Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of a signal transduction phosphorelay controlling Rhodobacter capsulatus gene transfer… Farrera Calderon, Reynold Gerardo 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2020_may_farreracalderon_reynold.pdf [ 2.39MB ]
JSON: 24-1.0389999.json
JSON-LD: 24-1.0389999-ld.json
RDF/XML (Pretty): 24-1.0389999-rdf.xml
RDF/JSON: 24-1.0389999-rdf.json
Turtle: 24-1.0389999-turtle.txt
N-Triples: 24-1.0389999-rdf-ntriples.txt
Original Record: 24-1.0389999-source.json
Full Text

Full Text

  CHARACTERIZATION OF A SIGNAL TRANSDUCTION PHOSPHORELAY CONTROLLING RHODOBACTER CAPSULATUS GENE TRANSFER AGENT (RcGTA) GENE EXPRESSION by  Reynold Gerardo Farrera Calderon  B.Sc., Instituto Politécnico Nacional, 2016  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Microbiology and Immunology)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2020   © Reynold Gerardo Farrera Calderon, 2020 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  CHARACTERIZATION OF A SIGNAL TRANSDUCTION PHOSPHORELAY CONTROLLING RHODOBACTER CAPSULATUS GENE TRANSFER AGENT (RcGTA) GENE EXPRESSION  submitted by Reynold Gerardo Farrera Calderon in partial fulfillment of the requirements for the degree of Master of Science in Microbiology and Immunology  Examining Committee: J. Thomas Beatty, Microbiology and Immunology Supervisor  Cara Haney, Microbiology and Immunology Supervisory Committee Member  Yossef Av-Gay, Microbiology and Immunology Supervisory Committee Member Rosemary Redfield, Zoology Additional Examiner     iii  Abstract The alphaproteobacterium Rhodobacter capsulatus mediates an unusual method of horizontal gene transfer within its population via the production of a small phage-like particle, a gene transfer agent (GTA). Although GTAs have been found to exist in a wide variety of microorganisms, the GTA of R. capsulatus (RcGTA) remains the best understood and has become a model for the study of GTAs. All known GTAs are under the control of bacterial regulators; specifically, in R. capsulatus the production of RcGTA is mediated by the CtrA response regulator. CtrA is mostly known as the master cell cycle regulator in Caulobacter crescentus, where it is the centerpiece of a regulatory phosphorelay that includes the CckA histidine kinase and the ChpT phosphotransferase. Homologues of CtrA, ChpT and CckA exist in R. capsulatus and their role in the regulation of RcGTA-mediated gene transfer has been demonstrated in different ways. CtrA is involved in both particle production and recipient capability, and CckA and ChpT are essential for particle maturation and release. With these roles in mind and the sequence similarity between the R. capsulatus and C. crescentus CckA, ChpT and CtrA proteins, it was thought that these proteins form a similar phosphorelay in R. capsulatus. However, although a growing body of evidence demonstrates the regulatory roles of these proteins in RcGTA-mediated gene transfer, no conclusive evidence existed to show that they form such a phosphorelay in R. capsulatus. Here, through a set of in vitro phosphorylation assays and PhosTag polyacrylamide gel electrophoresis analysis, I present evidence that the R. capsulatus CckA, ChpT and CtrA proteins form a phosphorelay. CckA autophosphorylates in the presence of ATP and transfers this phosphate to the ChpT phosphotransferase, which in turn transfers it to CtrA. I also show evidence that the CckA kinase responds to cyclic-di-GMP, switching its activity from a kinase to a phosphatase and inverting the phosphate flow in the phosphorelay. These data, coupled with the analysis of the effect of three different site-directed CckA mutants on the phosphorelay, support a model of how RcGTA gene expression is regulated differentially by CtrA and CtrA~P, the phosphorylated form of CtrA. iv  Lay Summary  A species of bacteria named Rhodobacter capsulatus is capable of transferring DNA between cells by producing particles that resemble a small virus; these particles are called gene transfer agents (GTAs). GTAs package randomly selected DNA from the cells that produce them, and these cells then burst and release the GTAs to their surroundings. Because the cells that produce GTAs die, they need to have very strict control on how and when they produce them. One of the control mechanisms they can use is called a phosphorelay: a set of proteins that send messages inside the cell by transferring a molecule of phosphate from one to another. Three proteins, named CckA, ChpT and CtrA, exist in Rhodobacter capsulatus and empirical evidence suggests that they might form a phosphorelay, however, solid biochemical proof is still lacking. Here I present experiments that confirm the existence and function of this phosphorelay.   v  Preface  This work contains contributions from former members of the Beatty lab, as well as members of Andrew Lang’s lab at St. John’s Memorial University as part of a long-term collaboration.  A number of figures in this work come from published sources, all are reproduced with permission from their copyright holders: Figures 1 (right panel), 3 and 8 are reproduced with permission from Elsevier under licenses 4807271303465, 4807280169240 and 4807270817054. Figure 2 is reproduced under authorization from the American Society for Microbiology. Figure 5 is reproduced under permission of Dr. Michael T. Laub. Figure 6 is reproduced by permission of the Federation of European Microbiological Societies under license 4807270252389. Figure 9 is reproduced from an open access article licensed under a Creative Commons Attribution 4.0 International License: All the sources for these figures are indicated in the figure captions. The leftmost panel in Figure 1 belongs to a manuscript submitted for publication by Bárdy et al (Pavol Bárdy, Tibor Füzik, Dominik Hrebík, Roman Pantůček, J. Thomas Beatty, Pavel Plevka. Structure and mechanism of DNA delivery of a gene transfer agent. submitted to Nature Communications). The pET28a::ctrA plasmid was constructed by Paul Fogg (former postdoc in the Beatty lab). The pET28a::chpT plasmid was constructed by J. Thomas Beatty. The truncated pET28a::cckA plasmids (wild type and mutant CckAs) were constructed by Purvikalyan Pallegar, a PhD student in the Lang lab at St John’s Memorial University. All the data presented in this thesis are from experiments carried out by me, although I gratefully acknowledge the assistance of undergraduate student trainees Liz Yang, Tina Tai and, especially, Jessie Zhang (who assisted in the initial testing of the PhosTag gel system, and in preparing samples for mass spectroscopy). Time-of-flight mass spectrometry analyses were done at the Proteomics Core Facility – Michael Smith Laboratories of the University of British Columbia. Experiments were designed by me, in consultation with Tom Beatty, and we jointly interpreted the results. This thesis was written by me, with editing by Tom Beatty.  vi  Table of Contents Abstract ................................................................................................................................iii Lay Summary ....................................................................................................................... iv Preface ................................................................................................................................... v Table of Contents ................................................................................................................. vi List of Tables ........................................................................................................................ ix List of Figures ....................................................................................................................... x List of Symbols .................................................................................................................... xii List of Abbreviations ......................................................................................................... xiii Acknowledgements ............................................................................................................ xiv Dedication............................................................................................................................ xv 1 Introduction. .................................................................................................................. 1 1.1 Rhodobacter capsulatus ........................................................................................... 1 1.2 Horizontal gene transfer in bacteria.......................................................................... 2 1.2.1 Conjugation ....................................................................................................... 3 1.2.2 Transformation .................................................................................................. 4 1.2.3 Transduction ...................................................................................................... 4 1.2.4 Other methods of HGT. ..................................................................................... 5 1.3 Gene transfer agents ................................................................................................ 5 1.3.1 General characteristics and role of gene transfer agents ...................................... 7 1.4 The Rhodobacter capsulatus Gene Transfer Agent .................................................. 9 1.4.1 RcGTA Structure ............................................................................................... 9 1.4.2 RcGTA mediated gene transfer ........................................................................ 10 1.4.3 RcGTA genes .................................................................................................. 14 1.5 Regulation of RcGTA gene expression .................................................................. 18 vii  1.5.1 Two-component systems and phosphorelays .................................................... 18 1.5.2 The CckA/ChpT/CtrA phosphorelay ................................................................ 20 1.5.3 The CckA histidine kinase ............................................................................... 23 1.5.4 The ChpT phosphotransferase .......................................................................... 24 1.5.5 The CtrA master response regulator ................................................................. 24 1.6 CckA/ChpT/CtrA in other organisms ..................................................................... 25 1.6.1 The CckA/ChpT/CtrA phosphorelay in the C. crescentus cell cycle. ................ 25 1.6.2 Other bacteria with similar regulatory networks. .............................................. 28 1.7 Quorum sensing and its role in RcGTA gene expression. ....................................... 29 1.8 Current model for the regulation of RcGTA gene expression ................................. 30 1.9 Structural insights into CckA ................................................................................. 34 2 Hypotheses and objectives .......................................................................................... 36 3 Materials and Methods ............................................................................................... 37 3.1 Strains used: .......................................................................................................... 37 3.2 Plasmids derived from pET28a (Novagen): ........................................................... 37 3.3 Growth conditions: ................................................................................................ 38 3.4 DNA extraction: .................................................................................................... 38 3.5 Sanger sequencing: ................................................................................................ 38 3.6 Induction: .............................................................................................................. 38 3.7 Protein extraction: ................................................................................................. 38 3.8 Protein purification: ............................................................................................... 39 3.9 Phosphotransfer profiling reactions: ....................................................................... 39 3.10 PhosTag PAGE: .................................................................................................... 41 3.11 Time-of-flight mass spectrometry: ......................................................................... 41  viii  4 Results .......................................................................................................................... 42 4.1 Creation of an R. capsulatus CckA/ChpT/CtrA phosphorelay in vitro. ................... 42 4.2 Phosphorylation of R. capsulatus CtrA can occur via both CckA-mediated and acetyl phosphate-mediated pathways, with different kinetics. ...................................................... 44 4.3 Time-of-flight mass spectrometry analysis of CckA and CtrA shows mass changes confirming phosphorylation. ............................................................................................. 46 4.4 CckA shifts from a kinase to a phosphatase in the presence of cyclic-di-GMP. ...... 48 4.5 CckA-mediated dephosphorylation of CtrA~P requires the presence of ChpT. ....... 51 4.6 CckA phosphatase activity is not altered by the presence of ADP nor activated by ADP alone in the absence of c-di-GMP. ............................................................................ 53 4.7 Modification of specific CckA amino acid residues results in altered kinase and phosphatase activities. ....................................................................................................... 54 4.7.1 CckA H399A lacks kinase activity and has normal phosphatase activity. ......... 55 4.7.2 CckA V443P lacks both kinase and phosphatase activities. .............................. 57 4.7.3 CckA Y589D presents normal kinase activity and lacks phosphatase activity. .. 58 5 Discussion .................................................................................................................... 60 5.1 The R. capsulatus CckA, ChpT and CtrA proteins form a phosphorelay in vitro. ... 