<p>ABSTRACTWILLIAMS, DANIEL. Analysis of RNase P Structure and Function. (Under the direction of James W. Brown.) The diversity of Archaea in municipal wastewater sludge was investigated by amplification of rRNA sequences from sludge DNA using archaeal-specific primers. The most common sequences were extreme halophiles; also found were sequence members of environmental euryarchaeal and crenarchaeal groups. Only distant relatives of Methanosarcina among the Methanomicrobiales were found.A detailed comparative analysis of archaeal RNase P RNA structure and a comparison of the resulting structural information with that of the bacterial RNA reveals that the archaeal RNase P RNAs are strikingly similar to those of Bacteria. The differences between the secondary structure models of archaeal and bacterial RNase P have largely disappeared, and even variation in the sequence and structure in the RNAs are similar in extent and type. The structure of the cruciform (P7-P11) has been reevaluated on the basis of a total of 321 bacterial and archaeal sequences, leading to a model for the structure of this region of the RNA that includes an extension to P11 that consistently organizes the cruciform and adjacent highly-conserved sequences. Archaeal and bacterial RNase P RNAs are very similar in sequence and secondary structure, but in the absence of protein, the archaeal RNAs are much less active and require extreme ionic conditions. In order to assess how readily the activity of the archaea RNA alone could be improved by point mutations in its sequence, in vitro selection was used to generate variants of the self-cleaving conjugant Methanobacterium formicicum: B. subtilus tRNAAsp (cpTP). Functional variants were generated with a broad spectrum of mutations that were predominately consistent with natural variation in this RNA. Variants generated from the selection were only comparable to wildtype in catalytic activity; more performed significantly faster at lower ionic strength. These results suggest that the archaeal RNase P RNA is globally optimized and that the protein may play larger compensatory roles in catalysis than previously thought. <P>

<p>RNase P is the ribonuclease responsible for the 5? maturation of precursor transfer RNA (pre-tRNA). In Bacteria RNase P is composed of a single 14kDa protein accompanied by a single catalytic RNA subunit that is capable of cleave pre-tRNA without the protein cofactor in vitro. The RNA subunit of archaeal RNase P resembles those found in Bacteria, however holoenzymes characterized from the Archaea indicate that the accompanying protein component is much larger than in the bacterial enzyme. Purified RNase P from the Euryarchaeon Methanococcus jannaschii possesses approximately eight protein subunits and has an RNA that is incapable of cleaving pre-tRNA alone in vitro. Four putative M. jannaschii RNase P proteins have been identified by their similarity to RNase P proteins of Methanothermobacter thermoautotrophicus. Here we confirm the presence of two of these proteins, MJ0464 and MJ1139, as well as the presence of several other hypothetical proteins (MJ0332.1, MJ0376, MJ1128, and MJ1625), two 30S ribosomal protein subunits (S6E and S8E), and a nicotinamide-nucleotide adenylyltransferase in M. jannaschii RNase P preparations. Eukaryotic RNase P holoenzymes also contain multiple protein subunits and it is thought that these "extra" proteins are required for function in the increased compartmentalization of the eukaryotic cell. Archaeal cells do not have the distinct compartmentalization seen in eukaryotic cells and it is unclear what purpose the protein subunits described here would have in vivo. One possibility is protein-substrate interactions to facilitate formation of the substrate-enzyme complex. M. jannaschii RNase P RNA lacks certain secondary structures present in E. coli RNase P RNA that have been shown in the bacterial model to interact with pre-tRNA substrate. Archaeal RNase P RNAs that possess these elements, such as that of M. thermoautotrophicus, are capable of cleaving substrate in vitro without protein, although only at very high salt concentration. M. jannaschii RNase P RNA does not have any extra secondary structural elements to compensate for the lack of substrate binding helices and it is possible that the protein component has evolved to assume these responsibilities. In order to test this hypothesis, circularly permuted transfer RNAs containing a photoagent positioned in the T loop were used in ultra-violet cross-linking reactions.<P>

