Bibliography: pages 229-241. / A trypsin inhibitor had previously been isolated from Erythrina caffra, a member of the Leguminosae family. The inhibitor, Erythrina trypsin inhibitor (ETI), is unique among plantderived inhibitors, in that in addition to trypsin, it inhibits chymotrypsin and tissue plasminogen activator (tPA). ETI was previously sequenced and its crystal structure determined from which it could be seen that ETI has good homology to soybean trypsin inhibitor (STI). However, STI does not inhibit tPA. From the three-dimensional structure of ETI it was known that the amino-terminus of the molecule forms a finger-like structure stabilized by hydrogen bonds and hydrophobic interactions. In addition, the N-terminal finger region is located in close proximity to the reactive site loop and the Nterminal residue (Val) is bound up in the finger region. In STI the N-terminal region is located in close proximity to the reactive site loop and is folded into a structure similar to that in ETI. While the crystal structure of STI was not detailed enough to determine secondary interactions such as hydrogen bonds it was hypothesized that the N-terminal region is stabilized as in ETI. It was further hypothesized that the N-terminal residue of STI (Asp), because of its hydrophilic nature, is not involved in the structured N-terminal finger region of this protein. This leaves this Asp residue of STI free to form an ion pair with Lys at position 60 in trypsin when STI and trypsin interact.

It is common practise in antisense technology to view third strand binding to be controlled by the same principles which are found to determine the stability of the double helix. In contrast to this view based on a general consideration of the various forces contributing to the binding energy of the third strand it was proposed that the dominant contributions will originate from electrostatic interactions. These electrostatic contributions can be subdivided into sequence independent repulsive forces between the negatively charged backbones and into sequence dependent attractive forces between the positively charged protonated Hoogsteen cytosines and the backbone phosphates. The observable changes in the stability of triple helices should be a reflection of the number (global composition) and distribution (local composition) of cytosines in the third strand. To this aim two families of 38-mer oligonucleotides were synthesized, which have as a common design feature a linear array of 10 homopurine bases followed by 10 homopyrimidine bases as Watson & Crick complementary strand to the homopurine region and ending in a 10 homopyrimidine residue stretch which binds to the W&C helix via Hoogsteen base-pairing. This arrangement of homopurine and homopyrimidine sections with connecting pyrimidine linkers allows the formation of intramolecular triple helices of predetermined stoichiometry and strand orientation. Physical (UV-spectroscopy, CD-spectroscopy and fluorimetry) and biochemical techniques (P1-nuclease digestion) have been used to show that the oligonucleotides undergo a stepwise folding process from a random coil into a hairpin with 3'dangling tail and then into a intramolecular triple helix. This folding occurs as a function of pH and/or ionic strength. The effect of local and global composition on the stability of the three conformational transitions has been evaluated from a comparison of the melting temperatures and the behavior of the phase boundaries of the different oligonucleotides. As the result of this thesis the following general rules emerge: The stability of the third strand depends on the particular combination of sequence, pH and ionic strength. At physiological conditions (pH 7.1, 150 mM Na⁺) thymines and cytosines contribute equally to the stability (global effect) provided that the cytosines are spaced by more than one thymine. (local effect). Below pH 7.1 (150 mM Na⁺) the stability increases linearly with the number of cytosines and at pH above pH 7.1 ( 150 mM Na⁺) it decreases. At ionic strength below 400 mM Na⁺ (pH 6. 75) the stability increases with the number of cytosine while above 400 mM Na⁺ (pH 6. 75) it decreases. Based on these results a rational approach for the design of oligonucleotide third strands and the choice of appropriate environmental conditions for the formation of a particular triple helix becomes feasible.