60 5.2 CckA changes its activity from a kinase to a phosphatase when incubated with c-di-GMP. 62 5.2.1 Effect of ADP on CckA phosphatase activity. .................................................. 64 5.3 Site specific-mutations of CckA alter its kinase and phosphatase activities. ........... 65 5.3.1 H399A ............................................................................................................. 65 5.3.2 V443P ............................................................................................................. 66 5.3.3 Y589D ............................................................................................................. 67 6 Conclusions and Future Directions ............................................................................ 69 References ........................................................................................................................... 72  ix  List of Tables Table 1. Components of the RcGTA genome. ....................................................................... 17 Table 2. Plasmids used in this work....................................................................................... 37   x  List of Figures Figure 1. Structure of the RcGTA. ........................................................................................ 10 Figure 2. Proposed model of RcGTA DNA insertion and integration compared to natural competence. .......................................................................................................................... 13 Figure 3. General description of RcGTA – mediated gene transfer. ....................................... 14 Figure 4. Genetic elements of RcGTA. .................................................................................. 16 Figure 5. Schematics of two-component systems (TCS) and phosphorelays. ......................... 19 Figure 6. Sequence identity between C. crescentus and R. capsulatus CtrA, CckA and ChpT proteins. ................................................................................................................................ 21 Figure 7. The CckA/ChpT/CtrA phosphorelay. ..................................................................... 22 Figure 8. The CckA/ChpT/CtrA phosphorelay in C. crescentus. ............................................ 27 Figure 9. Regulatory model of RcGTA gene expression. ....................................................... 32 Figure 10. The Rhodobacter capsulatus CckA/ChpT/CtrA phosphorelay. ............................. 43 Figure 11. CckA-mediated phosphorylation of R. capsulatus CtrA increases over time and occurs more rapidly than phosphorylation with acetyl-phosphate. ......................................... 45 Figure 12. Time-of-flight mass spectrometry analysis of CckA and CtrA samples corroborates phosphorylation. ................................................................................................................... 47 Figure 13. R. capsulatus CckA presents phosphatase activity. In addition to its kinase activity CckA can dephosphorylate CtrA~P in the presence of cyclic-di-GMP. .................................. 49 Figure 14. R. capsulatus CckA activity can be switched from kinase to phosphatase by addition of c-di-GMP, observed by dephosphorylation of CtrA. .......................................................... 50 Figure 15. Enzymatic dephosphorylation of CtrA~P requires the complete CckA/ChpT/CtrA phosphorelay. ........................................................................................................................ 52 Figure 16. CckA c-di-GMP-activated phosphatase activity is not affected by the presence of ADP nor activated by ADP alone in the absence of c-di-GMP. ............................................. 53 xi  Figure 17. CckA H399A mutant lacks kinase activity and presents normal phosphatase activity in the presence of c-di-GMP and ADP. The absence of ADP has little to no effect on CckA H399A phosphatase activity in response to c-di-GMP. .......................................................... 56 Figure 18. CckA V443P mutant lacks both kinase and phosphatase activities. ....................... 58 Figure 19. CckA Y589D mutant presents normal kinase activity but lacks phosphatase activity. ............................................................................................................................................. 59   xii  List of Symbols  α Alpha β Beta µ Mu / Micro ° Degrees λ Lambda    xiii  List of Abbreviations Ac~P Acetyl phosphate ADP Adenosine Di Phosphate ATP Adenosine Tri Phosphate bp Base pairs CA Catalytic Assisting c-di-GMP Cyclic Diguanylate Monophosphate CFU Colony Forming Units DHp Dimerization and Histidine phosphotransfer DNA Deoxyribonucleic Acid EDTA Ethylene Diamine Tetraacetic Acid GTA Gene Transfer Agent HGT Horizontal Gene Transfer HK Histidine Kinase HSL Acyl-Homoserine Lactone IPTG Isopropyl β-D-1-Thiogalactopyranoside kan Kanamycin kbp Kilo base pairs ORF Open Reading Frame PAGE Polyacrylamide Gel Electrophoresis PAS Per/ARNT/Sim Phage Bacteriophage QS Quorum Sensing RcGTA Rhodobacter capsulatus Gene Transfer Agent RD Receiver Domain RR Response Regulator SDS Sodium Dodecyl Sulphate T4SS Type IV Secretion System WT Wild Type   xiv  Acknowledgements I want to first of all thank the University of British Columbia and its staff for all the great experiences it has provided for me as an international graduate student. Most of all I offer my thanks to Dr. Tom Beatty, for being a magnificent supervisor, mentor and teacher for me during the last few years. Tom’s constant support, advice and willingness to help his students made my stay in the Beatty lab more than enjoyable. Without his support I would not have been able to pursue this project, and I am incredibly grateful to him for giving me this opportunity. Next, I wish to thank Darlene Birkenhead, for her endless support and aid in guiding us grad students through the journey of graduate life itself, from the most trivial to the most vital stuff. Her support has been unfaltering and immensely helpful. I am also grateful to Jeanette Beatty for her advice, her help in the lab and her great conversations. Her knowledge in all lab-related matters was invaluable to me and all time spent with her in the lab was always productive, amusing or often both. I would like to thank the members of my supervisory committee Dr Cara Haney and Dr Yossef Av-Gay for their advice and suggestions, and Dr Andrew Lang and Purvi Pallegar for their continued insight into the project. I would also like to thank the members of the Beatty lab, Amita Mahey, Daniel Jun and all the undergraduate students that assisted me during my project, with special thanks to Jessie Zhang. Additionally, I would like to thank the staff members of the Graduate Pathways to Success office and Dr Dave Oliver for their help in developing many of the key skills that I used to complete this project.  I would like to thank my parents Reynold and Georgina for their continued emotional, moral and financial support throughout the last two and a half years, and to my sister Karla, my cousin Omar and my friend Daniel for their support and ideas. And last, but definitely not least, I give my greatest and most sincere thanks to my wife Andrea, who has been unwavering in her support through everything, for her always great input, for being my fan, my counsel, my partner and my best friend.  xv  Dedication  To my love Andy,  For without your constant support I would have never taken the first steps of this years-long journey.  Thank you for everything you have done for me and us, this work is as much a fruit of your labor as it is of mine.    “Our lives are not our own. We are bound to others, past and present, and by each crime and every kindness, we birth our future.” ― David Mitchell, ‘Cloud Atlas’.  1  1 Introduction. 1.1 Rhodobacter capsulatus Rhodobacter capsulatus is a bacterial species belonging to the class alphaproteobacteria; otherwise it is also called a member of the purple bacteria, a diverse group of photosynthetic organisms widely distributed in nature, inhabiting mainly aquatic ecosystems (Madigan & Jung, 2009). These organisms have been an important focus for the study of the biochemistry of photosynthesis as they are capable of photoheterotrophic and photoautotrophic metabolism (Beatty & Gest, 1981; Weaver et al., 1975). From an environmental perspective they are important because they can oxidize hydrogen sulfide (H2S), degrade organic matter, and add organic compounds to anoxic environments (Madigan & Jung, 2009). R. capsulatus belongs to a specific group of purple bacteria: the purple nonsulfur bacteria, which from a physiological perspective are distinguished from purple sulfur bacteria by the fact that they tolerate only low concentrations of sulfide, and any S0 they produce is released and not stored intracellularly (Madigan & Jung, 2009).  R. capsulatus was originally studied to gain insights into the biochemistry of photosynthesis as well as into photosynthesis-related bacterial metabolism. It was used due to its capacity for rapid growth in simple synthetic media under a variety of conditions, and its hardiness when being studied in the lab (Weaver et al., 1975). Further study of R. capsulatus revealed the great metabolic flexibility presented by this species: it has the capacity to grow both phototrophically, using either CO2 or organic compounds as a carbon source, and chemotrophically, in darkness by respiration, fermentation or chemolithotrophy (Madigan & Gest, 1979; Madigan & Jung, 2009). This resulted in R. capsulatus being known as one of the most metabolically versatile bacteria (Madigan & Gest, 1979).  During some of the earlier studies into the genetics of R. capsulatus Barry Marrs discovered that some cells were capable of transferring genes between different strains through some unknown mechanism. Describing it as “unlike that of any previously described bacterial system” (Marrs, 1974), Marrs observed that strains of R. capsulatus were capable of transferring antibiotic resistance and wild type genes to other strains through some kind of extracellular vector that was both resistant to DNA-degrading enzymes and smaller than any known bacteriophage capable of transduction (see section 1.2.2). The study of this vector, which eventually became 2  known as a gene transfer agent (GTA) (Solioz & Marrs, 1977) gave rise to the discovery of an entirely new mechanism of horizontal gene transfer. 1.2 Horizontal gene transfer in bacteria. Horizontal gene transfer (HGT) is the mechanism through which organisms mobilize genetic material from the genome of one cell to the genome of another, which can happen across different species, phyla and even domains of life (Soucy et al., 2015). The name HGT is in contrast with vertical gene transfer, which is the passing of genes from progenitors to offspring (Daubin & Szöllősi, 2016). Although the first evidence for horizontal gene transfer was observed in 1928 (Griffith, 1928), it was until the late 1940s that it was identified as the transfer of genetic characteristics from one organism to another, being described as “the release of ‘transforming substances’ diffusing through the medium” by Joshua Lederberg in his study of Escherichia coli (Lederberg, 1947). Knowledge of HGT has increased drastically in the last few decades, revealing its vital role in the evolution and adaptation of modern organisms. The effect that HGT has had on the study of the genomes of living organisms, showing that genetic characteristics can be obtained by co-habitating organisms as well as from ancestral organisms has resulted in the model of the ‘tree of life’ to be reformed, being replaced with a model of a ‘web of life’ (Soucy et al., 2015). HGT tends to be more frequently successful between closely related organisms and was originally thought to only mediate the transfer of genes that provide some sort of selective advantage to the recipient organisms, in contrast with ‘selfish’ genes which enhance their own transmission without making any positive contribution to the fitness of the host. However, later studies have gradually shown that there is a lot of HGT involving the transfer of initially neutral genes that give no benefit to the host but do no harm either. These genes may later on provide benefits after the transferred material becomes “domesticated”, which is to say it produces a phenotype beneficial to the host (Soucy et al., 2015). Although occurring in all domains of life, both within and between themselves, HGT has been studied the most in prokaryotes. HGT occurs through different methods in prokaryotic cells, with three of them considered to be the classic mechanisms of HGT: conjugation, transformation and transduction (Daubin & Szöllősi, 2016; Soucy et al., 2015). 3  1.2.1 Conjugation Conjugation is the process through which one cell unidirectionally transfers DNA directly to another cell. Originally described as “the transfer of genetic material from one bacterial strain to another that is dependent upon cellular contact between members of the two bacterial strains” (Curtiss, 1969), conjugation is indeed dependent on cell-to-cell contact. This contact is carried out mostly by the donor cell, which interacts with a recipient cell using a pilus through which the DNA is then transported. Conjugation is mediated mainly by plasmids, but it can also be mediated by some types of bacteriophages, known as conjugative phages (Daubin & Szöllősi, 2016), and integrative and conjugative elements (ICEs), a type of transposon transferred by conjugation (Delavat et al., 2017). Here, I briefly summarize the main features of plasmid conjugation as representative of the diverse conjugative elements. The DNA molecule that is transferred through conjugation encodes the genes required for the transfer to take place and can be described as a selfish genetic element. For proper conjugation to occur the DNA molecule needs to contain an origin of transfer site (oriT) as well as a set of genes encoding the following proteins: a nucleoprotein complex called a selector or relaxosome, required for the selective initiation of the replicative process; a transmembranal conduit, a multiprotein complex that forms a type IV secretion system (T4SS) which is used to construct the pilus that will transport and insert the DNA into the recipient cell; a coupling protein which brings the replicative complex together with the secretion complex and actively pumps the DNA from the donor into the recipient (Llosa & de la Cruz, 2005). Once the donor cell has finished assembling these complexes the pilus will mediate cell-to-cell contact and create a bridge between the two cells. The molecular processes that follow are varied and have not been completely elucidated but are known to involve the donor protein complexes that pump the replicated DNA into the recipient cell which then uses its own energy and proteins to integrate the received genetic element (Cabezón et al., 2015). Although for this method of HGT to occur at least one self-transmissible element needs to be present, other genetic elements can be co-transported by it because there are non self-transmissible genetic elements that can mobilize in the presence of a self-transmissible plasmid (Thomas & Nielsen, 2005). 