Identification of the environmental reservoir of Mycobacterium ulcerans, the etiological agent of Buruli ulcer, within the aquatic ecosystem has been a salient research area within the last five years. Based on extensive environmental sampling and elegant laboratory models, associations have been made between the bacterial DNA and aquatic invertebrates, biofilms, plants, fish and detritus material captured on 0.2μm pore filters. These studies have suggested that M. ulcerans is widely distributed within many functional feeding groups and may be concentrated through different trophic links; however, the specific route of transmission to humans remains a mystery. In this study we have used laboratory models of infection to ascertain the role of aquatic invertebrates and fish in M. ulcerans transmission. A biologically relevant infection model in which M. ulcerans-infected mosquito larvae were fed to a species of predaceous hemiptera (African Belostomatidae) was used to demonstrate the persistent colonization of M. ulcerans and subsequent transmission of bacteria to naïve prey. The association of M. ulcerans with specific anatomical compartments showed that M. ulcerans accumulates preferentially on the exoskeleton. No difference was found between the ability of wild-type M. ulcerans and an M. ulcerans isogenic mycolactone-negative mutant to colonize belostomatids. These data show that African belostomatids can successfully be colonized by M. ulcerans and support the trophic transfer of M. ulcerans within the environment. We have shown that M. ulcerans with or without the toxin is not lethal to fish (Medaka) even at high doses following direct inoculation. Over time (23wks), infected Medaka do not exhibit any visible signs of infection or toxicity and histopathological sections do not reveal significant gross pathogenesis. M. ulcerans also appears not to replicate in infected Medaka. We also show that fish monocytes are susceptible to nanogram amounts of purified mycolactone. This is the first study to demonstrate the possibility of fish as a reservoir for M. ulcerans within the aquatic environment.

Construction of two recombinant luxI bacteriophage-based bioluminescent bioreporters was undertaken to develop detection and monitoring systems for Staphylococcus aureus and Salmonella spp. These systems take advantage of the high specificity of bacteriophage for their hosts and the Vibrio fischeri lux operon responsible for quorum sensing bioluminescence. The detection system is composed of two elements, a recombinant phage with luxI which is specific for the target pathogen, and an acylhomoserine lactone (AHL)-inducible bioreporter cell line carrying the reporter lux genes. The goal of this study was to construct Salmonella- and S. aureus-specific recombinant phages which contain luxI. The luxI expression in S. aureus was controlled by cloning the gene into S. aureus under the control of a Gram-positive promoter and ribosomal binding site (RBS). The same construct was also placed into Escherichia coli and Bacillus subtilis for comparison purposes. Although light was produced by E. coli and B. subtilis, there was no light from S. aureus. The presence of luxI transcript was shown in all strains using real time qRTPCR. Western blotting detected LuxI only in E. coli but not in B. subtilis and S. aureus. After the construct RBS was shown to be effective, the codon utilization of luxI was adapted to S. aureus phage P68 genome to eliminate a possible codon preference problem. Optimization had no effect on S. aureus while it resulted in a significant decrease of light from E. coli and a non-significant increase in B. subtilis. Additional studies are needed to determine reason(s) for luxI expression problem in S. aureus. Salmonella choleraesuis, which carries luxI under the control of Salmonella phage P22 promoter and a Gram-negative RBS, produced high levels of light. The luxI, under the control of the same promoter and RBS, was cloned into the phage P22 genome using homologous recombination. The recombinant phage P22luxI induced light from the bioreporter only when it was propagated using S. choleraesuis which harbored the homologous recombination plasmid. The bioluminescent bioreporter E. coli OHHLux was shown to produce light in the presence of P22luxI without the S. choleraesuis host. Bioluminescence tests showed P22luxI as the source of the problem. Escherichia coli OHHLux was shown not to be lysed by P22 or P22luxI and not to uptake the P22luxI genome during bioluminescence tests. P22luxI and S. choleraesuis in combination produced AHL by using an AHL bioreporter, Agrobacterium tumefaciens A136. Since A. tumefaciens A136 is not specific enough to use for quantification of AHL, another bioreporter was constructed by placing the E. coli OHHLux plasmid into Klebsiella strains. They produced light with P22luxI without the host like the previous bioreporter. Future studies should focus on constructing a bioreporter strain which does not produce light when mixed with P22luxI without its host. Optimization, and sensitivity and specificity determination studies of the system should be performed using this bioreporter strain.

