Estrogens

1. Estrogens You can’t live without ‘em: Reproductive failure Bone loss Vasomotor disturbances (hot flashes) Some cardiovascular system vulnerabilities Some cognitive declines, mood disorders Skin changes You can’t live with ‘em: Cancer (breast, uterus, colon, pituitary) Blood clots Nausea or eating disorders If you have the wrong ones (xenoestrogens) you’re in big trouble: endocrine disruption of many types

2. Xenoestrogens (XEs) are threats to both sexesPhysiologic E levels vs. sex, age, cycles, & pregnancy

3. Watson, STKE Dec 1999 How do estrogens and xenoestrogens signal in the cell? ER ER GPR30/ (ERα36) Both: From the nucleus (classical genomic mechanism) From the membrane (novel, nonclassical nongenomic mechanism)

4. Confocal views of E2-peroxidase binding in GH3/B6/F10 pituitary tumor cells Visualizing Membrane Estrogen Receptors (Binding) Nataliya Bulayeva

5. Detection of PROTEIN•PROTEIN epitope interactions – the proximity ligation assay (PLA) Either two ERα Abs or Abs for two differenent proteins mAb (AB-N09) R Ab - mAb + mAb R Ab

6. Slice 3D epitope proximity for ER Two epitopes from within ERα Guangzhen Hu

7. Slice 3D ERα partnering with either Gαi or Caveolin Guangzhen Hu

8. Estrogens (either physiologic or environmental) cause rapid disappearance of mERα Ab recognition (H151 hinge region epitope) while other epitope signals are not affected – rapid conformational change. GH3/B6/F10 pituitary tumor cells 3 min DES (Similar for: E2, dieldrin, endosulfan, nonylphenol) 15 min veh Such responses are both rapid and potent Celeste Campbell

9. Ever since then we have been examining mechanisms of rapid E and XE actions, related to - functional outcomes – looking at ways XEs can disrupt normal function.

10. So many estrogens… so little time! We picked these 16 originally to compare because of interests our interests in: ■ life-stage prevalence of physiological estrogens ■ xenoestrogen toxicities and disease associations ■ structural features/groupings ■ therapeutic uses All small molecules ~270-400 MW … and we are adding more compounds all the time to enhance our stucture-activity analyses

11. Some typical results using these assays: have taught us important concepts about how these compounds act via nongenomic mechanisms

12. We see non-monotonicdose-responsesfor E2-induced PRL release from GH3/B6/F10 pituitary cells at 6 min (also true for other estrogens) Prolactin (PRL) secretion PRL secretion Andrea Norfleet/Jennifer Jeng

13. Xenoestrogens also cause PRL release in a nonmonotonic pattern Bisphenol A * * PRL/CV (% of control) * 10-12 10-11 10-10 10-9 10-8 Big news for the toxicology world -- Nonmonotonic responses make it impossible to extrapolate back from a high dose to see what the lowest effective dose would be. Ann Wozniak

14. ERα is necessary for many of these responses Cells selected for very low mERα levels (D9 subclone) ….can’t elicit the PRL response to estrogens. D9 Cell Dose Dependent PRL Release at 3 Minutes D9 Cell Time Dependent PRL Release at 10-8M E2 PRL/CV (% of Control) 10-12M 10-11M 10-10M 10-9M 10-8M E2 Concentration (M) Time (minutes) Ann Wozniak

15. To efficiently identify specific receptors involved, we have also used --siRNA knockdowns of NRs ER primarily responsible, , and GPR30 sometimes modulatory --receptor-selective agonists --receptor-selective antagonists To prove action from the membrane we have used: --various impeded ligands (E2-BSA, E2-peroxidase, E2-dendrimers) --ERα Ab triggering of responses

16. What signaling mechanisms may be provoked by xenoEs acting on mERs Ca++ mitogen-activated protein kinases (MAPKs) cAMP and PKA G protein activations other kinases/phosphatases caspases downstream transcription factor post-translational modifications and how they relate to functional endpoints

17. Cell signaling is complicated - not just a linear signaling cascade. Many things must be coordinated. It is a web.

18. Signaling is like a Rube Goldberg Machine http://blueballfixed.ytmnd.com/ created by the Web site Something Awful

19.  - non-nuclear feeder signaling streams we have tested & found to be involved in E-induced downstream ERK signaling         Mitogen-activated protein kinases (MAPKs) like ERK: Are signal integrators. …..many signaling pathways funnel into the ERK “integrator”. Are associated with major cellular destinies, like cell proliferation   