4  1.2.2 Transformation Transformation can be described as the process of uptake, integration and expression of exogenous DNA by a cell. It is an active mechanism in which the recipient cell needs to be able to enter a metabolic state called ‘competence’, expressing a set of competence genes (com genes) required to produce proteins that mediate DNA uptake and integration into the genome (Mell & Redfield, 2014; Thomas & Nielsen, 2005). There are many bacterial species that are naturally competent, and the competence genes are widely distributed among bacteria. These genes are activated under certain conditions, with common inducers being elevated cell density, DNA damage, antibiotic-induced stress and nutrient starvation (Blokesch, 2016). DNA is continuously being released into the environment by the disruption of cells and viral particles, and even in some cases by active release from living cells. Once competent cells are exposed to exogenous DNA the transformation process begins. The first step is non-covalent binding of the DNA to cell surface proteins such as ComEA. The DNA is then introduced into the cell by a translocation process that varies among bacterial species but that can involve the synthesis of a pseudopilus (Klebsiella, Bacillus) or a pilus (Neisseria), and membrane translocation machinery (Haemophilus) (I. Chen & Dubnau, 2004; Mell & Redfield, 2014). The translocation process results in only one strand of DNA entering the cell, while the other one is degraded. Once inside the cell, the transported DNA strand can be processed in different ways depending on its characteristics: it might reassemble as double stranded DNA in the cytoplasm, reform if it is a plasmid, or be bound by a DNA-processing protein (DprA) and integrated into the host genome through homologous recombination (Johnston et al., 2014; Mell & Redfield, 2014; Thomas & Nielsen, 2005). For recombination to take place the cell needs to possess a RecA-dependent recombination system and the sequence needs to present a certain level of homology to a segment of the bacterial chromosome (I. Chen & Dubnau, 2004). 1.2.3 Transduction Transduction is a type of genetic transfer that is mediated by viruses, mostly through the effect of bacteriophages (phages) on bacteria. Normally phages hijack their host’s cellular machinery to construct more viral particles and pack them with copies of viral DNA. However, under certain conditions host DNA might be 5  erroneously packaged into the viral particle instead of viral DNA, which is known as generalized transduction. The transfer of DNA through generalized transduction could be considered a relatively rare event, occurring only once every 107 – 108 infection events, but given the numbers at which phages and bacteria interact in the environment this process becomes extremely frequent (Penadés et al., 2015).  Some phages, known as lysogenic phages, integrate into the bacterial genome as part of their lifecycle, during which they are known as prophages, and later excise before undergoing replication and liberation from the cell. During excision from the host genome the viral genes can rarely bring some of the genomic DNA along with them, and this DNA can then be packaged and transferred along with the rest of the viral genes in the viral particles; this is known as specialized transduction (Soucy et al., 2015). After either type of packaging transduction event has occurred, once cell lysis happens and viral particles are released, if a phage that contains bacterial DNA infects a new cell the bacterial DNA will be injected and it can then be incorporated into the new host genome, thus resulting in horizontal gene transfer (Daubin & Szöllősi, 2016).  1.2.4 Other methods of HGT. Other methods of HGT have been described in more recent years, like cell fusion, vesicle-mediated gene transfer, and gene transfer agents (GTAs). Cell fusion is a process in which cells become joined by small bridging structures with gene transfer occurring bidirectionally (Soucy et al., 2015), and vesicular gene transfer is mediated by nanoparticles composed of lipid membranes that are produced by both Gram-negative and Gram-positive bacteria (Domingues & Nielsen, 2017). Gene transfer agents are the main focus of my work and will be discussed shortly. Although there are other several genetic elements that are capable of moving from one genome to another like transposons and the previously mentioned ICEs, these use one of the three classical methods of gene transfer, conjugation, transformation or transduction to achieve mobility. 1.3 Gene transfer agents As mentioned earlier, an unusual method of horizontal gene transfer was first described by Marrs in 1974, when he identified a vector that mediated genetic recombination in R. capsulatus, 6  then known as Rhodopseudomonas capsulata (Marrs, 1974), through some kind of secreted particles. These particles, which were named gene transfer agents (GTA) by Solioz and Marrs (Solioz et al., 1975), are produced by cultures of R. capsulatus during the transition from logarithmic to stationary phase (Solioz et al., 1975). Further research into the characteristics of GTAs revealed that they are particles containing linear double stranded DNA of about 4.5 kbp in length. The particles resemble tailed bacteriophages of the family Siphoviridae, having a head and a tail of about 30 nm in diameter and 50 nm in length respectively (Yen et al., 1979). Having been characterised as a nucleoprotein particle resembling the shape of known phages, GTAs were suspected of being some kind of R. capsulatus – associated virus. However, some of the observations did not fit with this hypothesis: no assays showed the characteristic formation of plaques on solid cultures that accompanies viral production and release from bacteria, and GTA particles were much smaller than those of any known transducing phages. Furthermore, the DNA transferred seemed to be selected at random from the producing cells, and transfer required DNA sequence similarity between donor and recipient cells (Solioz & Marrs, 1977; Yen et al., 1979). Marrs and his associates proposed that GTA was a defective prophage that came from a virus that became defective early in the evolutionary history of R. capsulatus, as its activity was found to be widespread among wild-type strains of the bacteria, but the nature of how GTAs acted and were controlled remained a mystery (Yen et al., 1979). Their research also gave rise to the alternative theory that the genes coding for GTA production could have evolved from bacterial genes, and that GTA represented a phage progenitor (Marrs, 2002). However, although much remained to be discovered about the characteristics, mechanisms and roles of GTAs, their study stopped almost entirely during the 1980s and 1990s, possibly due to technological limitations at the time. This was until the year 2000, when Lang and Beatty (Lang & Beatty, 2000) performed a series of gene cloning and mutagenesis experiments that showed that, genetically, GTA resembles a defective prophage, and showed that its expression is regulated by bacterial genes. This pointed towards GTAs having evolved along with the bacterial genome and that they are highly integrated into the cell’s own regulatory processes. This work also revealed another important 7  facet of GTAs that was suspected since the first observations of the particle size were done: The GTA gene cluster was found to be about 15 kbp in length, which is not only too small to encode a typical DNA-containing tailed phage, but it is also too large to be packaged inside the GTA particles themselves, as they only package randomly selected fragments about 4.5 kbp in length (Lang & Beatty, 2000; Yen et al., 1979). These new insights gave rise to a plethora of questions and launched a new era of research into GTAs. 1.3.1 General characteristics and role of gene transfer agents As the study into GTAs continued, other organisms were found to present similar systems. A genome sequencing project revealed that gene clusters resembling those found for the GTA of R. capsulatus were present in many other species of the alphaproteobacteria (Lang et al., 2002), and GTA-like mechanisms of gene transfer were soon identified in an even wider variety of microorganisms. Therefore, because of the increase in the number of (sometimes putative) GTAs, at this time the GTA of R. capsulatus started being referred to as RcGTA (Lang & Beatty, 2006). By 2012 GTA production had been described in several other alphaproteobacteria: Ruegeria pomeroyi (Biers et al., 2008), Roseovarius nubinhibens (McDaniel et al., 2012) and several species of Bartonella (Anderson et al., 1994). But GTAs were not only found in alphaproteobacteria, but also in other classes of bacteria, with the Dd1 GTA of the deltaproteobacterium Desulfovibrio desulfuricans (Rapp & Wall, 1987). GTAs were found in different phyla, with the VSH-1 GTA of the spirochete Brachyspyrum hyodysenteriae (Humphrey et al., 1997) and even in archaea with the VTA (Voltae transfer agent) of the archaeon Methanococcus voltae (Bertani, 1999). Around this time the characteristics of GTA became firmly established: GTAs are defined as phage-like entities that mediate horizontal gene transfer by transporting a randomly selected piece of the genome of the producing cell and are incapable of transferring enough genetic material to encode the protein components of the particles themselves (Lang et al., 2012). Furthermore, for GTA-mediated gene transfer to occur the recipient cells need to integrate the transported DNA through a homologous recombination pathway (Brimacombe et al., 2014; Lang et al., 2012). GTA-mediated gene transfer is now considered to integrate aspects from both gene transduction and transformation (Westbye, Beatty, et al., 2017).  8  With this characterization in mind multiple questions about the evolution and role of GTAs came into focus. The main consensus was that GTAs evolved from phages that interacted with ancestral prokaryotes, and that they provide some level of selective advantage to the producing cells, mostly acting as a way of increasing diversity in the gene pool and repairing damaged DNA (Lang et al., 2012). In addition to this some researchers suggested that GTAs could have even had a role in the evolution of endosymbiotic organisms that gave rise to eukaryotic cells (Richards & Archibald, 2011). Some of these hypothesis are somewhat contested: GTAs could have descended from phages multiple independent times, maybe a single GTA-like element was present in the last common ancestor of these organisms and phages originated from it, or some undiscovered element led to the origin of both GTAs and phages (Lang et al., 2012). Some authors suggested that GTAs did not provide selective advantages but were merely a result of a damage-control mechanism looking to reduce the harmful effects of phages on its host (Redfield, 2001). Further studies support some parts of these theories while disproving others. Genetic studies have shown that the RcGTA genes are under purifying selection (Lang et al., 2012), indicating that they are under positive selective pressure to remain functionally conserved in the genome of the host cells. Phylogenetic analyses point against GTAs being ‘selfish’ or ‘decaying’ elements reinforcing the hypothesis that GTA maintenance occurs at a population level and contributes to the population survival either through an increase of the population genetic diversity or even as a nutritional component (Shakya et al., 2017). These same studies showed that it is unlikely that GTAs contributed to the origin of eukaryotic endosymbionts as GTAs seemed to have appeared at a later point in time than eukaryotes (Shakya et al., 2017). Most recently a computer simulation quantitative approach was used to model possible benefits to GTA – containing cell populations, using a variety of parameters to assess fitness costs and benefits (Redfield & Soucy, 2018). It was concluded that any possible benefits arising from genetic recombination are insufficient to overcome the detrimental effects of cell lysis accompanying production of GTAs, and that additional, unknown factors must exist.  9  1.4 The Rhodobacter capsulatus Gene Transfer Agent Throughout all of this, R. capsulatus remained the predominant model host for the study of GTAs, and RcGTA became the most well-studied GTA (Fogg, 2019). Due to the natural characteristics of GTAs the study of RcGTA was focused towards trying to decipher the mechanisms through which it is codified, regulated, expressed and transmitted. 1.4.1 RcGTA Structure RcGTA was originally characterized as a small nucleoprotein particle with a 30 nm head and a 50 nm tail that resembled a small bacteriophage (Solioz & Marrs, 1977; Yen et al., 1979). As research into it progressed more was learned about its structure and currently an RcGTA particle can be described as follows: RcGTAs are composed of an icosahedral head 38 nm in diameter with spikes attached to the vertexes of the capsid, these spikes are between 4 and 7 nm wide and can be up to 12 nm long. Attached to the head is a short striated cylindrical tail 48 nm long and ~8 nm wide, attached to the end of the tail is a baseplate structure that is between three and four times wider than the tail. Bound to the tail are also three to four fiber-like structures approximately 20 nm long (Figure 1) (Westbye et al., 2016) (Bárdy et al., manuscript submitted for publication). The RcGTA head capsid is packed with linear unnicked double-stranded DNA (Solioz & Marrs, 1977), that is a randomly selected 3.5-4.5 kbp–long fragment of the R. capsulatus genome (Hynes et al., 2012; Lang & Beatty, 2001). 