Estrogenic compounds can stimulate proliferation of the mammary epithelium, but also potentiate the activity of the p53 tumor suppressor protein. These contradictory activities of estrogenic compounds in mammary tissues may be mediated through activation of two estrogen receptor (ER) subtypes, ERα and ERβ. The following experiments were conducted to examine the roles of these receptors in regulating p53 activity in the mammary epithelium in vivo and in vitro. Selective agonist for ERα (PPT) and ERβ (DPN) were compared with 17β-estradiol to examine the roles of ERα and ERβ in potentiating p53 activity, radiation-induced apoptosis and proliferation in ovariectomized mice. DPN was sufficient to potentiate p53-dependent apoptosis in the mammary epithelium following irradiation without inducing proliferation. DPN was also 2.5-fold more potent in stimulating expression of Egr1 , a modulator of p53 activity. Introduction of ERβ into MCF-7 cells increased in the transcriptional activity of p53. As radiation-induced apoptosis was diminished in mice lacking ERβ (BERKO) mice, ERβ appears necessary for optimal activity of p53 in the mammary epithelium. The ability of DPN to maximally stimulate responsiveness of p53 to ionizing radiation in the absence of proliferation suggests that ERβ agonists may be an effective adjuvant therapy. Phytoestrogens are estrogenic compounds that are abundant in soy-based products, a key component in Asian diet associated with reduced breast cancer incidence in Asian women, and are preferential ligands for ERβ. However, the effects of soy differ greatly depending on the form and doses administered. Therefore, the effects of water-soluble extracts of non-fermented and fermented soy (NFSE and FSE, respectively) were compared. At physiological relevant doses both NFSE and FSE inhibited proliferation of cell lines from normal breast epithelium (76N-TERT) and breast cancers (21MT-1,MDA-MB-231). The FSE also increased the tumor-free survival of mice bearing xenografts of MDA-MB-231 cells. However, these effects of soy extracts were independent of both p53 and ERα. As both p53 and ERα are commonly lost in breast tumors, the pathways by which soy extracts antagonize tumor growth could provide valuable therapeutic targets for the treatment and prevention of breast tumors.

Programmed cell death (PCD) is an essential developmental process in all multicellular organisms. It serves multiple functions including selected removal of unneeded and/or deleterious cells, and regulation of cell numbers (reviewed in Milligan and Schwartz, 1996). The intersegmental muscles (ISMs) of the hawkmoth Manduca sexta provide an ideal model for studying PCD (reviewed in Schwartz, 1992). In response to a decline in the circulating titer of the steroid hormone 20-hydroxyecdysone (20-HE), these cells initiate a death program which includes both the up- and down-regulation of specific genes. Following eclosion, the ISMs undergo PCD that results in the complete destruction of the muscles during the subsequent 30 hours (Finlayson, 1956). This dissertation examines in detail the repression of actin and myosin heavy chain expression that occurs when the muscles become committed to die. At the protein level, actin expression was reduced by 84% at the time the muscles were committed to die, which presumably plays a role in the rapid dissolution of the muscles. When the ISMs became committed to die, there were dramatic increases in proteolytic activity that are correlated with an approximately eightfold increase in the absolute amounts of multicatalytic proteinase (MCP). At the time of commitment, four new MCP subunits were observed to be associated with the complex. Correlated with the addition of these new subunits was a dramatic increase in the levels of immunodetectable MCP throughout the cytoplasm and within the nuclei of dying muscles. These changes in MCP were regulated by the same hormonal signals that mediate cell death. Cells dying by PCD often display a characteristic set of features termed apoptosis. These features include chromatin condensation, DNA fragmentation, membrane blebbing and phagocytic removal of the dying cells. However, dying ISMs display few characteristics of apoptosis. Interestingly, apoptotic cell death does occur in Manduca embryogenesis. Evidently Manduca possesses the necessary biochemical machinery to undergo apoptosis and does so in specific developmental circumstances. These data suggest that more than one cell death mechanism is used during development.

