Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) offer an adaptive immune system that protects bacteria and archaea from nucleic acid invaders through an RNA-mediated nucleic acid cleavage mechanism. Our knowledge of CRISPR nucleic acid cleavage mechanisms is limited to three examples of widely different ribonucleoprotein particles that target either DNA or RNA. CRISPR Type I and II silencing complexes have been well studied and shown to target DNA using via a complex named Cascade and the single protein Cas9 respectively. Type III systems can target both DNA and RNA but the mechanism is still not well understood. The work presented here focuses on the silencing complexes in Type III CRISPR systems and shed some light on the activity of the largest subunit of these complexes. Staphylococcus epidermidis belongs to the Type III-A CRISPR system and has been shown to interfere with invading DNA in vivo. The Type III-A CRISPR system is characterized by the presence of Csm1, a Cas10 family of proteins member, that has a permuted histidine-aspartate (HD) and a nucleotidyl cyclase-like domain, both of which contain sequence features characteristic of nucleases. In chapter 2, we show in vitro that a recombinant S. epidermidis Csm1 cleaves single-stranded DNA exonucleolytically in the 3'-5' direction and in a divalent-metal dependent manner. We further showed that its DNA cleavage activity resides in the GGDD motif of the cyclase-like domain rather than the HD domain. Our data suggest that Csm1 might work in the context of an effector complex to degrade invading DNA. Type III-B Cmr complex from Pyrococcus furiosis has been shown to target RNA. Cmr2 is the largest subunit of the complex, and belongs to the Cas10 family; its domains organization is comparable to Csm1. In chapter 3, we present a structural and functional study of Cmr2 showing that Cmr2 is not the catalytic site of the Cmr complex for the RNA-guided RNA cleavage. However, exciting results of a DNA cleavage activity by Cmr2 suggest a possible dual silencing of DNA and RNA by type III-B systems. Taken in total, the work presented in this dissertation provides insights into the silencing mechanism of the effector complexes in viral defense and highlights the role of their largest subunit, the signature protein Cas10, in type III CRISPR systems. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2013. / June 26, 2013. / Includes bibliographical references. / Hong Li, Professor Directing Dissertation; Myra Hurt, University Representative; Scott Stagg, Committee Member; Thomas C. S. Keller, III, Committee Member; Fanxiu Zhu, Committee Member.

Troponin C (TnC), the Ca2+ binding subunit in striated muscle, is central to regulation of contraction. Binding of Ca2+ to TnC results in a series of conformational changes in the different regulatory proteins, which ultimately leads to muscle contraction. From structural studies about Troponin C, we hypothesize that binding of divalent cations causes a closing in the molecule, bringing the ends of TnC closer together. We designed a series of novel FRET constructs using human cTnC, to examine the relative positions of the N- and C-termini upon divalent cation binding. Full length cTnC was flanked by FRET pairs of fluorescent proteins (mCerulean/mVenus, mTurquoise/mNeonGreen, mTurquoise/cpVenus), varying the linker length between TnC and the FRET proteins. FRET, as quantified by changes in the fluorescence ratio (FR) of acceptor to donor, was measured in the presence and absence of saturating Ca2+ and/or Mg2+. FRET increased substantially and reversibly upon Ca2+ binding to cTnC. Similar FRET changes were observed upon saturation of the C-terminus with Mg2+, suggesting that the structural changes detected are primarily attributable to occupancy of the C-terminal sites. We chose to further characterize our most successful construct CTV-TnC by measuring conformational changes by AUC. We also confirmed the functionality of our construct with skinned fiber force measurements. In addition, we have conducted FRET and AUC experiment with mutant TnC where one or more EF hand was inactivated, to extract the contribution of each EF hand to the total signal. Analytical ultracentrifugation (AUC) confirmed that constructs undergo global conformational changes to a more compact structure upon Mg2+ binding, with further compaction when Ca2+ occupies all 3 sites of cTnC. Fiber experiments have shown that CTV-TnC is capable of reconstituting into the sarcomere and generating force under activating conditions, with a slight decrease in Ca2+ sensitivity. Finally, we fully titrated CTV-TnC with Ca2+ and Mg2+ to determine affinities of the divalent cation to the EF hands, using the FR signal. The full divalent cation titration generally agrees with known affinities of WT TnC. Finally, studies of the mutant TnC show that the N-terminus is associated with a negative FRET signal, which requires more experiments to be fully understood. The C-terminal sites show a difference in contribution when bound to Ca2+ vs. Mg2+, with the highest contribution of site III upon Ca2+ binding, and the highest contribution of site IV upon Mg2+ binding. In summary, we have successfully designed a Ca2+ sensor capable of indicating Ca2+ binding through large changes in FRET signal. This sensor is capable of reconstituting into troponin complex, as well as into sarcomeres, making it useful to the muscle research community. The FRET and AUC studies show that the ends of cTnC come closer upon binding divalent cations yielding a more compact structure, for CTV-TnC alone and within the troponin complex. Our study also indicates a difference in site III and IV, known as the C-terminal sites, in the contribution to the FRET signal when bound to each of the divalent cations Ca2+ or Mg2+. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2013. / June 26, 2013. / Cardiac muscle, FRET, Troponin C / Includes bibliographical references. / P. Bryant Chase, Professor Directing Dissertation; Michael Overton, University Representative; Piotr Fajer, Committee Member; Thomas C. S. Keller, III, Committee Member; Geoffrey Strouse, Committee Member.