Bibliography: p. 194-210. / Embryonic olfactory neuroepithelium provides a useful experimental system for the study of olfactory neurogenesis. As a substrate for experimental neural cell biology, olfactory neuroepithelium is of unique interest since, unlike other neural cells, olfactory neurons are continually replaced - a feature that is dictated by their direct exposure to the damaging external environment. Basal cells in the olfactory placode are the source of this replacement. Each olfactory neuron expresses only one or a few of the many olfactory receptors that are encoded by the large array of olfactory genes. Despite this limited cellular display of receptors, vertebrates are able to distinguish many thousands of different odorants, implying a complicated need for perceptive neurological processing of signals coming from individual olfactory neurons. To study the events that take place during the differentiation of neuronal precursors - a process that sustains a diverse receptor repertoire - I felt that lines of conditionally immortalised cells that could be induced to differentiate would provide useful reagents. In this thesis I describe my successful attempts to immortalise olfactory cell lines from the neuroepithelium of E 10.5 mouse embryos. I used a conditionally immortalising retrovirus that included the coding sequence for the temperature-sensitive SY 40 large T antigen. Integration of this retrovirus into the genome of cells allowed continuous proliferation at the permissive temperature of 33°C. A shift to the nonpermissive temperature of 39°C inactivated the SV40 large T antigen, the cells ceased proliferation and differentiation commenced. Sixty cell lines were derived of which four were chosen for further characterisation. These four cell lines (OP6, OP27, OP47 and OP55) were clonally derived and were immortalised rather than transformed. They continued to express the SV40 large T antigen at 33°C but lost expression at 39°C concomitant with cessation of proliferation. When the OP cells were shifted to 39°C in the absence or presence of the morphogen, retinoic acid, morphological changes ensued that were consistent with the development of neuronal characteristics. The OP6, OP27 and the OP47 cells became phase-bright with neuritic extensions. The OP55 cells were the exception in that they did not develop extensions but instead differentiated to form compact epithelial islands when grown in DM-10 medium but not in RA medium. Differentiation of the OP cells at 39°C was further documented by the induced expression of a number of markers demonstrated by RT-PCR and/or immunocytochemistry. The OP cells differentiated at 39°C in DM-10 and in retinoic acid-containing medium to express olfactory receptor transcripts. Cloning and sequencing showed that each cell line expressed a single receptor type but that different receptors were expressed by different cell lines. Sequencing revealed that the receptors cloned from the OP27 cells were 98% homologous to the mouse-M65 olfactory receptor whereas OP55 had greatest homology to rat-Olf3 olfactory receptor. The transcripts induced in OP6 and the OP47 cells showed greatest homology with Gus58 - a taste receptor homologous to olfactory receptors. Sequences obtained from OP6, OP47 and OP55 cells were not 100% identical to published receptors and could thus represent members of different subfamilies. Interestingly, induced OP55 cells also expressed mRNA for clusterin - a molecule that has no homology with olfactory receptor transcripts but is involved in differentiation during embryogenesis.

Watson-Crick CWC) and Hoogsteen (HG) triple-helical three-way junctions were constructed from three 33-mer oligonucleotides. The same sub-set of sequences have been used in the arms of the junctions. The junctions differ primary in the arrangement of the branch point and the ends of the arms. In the case of the HG triple-helical three-way junction, the three 33-mer oligonucleotides can fold into hairpin structures, linked by a four membered cytosine loop. Each of the hairpins contain a homo-pyrimidine 10-mer single strand extension which interacts with a neighboring hairpin to form a triple-helix on lowering the pH (between 6 and 4), via Hoogsteen (HG) hydrogen bonding. Collectively this process results in the formation of the branch point and the triple-helical arms. In the case of the WC triple-helical three-way junction, the three 33-mer oligonucleotides interact with a neighboring oligonucleotide to form a duplex. Collectively this process leads to the formation of a double-helical three-way junction with each arm containing a 9-mer homo-pyrimidine extension connected by four cytosines. Each of the single strand extensions can mutually fold back onto the duplex arms converting each arm into a triplex. The WC and HG triple-helical three-way junctions were characterized by gel electrophoresis, temperature gradient gel electrophoresis, circular dichroism (CD), uv melting, and differential scanning calorimetry (DSC). In both structures; arm A contained exclusively TAT triad bases, while arms B and C contained an increasing number of CGC+ triads, respectively. To counteract possible crowding at the branch point the 5' sequence was shortened by one base. The assembly of the completely folded structure was found to be spontaneous if an appropriate ionic strength and pH range was chosen. A separate set of isolated arms has been investigated to elucidate the role each arm plays in the complete structure. Comparing the summed-up properties of these arms with the data obtained for the integral three-way junctions, it is obvious that the we three-way junction is partly distorted at the branch point, in line with observation obtained from double-helical three-way junctions, while the HG threeway junction is completely ordered. A set of mathematical models has been developed to describe the thermal unfolding of the multi-strand DNA structure and to identify the intermediate states. Presented is a formalism, starting from the grand partition function, that describes the effects of pH on the thermal stability of triple-helices. The formalism can be used over a wide pH range. It covers nearest neighbor electrostatic effects of closely spaced cytosines in the Hoogsteen and Watson & Crick strands. A procedure is employed to predict enthalpy and entropy changes for triplex formation. The obtained values are in good agreement with the results obtained by differential scanning calorimetry. It is the first time that multistrand, branched DNA structures of this complexity were constructed, completely described, and characterized thermodynamically.