The pituitary gland is the major endocrine gland in the forebrain. The hormones secreted from this gland regulate vital processes such as reproduction, growth and stress response. Distinct endocrine cells arise from pituitary precursors cells. The endocrine cells are spatially organized along the anterior-posterior axis within the anterior lobe of the pituitary, the adenohypophysis. Several signaling molecules have been shown to play roles in the development of this endocrine gland. Previous work in our lab showed that Sonic Hedgehog (Shh) is required for induction and patterning of the adenohypophysis. Hedgehog (Hh) signaling is involved in many developmental processes including induction, patterning and differentiation of many tissues. In addition, independent studies show that Fibroblast growth factor (Fgf) signaling also plays a role in the development of zebrafish adenohypophysis. ^ One of the main aims of my dissertation was to determine how the Hh and Fgf signaling pathways specify the functional patterning of the adenohypophysis. Using small molecule inhibitors I show that high levels of Hh signaling are required for the formation of the anterior adenohypophysis, the pars distalis (PD) and high levels of Fgf signaling are required for the formation of the posterior adenohypophysis, the pars intermedia (PI). My dissertation work also shows that high Hh levels are required for differentiation of the endocrine cells in the PD, and in contrast high Fgf levels are required for differentiation of the endocrine cells in the PI. Using live-imaging of a transgenic zebrafish line, I show that the PD and PI originate from distinct regions. My analyses has revealed that graded Hh and Fgf signaling help pattern the adenohypophysis along anterior-posterior axis by guiding endocrine cell differentiation in a dose dependent manner. ^ A related aim of my research was to analyze the role of Hh signaling in zebrafish neural tube patterning. The transcriptional response to varying Hh levels is well characterized in chick and mouse neural tube, and this transcriptional response has been partially described in zebrafish. The analysis of the Hh transcriptional response in wild type, Hh mutant and Hh over-expressing embryos show that there is a conserved transcriptional response to Hh signaling in the zebrafish neural tube. My comprehensive analyses of the Hh transcriptional response in the zebrafish neural tube provides a useful tool for the characterization of Hh signaling in zebrafish. ^

Pollen tube elongation is a polarized cell growth process that directionally transports the male gametes from the stigma to the ovary for fertilization inside the ovules. Actin cytoskeleton is known to support this growth process and Rac-like GTPases have been shown recently to be important to regulate actin organization in elongating pollen tubes. Actin depolymerizing factor/cofilins (ADF/cofilins) are actinbinding proteins that increase actin dynamics by enhancing actin depolymerization. They are also responsible to regulate actin organization in Rac-mediated signaling. My thesis research focuses on establishing a signaling pathway from Rac GTPase to the actin cytoskeleton via the regulation of ADF/cofilins in tobacco pollen tubes. I have isolated and characterized cDNAs for tobacco pollen ADF, NtADFs, to study their function in pollen tube growth. First, I showed the activity of Rac-like GTPase is essential for pollen germination. Tobacco pollen germination and early pollen tube growth stimulates the activation of these small GTPases, the phosphorylation of NtADFs and an increase in the ratio of F- to G-actin. Moreover, over-production of a pollen-expressed Rac-like GTPase, NtRac1 from tobacco, induces increased ADF phosphorylation in transformed pollen and diminished the binding of GFP-tagged NtADF1 (GFP-NtADF1) to actin filaments in growing pollen tubes. These observations are consistent with the presence of a signaling pathway in pollen whereby Rac-like GTPase are stimulated by germination to activate a phosphorylation cascade that down regulates the activity of ADFs. This ultimately affects actin dynamics and growth characteristics in pollen tubes. Second, I also showed that NtADF activity is important for actin organization in pollen tube growth. When expressed in a moderate level in pollen tubes, GFP-NtADF1 associated prominently with a sub-apical actin mesh comprised of short dynamic actin filaments and with long dynamic actin cables in the shank. Over-producing NtADF1 resulted in the reduction of fine, axially oriented actin cables in transformed pollen tubes. Pollen tube growth was also inhibited by over-expressed NtADF1 in a dosage-dependent manner, suggesting proper regulation of actin turnover by NtADF1 is critical for the pollen tube growth process. In addition, NtADF1 activity is regulated by phosphorylation and pH. By creating mutants on the serine 6 residue on NtADF1, the result showed that the charge characteristics on serine 6 is important for NtADF1 interaction with actin and for its activity on pollen tube growth. By an in vitro depolymerization assay, recombinant NtADF1 depolymerizes actin more efficiently at pH 8 than pH6. This and localization of a NtADF1-rich actin mesh in the sub-apical region of elongating pollen tube, which is known to have a more alkaline cytoplasmic condition relative to the apex, suggest that interaction between NtADF1 and actin in this vicinity maybe critical for the pollen tube growth process. Finally, to examine a signaling pathway from Rac to ADF, I showed that overexpression of NtADF1 suppressed Rac-induced isotropic pollen tube growth. These observations demonstrating biologically that pollen ADFs mediate signaling activated by Rac-like GTPase to the actin cytoskeleton in pollen tube growth.* *This dissertation is a compound document (contains both a paper copy and a CD as part of the dissertation). The CD requires the following system requirements: QuickTime.