20. Signaling starting at the membrane funnels into the downstream kinases ……and then fans out again to coordinate major complex downstream functions.The MAP kinases are the integrators that separate these two phases

21. Proliferation cell cycle receptors DNA and protein synthesis for replication MAPKs ERKs JNKs p38 2nd messengers and signaling enzymes (kinases, and phosphatases, lipases, NT-cyclases) and scaffolds new proteins caspase cascades Differentiation Death Migration Secretion

22. One of our most useful techniques is a fixed-cell 96-well plate immunoassay which: --is optimized for each cell type, epitope-Ab, and cell compartments (based on permiabilization); makes use of the many Abs now available to the activated (usu. p’ated) form of proteins-- is relatively high throughput allowing us to compare multiple xenoestrogens at a wide range of concentrations acting on multiple antigens-- allows us to do physiologic Es + XEs in incremental mixtures, the real-life scenarios of XE toxic exposures

23. Responses oscillate with time; XEs cause different phasing than E2 Nonylphenol DDE E2 Coumestrol ERK activation - % of control) DES Dieldrin Endosulfan E2-Peroxidase time, min 96-well plate immunoassays for phospho-ERK ligands all at 1nM Nataliya Bulayeva

24. Responses are often non-monotonic: U-, V-, or M-shaped dose-responses ERK activation Concentration Nataliya Bulayeva

25. Why do MAPK responses oscillate oscillate with time and concentration.

26. The sum of pathways A & B is an oscillating curve over time or concentration

27. Some responses are very rapid – seconds Here is a new example of one of the most rapid -- Gi activation(charged with GTP and assayable with an Ab to the charged form). Because it is so proximal to receptor-binding it is much more rapid (peaks at 15 seconds) than downstream signal integrator responses like for MAPKs (several minutes). Again see differences between E2 and XEs Guangzhen Hu

28. Proliferation function downstream of MAPKs are affected by XEs Both physiologic estrogens and xenoestrogens cause pituitary cells to proliferate (just as pituitaries and their tumors grow in response to estrogens)… Jennifer Jeng

29. Mechanisms that affect cell number – apoptosis – are elicited by XEs Rene Vinas

30. …Es and XEs (in this case phytoEs) can also rapidly activate (phosphorylate) transcription factors downstream of ERKs Examples: Elk and ATF2, with some differences in response pattern Jennifer Jeng

31. To summarize the 1st part…endocrine disruption by xenoestrogens includes:  altered timing (phasing)  altered dose-response patterns and levels  imperfect mimicry or inhibition of multiple estrogenic actions activated via the nongenomic pathways Can a structure/chemical characteristic of an estrogen predict its activity in a nongenomic response?

32. Here again are the ones we have studied They can all be described by their physical characteristic of lipophilicity – the partition coefficient between octanol and water Short chain alkylphenols are less lipohillic than long chain ones

33. 13 estrogens – their activities in various functional and signaling responses compared to their hydrophobicity Best correlations are for structures that vary in simple ways (side-chain length)  E2 1 nM  E1 1 nM  E3 1nM Cou 10 nM  Dai 100 nM  Gen 100 nM  Res 100 nM  EP 1 nM  PP 1 nM OP 1 nM  NP 1 nM  BPA DES The most complicated - both genomic and nongenomic pathway control (Chlorinated compounds were left out as they significantly reduced the correlations) Chemists can use such information to design less estrogenic versions of these compounds Jennifer Jeng Mikhail Kochukov

34. But in real life these estrogens are rarely present by themselves You have physiologic estrogens on board already – and then you get exposed to xenoestrogens on top of that. Do xenoestrogens interfere with physiologic estrogens? Do they disturb the normal signaling patterns? How? In different ways - examples for ERK signaling follow that demonstrate some principles of estrogenic endocrine disruption.

35. Temporal phasing of response….XenoEs can delay the response after causing an initial dephosphorylation pERK Time in Minutes Jennifer Jeng

36. XEs can enhance an estrogenic response at concentrations where it is less effective by itself; inhibit at higher concentrations – this is the most typical effect of combinations (physiologic estrogens all at 1 nM for these studies; 5 min responses) Have assessed for 3 physiologic estrogens challenged by 5 different XEs Jennifer Jeng

37. Extreme Disruption by Multiple XEs - can completely wipe out a physicologic E response – we have seen this now with multiple combinations of XEs. Rene´ Viñas