Proteins are the most versatile macromolecules in the cell. Major biochemical functions of proteins include binding; catalysis; and serving as structural components of cells and organisms. Proteins are made of amino acids that are linearly connected by peptide bonds. The linear protein sequence, determined by a gene, can fold into higher-level structures. Proper protein structure/conformation is important to maintain its correct biological function. Generally used methods to determine protein and protein complex conformations are X-ray crystal diffraction and high resolution solution phase nuclear magnetic resonance (NMR), because both of them provide the highest sequence resolution for proteins. However, both methods require large (milligram) quantities of proteins. X-ray diffraction requires large crystals formed from protein solution, but it is usually impossible to form crystals for some proteins, e.g., proteins with very flexible regions. NMR requires high sample concentration (often causing solubility problems) and is limited by the size of the protein. Hydrogen/deuterium exchange coupled with mass spectrometry (HDX MS) has become a powerful tool to study protein and protein complex conformations, at biological concentrations (usually at pico-molarity level). High-resolution mass analysis (such as Fourier transform ion cyclotron resonance mass spectrometer) is particularly advantageous for the HDX MS method, by accurately assigning the peptide fragments after protease digestion and in resolving overlap of peptide isotopic distributions, both before and after HDX MS. Chapter 1 introduces the theory of the instrument for mass measurement, ionization methods for biological samples. An introduction of several biophysical methods to study protein conformation and protein-ligand interactions is also covered. The last portion of this chapter gives an overview of the HDX MS method, its history, development and applications. We have been continuously optimizing the HDX MS methods in our lab. To investigate conformational change of a protein, same proteolyzed fragments from two states of a protein are compared. Pepsin has previously been the enzyme of choice for such experiments because of its activity at low pH and its broad specificity (yielding a wide variety of peptides). However, incomplete cleavage of some proteins and limited number of proteolytic peptides has restricted its range of applicability. In Chapter 2, a new enzyme, protease type XIII from Aspergillus saitoi, is characterized. This enzyme prefers to cleave on the C-terminal end of basic amino acids (i.e., histidine, arginine and lysine). Under positive-ion electrospray conditions, basic amino acid side chains are protonated, to yield abundant positive ions during ESI, resulting in increased signal-to-noise ratio for protease type XIII cleaved fragments. Increased signal-to- noise ratio is especially useful for deuterium incorporation assignment for HDX analysis, due to reduction in signal for ion isotopic distributions after deuterium incorporation. This enzyme is also proved to have much less self-digestion than pepsin, which results in less interrupting fragments from the enzyme itself. These results were recently published (Zhang H.-M. et al., Anal. Chem., (2008), 80 (23), 9034-9041.) Back exchange is the single biggest challenge for HDX MS and results in loss of information, because with back exchange, a fast exchanged D quickly exchanges back to H after quench and appears to be no exchange. We have been addressing the back exchange problem by reducing H2O content during peptides separation with salts. We replaced high performance liquid chromatography (HPLC) (H2O mobile phase) by on-line supercritical fluid (a state between gas and liquid) chromatography (SFC). The mobile phase (to wash off salts and peptides at different times for their separation) of SFC is supercritical fluid CO2, which cannot contribute to hydrogen back-exchange due to lack of H. SFC greatly reduced back exchange, however, robustness of the columns and efficiency of the SFC separation are being improved. In parallel with SFC, Chapter 3 describes a fast liquid chromatography method to reduce back exchange. Back exchange is reduced by at least 25% compared to conventional LC separations in the HDX community. A sham digestion method is employed to prove that no backbone hydrogen back exchange occurs during digestion and it mainly occurs during following LC separations. A manuscript describing these results has been published in J. Am. Soc. Mass Spectrom. (Zhang H.-M. et al. J. Am. Soc. Mass Spectrom., 20, 520-524 (2009). Our entire HDX MS experiment has been automated by a Leap robot. All our experiments are done in triplicate to verify reproducibility. This automation generates a large amount of data, usually more than 110, 000 peaks for a 50, 000 Da protein. The data amount is even larger for a larger protein or protein complex. Therefore, it is a big challenge to analyze the data efficiently and accurately. A program (written in Visual Basic) automatically identifies all digested fragments and determines deuterium incorporation time-course profiles for each proteolytic fragment (manuscript submitted to JASMS). Chapter 4 describes a Python program to sort data for each fragment under different conditions, and plot time-course deuterium uptake profiles and bar graphs for comparison. This programming demonstrates dramatically improved analysis accuracy