Cells respond to changes in their environment, or to mitogenic events at the cell surface by generating signals. Intracellular signals induced as a result, pass from the cell surface to the nucleus via a multitude of mediators, and terminate in cellular proliferation, differentiation or even cell death. Protein phosphorylation is an integral part of signal transduction, and protein kinases are important second messengers in signal transduction pathways. Binding of SV40 to its receptor on CV-1 cells induced signals that led to the upregulation of the primary response genes, c-myc and c-jun. The major histocompatibility complex class I proteins are an essential component of the SV40 receptor. Antibodies against the MHC class I proteins also led to the upregulation of c-myc and c-jun. The serine/threonine kinase Raf and the mitogen activated protein kinases (or MAPK), are highly conserved and play an integral role in signal transduction. Along with the GTP-binding protein p21Ras, the activation of Raf and MAPK form a central paradigm of cell signaling in eukaryotic cell. Activation of protein kinase C, another serine/threonine kinase has also been implied in certain systems. A study of the signaling pathways been elicited by extracellular SV40 imply protein kinase C to be involved in the upregulation of c-myc and c-jun, but provide no evidence for the involvement of either Raf, or MAPK. Antibodies against the MHC class I proteins, on the other hand, indicate the activation of protein kinase C, Raf and MAPK. Moreover, experimental data suggests the anti-class I-induced activation of protein kinase C to be dependent on tyrosine kinases that lie upstream of protein kinase C. Analysis of the signaling pathways being evoked by SV40 and antibodies against class I proteins suggests the existence of separate and distinct signaling pathways being induced by SV4O, and anti-MHC class I antibodies in CV-1 cells.

Programmed cell death (PCD) is an essential component of animal development and serves a variety of functions. The intersegmental muscles (ISMs) of the tobacco hawkmoth Manduco sexta provide a useful model system to study the molecular mechanisms that mediate PCD. The ISMs participate in the emergence behavior of the adult moth at the end of metamorphosis and then die during the subsequent 30 hours. In addition, several populations of interneurons and uniquely identified motor neurons also die following adult emergence. The trigger for this death is a decline in the circulating titer of the molting hormone 20-hydroxyecdysone. Previous work has demonstrated that the ability of the ISMs to die is dependent on new gene expression. Using a differential hybridization cloning strategy, a cDNA library generated from condemned ISMs was screened, and four up-regulated clones were isolated. One of these recombinants was found to encode apolipophorin III (apoLp-III), a component of lipophorin. The expression of apoLp-III was dramatically elevated with the death of the ISMs, some interneurons and identified motor neurons. ApoLp-III was not detectably associated with apolipophorin I and II, required components of lipophorin, or with other molecules in the dying cells, suggesting that apoLp-III has activities independent of lipid transport that may play a role in programmed cell death. Another clone, 18-56, encodes a phylogenetically conserved ATPase domain-containing (CAD) family member related to putative proteasomal subunits and transcriptional regulators. While clone 18-56 was expressed in all tissues examined and during every stage of ISM development, there was a dramatic increase in its expression at both mRNA and protein levels when the ISMs became committed to die. Furthermore, the mouse homolog of 18-56, m56, was cloned and its expression pattern was examined in the mouse. No correlation was detected between enhanced m56 expression and apoptosis in mammalian cells, suggesting that the molecular pathway used by the ISM death may be distinct from that of apoptosis. Rat-1 fibroblast cell lines that over or under express m56 were generated, thus providing tools for further functional study of m56 in mammalian cells.