10   Figure 1. Structure of the RcGTA. Left: RcGTA cryo-EM structure (personal communication, J. Thomas Beatty). Right: RcGTA electron micrographs, taken from (Westbye et al., 2016). Arrows indicate head spikes, arrowheads indicate tail fibers. 1.4.2 RcGTA mediated gene transfer The understanding of how RcGTA mediates gene transfer has come a long way since it was first described in 1974. Early studies identified that RcGTAs were mostly produced during the transition from exponential growth phase into stationary phase in an R. capsulatus culture (Solioz et al., 1975), but could not elucidate the conditions that caused its production, nor the way in which RcGTA particles were released (Lang & Beatty, 2001). This became one of the longer standing hurdles of understanding RcGTA-mediated gene transfer, as researchers could not identify the process through which RcGTAs were released from the producing cells. This process was thought to possibly not involve cell lysis at all, as assays for plaque formation, and measurement of liquid culture turbidity yielded undetectable lysis (F. Chen et al., 2009). It was not until an ‘overproducer’ mutant was studied that lysis was detected, and several genetic analyses showed that in fact the production of holin and endolysin proteins was essential for particle release (Fogg et al., 2012; Westbye et al., 2013). The reason why lysis was not observed in WT cultures is that RcGTA–producing cells are only a very small subpopulation, between 11  1.5 to 3% of viable cells (Hynes et al., 2012). Later on it was found out that stress factors, specifically nutrient starvation, could induce the production of RcGTA in this subpopulation (Westbye, O’Neill, et al., 2017), and that the ‘activation’ of producing cells was essentially a random process (Ding et al., 2019). Once an R. capsulatus cell has had its RcGTA production turned on it begins synthesizing the structural components of the RcGTA particle, DNA is randomly selected and inserted into the capsid by what has been proposed to be a headful-packaging mechanism (Westbye et al., 2018), a non-specific method in which DNA is inserted until the capsid is full. The packaging process is assisted by the action of RcGTA portal proteins and an RcGTA – encoded terminase complex composed of the TerL large terminase and the TerS small terminase. TerS assists with DNA recognition and binding, TerL then creates a free DNA end via nuclease activity and translocates DNA into the preformed capsid using an ATPase domain (Sherlock et al., 2019). Tail attachment to the head and particle maturation occur afterwards; these steps of the process are still not fully elucidated, but are believed to be highly similar to the processes that occur during the assembly of phages of the Siphoviridae family such as phage λ (Westbye, 2016). Once particles are fully assembled and matured, the producing cell undergoes lysis by the activity of the holin and endolysin, releasing the RcGTA to the surrounding environment (Fogg et al., 2012). After RcGTA particles have been released they still need to be able to selectively target and bind to the recipient cells. It was found out that this process is not solely mediated by the RcGTAs, as recipient cells also take an active role in being able to receive the DNA being transferred. First of all, recipient cells must produce a capsular polysaccharide that acts as a receptor for the RcGTA particles (Brimacombe et al., 2013). The particles bind to the capsular receptor through the head spikes (Westbye et al., 2016), after which the phage introduces the DNA into the recipient’s cell periplasm through a process that as of date has not been well described, although there is evidence that tail fibers play a role in helping stabilize the RcGTA – cell union, and may be involved keeping the DNA inside the capsid (Hynes et al., 2016) (Bárdy et al., manuscript submitted for publication). 12  Next, for DNA entry to the cell to continue, the recipient cell must be expressing the competence genes coding for the ComEC, ComF and ComM proteins. The RcGTA releases the DNA through the outer membrane into the periplasm, where it traverses the peptidoglycan layer by the action of the RcGTA p14 cell-wall protease. After reaching the inner membrane it is then taken by the ComEC/ComF complex and presumably transported as single stranded DNA into the cytoplasm. The specific function of ComM is unknown but genetic evidence shows that it is required for transfer to take place (Brimacombe et al., 2015). Once the single stranded DNA has entered the cytoplasm it is targeted and bound by the DNA protecting protein DprA (Brimacombe et al., 2014) which helps initiate a RecA mediated recombination through which the DNA is finally integrated into the recipient cell’s genome (Brimacombe et al., 2014, 2015). The active role that the recipient cell must take and the types of genes it needs to express has led to this stage of RcGTA-mediated gene transfer being compared to natural transformation, and a comparison between the two is presented in Figure 2. The two processes differ in the way in which the DNA is inserted into the periplasm, but after that the two are highly similar (Brimacombe et al., 2015), and as has been mentioned before there are various ways through which the first steps of natural competence take place (I. Chen & Dubnau, 2004). 13   Figure 2. Proposed model of RcGTA DNA insertion and integration compared to natural competence. CPS: Capsule polysaccharide. PG: Peptidoglycan. P14: RcGTA cell wall protease. Altered from (Brimacombe et al., 2015). This similarity between the second half of RcGTA – mediated  gene transfer and transformation, along with the noticeable resemblance that the first half of the process presents to the production of generalized transducing phages has led many researchers to explain the whole process of RcGTA transfer as a mix of generalized phage transduction and natural transformation (Figure 3) (Westbye, Beatty, et al., 2017). 14   Figure 3. General description of RcGTA – mediated gene transfer. Taken from (Westbye, Beatty, et al., 2017) However, unlike generalized transduction RcGTA production is not regulated by phage genes but by bacterial regulators, and it has been discovered that both the production of RcGTA and the recipient capability are controlled by the same network of regulatory proteins (Westbye, Beatty, et al., 2017). Understanding this network has thus been the main focus of RGT related research during the last decade. 1.4.3 RcGTA genes The first forays into the genetics of RcGTA identified at least 3 structural genes due to their similarity to DNA phages, and these were present within a 15 kbp long gene cluster which was identified as likely an operon of RcGTA structural genes (Lang & Beatty, 2000). A deeper analysis soon showed that this operon contained 15 ORFs and that these included the genes encoding for the head and tail proteins (Lang & Beatty, 2001). These early studies also identified the first genes thought to be involved in the regulation of RcGTA expression: encoding homologues of the Caulobacter crescentus CckA and CtrA proteins (Lang & Beatty, 2000). These proteins, along with a homologue of the C. crescentus ChpT protein (Mercer et al., 2012) were eventually revealed to be the core of the regulatory systems of RcGTA gene expression, on the basis of the effects of gene knockouts on RcGTA production. The CckA, ChpT and CtrA protein homologues are the main focus of my work and will be discussed at length in later sections. Further studies progressively revealed more about the genes coding for RcGTA production, Chen et al. identified at least 9 different proteins within the RcGTA structural gene cluster, by 15  isolating RcGTA particles and showing they contained these proteins. Furthermore they co-purified 4 other proteins that came from other sites besides the RcGTA structural gene cluster but did not make headway into their identification, nor could they identify any genes related to cell lysis (F. Chen et al., 2009). A significant advance into the understanding of the RcGTA genes came from a study by Hynes et al. that discovered that the RcGTA genome consisted of at least 24 genes located in five different loci across the R. capsulatus genome. The five loci consisted of the already described RcGTA structural gene cluster, the lytic genes and the head spike genes, plus two additional loci that encoded proteins thought to have a role in attachment to the recipient cells (rcc00171) and RcGTA regulation and maturation (rcc01865-66) (Hynes et al., 2016). The latest breakthroughs in the study of the RcGTA genes came from Sherlock, Long and Fogg by confirming that rcc01682, the first gene of the structural gene cluster, is a small terminase TerS involved in DNA binding and packaging, as well as identifying rcc01865 (discovered by (Hynes et al., 2016)) as a direct regulator of the expression of the RcGTA genes, renaming it as GafA (Fogg, 2019; Sherlock et al., 2019). Figure 4 and Table 1 summarize the components of the RcGTA genome. Please note that these are only the RcGTA–specific genes and exclude most of the regulators of RcGTA expression as well as the genes required for recipient capability. That is because, as mentioned earlier, these genes are of bacterial origin and their analysis merits a section of their own.16   Figure 4. Genetic elements of RcGTA. Genes are drawn to scale; intergenic regions are not. Orange: structural genes Blue: DNA packaging and maturation genes. Pink: Lysis genes. Red: Regulatory gene. Gray: unknown function. RcGTA genes required for proper particle synthesis, maturation and release, bacterial regulatory genes and genes required for recipient capability are not pictured. Protein products are elaborated upon in Table 1.   17  Table 1. Components of the RcGTA genome. ORF Protein / Predicted protein ORF Protein / Predicted protein rcc00171 Tail fiber rcc01689 Stopper protein rcc00555 Endolysin rcc01690 Tail terminator protein rcc00556 Holin rcc01691 Tail tube protein rcc01079 GhsA Head spike protein rcc01692 Tail chaperone 1 rcc01080 GhsB Head spike protein rcc01693 Tail chaperone 2 rcc01682 TerS Small terminase rcc01694 Tail tape measure protein rcc01683 TerL Large terminase rcc01695 Distal tail protein rcc01684 Portal protein rcc01696 Tail hub protein rcc01685 HP rcc01697 p14 Cell wall protease rcc01686 Prohead protease rcc01698 Megatron protein rcc01687 Major capsid protein rcc01865 GafA Transcriptional regulator rcc01688 Head-tail adaptors rcc01866 HP HP: Hypothetical protein. (F. Chen et al., 2009; Fogg, 2019; Hynes et al., 2012, 2016; Lang & Beatty, 2000, 2001; Sherlock et al., 2019; Westbye et al., 2013, 2016) (Bárdy et al., submitted for publication)   18  1.5 Regulation of RcGTA gene expression As the expression of RcGTA occurs under specific environmental conditions (Westbye, O’Neill, et al., 2017), at a specific point of the growth phase (Solioz et al., 1975), and only in a subsection of the bacterial population (Hynes et al., 2012) that undergoes cell lysis as a result of this process (Fogg et al., 2012), the expression of RcGTA genes must be very tightly regulated. A variety of factors are part of these control mechanisms, including a quorum sensing (QS) system (Leung et al., 2012; Schaefer et al., 2002) and a protein phosphorelay network (Lang et al., 2017).  In the next sections I will elaborate on the nature of some of these mechanisms, in order to adequately describe how they fit together to form the current model for the regulation of RcGTA gene expression, and similarities with and differences from homologous systems. 1.5.1 Two-component systems and phosphorelays One of the most common methods cells use to regulate gene expression in response to environmental signals are two-component systems (TCS). These systems are widespread among living organisms, being present in bacteria, slime molds, fungi and plants, although absent in metazoans (Skerker et al., 2005). To perform their function TCS are composed of two different proteins: first a membrane–bound sensor histidine kinase (HK) that usually possesses an extracellular sensor domain and is capable of ATP–dependent autophosphorylation. These HKs respond to signals by autophosphorylating a histidine (His) residue in their dimerization and histidine phosphotransfer (DHp) domains and then transferring that phosphate molecule to the second component of the TCS: response regulators (RR), which become activated by changes in their phosphorylation profile (Casino et al., 2010). Most RRs regulate their activity by acting as transcription factors, binding to DNA and altering the expression of target genes, although some have been shown to mediate protein interactions or enzymatic functions. RRs become phosphorylated on a conserved aspartyl residue on their receiver domains (RD) and are further controlled by the action of phosphatases and proteolytic enzymes (Skerker et al., 2005). TCS are the simpler forms of signal transduction systems and are limited in their regulatory capabilities; a more complex form of such systems that allows for more flexible regulatory capabilities is the phosphorelay (Burbulys et al., 1991). 19  Phosphorelays follow the same structure of TCS, but instead of a two–step reaction phosphorelays have four steps. Phosphorelay HK proteins respond the same way to signals, by autophosphorylating a His residue. However, these proteins then transfer the phosphate molecule to an aspartyl residue in a RD within their own structure. This is then followed by transferring the phosphate to an intramolecular His or to another protein called a histidine phosphotransferase. This second protein then transfers the phosphate to the final RR protein, which acts in the same way as RRs in TCS (Stock et al., 2000). A schematic depiction of these systems is presented in Figure 5.  