Vertically aligned carbon nanofiber (VACNF) arrays have been developed as a novel tool for direct physical introduction and expression of DNA in mammalian cells (termed impalefection). This study describes the optimization of impalefection, the quantification of immobilized DNA on VACNFs, and the application of VACNFs in analysing gene expression in mammalian cells. Mechanical, chemical and biological parameters were optimized for impalefection. Alterations in a majority of the parameters resulted in no significant difference in impalefection efficiency, including nanofiber composition, DNA precipitation, cell confluency, cell concentration and sodium butyrate. The optimal DNA concentration ranged between 100 nanograms and 1 microgram, and the optimal impalefection substrate proved to be a Durx filter pad on plastic surface. High levels of efficiency in a wide range of mammalian cell lines demonstrated the versatile applicability of the impalefection method. Polymerase chain reaction (PCR) and in-vitro transcription (IVT) were used to investigate the transcriptional accessibility of immobilized DNA on VACNF arrays by correlating the yields of both IVT and PCR to that of non-immobilized DNA. Quantitative PCR was used to quantify the number of accessible yfp reporter gene copies immobilized to nanofiber arrays. DNA yields decreased dramatically in the non-immobilized control over time, while the majority of immobilized DNA was retained on VACNF arrays. These data demonstrated the development of methods for monitoring DNA immobilization techniques. To validate the applicability of VACNF arrays for controlling and monitoring mammalian gene expression, a tetracycline-inducible shRNA vector system was designed for silencing CFP expression and was impalefected into mammalian cells. VACNF arrays provided simultaneous delivery of multiple genes, subsequent adherence and proliferation of cells, and repeated monitoring of single cells over time. Following impalefection and tetracycline induction, 53.1% ± 10.4% of impalefected cells were fully silenced by the inducible shRNA vector. Additionally, efficient CFP-silencing was observed in single cells among a population of cells that remained CFP-expressing. This effective transient expression system enabled rapid analysis of gene silencing effects using RNAi in single cells and cell populations.

Mycobacterium ulcerans is the causative agent of Buruli ulcer, a necrotizing skin disease prevalent in 30 countries of West Africa and Australia. The disease begins as a painless nodule or papule that, if left untreated, can lead to ulceration that can cover much of the body. Though not usually fatal, morbidity is high, especially in rural areas where the disease is prevalent. Epidemiological evidence has linked incidence of Buruli ulcer to slow moving or stagnant waters, but despite this, the mode of transmission is not known. Evidence for M. ulcerans in aquatic environments has relied on detection based upon PCR targeting IS2404, an insertion sequence once thought to be specific for M. ulcerans. In the past few years a growing body of evidence has shown IS2404 present in a number of aquatic mycobacterial species. Although the use of PCR primers targeting the mycolactone toxin has increased the sensitivity of M. ulcerans PCR, many ulcerans-like mycobacteria also contain the mycolactone genes. In the studies presented in this body of work we applied variable number of tandem repeat (VNTR) typing methods to environmental samples in order to map the distribution of M. ulcerans in aquatic environments in Ghana both endemic and non-endemic for Buruli ulcer. Environmental samples were collected using standardized ecological methods of sampling from 2004 until 2007. VNTR profiling was also used to genotype tissue samples of patients with Buruli ulcer in an effort to link environmental samples to human cases. Finally, the ability of M. ulcerans to associate and survive within amoeba was also tested. Results from this work demonstrate the presence of M. ulcerans in both endemic and nonendemic aquatic environments in Ghana and the association with protists. Considerable M. ulcerans heterogeneity was also found between patient and environmental samples. Results also support the use of VNTR profiling for confirmation of M. ulcerans in environmental samples, and for molecular epidemiology.