Directional cell migration, chemotaxis, requires a polarized cell morphology in which the cell extends pseudopodia at the front and contracts the rear to move towards a stimulus. PI(4,5)P2 levels set up a threshold for the activity of signaling molecules at the rear and the leading-edge of a cell. To further demonstrate the importance of plasma membrane (PM) PI(4,5)P2 in maintaining cell morphology during chemotaxis, we used a mutant strain of the eukaryotic model system, Dictyostelium discoideum. This mutant strain lacks the type I PIP5 kinase, which is the main enzyme synthesizing PI(4,5)P2. These cells, designated pikI-, have highly reduced PI(4,5)P2 levels and higher Ras GTPase activity compared to wildtype cells. Leading-edge biosensors diffuse to the cytosol when the pikI- round-up and translocate back to the PM when the cells spread. These observations propose that PI(4,5)P2 levels elevate as cells round-up and decrease as cells spread. This interesting phenotype resembles the front and rear of a migratory cells. Interestingly, pikI- resemble similar cell morphology and biosensors dynamics observed when we use an inducible system to deplete PM PI(4,5)P2 levels. We, also, observed the dynamics of a biosensor for an F-actin polymerization protein called formin A (ForA). ForA has been shown to localize at a polarized cell’s rear and in the cleavage furrow of dividing cells. In addition, ForA have a PI(4,5)P2 binding motif and binds to PI(4,5)P2 preferentially in vitro. Our results support a role for PI(4,5)P2 in regulating ForA with the plasma membrane. Taken together, we proposed that local levels of PI(4,5)P2 contribute to the electrostatic interactions of regulatory proteins controlling actin dynamics and membrane protrusions. PM PI(4,5)P2 below a threshold activate regulatory proteins that excite the signaling that promotes protrusions, while below threshold levels would inhibit those proteins activity. The change in PI(4,5)P2 levels would be predicted to affect the membrane’s charge, which in turn changes the interaction and disassociation of many anionic regulatory proteins involved in the signaling pathway and cytoskeletal rearrangements. Additionally, we show for the first time, a correlation between the PM PI(4,5)P2 threshold and rates of phosphatidylserine (PS) exposure in cancer therapeutics. Receptor-mediated cell stimulation triggers PS exposure to the outer leaflet of the plasma membrane. Interestingly, pik1- and cells using an inducible system to deplete PM PI(4,5)P2 levels depicted the same responses. In addition to PI(4,5)P2, PS exposure affects the membrane’s charge which impacts the signaling molecules activity in the pathway. Altogether, PI(4,5)P2 and PS are proposed to be novel therapeutic targets in cancer treatments. Chemotaxis is a feature of metastatic cancer cells and is regulated by various regulators including actin cytoskeleton. Actin cytoskeleton reorganization during chemotaxis is regulated by actin-binding proteins including those that interact with the PM PI(4,5)P2. Energy production regulates cell migration as well, through glycolysis pathway. A previous study proposed that actin reorganization releases Aldo A enzyme which enters glycolysis, through activating PI3K signaling pathway. However, the mechanism of action remains unclear. We speculate that local PI(4,5)P2 levels regulate Aldo Activity through regulating actin-severing proteins activity including cofilin and gelsolin, and actin polymerizing protein including ForA. PI(4,5)P2 levels below a threshold release actin-severing proteins to the cytosol triggering the severing of actin and the release of Aldo A. while, above PI(4,5)P2 threshold activates and localizes for a on the PM promoting the F-actin polymerization and the sequester of Aldo A into F-actin. The goal of this work is to discover a new model for actin cytoskeleton regulation during migration, as its linkage to glycolysis and metabolism has important implications for cancer.