Figure 5. Schematics of two-component systems (TCS) and phosphorelays. A: TCS B: Phosphate flow in a TCS. C: Phosphorelay D: Phosphate flow in a phosphorelay. Taken from (Laub et al., 2007) The multiple steps in phosphorelay systems allow for more control points to integrate modulatory signals, by providing alternative pathways for the phosphate within the system, and allow for branching into networks. An additional factor that adds complexity to this type of system is the fact that RRs may have high autophosphatase activity and HKs tend to be bifunctional, having both kinase and phosphatase activities, and as such being able to bidirectionally regulate phosphate flow within the system (Skerker et al., 2005). 20  Phosphorelays are thus often involved in regulating key decisions of the bacterial cell fate, such as cellular differentiation, the expression of very metabolically-demanding systems such as flagellar mobility and the control of DNA replication (Y. E. Chen et al., 2009). These characteristics fit the expression of RcGTA genes, which leads to cell death. Therefore, it may not be surprising that some of the first proteins discovered to regulate RcGTA gene expression were CtrA and CckA (Lang & Beatty, 2001), which are homologues of C. crescentus proteins that form a phosphorelay (Biondi et al., 2006). 1.5.2 The CckA/ChpT/CtrA phosphorelay At the time of their discovery, it was not known whether the R. capsulatus homologues of the histidine kinase CckA and the response regulator CtrA formed a TCS or a phosphorelay, but experimental evidence showed that both were essential for RcGTA production(Lang & Beatty, 2000), with further studies showing that they were also responsible for the regulation of motility genes in R. capsulatus (Lang & Beatty, 2002). In C. crescentus, they are essential proteins that help regulate the bacterial asymmetrical cell cycle and expression of motility genes, among other functions (Quon et al., 1996). Further study in C. crescentus revealed a third essential component of the system, hinting at its nature as a phosphorelay: the histidine phosphotransferase ChpT (Biondi et al., 2006). The biochemistry of the interactions between the C. crescentus proteins was studied through a series of experiments that finally cemented the nature of this system as a phosphorelay (Biondi et al., 2006; Y. E. Chen et al., 2009). Some of these experiments also revealed additional information on the enzymatic activity of CckA, observing that it presented not only kinase activity but also phosphatase activity (Y. E. Chen et al., 2009). Genetic analyses eventually found a homologue for ChpT in R. capsulatus (Figure 6), and although the evidence was not conclusive, a model was presented in which the R. capsulatus proteins formed a phosphorelay as in C. crescentus (Mercer et al., 2012). 21   Figure 6. Sequence identity between C. crescentus and R. capsulatus CtrA, CckA and ChpT proteins. REC: Receiver domain (RD). HK: Histidine kinase domain, also called dimerization and histidine phosphotransfer (DHp) domain TR: Transcriptional regulator. HA: Histidine kinase-like ATPase, also known as catalytic assistance (CA) domain. COG: Clusters of orthologous genes. Modified from (Mercer et al., 2012). The C. crescentus model for this phosphorelay is that CckA responds to a (currently unknown) signal and undergoes ATP-dependent autophosphorylation, becoming phosphorylated on a histidine residue of its DHp domain. Next intramolecular phosphate transfer occurs and CckA becomes phosphorylated on an aspartyl residue of its RD. ChpT is then specifically phosphorylated on a histidine residue by the RD of CckA, and it in turn mediates the phosphorylation of CtrA on an aspartyl residue. CtrA changes its DNA binding specificity depending on its phosphorylation state and thus mediates differential gene expression in response to the signals received by CckA. Additionally CckA is capable of having its enzymatic activity affected by proteins such as DivL (Tsokos & Laub, 2012; Westbye et al., 2018). Recently it was found that the secondary messenger cyclic diguanylate monophosphate (c-di-GMP) allows CckA to switch from a kinase to a phosphatase (Lori et al., 2015) and reverse the transfer of phosphate in the phosphorelay (Figure 7). In the phosphatase phosphorelay ChpT takes the phosphate from CtrA onto its DHp domain and transfers it to the RD of CckA where it is hydrolyzed and released as inorganic phosphate (Mann et al., 2016).  22   Figure 7. The CckA/ChpT/CtrA phosphorelay. A: Phosphorelay with CckA kinase activity. B: Phosphorelay with CckA phosphatase activity. DHp: Dimerization and histidine phosphotransfer domain. Black ovals represent transmembrane segments. CA: Catalytic assisting domain. RD: Receiver domain. TR: Transcriptional regulator. PAS: Per-Arnt-Sim domain. P: phosphate. H: Histidine. D: Aspartic acid. As mentioned previously, phosphorelays are great regulatory mechanisms due to their ability to integrate multiple regulatory signals and to control the flow of phosphate within and between systems. To gain more insight into how the CckA/ChpT/CtrA phosphorelay performs its function it is important to analyse the structures of these three proteins and how they relate to their functions. Although most of the studies done on these proteins have been done on C. crescentus the structural similarity with the R. capsulatus homologues allows for many comparisons. The following sections focus on the known properties of the C. crescentus enzymes, with key similarities to and differences from the R. capsulatus proteins noted as appropriate. 23  1.5.3 The CckA histidine kinase CckA is the largest of the CckA/ChpT/CtrA phosphorelay proteins, with the C. crescentus CckA being composed of 691 amino acid residues and the R. capsulatus CckA of 767. The two proteins have a sequence identity of 44%, focused around the domains involved with phosphorylation and catalytic activity (Figure 6). The CckA protein is composed of the following regions: an N-terminal transmembrane domain made up of two separate transmembrane segments separated by a short periplasmic loop,  a sensory region composed of two PAS domains, a catalytic core composed of the DHp domain and the ATP-binding CA domain, and a C-terminal receiver domain that mediates the transfer of phosphate to CtrA through ChpT (Jacobs et al., 1999; Mann et al., 2016). Like most other HKs CckA is homodimeric (Casino et al., 2009), although it can form tetramers as well under certain conditions (Mann & Shapiro, 2018). The signal(s) that can be detected by CckA are thought to be sensed by the N-terminal PAS domains, which then drive conformational changes within the protein that alter its enzymatic activity and binding capabilities (Narayanan et al., 2018). These changes can lead to different results such as protein oligomerization which results in enhanced kinase activity (Mann et al., 2016; Mann & Shapiro, 2018). The catalytic core consists of the DHp and CA domains. The DHp domain contains the conserved histidine residue to which the phosphate bonds (His322 in C. crescentus, His399 in R. capsulatus), and loss of this histidine results in loss of kinase activity (Y. E. Chen et al., 2009). Several other residues in the DHp and CA domains assist in binding to ATP and c-di-GMP (Dubey et al., 2016), which as mentioned earlier is the messenger involved in switching the enzymatic activity of CckA from a kinase to a phosphatase (Lori et al., 2015). The binding of c-di-GMP to CckA occurs through the CA domain, is assisted by conformational changes mediated by the PAS-B domain, and binding of ADP appears to stimulate the effect of c-di-GMP (Dubey et al., 2016; Mann et al., 2016). The receiver domain is in the C-terminus and contains the conserved aspartyl residue to which phosphate is transferred from the DHp domain (Asp623 in C. crescentus, Asp696 in R. capsulatus). Loss of this residue renders the protein unable to transfer phosphate to ChpT and to loss of phosphatase activity as well (Y. E. Chen et al., 2009). 24  1.5.4 The ChpT phosphotransferase ChpT is the histidine phosphotransferase that shuttles the phosphate from the RD of CckA to the RD of CtrA. The C. crescentus ChpT is 253 amino acids long, while the R. capsulatus ChpT is 211. Like HKs, ChpT adopts a homodimeric or homotetrameric structure in which each monomer consists of a DHp domain and a CA domain (Blair et al., 2013). The phosphorylation site of ChpT is His33 in C. crescentus and His31 in R. capsulatus . (Fioravanti et al., 2012). Although it presents some structural similarities to HKs, ChpT is incapable of autophosphorylation, as it lacks autokinase activity (Fioravanti et al., 2012). 1.5.5 The CtrA master response regulator Out of the three proteins composing the phosphorelay, CtrA is probably the most well-studied one; this due to the role it has as the master regulator of the C. crescentus cell cycle, where it controls the expression of at least 144 of the 553 cell-cycle related genes (Laub et al., 2002). C. crescentus CtrA is 237 amino acids long while R. capsulatus is 231 amino acids long. The two proteins have a 71% sequence identity, which goes up to 100% in the DNA-binding regions of the protein (Leung et al., 2013). CtrA belongs to the OmpR/PhoB family of response regulators with which it shares its basic structure and DNA binding domain (Spencer et al., 2009). CtrA consists of an N-terminal receiver domain and a C-terminal effector domain, also known as the transcriptional regulator (TR). The receiver domain interacts with the DHp domain of ChpT, catalyzing the transfer of phosphate onto the conserved Asp51 (in both C. crescentus and R. capsulatus) and is also capable of dephosphorylation, thus allowing for the transfer of phosphate back to ChpT (West & Stock, 2001). As in the effects of CckA and ChpT, loss of the conserved Asp residue results in CtrA becoming unable to receive the phosphate molecule (Quon et al., 1996). The last function of the RD is to mediate conformational changes that affect the effector domain in response to changes in phosphorylation. It is important to note that, despite their name, receiver domains are not passive partners in the phosphotransfer reaction as they are capable of catalyzing autophosphorylation from small molecules such as phosphoramidate and acetyl phosphate (West & Stock, 2001). The active state of the transcriptional regulator depends on the phosphorylation state of the RD. Usually, phosphorylation activates protein dimerization and causes the effector domain to bind 25  to DNA sequences that present direct repeats with tandem symmetry. Once bound to DNA RRs usually induce transcription, although they can also act as repressors (Bachhawat et al., 2005). However, although the active result of RR activation is DNA binding, some RRs can have effects on gene expression when not phosphorylated. This can be observed in R. capsulatus, where CtrA mutants that are not capable of undergoing phosphorylation have been shown to still have regulatory effects (Brimacombe et al., 2014). Phosphorylated and non-phosphorylated CtrA have different effects in the gene expression of the RcGTA genes, with CtrA being involved in the expression of the RcGTA structural gene cluster while CtrA~P activates the expression of lysis and maturation genes (Westbye et al., 2018). 1.6 CckA/ChpT/CtrA in other organisms Homologues of the CckA/ChpT/CtrA phosphorelay are found in many organisms, being present in most alphaproteobacteria, where it controls a variable range of functions, although some of these are conserved in most species: cell division, motility and signal transduction (Brilli et al., 2010). Understanding the similarities and differences in function of these systems between different organisms helps us gain a better understanding of such regulatory processes as a whole. One key example of this is how the CtrA gene is essential in the Rhizobiales and Caulobacterales but is not essential in the Rhodobacterales. A better understanding of the reasons for such differences can only improve the understanding of this type of regulatory system in all organisms (Wang et al., 2014). 1.6.1 The CckA/ChpT/CtrA phosphorelay in the C. crescentus cell cycle. C. crescentus has long been studied as a model of bacterial cell cycle regulation due to the peculiarity of its own cell cycle. C. crescentus undergoes asymmetric cell division where each division produces two morphologically distinct cells: a sessile ‘stalked’ cell that is capable of continuing its cell cycle immediately after division occurs, and a mobile ‘swarmer’ cell that arrests its cell cycle at the G1-S phase transition, and cannot replicate until it has differentiated into a stalked cell (Y. E. Chen et al., 2009; Curtis & Brun, 2010). As mentioned before CtrA is known to control the expression of at least 25% of the genes involved in the C. crescentus cell cycle, as well as genes involved in flagellar biosynthesis (Laub et al., 2002), and as such the control of CtrA sits at the center of a complex regulatory network that must be controlled both temporally and spatially for proper cell division to take place. 26  CtrA is directly controlled by the action of the CckA/ChpT/CtrA phosphorelay, which is bifurcated as ChpT has a second target, the protein CpdR (Y. E. Chen et al., 2009), unlike in R. capsulatus which lacks a CpdR homologue. The phosphorelay works as described previously, with CckA being controlled to enhance either its kinase or its phosphatase activity. When its kinase activity is upregulated CckA increases the flow of phosphate towards CtrA and CpdR, thus increasing the concentrations of their phosphorylated forms CtrA~P and CpdR~P. CtrA~P has an increased affinity for the C. crescentus origin of replication (Cori), binding to it and arresting the cell cycle before the S phase. This is the cell state in which swarmer cells are, CckA kinase activity is upregulated and this results in the cell cycle becoming arrested at the G1 phase. For the cell cycle to proceed CtrA must be deactivated, and although this is directly mediated by CckA-mediated dephosphorylation, CpdR adds a second regulatory prong to CtrA deactivation (Y. E. Chen et al., 2009; Curtis & Brun, 2010). CpdR~P is the inactivated form of the protein, and so when the flow of phosphate within the phosphorelay is reversed CpdR becomes activated and helps localize the protease ClpXP to CtrA. ClpXP then proceeds to degrade CtrA to help reduce CtrA levels overall and facilitate the release of activated CtrA~P bound to the Cori site (Heinrich et al., 2016).  