and efficiency (analysis time reduced from several days to less than 10 min!). Certain cancers such as gastrointestinal stromal tumors (GISTs) exhibit elevated expression and/or activating mutations of the receptor tyrosine kinase, KIT. The efficacy of some small-molecule inhibitors (e.g., sunitinib) in GIST patients is thought to be based on the ability of the drug to inhibit the KIT proteins. Acquired resistance to systemic therapy is a critical problem in treatment of metastatic cancers. Like many kinases, KIT has a flexible kinase insertion domain (KID). The presence of the KID region inhibits crystal formation of KIT protein. All X-ray crystallography has been performed in the absence of the KID region. Chapter 5 discusses the use of solution-phase HDX experiments to verify that the KID has no conformational effect on KIT either in the presence or absence of sunitinib. By comparing the conformation of a sunitinib resistance mutant (D816H) with wild-type KIT, the drug resistance mechanism for KIT is elucidated. A manuscript of our HDX results along with extensive X-ray crystallography and enzyme kinetics measurements has been published on Proc. Natl. Acad. Sci. (Proc. Nat. Acad. Sci. 106 (5), 1542-1547 (2009)). HDX studies of the drug imatinib and several other KIT mutants (with different sensitivity to imatinib and sunitinib) also revealed different molecular mechanisms of sunitinib and imatinib binding and inhibition, and provided insight for future anit-GISTs drug design to specifically target the resistance. A manuscript of these results will soon be submitted to Protein Sci. The RAGE (Receptor for Advanced Glycosylation End Products) protein has been proposed to be involved in the active transport of beta-amyloid protein across the blood-brain barrier, which is thought to be one of the major contributing factors of Alzheimer's disease. Consequently, RAGE is being targeted to prevent the passage of amyloid beta. The wild type RAGE protein is glycosylated. Two expression systems are currently being used to engineeringly produce sRAGE (extracellular soluble portion of RAGE): BacMam (insect cells) and E. coli (bacterial cells). BacMam- expressed sRAGE has sugars on it (glycosylated), because it is the same as the human RAGE. E. coli expressed RAGE is not glycosylated. Both constructs of RAGE are difficult to concentrate and thus can only be studied at low concentrations. Due to the glycosylation(s) and low solubility, RAGE is not suitable to be studied by X-ray crystallography or NMR. Chapter 6 discusses use of solution-phase HDX to study the conformational changes in RAGE induced by the presence of glycosylation. A region of binding beta-amyloid was found to be more flexible in the BacMam expressed RAGE than the E. coli expressed RAGE due to the opening effect by glycosylation on BacMam-RAGE. This change in structure has great potential to facilitate the design of therapeutics. A manuscript of these results will soon be submitted to J. Mol. Biol. L-leucine is an essential amino acid that can be synthesized by Mycobacteria, which cause the deadly disease tuberculosis (TB), but not mammals. Humans need to obtain this amino acid from food. The enzyme, '-isopropylmalate synthase from Mycobacteria tuberculosis (MtIPMS) is essential to produce leucine for mycobacteria. Inhibition of the enzyme activity of IPMS helps to kill Mycobacteria, therefore MtIPMS is a therapeutic target for treatment of TB. As an effector, L-leucine binds on the regulatory domain and causes allosteric change on the active site of MtIPMS. Chapter 7 talks about use of HDX MS to map the allosteric regulation pathway in the IPMS protein upon L-leucine binding. An L-leucine - insensitive mutant Y410F MtIPMS was used to verify the inhibition transduction pathway. These results demonstrate one of the first reports of an experimentally mapped allosteric mechanism in a protein of this size. These results will facilitate the design of new anti-TB therapeutics. A manuscript of these results has been recently submitted to Biochemistry. The appendixes include the papers I have published during my graduate study. Appendix A is a publication that covers the collaboration work with Dr. Anke Meyer-Baese (FSU School of Engineering) and Ernst Althaus (Max-Planck Institute) to develop probability analysis algorithms (Althaus et al., Symposium on Applied Computing, 2008, 1273). This research has great potential to resolve H/D exchange rate information for single amino acids or small local regions of the peptide, and thus greatly improve sequence resolution for HDX MS. Appendix B is the paper we published in Analytical Chemistry, covering the characterization of the protease type XIII, which is proven to be superior to pepsin. Appendix C is the paper we published in J. Am. Soc. Mass Spectrom., which covers the use of fast reversed-phase LC to reduce back exchange for HDX MS. Appendix D is the paper we published in Proc. Natl. Acad. Sci., which is collaborated with Pfizer, La Jolla for the application of HDX MS to elucidate the drug resistance mechanism in Gastrointesinal Stromal Tumors (GISTs) patients. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2009. / June 15, 2009. / Includes bibliographical references. / Alan G. Marshall, Professor Directing Dissertation; Igor Alabugin, Outside Committee Member; Mark R. Emmett, Committee Member; Timothy M. Logan, Committee Member; Michael Roper, Committee Member; Hengli Tang, Committee Member.