No description available.

No description available.

No description available.

No description available.

The Hippo pathway is a well-conserved signaling pathway composed of a series of kinases, ending in the LATS1 and LATS2 kinases (LATS1/2), that control the activity of the transcriptional effectors TAZ and YAP (TAZ/YAP). The Hippo pathway is responsive to several external cues, including mechanical stiffness and cell-cell contact. The transcriptional targets of TAZ/YAP have a wide range of effects, including promotion of cell growth, inhibition of apoptosis, regulation of cell fate, and secretion of growth factors. Due to their wide-ranging effects, in these studies we investigated the roles for TAZ and YAP in several disease areas. First, we explored the roles of TAZ/YAP in glutamine addiction, a phenomenon in which cancer cells rely on glutamine for cell growth, in breast cancer. We demonstrated that breast cancer cells with high TAZ/YAP expression exhibit more reliance on glutamine as an energy source than those with low TAZ/YAP. Depletion of TAZ/YAP in high-TAZ/YAP breast cancer cell lines reduced their reliance on glutamine. We showed that TAZ/YAP promote increased transcription of the transaminases GOT1 and PSAT1, which allow for the carbon from glutamine to enter the tricarboxylic acid cycle and generate energy, providing a mechanism by which TAZ/YAP allow for increased processing of glutamine. Second, we explored whether and how loss of Hippo pathway activity contributes to the formation of melanoma. The most common genetic mutation found in melanoma patients is BrafV600E, but this mutation alone is insufficient to drive melanomagenesis, instead promoting an initial proliferative burst that ends in senescence. We found that BrafV600E activates the Hippo pathway to inhibit TAZ/YAP, contributing to the observed growth arrest phenotype in precancerous nevi. We found that deletion of Lats1/2 in melanocytes alone or in combination with BrafV600E expression lead to the rapid development of melanoma in mice. Third, we assessed how TAZ/YAP mediate homeostasis, injury response, and fibrosis in the lung. We explored TAZ/YAP responses in tissue regeneration and fibrosis using the bleomycin injury model in mice. We found that nuclear levels of TAZ/YAP dynamically increase in the epithelium and mesenchyme of the lung after injury; nuclear TAZ/YAP decreases in these populations as the injury resolves. We conditionally deleted Taz and/or Yap in a subset of mesenchymal cells marked by Platelet-Derived Growth Factor Receptor β (PDGFRβ) in adult mice and found that TAZ and YAP play essential roles in lung homeostasis and responses to bleomycin-injury, with TAZ/YAP-deleted animals showing reduced survival. Notable defects included disorganization of lung epithelial and endothelial cells, indicating that TAZ/YAP in PDGFRβ-expressing cells direct signals that coordinate cellular homeostasis in the lung. In total, these studies detail new mechanisms for TAZ/YAP and Hippo signaling in lung homeostasis and injury repair, melanoma development, and metabolic reprogramming in breast cancer.

No description available.