CMV is the leading cause of both non-hereditary mental retardation and hearing loss and CMV infection following transplantation carries a serious risk for complications. The development of a CMV vaccine or better therapeutic treatment is desired but to develop these a more complete understanding of CMV pathogenesis is necessary. Sequence comparisons between attenuated and virulent strains of HCMV map major differences to a 15kb region (ULb’) containing the chemokine homolog, vCXCL-1. The vCXCL-1 protein of the Toledo (Tol) strain was previously shown to function in vitro as a CXC chemokine. Murine CMV (MCMV) also encodes a viral chemokine, MCK2. Prior research has shown that infection of mice with a recombinant MCMV, RM4511, which lacks a functional MCK2, resulted in decreased inflammation at the site of infection and decreased dissemination to the salivary gland. Analysis of the role of vCXCL-1 in the context of HCMV infection is limited by the species specificity of the CMVs. One possible model system for analyzing the function of this chemokine is the chimpanzee model of CMV infection. Chimpanzee CMV (CCMV) has at least one gene with similarity to the vCXCL-1 gene of the Toledo strain of HCMV. The hypothesis of this study was that vCXCL-1CCMV is a functional CXC chemokine that contributes to viral dissemination. To address this we initially compared the functional and biochemical characteristics in vitro of vCXCL-1 from Toledo HCMV (vCXCL-1Tol) and CCMV (vCXCL-1ccmv) by analyzing receptor binding, activation, chemotaxis, signaling, and anti apoptotic suppression in neutrophils. The CCMV chemokine had a ~70-fold lower affinity for human CXCR2 compared to (vCXCL-1Tol. Nevertheless, the homologs functioned very similarly and we only observed differences in integrin induction and the ability of the two chemokines to reduce neutrophil apoptosis. We evaluated the in vivo function of vCXCL-1ccmv using MCMV infection in mice. Due to the strict species specificity of CMV we used MCMV RM4511 to produce recombinant MCMVs expressing vCXCL-1ccmv or a control host chemokine, mCXCL1, under the control of the HCMV IE promoter. RMvCXCL1ccmv and RMmCXCL1 grew with similar kinetics to RM4511 in cell culture and expressed vCXCL-1ccmv and mCXCL1, respectively. Primary dissemination of RMvCXCL1ccmv and RMmCXCL1 was similar to RM4511 but neither recombinant was recovered from the salivary gland at any time point. Maintenance of this phenotype in SCID mice showed the recombinant viruses were not cleared from the salivary gland by the adaptive immune response. RMvCXCL1ccmv and RMmCXCL1 were also not recovered in neutrophil-depleted mice, although RM4511 titers were reduced in these mice indicating neutrophils may play a role in dissemination of MCMV to the salivary gland. This study is the first to characterize the CCMV viral chemokine, vCXCL-1, and show it is functionally similar to vCXCL-1Tol This provides evidence that the study of vCXCL-1ccmv may contribute to a better understanding of its HCMV homolog. Furthermore, this research is the first to evaluate the role of a CMV CXC chemokine in vivo Future work aimed at evaluating the role of other viral and host chemokines in vivo using the mouse model will likely advance our understanding of viral chemokines and their role in virus-host interactions.

G protein-coupled receptors (GPCRs) are a class of integral membrane receptor proteins that are characterized by seven-transmembrane (7TM) domains connected by intracellular and extracellular loops, an extracellular N-terminus, and an intracellular Cterminus. GPCRs recognize neurotransmitters, sensory molecules and chemotactic agents and are involved in the control of many aspects of metabolism. Since GPCRs play important roles in diverse processes such as pain perception, growth and blood pressure regulation, and viral pathogenesis, GPCRs became important target for therapeutic agents. The tridecapeptide α-factor pheromone (W1H2W3L4Q5L6K7P8G9Q10P11M12Y13) of Saccharomyces cerevisiae and Ste2p, its cognate GPCR, have been used extensively as a model for peptide ligand-GPCR structure and function. The power of yeast genetics has been used to examine the structure and function of Ste2p. Recently, GPCR homodimerization has been demonstrated for many GPCRs, although the role(s) of dimerization in receptor function is disputed. In this dissertation, Ste2p has been used to investigate GPCR dimerization. Part I of this dissertation is an overview of the GPCR structure and its ligandinduced conformational change with specific emphasis on the peptide pheromone α- factor and its receptor Ste2p. Part II of this dissertation is a study originally designed to probe inter-helical interaction between TM1 and TM7 of Ste2p. Site-directed mutagenesis and cysteine cross-linking with targeted residues of Ste2p were carried out. Although the anticipated inter-helical interactions were not identified from this study, the results provided strong evidence for Ste2p dimerization. Part III of this dissertation describes dimer interfaces including TM1 and TM7 of Ste2p. By using the disulfide cross-linking methodology, we studied the participation of specific residues at the intracellular boundary between TM1 and intracellular loop one and the entire TM7 in Ste2p dimerization. The final part of this dissertation contains a study of the participation of the Ste2p N-terminus in homo-dimer formation and the effect of ligand binding on this interaction. This part also includes overall conclusions and suggestions for future experiments that could contribute to an understanding of the dimer interfaces in Ste2p and the role of dimerization in the function of this receptor.