The Clustered Regularly Interspaced Short Palindromic Repeat loci found in most archaea and some bacteria contain DNA sequences (spacers) that originate from genetic invaders like viruses, transposons, and plasmids. The CRISPR clusters are transcribed into RNA and then processed into short guide CRISPR RNAs (crRNA) that are incorporated into ribonucleoprotein complexes to recognize invaders through complementary base-pairing. The Cmr complex of Pyrococcus furiosus, is an example of a ribonucleoprotein effector complex that uses crRNAs to recognize and cleave target RNA. This complex is composed of six subunits Cmr1-6 that can use diverse crRNAs of 39nt or 45nt lengths to recognize and destroy diverse target RNA sequences. The RNA cleavage activity of the Cmr complex follows a ruler mechanism by which the cleavage occurs at the 14th nucleotide form the 3' end of the crRNA in a metal dependent manner. Although the biological function of the Cmr complex is now well understood, the mechanistic details of its activity such as how the complex assembles and the roles of the different subunits, particularly the identity of the catalytic site, remain unknown. This work uses biochemical assays of RNA cleavage, in-vitro assembly studies, and structural studies to address those questions. I first addressed the role of the largest subunit of the Cmr complex Cmr2 alone and in complex with another subunit, Cmr3. Initial predictions suggested that Cmr2 may harbor the active site of the complex. However, through structural and mutagenesis studies I showed that Cmr2 does not play a direct role in the RNA-cleavage catalysis of the Cmr complex. The interaction between Cmr2 and Cmr3 results in a highly positively charged region between the two proteins that contains the nucleotide-binding site of Cmr2, as determined by solving the structure of Cmr2-Cmr3. Although Cmr3 contains multiple conserved structural elements and potentially catalytic residues, mutagenesis showed that it does not play a role in cleaving target RNAs either. To further characterize the assembly of the Cmr complex and determine the roles of its subunits, I worked in collaboration with Michael Spilman from the Scott Stagg laboratory. We obtained a low-resolution structure of the crRNA- and target RNA-bound Cmr1-6 complex. The results show that the complex has a helical architecture comprised of a Cmr2-3 foot, a Cmr4-5 twisted ladder composed of 3 Cmr4-5 steps, and a Cmr6-Cmr1 head. The crRNA-target RNA duplex binds vertically along the length of the Cmr complex. The results corroborated the previous findings on the roles of Cmr2 and Cmr3 as they showed that the two proteins are involved in specific crRNA binding. While the nature of the complex's active site remains elusive, the results provide structural support for the ruler mechanism of catalysis of the Cmr complex. The helical architecture of the complex revealed an unexpected similarity to the type I CRISPR effector complexes and suggested potential functional and structural similarities among all CRISPR effector complexes. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2013. / June 28, 2013. / adaptive immunity, CRISPR, interference, prokaryote, pyrococcus furiosus, RNA / Includes bibliographical references. / Hong Li, Professor Directing Dissertation; Joseph Travis, University Representative; Scott Stagg, Committee Member; Jamila Horabin, Committee Member; Wu-min Deng, Committee Member.

