as well as directing their differentiation. McDonnell showed that for genes in the proliferation/migration class, HOXB13 functioned to tether AR to DNA, whereas for genes in the differentiation class, it functioned as a repressor by physically blocking the interaction of AR to DNA by binding to the AR DNA-binding domain. In a final twist, NKX3.1, a Class II gene, contained both an ARE and a Hox-binding element, and was dependent on both elements for its transcription, indicating that both factors bound DNA and co-operated in the regulation of the gene.

Geoff Greene (University of Chicago) turned the spotlight on how X ray crystallography has been used to divine the mechanisms by which selective estrogen receptor modulators (SERMs) induce unique receptor conformations that correlate with different behaviors in hormone-responsive tissues and cancers. He first drew a distinction between different classes of ER ligands including, on the one hand, full agonists (E2 and diethylstilbestrol) and partial agonists (genistein in the case of ERβ) and, on the other, pure antagonists such as Fulvestrant and the context-dependent mixed agonists/antagonists such as 4-hydroxytamoxifen (4HT) and raloxifene (Ral). He then reviewed work in his laboratory that showed that agonists and antagonists stabilized distinct conformations of helix 12 in the ERα LBD, and that SERMs prevented formation of an active AF-2. Greene then summarized a series of experiments involving hydrogen-deuterium exchange mass spectrometry of the E/F domains of ER bound to Ral and 4-HT. He showed a heat map that represented the degree of exchange that can take place between the hydrogens on the protein and the deuterium in solution; in the presence of Ral, the entire E/F domain is stabilized relative to its complex with 4HT.

Greene went on to shed light on the basis for the properties of less well characterized ER ligands such as R,R-THC, an ERα agonist and an ERβ antagonist. His group solved the structures for both ER subtypes bound to R,R-THC and demonstrated that in the case of ERβ, helix 12 is in the “antagonist conformation” whereas in the case of ERα, it is in the “agonist conformation”. What is the basis for two distinct conformations with one ligand? It appears to be due to a difference in the position of helix 11, such that in the ERβ crystal, H11 was pushed away from the ligand binding pocket and into the space where H12 preferentially docked. In the case of genistein, previous work from another group had shown that when it is bound to ERβ, H12 was in the antagonist conformation. Greene’s group recently solved the structure for genistein bound to ERα in the presence of an SRC-2 peptide, and showed that H12 was in the agonist conformation. He next summarized experiments involving mutational analysis and an ERα subtype-specific ligand, PPT, which illuminated the role of multiple amino acid residues both in and around the ligand-binding pocket in defining the difference in ER subtype pharmacology. Moving on to a series of structures of ERα with a panel of E2 metabolites, he demonstrated that the major factor in determining the lower binding affinity of these molecules compared to E2 was the absence of a hydroxyl group in the D ring of E2 which formed a stabilizing hydrogen bond with H524. In conclusion, Greene reiterated that H12 can adopt multiple conformations and that the presence of an appropriate partner, such as a coactivator, can influence the formation of the agonist conformation.

Day 1 | Day 2 | Day 3