The key factor for these changes to occur is the modulation of the activities of CckA, which are in turn regulated by both cellular levels of c-di-GMP and other proteins, including another phosphorelay. Stalked (divisionally-active) cells have an active HK called DivJ that mediates the phosphorylation of the single-domain RR DivK. When phosphorylated, DivK blocks the activity of the protein DivL, a non-conventional HK that increases CckA kinase activity. Stalked cells also present a diguanylate cyclase (DGC) protein called PleD, which increases concentrations of c-di-GMP. The lack of DivL activity and the presence of c-di-GMP switch CckA to its phosphatase form thus resulting in decreased levels of CtrA~P and allowing for DNA replication to take place (Jenal et al., 2017; Lori et al., 2015; Tsokos & Laub, 2012). Swarmer (non-divisional) cells lack DivJ and in its place possess PleC, an HK with primarily phosphatase activity. PleC mediates the dephosphorylation of PleD and DivK, thus simultaneously reducing c-di-GMP levels and allowing DivL to interact with CckA to activate 27  its kinase activity. The levels of c-di-GMP are further reduced by the activity of the PdeA phosphodiesterase. These factors favor the kinase activity of CckA and the phosphorylated state of both CtrA and CpdR so that CtrA~P is not degraded and is bound to the Cori site (Jenal et al., 2017; Tsokos & Laub, 2012). In pre-divisional cells DivJ and PleD localize to the stalked cell pole, and PleC and PdeA localize to the swarmer cell pole to allow for the temporal and spatial control of CtrA~P levels (Figure 8); this helps newly divided, differentiated cells acquire their regulatory roles immediately after separation occurs (Curtis & Brun, 2010; Jenal et al., 2017). The specific role that the CckA/ChpT/CtrA phosphorelay takes is to function at the center of this network to mediate and integrate the inputs of the other regulatory signals to control the levels of CtrA and CtrA~P, which will in turn affect DNA replication and the levels of gene expression.   Figure 8. The CckA/ChpT/CtrA phosphorelay in C. crescentus. Modified from (Tsokos & Laub, 2012) 28  1.6.2 Other bacteria with similar regulatory networks. Another organism that contains a homologous phosphorelay is the pathogen Brucella abortus. This phosphorelay, as in C. crescentus phosphorelay is a CckA/ChpT/CtrA/CpdR system and a major observed function is to arrest the cell cycle at the G1 stage. However, the phosphorelay is also involved in the expression of several virulence factors. Being an intracellular pathogen B. abortus needs to tightly regulate gene expression to properly reach its intracellular niche, some of such genes are those required for production of a T4SS that allows them to infiltrate the cell, and genes required for vacuole acidification and long-term survival. Additionally, the B. abortus CtrA is involved in production of the bacterial envelope, specifically with the production of LPS and outer membrane proteins. This last type of regulation is also observed in the pathogen Ehrlichia chaffeensis (Poncin et al., 2018). One difference between C. crescentus and B. abortus CtrA regulatory roles is that chromosomal replication in the latter is regulated by DnaA instead of CtrA. Another member of the Rhizobiales that possesses a CckA/ChpT/CtrA regulatory system is the soil bacterium Sinorhizobium meliloti. The system in this organism is similar to the one in C. crescentus as it also expresses a DivJ/PleC/DivK system with an additional component, the HK CbrA. This system also mediates the symbiotic activity between S. meliloti and the plant Medicago sativa, and in this organism the DivK phosphorylation branch of the phosphorelay is essential, unlike C. crescentus (Pini et al., 2013). This shows that the additional factors that control the phosphorelay can vary, and change the output effects between different organisms. Agrobacterium tumefaciens is another plant-related organism that contains this system. A. tumefaciens presents two additional kinases homologous to PleC and DivJ, called PdhS1 and PdhS2, that are involved in the regulation of motility and biofilm formation. A study of the PleD and DivK proteins of this organism, which are involved in c-di-GMP modulation, produced a  CckA Y674D mutant that is resistant to the phosphatase-stimulating effect of c-di-GMP (Kim et al., 2013). A. tumefaciens regulates its chromosomal replication and segregation through CtrA, but not the cell cycle (Poncin et al., 2018). In contrast to the previous examples, the bacterium Dinoroseobacter shibae possesses the phosphorelay but it is not essential for its survival. It contains the ctrA, chpT and cckA genes, which are involved in regulating the morphology of its different cellular subtypes during cell 29  division, as well as in the regulation of motility genes. However, the expression of the phosphorelay genes in D. shibae is regulated by quorum sensing (QS) genes, specifically through a LuxR/I system, which is integrated into the phosphorelay (Wang et al., 2014). As described in the following section, RcGTA gene expression in R. capsulatus was found to be induced by QS (Leung et al., 2012; Schaefer et al., 2002), which highlights how these regulatory systems may be integrated in similar ways in different microorganisms. 1.7 Quorum sensing and its role in RcGTA gene expression. Quorum sensing can be described as the process through which gene expression in a bacterial population is regulated at a community level, by the production and release of chemical signalling molecules that increase in concentration as cell density becomes higher (Miller & Bassler, 2001). The most well-known type of QS is the acyl-homoserine lactone (HSL) signalling system mediated by the LuxI/LuxR proteins, also known as the LuxI/R-type QS system. The LuxI protein is an HSL-synthase and LuxR is a transcription factor that becomes activated when bound by HSL, increasing expression of LuxI and any target genes regulated by the system, as well as reducing transcription of its own gene. When the production of HSL in the population reaches a threshold level LuxR becomes activated, further increasing the production of LuxI and consequently of HSL. (Miller & Bassler, 2001). R. capsulatus synthesizes a 16-carbon tail-long HSL (C16-HSL) produced by a LuxI-type protein named GtaI (Schaefer et al., 2002). This C16-HSL was shown to both increase the expression of the RcGTA structural gene cluster and induce the production of RcGTA particles in R. capsulatus cultures, which is consistent with the observations that RcGTA production peaks at the start of the stationary phase (Solioz et al., 1975). Further studies identified the LuxR-type component of the system and named this protein GtaR. GtaR is co-transcribed with GtaI and in the absence of C16-HSL represses the transcription of gtaRI. This system differs from typical LuxI/R-type systems in that it amplifies the HSL signal not by activation but by de-repression of transcription. This system was confirmed to regulate RcGTA expression through transcription experiments, but no GtaR-binding sequences were found near the RcGTA promoters (Leung et al., 2012). Furthermore, in a QS (gtaI) mutant the magnitude of ctrA expression was about 50% of the WT strain. It would later be found that QS 30  and CtrA regulate RcGTA expression through the expression of gafA, the direct activator of the RcGTA operon (Fogg, 2019). The GtaI/R system was also shown to be important not only for the expression of the RcGTA structural genes but also for recipient capability; it regulates the expression of the genes required for the synthesis of the capsular polysaccharide receptor (Brimacombe et al., 2013), as well as the expression of dprA (Brimacombe et al., 2014), one of the genes required for RecA-mediated recombination to occur (section 1.4.2). All of this reinforces the suggestion that the RcGTA system evolved in concert with R. capsulatus, as both the production and recipient capabilities are induced by high concentrations of cells. Therefore, this ensures that when RcGTA production occurs recipient cells are expressing the genes required for gene transfer, increasing the exchange of alleles in the population. 1.8 Current model for the regulation of RcGTA gene expression Today, many pieces of the RcGTA regulatory model have been discovered and characterised, with the CckA/ChpT/CtrA phosphorelay playing a central role that ensures the differential expression of RcGTA-structural genes, lysis and maturation genes and recipient capability. Earlier research could not explain how CtrA mediated the expression of these genes, as the RcGTA core cluster lacked CtrA-binding sequences (Lang & Beatty, 2006). The recent work of the former Beatty lab member Paul Fogg has cleared up this matter by identifying the rcc01865 gene product (GafA) as the inducer of the RcGTA structural genes. Lang’s lab had shown that rcc01865 is regulated by CtrA, but Fogg additionally implicated the GtaI/R system, and showed that GafA binds to the RcGTA promoter (Fogg, 2019). The discovery of GafA has helped tie together the QS system and the phosphorelay in regards to the expression of RcGTA genes. Homologues of the C. crescentus DivL atypical HK and the ClpXP protease were identified in R. capsulatus, and these were found to be involved in the regulation of the CckA/ChpT/CtrA phosphorelay in a similar manner to C. crescentus. Loss of DivL was correlated with reduced CckA kinase activity and increased phosphatase activity, which led to a decrease in the expression of  genes activated by CtrA~P. Loss of ClpX resulted in increased levels of CtrA, a result consistent with the expected role of ClpXP in downregulation of CtrA through protease activity (Westbye et al., 2018). 31  The LexA protein, master regulator of the SOS response DNA repair pathway, was also discovered to be involved in the regulation of RcGTA genes. Mutants lacking the lexA gene were found to have increased expression of cckA that led to decreased RcGTA production. Further analysis showed that these mutants presented elevated CckA levels with increased phosphatase activity, likely a result of an excess of CckA in relation to DivL. Intriguingly these cells did not produce RcGTA capsids, the production of which is believed to be upregulated by non-phosphorylated CtrA. This suggests that the CckA/ChpT/CtrA phosphorelay controls other targets besides the phosphorylation state of CtrA (Kuchinski et al., 2016). A current model for the regulation of RcGTA gene expression is presented in Figure 9. The total amount of CtrA is decreased by the ClpXP protease and increased by QS (not shown). The balance of CtrA and CtrA~P is regulated by the activity of the CckA/ChpT/CtrA phosphorelay, specifically through the balance of kinase/phosphatase activities of CckA, which is itself regulated through the effects of DivL and c-di-GMP. Non phosphorylated CtrA, along with the GtaR/I system, mediates the activation of gafA which induces the transcription of the RcGTA structural gene cluster and the lysis genes. When the balance of the CckA/ChpT/CtrA phosphorelay activity shifts to kinase it leads to the accumulation of CtrA~P, which activates the expression of the accessory structural genes, maturation genes, and cell lysis genes.  32   Figure 9. Regulatory model of RcGTA gene expression. Taken from (Fogg, 2019) The biggest and yet unresolved question regarding RcGTA expression is the following: What controls which cells become producers and which cells become recipients? We know that quorum sensing is involved and that the process is regulated through the CckA/ChpT/CtrA phosphorelay, but the exact signal that controls the “switching on” of RcGTA production in a small subset of cells is still unknown. A recent advance has occurred regarding this question. An RcGTA overproducing R. capsulatus mutant has been used to study RcGTA since the 1970s (Yen et al., 1979), but the exact nature of the mutation had never been identified. A 2019 work by Ding et al. identified the mutation, present in the rcc00280 gene, and through bioinformatic analysis predicted its protein product to be an extracellular calcium-binding protein (Ding et al., 2019). Additionally, this research 33  confirmed that the activation of the RcGTA genes is essentially a random process and that the Rc280 protein is an extracellular signal that initiates a pathway that leads to the activation of gafA through the CckA/ChpT/ CtrA phosphorelay. However, the nature of the Rc280 pathway remains unknown, as well as the exact mechanisms by which the phosphorelay is regulated to modify its regulatory output to distinguish between producing and recipient cells. Although plenty of experimental evidence points towards the R. capsulatus CckA/ChpT/CtrA phosphorelay behaving as the C. crescentus one does, this has not been biochemically verified (Westbye et al., 2018). Within the phosphorelay the CckA protein is where most of the controlling inputs and changes occur in C. crescentus and other organisms (Dubey et al., 2016; Kim et al., 2013). Although experimental evidence has shown that site-specific mutations in the R. capsulatus cckA affect expression of the RcGTA genes in vivo (Wiesmann, 2016) there are no data showing the biochemical effects of these mutations on kinase or phosphatase activity.   34  1.9 Structural insights into CckA The structural study of CckA in C. crescentus and R. capsulatus has given much information towards understanding how this protein can be regulated to alter its catalytic output. Some of the most important amino acid residues in CckA have already been mentioned in section 1.5.3:  The phosphate receiving group in the HK domain (C. crescentus H322 and R. capsulatus H399) and the RC domain (D623/D696), and how their loss affects the activity of CckA in C. crescentus, with loss of H322 causing a loss of autokinase activity and loss of D623 causing a loss of both kinase and phosphatase activities. However, several other regions and residues have been implicated in the regulation of CckA activity. A CckA V366 mutant (V443 in R. capsulatus) was shown to present normal kinase activity but lack phosphatase activity. The exact mechanism behind this effect is unknown but has been suggested to be a result of affected domain-domain interactions (Y. E. Chen et al., 2009). Other mutants have been created that affect CckA activity in different ways: C. crescentus mutants G318T and G319E (G395T and G396E in R. capsulatus) present increased kinase activity but lack phosphatase activity; the mechanism behind this effect has not been fully elucidated but the lack of phosphatase activity is likely a result of hyperactive kinase activity (Y. E. Chen et al., 2009). Analysing how c-di-GMP binds to and alters the CckA structure is also highly important to understand how CckA activity is regulated, as c-di-GMP is the main molecule that switches CckA from its kinase to its phosphatase activity. An X-ray crystallographic structural analysis of CckA revealed that the catalytic region (CA domain) contains most of the amino acid residues responsible for binding of both c-di-GMP and ATP, although several residues in the DHp domain assist with this function as well (Dubey et al., 2016). When binding to CckA one c-di-GMP molecule forms two hydrogen bonds with the atoms of the main chain of I524 (I599 in R. capsulatus) and stacks with Tyr514 (Y589) and Trp523(Y598). A Y514D mutant presents normal kinase activity and basal phosphatase activity, but is incapable of binding c-di-GMP and thus unable to fully activate its phosphatase activity, an effect which is also observed on the analogous A. tumefaciens CckA Y674D mutant (Kim et al., 2013; Lori et al., 2015). 