Nucleosome distributions are critically important in regulating access to the eukaryotic genome. Cells with different physiologies have strikingly similar nucleosome distributions. Few studies in human cells have measured genome-wide nucleosome distributions at high temporal resolution during a response to a common stimulus. Factors regulating the maintenance of the basal state as well as changes in nucleosome distribution following a response must be investigated. In our first set of experiments we used the reactivation of Kaposi's sarcoma-associated herpesvirus (KSHV) as a model system for stimulus-induced nucleosome distribution changes. We measured nucleosome distribution at high temporal resolution in human cells at the 2 kb flanking the transcription start sites (TSSs) of hundreds immunity-related loci, using microarray technology, during the reactivation of KSHV. We show that nucleosome redistribution peaks at 24 hours post KSHV reactivation and that the nucleosomal redistributions are widespread and transient. To clarify the role of DNA sequence in these nucleosomal redistributions, we compared the genes with altered nucleosome distribution to a sequence-based computer model and in vitro assembled nucleosomes. We demonstrate that both the predicted model and the assembled nucleosome distributions are concordant with the majority of nucleosome redistributions at 24 hours post KSHV reactivation. We suggest a model in which loci are held in an unfavorable chromatin architecture and "spring" to a transient intermediate state directed by DNA sequence information. We propose that DNA sequence plays a more considerable role in the regulation of nucleosome positions than was previously appreciated. The surprising findings that nucleosome redistributions are widespread, transient, and DNA-directed shift the current perspective regarding regulation of nucleosome distribution in humans. We next wanted to affirm and extend our previous observations regarding the widespread and transient nature of nucleosome redistributions during viral reactivation. We tested if this widespread nucleosome remodeling was a genome wide event or limited solely to the hundreds of immunity-related loci measured by microarray. We measured nucleosome distributions at high temporal resolution following KSHV reactivation using our newly developed mTSS-seq technology, which maps nucleosome distribution at the TSS of all human genes. Nucleosomes underwent widespread changes in organization 24 hours after KSHV reactivation and returned to their basal nucleosomal architecture 48 hours after KSHV reactivation. 72% of the loci with translationally remodeled nucleosomes have nucleosomes that moved to positions encoded by the sophisticated underlying DNA sequence. We demonstrated that these widespread alterations in nucleosomal architecture potentiated regulatory factor binding. These descriptions of nucleosomal architecture changes have allowed us to propose a new hierarchical model for chromatin-based regulation of genome response. Given that we discovered that nucleosome distributions are widespread and transient, it was important for us to understand the forces maintaining the basal state. We It would be interesting to understand the forces and factors that maintain nucleosome architecture in a basal state and regenerate it following a response, such as KSHV reactivation. An appealing group of candidates that might maintain and regenerate the nucleosome architecture in its basal state is the transcriptional machinery. We identified RNA polymerase II (RNA Pol II) as a likely candidate as it is found throughout the genome and not always associated with transcription. We next were interested in the role RNA Pol II plays in the maintenance of chromatin structure. We measured nucleosome distributions in response to RNA Pol II inhibition by δ-amanitin treatment. Nucleosome distribution changes, following RNA Pol II inhibition, were widespread and the TSSs with nucleosome distribution changes were enriched for RNA Pol II independent of it's role it plays in active transcription. This work gives new insight into understanding the role of chromatin structure regulates and maintains the human genome. / A Dissertation submitted to the Department of Biological Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester, 2015. / April 9, 2015. / Chromatin, KSHV, microarray, next generation sequencing, nucleosome, RNA Pol II / Includes bibliographical references. / Jonathan H. Dennis, Professor Directing Dissertation; Michael G. Roper, University Representative; Hank W. Bass, Committee Member; P. Bryant Chase, Committee Member; Debra A. Fadool, Committee Member.

Using the fruit fly Drosophila melanogaster as a model organism, this dissertation dissects molecular and biological functions of three proteins: Rootletin (Root), testis-specific Centrosomin (CnnT) and Spermitin (Sprn). Centrosomes are the major microtubule-organizing centers (MTOCs) in animal cells and each one consists of a pair of centrioles, a mother and a daughter centriole, surrounded by the pericentriolar material (PCM). Centrosomes play critical roles in cell division, cell polarization and intracellular trafficking, etc. In some cell types, the mother centriole matures to organize the primary cilium that is important for cell signaling and sensory perception. Cilia are linked at their base to the cell body by a cytoskeletal structure called the rootlet, whose detailed functions remain largely unknown. We have identified a protein called Rootletin (Root) that localizes at the rootlet in ciliated neurons. Root is essential for normal neuron-specific behaviors of the flies, including locomotion, mechanosensation, chemosensation and hearing. Furthermore, ultrastructure studies revealed that Root is required for organizing the rootlet, and we found that rootlet assembly is centriole-dependent. Altogether, we will define the important role of Root in rootlet organization and its requirement for ciliary functions. Mitochondria are energy centers in cells. In Drosophila, they also participate in sperm tail elongation by providing a structural platform for microtubule (MT) organization to support the elongating tail. Centrosomin (Cnn) has several variants, its centrosomal forms (CnnC) are essential for functional centrosomes. We discovered that the other non-centrosomal class of Cnn splice products (CnnT) is mainly expressed in fly testes. And unlike CnnC, which localizes at centrosomes, CnnT localizes to spermatid mitochondria (nebenkern). Cell culture and in vivo studies show that CnnT is necessary and sufficient to recruit MT-nucleating factors to nucleate MTs on mitochondria. Our study also suggested CnnT is required for normal sperm elongation and male fertility. Overall, we propose that CnnT promotes assembly of unique MTOCs on the surface of mitochondria where, in elongating sperm cells, it facilitates sperm tail growth. In the last chapter, we present our study on a novel testis-specific mitochondrial protein named Spermitin (Sprn). Using fluorescence microscopy, we found that Sprn is a mitochondrial protein localizing in the matrix of the spermatid nebenkern. However, Sprn is dispensable for normal sperm development and male fertility. / A Dissertation submitted to the Department of Biomedical Sciences in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester, 2014. / October 30, 2014. / centrosome, cilia, mitochondrion, MTOC, rootlet, spermatogenesis / Includes bibliographical references. / Timothy L. Megraw, Professor Directing Dissertation; Branko Stefanovic, Committee Member; Yanchang Wang, Committee Member; Yi Zhou, Committee Member.