The Saccharomyces cerevisiae alpha-factor receptor, Ste2p, belongs to the G proteincoupled receptors (GPCRs), a class of integral membrane proteins that are characterized by seven-transmembrane (TM) domains. Ste2p-alpha-factor pair has been used extensively as a paradigm for investigating GPCRs structure and function. Upon binding of alpha-factor to Ste2p, a signal is transduced via an associated guanine-nucleotide binding protein, Gpa1p, initiating a cascade of events similar to those for mammalian GPCRs signal transduction. GPCRs are essential in many physiological processes associated with human diseases. Many aspects of structure and function are highly conserved across GPCRs, irrespective of primary amino acid sequence. This dissertation investigated the interactions of Ste2p with alpha-factor and Gpa1p. An overview of GPCRs in general with specific emphasis on Ste2p interactions with alpha-factor and Gpa1p are discussed in part I. Cross-linking studies of alpha-factor analogs containing 3,4-dihydroxylphenylalanine (DOPA) at positions 1 and 13 indicated that Trp1 and Tyr13 of α-factor are in close proximity to Lys269 and Cys59 of Ste2p, respectively when alphafactor is bound to Ste2p. An alpha-factor synergist lacking the last two amino acids required for binding could only inhibit the cross-linking of DOPA at position 1 suggesting that the alphafactor synergist interacts with the N-terminus of alpha-factor (described in part II). Part III describes the first report of an unnatural amino acid, p-benzoyl-L-phenylalanine (Bpa), replacement in a GPCR expressed in its native environment, and the use these receptors to photocapture a peptide ligand. Many of the Bpa-substituted Ste2p receptors exhibited biological activity and two of them, Phe55-Bpa and Tyr193-Bpa, were able to selectively capture alphafactor in the ligand binding site after photoactivation; indicating that these residues may be in direct contact or in close proximity to alpha-factor when bound to Ste2p. Part IV reports for the first time the involvement of the third intracellular loop (IL3) in Ste2p homo-dimer formation, and also conformational changes at the cytoplasmic ends of TM5-TM6 of Ste2p induced by alpha-factor binding. Conformational changes in the C-terminus of Gpa1p occurring during Ste2p and Gpa1p activation are also discussed for the first time in part V. Variants of Ste2p- Gpa1p fusion proteins that displayed activities similar to Ste2p and Gpa1p are described in part VI. The final part of this dissertation discusses the overall conclusions and contains suggestions for future studies. The results obtained during this study should provide very important information about the mechanisms underlying the activation of GPCRs and G proteins.

The inauguration of the International Human Genome Initiative in the later part of the 1980’s coincided with the development of scanning probe microscopy (SPM). SPM was a good fit as one of the new technologies that might be implemented to sequence or map DNA and perhaps make a major contribution toward the goal of sequencing the entire human genome. Although the scanning tunneling microscope (STM) was invented in 1982 [Binnig 1982] and the atomic force microscope (AFM) in 1986 [Binnig 1986], it was not until 1987 that the first STM became commercially available; the AFM became available in 1989. Our group entered this exciting new scientific adventure in genome research towards the end of 1987. An interdisciplinary team was assembled that capitalized on our expertise in the fabrication and use of scanning tunneling microscopes. Additionally, we purchased one of the first commercially available scanning tunneling microscopes. Rounding out the team was expertise in imaging biomolecules such as DNA using electron microscopy. Our initial research focused on STM imaging and scanning tunneling spectroscopy (STS) of 1) tobacco mosaic virus adsorbed to gold surfaces [Mantovani 1990] and 2) DNA passively mounted on highly ordered pyrolytic graphite (HOPG) surfaces [Allison 1990]. Our spectroscopy results with DNA encouraged speculation that the electronic signatures of nucleotide bases might be used to sequence DNA. However, in this early work we also discovered that simply adsorbing either virus or DNA to gold or HOPG surfaces resulted in substantial amounts of the sample being removed by the STM tunneling tip. In order to more firmly immobilize negatively charged DNA molecules onto surfaces for STM scanning, we created gold sample surfaces with positive functionality mediated by a self-assembled monolayer of 2-dimethylaminoethane thiol. Using this method, we produced the first reported images of entire genetically functional plasmid DNA molecules obtained by STM [Allison 1992a, Allison 1992b Allison 1993, Bottomley 1992]. Efforts initiated by our laboratory to restriction map DNA molecules by AFM imaging were successful. This was accomplished by physically mapping the location of a mutant EcoRI endonuclease that binds to but does not cleave large DNA clones. Our new AFM technology was pioneered as an alternative to conventional gel-based restriction mapping; it was first demonstrated on plasmid DNA molecules [Allison 1996] and later on larger molecules including cosmid clones [Allison 1997]. This technology should prove to be more effective than conventional mapping methods because by using AFM mapping neither the number nor the proximity of restriction sites to one another is problematic.