35  An interesting observation is that although the Trp523 residue interacts with c-di-GMP directly, there is no Trp residue in the DHp nor the CA domains of the R. capsulatus CckA, having a tyrosine at the equivalent position instead (Tyr598). The implications that this might have on the function of the R. capsulatus CckA are unknown. In vivo experiments done on the respective CckA mutants of R. capsulatus have yielded data that support the RcGTA model of regulation in the context of CckA activity. R. capsulatus cells expressing CckA H399A present an increase in intracellular RcGTA production with diminished release, attributed to decreased levels of CtrA~P. These same cells also show increased levels of recipient capability (Wiesmann, 2016). However, the results observed in V443P and Y589D mutants are not as clear: a V443P mutant presents increased expression of genes with a CtrA~P promoter as expected, but also presents a slight increase in production of RcGTA particles and decreased release, which contradicts what would be expected from increased CtrA~P levels , as well as showing no decrease in recipient capability. A Y589D mutant fails to produce significant increase in the expression of genes with a CtrA~P promoter, but also shows a decrease in RcGTA production and lacks observable recipient capability (Wiesmann, 2016). These results suggest that there is still much to be discovered regarding the regulatory capabilities of the CckA/ChpT/CtrA phosphorelay. Further study into it will allow for a better and more complete understanding of the role it has in the control of RcGTA gene expression.   36  2 Hypotheses and objectives  The current model of RcGTA regulation has non-phosphorylated CtrA activating gafA, allowing for transcription of the core RcGTA genes. Subsequently, CtrA is phosphorylated via CckA/ChpT, and CtrA~P activates the expression of genes required for particle maturation and cell lysis. Although much research supports the CckA/ChpT/CtrA proteins forming a phosphorelay in R. capsulatus as they do in C. crescentus, solid biochemical evidence confirming the existence of this phosphorelay is still lacking. Previous in vivo studies on the effects of site-directed CckA mutants on the expression of RcGTA in respect to this model were inconclusive, mainly because it was unclear if the CckA residues targeted function in R. capsulatus in the same manner as in C. crescentus. I hypothesized that the CckA/ChpT/CtrA proteins form a phosphorelay in R. capsulatus and that this phosphorelay can be replicated through phosphorylation assays in vitro. If the phosphorelay can be replicated in vitro then the effects of the R. capsulatus CckA mutations on CckA activity as a kinase and a phosphatase can be determined by phosphorylation assays.  To test these hypotheses my objectives were to: Purify His-tagged R. capsulatus CckA, ChpT, and CtrA proteins: I use previously constructed E coli strains containing plasmids encoding 6 His-tagged recombinant proteins, to obtain purified samples of R. capsulatus CtrA, ChpT and the cytoplasmic domain of CckA (Table 2). Protein identity was verified through SDSPAGE and time of flight (ToF) mass spectrometry. Evaluate phosphate transfer between CckA, ChpT and CtrA proteins by PhosTag PAGE analysis: Kinase, phosphatase and phosphotransfer assays were designed, implemented and optimized to evaluate both the putative kinase and phosphatase activities of CckA within the phosphorelay. I adapted a PhosTag SDSPAGE method (Barbieri & Stock, 2008; Gao & Stock, 2013) to determine the phosphorylation state of the purified proteins after phosphorylation assays, allowing me to verify the type of activity presented by the proteins.  Purify and analyse CckA mutants H399A, V443P, and Y589D, and compare their activities to the WT CckA: These mutant proteins were subjected to the same extraction and purification protocol as the WT proteins. Using the optimized phosphorylation protocols, I ran a set of assays to identify the kinase and phosphatase capabilities of these mutants.  37  3 Materials and Methods 3.1 Strains used: Escherichia coli BL21(DE3) [New England Biolabs]  3.2 Plasmids derived from pET28a (Novagen): Table 2. Plasmids used in this work. Name Description pET28a::chpT Contains the R. capsulatus chpT gene modified to encode 6 His residues at the N terminus. The chpT sequences extend from a NcoI site located 56 bp before the start codon to an EcoRI site 44 bp beyond the stop codon. pET28a::ctrA Contains the R. capsulatus chpT gene modified to encode 6 His residues at the N terminus. The ctrA sequences extend from a NcoI site located 56 bp before the start codon to an EcoRI site 16 bp beyond the stop codon. pET28a::cckA Contains the R. capsulatus cckA gene modified to encode 6 His residues at the C terminus, the N terminus region corresponding to the transmembrane domains has been removed (207 bp). The cckA sequences extend from a NcoI site located 2 bp before the start codon to a HindIII site 39 bp before the stop codon. pET28a::cckA H399A Same as pET28a::cckA but with a CA to GC change 991 and 992 bp after the start codon, this changes the codon from CAT to GCT thus changing the translation from H to A. pET28a::cckA V443P Same as pET28a::cckA but with a GT to CC change 1123 and 1124 bp after the start codon, this changes the codon from GTC to CCC thus changing the translation from V to P. pET28a::cckA Y589D Same as pET28a::cckA but with a T to G change 1561 bp after the start codon, this changes the codon from TAC to GAC thus changing the translation from Y to D. 38  3.3 Growth conditions: Escherichia coli strains containing the recombinant plasmids were grown in Luria-Bertani (LB) media supplemented with kanamycin sulfate (30 µg/mL) (Sambrook, J. ; Fritsch, E. F. ; Maniatis, 1989) at 37°C and shaking at 200 rpm. 3.4 DNA extraction: Plasmid DNA was extracted from E. coli cells using an alkaline lysis protocol [Qiagen protocol]. Extracted DNA was quantified by measuring absorbance at 280 nm in a Thermofisher NanoDrop spectrophotometer. 3.5 Sanger sequencing: DNA sequencing was done via Sanger sequencing using Genewiz sequencing services using the T7 terminase primer provided by the service and for cckA additional primers designed to bind in the coding region. 3.6 Induction: For protein expression E. coli strains were cultured overnight in tubes containing 8 mL of medium and then used to inoculate 300 mL of medium. These cultures were incubated until reaching an OD of approximately 80 Klett units (100 Klett units = ~4 x 108 CFU/mL) at which point IPTG was added to the cultures to a final concentration of 1 mM. Cells were then incubated for 3 more hours. After incubation was finished cultures were transferred to 50 mL tubes and centrifuged at 4000 x g for 10 minutes. The supernatant liquid was discarded and pellets were frozen overnight at -20°C. 3.7 Protein extraction: Cell pellets were resuspended in 4 mL of 1x Native Purification Buffer (50 mM Na2HPO4, 0.5 M NaCl, pH 8.0) and pooled together in a single tube. Lysozyme and DNase I (a few crystals) were then added to this sample and incubated on ice for 10 minutes. The cell suspension was then passed through a chilled French press at a pressure of ~3500 psi twice to lyse the cells (Kuchinski et al., 2016). The lysate was then centrifuged at 4000 x g for 15 minutes to pellet cell debris and the supernatant was transferred to a fresh tube with a 50 µL aliquot for later SDSPAGE analysis, and frozen at -20°C. 39  3.8 Protein purification: Columns were packed with 2 mL of Ni-NTA agarose and washed once with water and once with 1x Native Purification Buffer. The cell lysate was then added to the column and let bind for 1 hour at room temperature, with mixing every 15 minutes. After incubation the column was opened and flow-through was collected for analysis. Columns were then washed with 40 mL of 20 mM imidazole wash buffer (50 mM Na2HPO4, 0.5 M NaCl, 20 mM imidazole, pH 8.0), and wash fractions were also collected for later analysis. Finally, protein was eluted using 5 mL of 250 mM imidazole elution buffer (50 mM Na2HPO4, 0.5 M NaCl, 250 mM imidazole, pH 8.0). These 5 mL were collected as 1 mL fractions and 50 µL were taken from each for later analysis. In order to remove the imidazole from the solution and do a buffer exchange; the elution fractions were mixed together and diluted with kinase activity buffer (10 mM HEPES-KOH, 50 mM KCl, 10% glycerol, 5 mM MgCl2, pH 8.0) (Skerker et al., 2005) to a final volume of 12 mL and centrifuged in a 10 kDa-molecular weight cut-off Macrosep Advanced Centrifugation Device (Pall Corporation) at 5000 x g for 30 minutes. The non-filtered solution was then resuspended in kinase buffer to a volume of 4 mL and purified in a second step through an Amicon Ultra Centrifugation Device (Amicon) of the appropriate molecular weight cut-off (50 kDa for CckA, 10 kDa for ChpT and CtrA) at 4000 x g for 20 minutes (Mónico et al., 2017). This step was repeated 3 times, resuspending the upper fraction in kinase buffer at a final volume of 2 mL. Protein concentration was measured through absorbance at 280 nm (Hitachi U-2000 spectrophotometer). Purified protein samples were separated into 50 µL aliquots and frozen at -80°C. 3.9 Phosphotransfer profiling reactions: Proteins of interest were obtained in a kinase reaction buffer (10 mM HEPES-KOH, 50 mM KCl, 10% glycerol, 5mM MgCl2, pH 8.0) modified from (Lori et al., 2015). Phosphorylation reaction mixes were done by adding CtrA and ChpT at equimolar concentrations 4x the molar concentration of CckA (this was done to add approximately the same mass of each protein to facilitate visualization of gel bands). When used, additional reagents were present at the following concentrations: ATP/ADP were added at ~100x the molar 40  concentration of CckA, acetyl phosphate was added at ~2000x the molar concentration of CtrA (Lukat et al., 1992). Cyclic di-GMP was added at equimolar concentration of CckA. Reaction mixes were started by addition of phosphate donors (ATP or Ac~P) and incubated at 30°C for the designated periods of time (30,60,90,120 min). Reactions were stopped by adding 5 µL of 4X SDS-PAGE loading buffer to the mixes and placing samples on ice. This protocol was based on the protocols presented by (Gao & Stock, 2013; Laub et al., 2007). For the evaluation of phosphate transfer within the phosphorelay reactions were prepared as follows: 15 µL reaction mixes were prepared to have the following final concentrations: CtrA 22 µM, ChpT 27 µM, CckA 7 µM. When ATP was used it was added at a final concentration of 800 µM. Reaction mixes were incubated at 30°C for 90 minutes.  3.10 CckA activity assays For CckA kinase-phosphatase activity assays samples were prepared as follows: a master mix was made that consisted of 5 µM CckA and 15 µM CtrA and ChpT. Final volume of the mix was determined by the number of timepoints to be taken, with 15 µL samples being taken at each timepoint. Phosphate donors (Ac~P or ATP) were added to start the reaction and samples were incubated at 30°C. At the indicated timepoints (every 7.5 minutes for ATP and every 30 min for Ac~P) 15 µL of sample were taken out of the master mix, mixed with 5 µL of 4X SDS-PAGE loading buffer and placed on ice. After the indicated time for the kinase reaction had passed (15 min for ATP, 60 min for Ac~P) and the sample for that timepoint was taken, the remaining volume was separated into two different tubes: one to act as a negative control, and one as the experimental sample. Experimental samples were supplemented with either c-di-GMP (7 µM final concentration), ADP (5 mM final concentration), or both depending on the reaction. Control samples were not supplemented and all samples continued incubation at 30°C with 15 µL being taken at the same intervals as before. After all timepoints were taken, and reactions were stopped and placed on ice, samples were run in PhosTag gels as indicated in the following section. 41  3.11 PhosTag PAGE: Polyacrylamide gels were prepared at a 10%-12% concentration of acrylamide/bis-acrylamide (10% for CckA-only gels, 12% for gels with other proteins present). PhosTag was added at a 50 µM concentration along with 100 mM MnCl2. SDS was not added to the gels but it was present in the running buffer (0.4% (w/v) SDS, 25 mM Tris, 192 mM glycine). All gels were run for 90-120 minutes at 100 volts.(Barbieri & Stock, 2008; Gao & Stock, 2013). 3.12 Time-of-flight mass spectrometry: CckA and CtrA proteins were overexpressed and purified from corresponding E. coli strains. Each individual purified protein was then treated with a phosphate donor in a 50 µL reaction. CtrA was incubated with acetyl phosphate at ~2000x the molar concentration of CtrA, CckA was incubated with ATP at ~100x the molar concentration of CckA. Reaction mixes were incubated for 2 hours at kA. Reaction mixes were incubated for 2 hours at 30°C and placed on ice immediately after incubation. Samples were then diluted to 1 mg/mL protein. Samples at a concentration of 1 mg/mL were diluted 1:250 in a solution of 5% acetonitrile, 0.1% formic acid. 5 µL of this sample were then injected into a 5 mm C4 Pepmap desalting column (LC Packings) connected to a Waters Xevo GS-2 QToF mass spectrometer via a NanoAquity UPLC system. Samples were eluted in a 2’minute gradient from 5-100% acetonitrile at 20 µL/min. The spectra were summed and deconvoluted using Waters’ MaxEnt algorithm (Cottrell & Green, 1998). These analyses were done at the Proteomics Core Facility of the Michael Smith Laboratories at the University of British Columbia.  