Centromeres are the site on a chromosome that allows for proper chromosome segregation during cell division. The kinetochore, a DNA-protein complex that forms at the centromere, attaches the chromosome to the spindle fiber and therefore ensures faithful separation of the chromosomes. Cse4p was identified in the budding yeast S. cerevisiae and is protein that resembles histone H3. Due to its centromere specific localization led to the proposal of the centromere specific nucleosome(s). Further identification of Cse4p homologues in many diverse organisms, including mammalian systems, supported the idea that the centromere specific nucleosome is an evolutionarily conserved structure. The localization of the centromere specific histone H3-like proteins (CenH3) to the centromere is one of the remaining questions in centromere biology. It is known that the CenH3 proteins have two domains, the conserved Histone Fold Domain (HFD), and a unique N-terminus, and that each domain has a distinct function. I analyzed the function of the Cse4 N-terminus and HFD in centromere localization and showed that the N-terminus has no function in centromere localization, but rather is involved in regulating Cse4 protein turnover. The HFD, therefore, contains all the centromere targeting information and is necessary and sufficient for Cse4 function. Further analysis revealed that the specific DNA contacts of the Cse4 HFD are important for localization to the centromere. In the budding yeast, S. cerevisiae, the centromere DNA is defined by a specific DNA sequence that is divided into three distinct regions, CDEI, II, and III. Mutations of these regions affect centromere function to varying degrees. Analysis of Cse4p localization to mutant centromere constructs revealed that CDEII is necessary for Cse4 centromere localization, but that the intrinsic bend of this element is not involved in Cse4p centromere localization. Cse4p localization to two other centromere mutants, cen3-X35 and cen3-X50, was analyzed. Cse4p could recognize cen3-X35, a construct with a CDEII deletion, but failed to recognize cen3-X50, a CDEII and CDEIII mutant centromere construct. Furthermore, the CBF3 complex was present at both of these centromere constructs, indicating that the CDEIII by deleted in cen3-X50 directly affects Cse4p centromere localization.

The earliest steps of nascent protein folding are critical to the overall folding efficiency. Folding events start as soon as the protein is translocated into the endoplasmic reticulum lumen, where the co-translational machinery ensures the fidelity of protein folding by coupling molecular chaperones, foldases and folding sensors. We seek to investigate the co-translational maturation events of disease-related glycoproteins containing different membrane topology to determine the generality and substrate specificity of this process, to hopefully provide new insight into therapeutic methods by targeting protein maturation at early stages. A variety of cell biological, biochemical, and molecular biological approaches using cell-free assays, isolated organelles and live cells have been applied in this study. The work on co-translational maturation of a type I membrane protein human tyrosinase has shown that Hsp70 family member BiP handed off tyrosinase to the lectin chaperones calnexin/calreticulin as glycans were added. The maturation pathway of the albino mutation of tyrosinase (C71R) diverges from that of the wild type co-translationally through its recognition by the oxidoreductase ERp57. The work on a type II membrane protein influenza neuraminidase (NA) subtype N9 has shown that calnexin co-translationally interacted with NA prior to calreticulin. This sequential manner was found to be a common feature of the ER assembly line determined by the membrane localization and soluble characteristics of calnexin/calreticulin, respectively. These interactions were required for the proper maturation of NA as NA aggregated if calnexin/calreticulin interaction was abolished by glycosylation inhibition or removal of specific glycans. Surprisingly, a subset of NA molecules can form intermolecular disulfides co-translationally supporting NA homodimerization. NA co-translational dimerization also occurs for a NA mutant lacking the critical large loop disulfide bonds, indicating that the dimerization of the stem domain does not require proper folding of the top globular domain of NA. This represents an exception to the general rule that protein oligomerization happens after the folding of individual domains. Future work on a variety of other substrates will help illuminate a global pathway of glycoprotein co-translational maturation.

Nur77, an orphan nuclear receptor, plays a key role in T cell apoptosis. As a transcription factor, Nur77 is assumed to exert its functions by driving the expression of target genes. However, Nur77 targets in T cell apoptosis are unknown. In cancer cell lines, Nur77 can induce apoptosis through the intrinsic apoptotic pathway but it remains controversial how Nur77 kills T cells. In this study, we provide biochemical, pharmacological and genetic evidence demonstrating that Nur77 induces apoptosis through the activation of the intrinsic pathway in T cells. We also show that Nur77 is a physiological substrate of the MEK-ERK-RSK-cascade. Specifically, we demonstrate that RSK phosphorylate Nur77 at serine 354 and this modulates Nur77 nuclear export and intracellular translocation during T cell death. Our data reveal that Nur77 phosphorylation and mitochondrial targeting, regulated by RSK, may define a role for the MEK1/2-ERK1/2 cascade in T cell apoptosis.