Small helical membrane proteins are functionally very important. They are involved in many cellular transport and signal transduction pathways maintaining the chemical, mechanical and electrical potentials of cells. They constitute approximately 30% of all expressed genes in bacteria, yeast and in the human genome yet, less than 1% of the protein structures in the Protein Data Bank (PDB) are integral membrane proteins. For the Mycobaterium tuberculoisis genome, approximately 1100 genes were predicted to code for α-helical integral membrane proteins and 60% of them were predicted to be small helical membrane protein containing one, two or three helices. The structural characterization of small helical membrane proteins is particularly challenging, since their tertiary structural stability is low due to the hydrophobic amino acid composition and the uniform low dielectric environment of the membrane interstices resulting in only vander Waals interactions and a few weak electrostatic interactions to stabilize the structure. Increasingly, it is recognized that this class of proteins needs to be characterized in a sample environment that accurately reflects the properties of the native membrane environment, such as lipid bilayers. In this thesis, I describe the detailed development of small helical membrane protein over expression, purification, reconstitution and their structure characterization methods in lipid bilayers by studies of three membrane proteins from M. tuberculosis. The three membrane proteins were CwsA (Rv0008c), CrgA (Rv0011c) and FtsX (Rv3101c), all of which are associated with cell division and the three dimensional structure was determined for CrgA by solid state NMR spectroscopy. Much effort has been devoted to the development of Oriented Sample (OS) and Magic Angle Spinning (MAS) ssNMR technology for characterizing membrane protein structures with atomic resolution in lipid environments. But the combined application of the two methods has become very popular recently. This thesis describes the advantages and application of this combined approach by determining the tertiary structure of the M. tuberculosis CrgA transmembrane domain in liquid crystalline phospholipid bilayer at atomic resolution. Anisotropic orientational restraints and sparse distance restraints were used to calculate the initial CrgA structure followed by a restrained molecular dynamics simulation refinement process in explicit lipid bilayers. As a collaborative work, the function of full length CrgA membrane protein was pursued in vivo and in vitro. Protein-protein interactions with CrgA, CwsA and FtsZ (a water soluble protein) were performed using a pull down assay in a nickel affinity column. All together an analysis of CrgA structure and function study is presented that sheds light on how FtsI and CwsA may bind to CrgA. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester, 2014. / March 20, 2014. / CrgA, CwsA, FtsX, Membrane protein, M.tuberculosis cell division, Solid state NMR / Includes bibliographical references. / Timothy A. Cross, Professor Directing Dissertation; Naresh Dalal, University Representative; Timothy M. Logan, Committee Member; Bryant P. Chase, Committee Member; Brian G. Miller, Committee Member.