38. Any of these alterations “disrupts” the normal pattern or extent of physiologic estrogen signaling.

39. In summary we have learned that Es/XEs acting via nongenomic pathways: ► signal more potently via the membrane versions of estrogen receptors ►activate multiple signaling pathways ► responses often oscillate with time and are non-monotonic (U- V- or even M-shaped dose-responses and XEs alter this to disrupt signaling – a difficult target for regulation ► mediated primarily via ER (depending on tissue); modulated by other ER subtypes ► XEselicit tissue-specific disfunctions (secretion, differentiation, cell proliferation, enzymatic cascades, components of inflammatory responses, behavior, ROS)in different nontransfected cell types: pituitary, breast, prostate, neuronal, cells of the immune system…… ► have particular developmental window / stage-specific vulnerabilities to XEs ► chemical structures of Es/XEs can predict strength/potency of nongenomic responses ► XE mixtures (which is what we are exposed to) dramatically disrupt physiologic E responses

40. How we would like to use our newly determined principles learned via high throughput signaling assays: Green Chemistry Design nongenomic cell signaling - later assigned to a NR remediation of existing contamin-ants animal develop-ment - later assigned to a NR NR binding and NR-based genomic response in silico chemical- receptor docking REDESIGN We are currently collaborating with labs who view endocrine disruption at different (increasingly more complicated) levels. + - + + +

41. Acknowledgements Acknowledgements Watson lab – over the years working on nongenomic E signaling projects Todd Pappas Celeste Campbell (Finnerty) Andrea Norfleet Adrian Jakubas Bridget Hawkins Gaga Zivadinovic Nataliya Bulayeva Teresa Reed Leanne Lash Ann Wozniak Rebecca Alyea Yow-Jiun Jennifer Jeng Mikhail Kochukov Stephanie Laurence Anannya Banga Luke Koong June Guptarak Guangzhen Hu Rene Vinas Manish Saraf Collaborators: Bahiru Gametchu, Mary Thomas, Kathryn Cunningham, Randy Goldblum & Terumi Midoro-Horiuti Green Chemistry Collaborators: Ruben Abyuan & Fiona McRobb, Bruce Blumberg, Susan Jobling, Terry Collins, John Warner, Pete Myers & Karen O’Brien Funding NIEHS (R01s and training grant) NIDA (P20 Center and training grant) Am. Inst. for Cancer Research UTMB Center for Addiction Research Sealy Center for Environ. Health & Medicine Passport Fndn. Innovator Award

43. Immuno-assay for quantitation of membrane vs. intracellular proteins, their trafficking, and their activation state outside inside Fix cells: various methods to match epitope and cell type -- unpermeabilized measures membrane proteins -- permeabilized (usu. with detergent) measures intracellular proteins (We have used it for various ERs, DAT, pMAPKs, and p-transcription factors, and GTP-G proteins in many different cell types) Incubate with 1°Ab; biotinylated 2° Ab; avidin-conjugated alk. phos. Incubate with pNpp →pNpat 37° in dark, & read at 405 nm Ag quantitation Wash off reagents Stain with 0.1% crystal violet, wash, extract, read at 562 nM cell number (normalization for each well)

44. molecules- receptors -- cell signaling - organisms- remediation genomics nongenomics In silico – Ruben Abyuan and Fiona ….. Receptor binding and genomic activation – Bruce Blumberg Nongenomic cell-type specific signaling – Cheryl Watson Developing organisms – Susan Jobling Remediation – Terry Collins, John Warner Cementing it all together – Pete Myers and Karen O’Brien

45. In silico cell-free cell dependent development chemical capture of amphib & fishes and degradation

46. Do TAML activators activate ERK – NO!!!!!

47. www.secondlife.com CW searching for receptors, wherever they are in the cell Genome Island, Genome (139,145, 37)

48. Crucial difference between our studies and past studies on xenoestrogens: ●we look at rapidnongenomic effects ●we use very low (fMnM) concentrations (very relevant to those found commonly in the environment) within very wide concentration ranges ….and sensitive, quantitative cellular response assay systems ●we use non-transfected cell systems to avoid overexpression and heteroexpression artifacts

49. The assays we plan to do in GH3/B6/F10 pituitary cells for the Green Chem Center Project • Signaling assays --G protein activation (proximal signaling) --ERK (+JNK & p38) kinase activations (downstream integrated signaling) • Function assays --PRL release (nongenomic functional endpoint) --Cell proliferation (complex functional endpoint)

50. If the money got bigger I could add: • Other cell types in which we have already monitored nongenomic signaling effects (breast cancer, prostate cancer, neuronal, immune system) • Other myriad signaling pathways (IPs, ions, other kinases, activation of transcription factors) • Other functional endpoints