Helical membrane proteins have been difficult targets for structural characterization but are ~30% of the typical genome. This dissertation is focused on the extension of solid state nuclear magnetic resonance methodology to be able to routinely obtain structural restraints and calculate high resolution structures for membrane proteins in a bilayer environment. Rv1861, a M. tuberculosis protein involved in the regulation of transglycosylase activity is used as a model system. One bottleneck for obtaining structure restraints is the preparation of high quality proteoliposome samples. Rv1861 is readily purified in one detergent but the same detergent fails to produce samples of sufficient quality for structural studies. Based on the physical properties of the original detergent, new detergents were selected and a screening process was devised which improves the reconstitution into liposomes and results in high quality samples. To date, a general procedure for optimizing the detergent used in the preparation of proteoliposome samples of helical membrane proteins for solid state nuclear magnetic resonance has yet to be published. Another bottleneck for obtaining structural restraints is the assignment of resonances. For both oriented sample and magic angle spinning techniques methods are devised that allow certain crucial assignments to be made. Enough assignments are obtained allowing the measurement of a limited set of structural restraints for Rv1861. The restraints obtained from both techniques are used together resulting in a high quality structure using fewer restraints than required using either technique alone. The methods illustrated here for Rv1861 are generally applicable to all membrane proteins and are expected to drastically increase the rate at which atomic resolution structures are obtained for helical membrane proteins in a lipid bilayer environment. / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester, 2014. / March 10, 2014. / Helix, Lipid Bilayer, Membrane Protein, solid state NMR, Structure, Tuberculosis / Includes bibliographical references. / Timothy Cross, Professor Directing Dissertation; Rufina Alamo, University Representative; Brian Miller, Committee Member; Rafael Bruschweiler, Committee Member; Debra Fadool, Committee Member.

Myosin V, as a member of the myosin family, can "walk" along F-actin to transport many "cargos", such as mRNA and secretary vesicles, to its destination. This study is part of a research project that aims to understand the interaction between myosin V with F-actin through direct visualization of biochemical states of myosin V that bind weakly to F-actin and are thus conformationally heterogeneous. Because of its processivity, it's a good model to illustrate how myosin interacts with F-actin to fulfill its function (Chapter 1). Specimens of full-length myosin V with F-actin were cryogenically vitrified to maintain the close-to-nature conformation. ADP·Pi was added into the solution to keep myosin V in the inhibited state in which the heads are folded back onto the cargo binding domain thereby preventing both heads from binding to a single actin filament at the same time. Electron Tomography rather than single-particle method was used because the relatively low affinity of inhibited myosin V to F-actin which makes the decoration far away from saturation. With subsequent sub-volume processing, ET can provide molecular-level information for relatively heterogeneous and/or sparse macromolecule complexes (Chapter 2). New strategies and methods for processing tomographic sub-volumes with heterogeneity were designed to extract homogenous and meaningful class-averages. F-actin repeats with or without myosin V decoration were extracted to do the sub-volume averaging. Multivariate Data Analysis (MDA) and cluster analysis were used to deal with the heterogeneity issue. More homogeneous repeats were clustered into the same classes and corresponding class-averages were generated to improve signal-to-noise ratio (SNR). Repeats with myosin V decoration were later identified and grouped together to get the conformational information. Focused classification was used to further separate different conformations of the bound myosin V (Chapter 3). The new data processing methods present much conformational information of inhibited myosin V on F-actin. The enumeration of bound myosin V shows that inhibited myosin V mainly is bound to F-actin with only one head. Analysis of the binding angle of the lever-arm with respect to the F-actin filament shows the lever-arm angle of most bound myosin V is not close to 90º, which means that most bound myosin might be not in the transition state. This could be due to Pi release after myosin V binds to F-actin such that Myosin V is in the rigor-like state, even though ADP·Pi is present in the solution. However, for some bound myosin V, the lever-arm angle is really close to 90º, which means the lever-arm is in the "up" position or that the myosin head is binding to actin weakly in a previously unidentified orientation. One explanation is that even though the binding to F-actin could accelerate the release of Pi, some bound myosin V might be still frozen with ADP·Pi during fast-freezing. In addition, copies of double-head bound myosin V were found. The existence of double-head binding requires that the two heads be bound to actin in different conformations; this could be because of random Brownian motion of the second head and subsequent collision. Further quasi-atomic models were built and docked into class-averages to determine the possible nucleotide condition and provide structural information beyond the resolution of the density map. The improved procedure, with quantitative analysis and simulated data verification provides integrated and detailed structural information of inhibited myosin V on F-actin (Chapter 4). Here we identified and characterized the conformation of inhibited myosin V bound to the F-actin filament. We found most inhibited myosin V bound to F-actin with one head. Also we found the lever-arm of most myosin V could be in a rigor-like position rather than the transition position. In addition to single head binding, we found some copies of double-head binding, one head in rigor-like binding state and the other in the transition state. It's the first time to obtain the molecular-level structural information of inhibited myosin V on F-actin, which could fix the missing structural gap for the myosin V ATPase cycle, thus helping us better understand the mechanism of myosin (Chapter 4). / A Dissertation submitted to the Institute of Molecular Biophysics in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2014. / July 18, 2014. / ADP.Pi, Conformation, Cryo-ET, F-Actin, Myosin V / Includes bibliographical references. / Kenneth A. Taylor, Professor Directing Dissertation; Xiaoqiang Wang, University Representative; Hong Li, Committee Member; Thomas C. S. Keller, III, Committee Member; Scott Stagg, Committee Member.

In eukaryotes, each individual chromosome is one large DNA molecule packed by histone proteins into a compact nucleoprotein filament. Two molecules each of core histone proteins H2A, H2B, H3 and H4 assemble to form an octamer protein core around which 147 base pairs of DNA is wrapped to form the nucleosome core particle and this structure is repeated to form chromatin. Histones are essential proteins as they package the genomic DNA to fit it inside the relatively tiny nucleus and regulate DNA accessibility. However, when present in excess, the positively charged histones can bind non-specifically to negatively charged DNA and affect all forms of DNA metabolism such as transcription, replication, repair and recombination. The DNA of all organisms is under constant threat of damage from both exogenous and endogenous agents that can contribute to genomic instability, which is characterized by the increased rate of acquisition of alterations in the genome and is a hallmark of cancer cells. Hence, cells have evolved multiple mechanisms to ensure genomic stability. Since DNA damage and repair occurs in a chromatin context in eukaryotes, chromatin structure and histones may affect genomic stability. Not surprisingly, scarcity of histones during DNA replication results in spontaneous DNA damage. On the other hand, accumulation of excess histones leads to genomic instability in the form of excessive chromosome loss, enhanced sensitivity to DNA damaging agents and cytotoxicity. Therefore, histone synthesis is tightly regulated at transcriptional, posttranscriptional, as well as posttranslational levels. Here, we have investigated the mechanism/s via which excess histones exert their deleterious effects in vivo in the budding yeast. We find that the presence of excess histones saturates certain histone modifying enzymes, potentially interfering with their activities. Additionally, excess histones appear to bind non-specifically to DNA as well as RNA, which can adversely affect their metabolism. Microarray analysis revealed that upon overexpression of the histone H3 and H4 gene pair or all four core histones but not individual histones, about 240 genes were either up or downregulated by 2-fold or more. Interestingly, histone overexpression does not affect the bulk chromatin structure, but alters the fine structure of chromatin. Overall, we present evidence that excess histones are likely to mediate their cytotoxic effects via multiple mechanisms that are primarily dependent on inappropriate electrostatic interactions between the positively charged histones and diverse negatively charged molecules in the cell. We have also investigated how changes in histone gene dosage affects the DNA damage sensitivity of budding yeast cells that have two copies of each histone gene when only one copy is needed for survival. We found that overexpression of histones led to an increase in DNA damage sensitivity. Next, we deleted the second copy of the gene pair (HHT2-HHF2) encoding histones H3 and H4 that contributes 6-8 fold more histone mRNA than the first gene pair (HHT1-HHF1), to create an experimental system to study the effects of reduced histone levels in vivo . A reduction in the dosage of histone H3-H4 resulted in a significant decrease in DNA damage sensitivity. By taking the advantage of a HO endonuclease induced DNA double stand break (DSB) at the budding yeast mating type (MAT) locus, we were able to study the DSB repair process in detail in strains with a reduced histone gene dosage. We found that the efficiency of Homologous Recombination (HR) at a DSB, as well as genome wide HR, was elevated in hht2-hhf2 deletion strain, while Non-Homologous End Joining remained unchanged. These effects were not associated with global changes in the expression of DNA repair genes or DNA damage checkpoint responses. We also found that there was no alteration of gross chromatin structure in response to changes in histone gene dosage. One mechanism by which reduced histone dosage leads to elevated HR mediated repair of a DSB at the MAT locus is through enhanced recruitment of the HR factors, as determined by the Chromatin Immunoprecipitation (ChIP) assay. Concomitant with this, cells experience a greater histone loss around this DSB upon a reduction in histone gene dosage. We propose that high levels of endogenous histones generated by multiple genes in eukaryotes compete with HR factors, thereby reducing HR efficiency and may normally function to restrain potentially excessive HR activities during S-phase. Overall, our findings help to explain the basis for the existence of multiple mechanisms that regulate histone levels and highlight their role in maintaining genomic stability and cell viability. Our findings could have major implications for DNA repair, genomic stability, carcinogenesis and aging in human cells that have dozens of histone genes. / A Dissertation submitted to the Department of Biomedical Sciences in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester, 2011. / July 1, 2011. / Double Strand Break, Epigenetics, Homologous Recombination, Budding Yeast, Genomic Stability / Includes bibliographical references. / Akash Gunjan, Professor Directing Dissertation; David Gilbert, University Representative; Myra Hurt, Committee Member; Johanna Paik, Committee Member; Yanchang Wang, Committee Member.