42  4 Results 4.1 Creation of an R. capsulatus CckA/ChpT/CtrA phosphorelay in vitro. To determine whether the R. capsulatus CckA, ChpT and CtrA proteins form a phosphorelay as the one in C. crescentus, I carried out a set of phosphorylation assays based on the phosphotransfer profiling protocols presented by Laub et al. and Barbieri & Stock (Barbieri & Stock, 2008; Laub et al., 2007). These methods allow for in vitro systematic identification of the components of TCS pathways and phosphorelays, purified HKs are evaluated for their ability to autophosphorylate and transfer phosphoryl groups to other components of the system. I overexpressed 6-His-tagged CckA, ChpT and CtrA (Table 2.) in E. coli cultures and purified them through Ni-NTA column chromatography, these purified proteins were then used for phosphotransfer profiling assays as described in section 3.9. Normally, phosphotransfer profiling assays involve radiolabelling of ATP with 32P (Laub et al., 2007). Radiolabelling however can carry certain limitations, as only a fraction of the phosphoryl groups is labelled and this fraction is difficult to properly quantify (Gao & Stock, 2013). A more recent method that has been developed to evaluate phosphate transfer is PhosTag polyacrylamide gel electrophoresis (PhosTag PAGE). PhosTag is a dinuclear metal complex that acts as a phosphate-binding chelator; when added to a polyacrylamide gel in the presence of Zn2+ or Mn2+ it interacts with phosphate groups in proteins running through the gel during electrophoresis, slowing their migration (Barbieri & Stock, 2008).  In my phosphotransfer profiling analysis I used PhosTag Mn2+ gels to identify transfer of phosphate between the purified proteins. I first evaluated if CckA, ChpT or CtrA present autophosphorylation activity by incubating them with ATP. Phosphorylation reactions were then assayed in a PhosTag PAGE, and phosphorylation of individual proteins was observed as the appearance of additional bands or a shift in the mobility of existing bands (Gao & Stock, 2013), with slower-migrating bands corresponding to phosphorylated forms of the proteins being analyzed. 43   Figure 10. The Rhodobacter capsulatus CckA/ChpT/CtrA phosphorelay. Phosphate transfer occurs in vitro between R. capsulatus proteins CckA, ChpT and CtrA. CckA autophosphorylates in the presence of ATP, then transfers phosphate to ChpT which finally transfers phosphate to CtrA. A. PhosTag SDS-PAGE of phosphorylation reaction mixes of the indicated proteins. B. Standard SDS-PAGE of same set of samples. Differential band migration in PhosTag SDS-PAGE results from differences in both molecular weight and phosphorylation state. Band shift in reaction mixes with ATP present is only observed in PhosTag gel indicating that it is due to the formation of a phosphorylated form of the protein.  As shown in Figure 10A, when either CtrA or ChpT were incubated in the presence of ATP there was no difference in banding pattern compared to their untreated controls. However, CckA did present a band shift, indicating that CckA is capable of autophosphorylation. Although the original band remained present, indicating only partial phosphorylation. Additionally, CckA as purified from E. coli consisted of a mixture of predominantly non-phosphorylated CckA and a small amount of CckA~P, indicating a small amount of partial autophosphorylation present before any in vitro treatment.  44  Next, to determine if phosphate transfer occurs between pairs of proteins, and to determine the directionality of the phosphate transfer, the three different proteins were mixed pairwise (CtrA+ChpT, CtrA+CckA and CckA+ChpT) and incubated with ATP (Figure 10A). when incubated together CtrA and ChpT showed no changes resulting from the presence of ATP. These data confirmed the results obtained when each of these proteins were incubated alone with ATP and showed that CckA is needed for phosphorylation. CtrA was not phosphorylated when incubated only in the presence of CckA, although CckA did autophosphorylate in this reaction. ChpT presented almost complete phosphorylation when incubated with CckA, showing that ChpT obtains phosphate from CckA in the absence of CtrA. Finally, to determine if phosphate transfer occurs between the three proteins, I incubated a mix of all three of them with ATP. The results (Figure 10A) showed the formation of a new band for both CckA and ChpT, indicating partial phosphorylation. Meanwhile CtrA shifted almost in its entirety, indicating almost complete phosphorylation. In summary, these results show that the phosphate transfer in this system occurs through ATP-dependent autophosphorylation of CckA, followed by phosphate transfer from CckA~P to ChpT and finally from ChpT~P to CtrA. I also tested the same samples in a non-PhosTag SDS-PAGE, to verify that the change in band position observed in the PhosTag gels was not a result of changes in molecular weight of the purified proteins caused by dimerization. As shown in Figure 10B, none of the band shifts observed in the PhosTag gels were present in the standard polyacrylamide gels, thus confirming that they are a result of phosphorylation.   4.2 Phosphorylation of R. capsulatus CtrA can occur via both CckA-mediated and acetyl phosphate-mediated pathways, with different kinetics. To properly evaluate and compare the kinase and phosphatase activities of CckA I needed to determine more specific incubation time points for CckA/ChpT-mediated phosphorylation of CtrA and to obtain phosphorylated CtrA without using CckA. In order to do this, I did timepoint assays for CckA-mediated phosphorylation of CtrA, and CtrA autophosphorylation with acetyl-phosphate (Lukat et al., 1992), analysing results in a PhosTag gel (Figure 11). 45   Figure 11. CckA-mediated phosphorylation of R. capsulatus CtrA increases over time and occurs more rapidly than phosphorylation with acetyl-phosphate. The no acetyl phosphate control (No Ac~P) was incubated for 120 minutes A. CtrA is phosphorylated via the CckA/ChpT/CtrA phosphorelay when incubated in the presence of ATP, presenting a higher amount of CtrA~P than non-phosphorylated CtrA within 60 minutes of starting the reaction. B. Autophosphorylation of CtrA via a low molecular weight phosphate donor such as acetyl phosphate can occur in absence of CckA and ChpT. The amount of CtrA~P generated within 60 minutes of starting the reaction is lower than the amount of non-phosphorylated CtrA. These results show that the CckA-mediated phosphorylation of CtrA is more efficient than CtrA autophosphorylation, having phosphorylated more than half the sample by the 1-hour mark. Although CtrA autophosphorylation was slow, and partial even after 2 hours, the amount of 46  CtrA~P obtained was sufficient to have a phosphorylated sample to be used in further experiments. 4.3 Time-of-flight mass spectrometry analysis of CckA and CtrA shows mass changes confirming phosphorylation. The R. capsulatus CckA and CtrA proteins are homologous to proven phosphorelay proteins (Jacobs et al., 1999; Lang & Beatty, 2002; Quon et al., 1996), and the PhosTag gel system has been shown to retard the migration of a variety of phosphorylated proteins relative to their non-phosphorylated forms (Barbieri & Stock, 2008; Gao & Stock, 2013). Nevertheless, it was thought that an alternative measurement of phosphorylation, a change in mass, would help confirm that the changes in the protein banding patterns observed in the PhosTag gels are due to phosphorylation of the respective proteins. Mass spectrometry analysis of samples of CtrA and CckA, in both non-phosphorylated and phosphorylated states (autophosphorylation with ATP for CckA, acetyl phosphate phosphorylation for CtrA), revealed that the phosphorylated samples presented shifts in mass corresponding to the covalent attachment of an 80 Da phosphate moiety (Figure 12). 47   Figure 12. Time-of-flight mass spectrometry analysis of CckA and CtrA samples corroborates phosphorylation. A: Untreated CtrA (left) and acetyl phosphate – treated CtrA (right). The second highest peak (28854.80 Da) in the treated sample presents a molecular weight 80 Da greater than that of the highest peak in the untreated sample. B: Untreated CckA (left) and ATP-treated CckA (right). The peak molecular weight shifts from 77563.70 Da to 77643.10 Da, a 79.4 Da difference. The untreated CtrA sample (Figure 12A – left) has one noticeable peak with a molecular weight of 28774.7 Da, close to the expected mass of CtrA (28905.25 Da). After phosphorylation (right), a smaller peak appears, corresponding to a molecule with a mass of 28854.8 Da, consistent with 48  change in molecular weight occurring as a result of phosphorylation. The low relative amount of CtrA~P to CtrA corresponds to the observed low efficiency of CtrA autophosphorylation (Section 4.2, Figure 11B). The CckA sample (expected mass of 77689.7 Da) incubated with ATP (Figure 12B – right) shows a high peak with a 79.4 Da shift in molecular weight from the highest peak in the untreated sample (Figure 12B – left), a change from 77563.7 Da to 77643.1 Da. The peak composition and distribution indicate that a high percentage of the CckA present in the sample was phosphorylated, with CckA~P being the most abundant form which supports the results from the PhosTag assays (Figure 12B).  The small differences in mass observed (for example the difference from 79.4 Da to the expected 80 Da shift in the CckA analysis) are probably due to differences between the theoretical and measured masses, although they can also be attributed to experimental errors and can be caused by poor data quality or overlapping signals. 4.4 CckA shifts from a kinase to a phosphatase in the presence of cyclic-di-GMP. To determine if R. capsulatus CckA exhibits a change from kinase activity to phosphatase activity in the presence of c-di-GMP, I used pre-phosphorylated CtrA to run a phosphatase assay. CtrA was autophosphorylated with acetyl phosphate, which was then removed from the sample, and the resulting mixture of CtrA and CtrA~P was added to a solution of CckA, ChpT, and c-di-GMP. Samples were taken at various timepoints and run in a PhosTag gel to observe changes in phosphorylation. 49   Figure 13. R. capsulatus CckA presents phosphatase activity. In addition to its kinase activity CckA can dephosphorylate CtrA~P in the presence of cyclic-di-GMP. Controls are CtrA/CtrA~P incubated in the absence of CckA, ChpT and c-di-GMP This (Figure 13) shows that the phosphatase reaction does occur as a result of c-di-GMP addition, and that it took place much faster than the phosphorylation reaction, with almost complete dephosphorylation occurring within the first 15 minutes. To corroborate the results from the previous experiment an assay was designed in which c-di-GMP was added directly to the reaction mix during incubation and taking samples before and after this addition.  50   Figure 14. R. capsulatus CckA activity can be switched from kinase to phosphatase by addition of c-di-GMP, observed by dephosphorylation of CtrA. This effect is observed whether CtrA is phosphorylated by the CckA/ChpT/CtrA phosphorelay (A), or autophosphorylated with acetyl phosphate (B). As seen in Figure 14, CckA switches from kinase to phosphatase following addition of c-di-GMP and dephosphorylation progresses over time. This occurred whether CtrA was phosphorylated by the CckA/ChpT/CtrA phosphorelay or by acetyl-phosphate-assisted autophosphorylation. 51  Although the phosphorylation and dephosphorylation rates were slightly different from those observed in previous results, phosphatase activity remained noticeably faster than kinase activity (Figure 14A). When being autophosphorylated with acetyl phosphate (Figure 14B) some CtrA~P was still observed at the final point of the assay (1-hour after c-di-GMP addition), likely because CtrA autophosphorylation is not stopped by the addition of c-di-GMP. In contrast, in the assay using CckA-mediated phosphorylation (Figure 14A) almost all the CtrA~P had been dephosphorylated after 30 minutes. This indicates that the addition of c-di-GMP halted CckA kinase activity and acts as a switch from CckA kinase to phosphatase activity. 4.5 CckA-mediated dephosphorylation of CtrA~P requires the presence of ChpT. As seen in Figure 10, the complete phosphorelay is needed to phosphorylate CtrA via CckA, however, in the phosphatase assays described above it is possible that ChpT was not needed for CckA-mediated dephosphorylation of CtrA~P, or that c-di-GMP is affecting the phosphorylation state of CtrA regardless of CckA activity. To address these questions an additional phosphatase assay was done. In this assay a reaction lacking either ChpT, CckA or both was used to evaluate whether the addition of c-di-GMP resulted in the dephosphorylation of CtrA~P. 52   Figure 15. Enzymatic dephosphorylation of CtrA~P requires the complete CckA/ChpT/CtrA phosphorelay. Addition of c-di-GMP does not result in dephosphorylation of CtrA~P in the absence of both ChpT and CckA (A) or either (B). Samples were added with c-di-GMP after 60 min of incubation, (*) marks samples into which c-di-GMP has been added.  As shown in Figure 15 all three proteins are required for the phosphatase reaction to take place; this shows that the entire phosphorelay needs to be present for phosphate transfer both from CckA to CtrA as well as from CtrA to CckA.  53  4.6 CckA phosphatase activity is not altered by the presence of ADP nor activated by ADP alone in the absence of c-di-GMP. Other research has resulted in the proposal that it is ADP bound to HKs that the causes the shift from kinase to phosphatase (Casino et al., 2014), and that the C. crescentus CckA binding of ADP and c-di-GMP is synergistic (cooperative), and maximal phosphatase activity is obtained in the presence of both (Dubey et al., 2016). Therefore, ADP addition was tested to determine whether ADP could stimulate or otherwise affect CckA phosphatase activity in my experimental model.  Figure 16. CckA c-di-GMP-activated phosphatase activity is not affected by the presence of ADP (A) nor activated by ADP alone in the absence of c-di-GMP (B). 54  As shown in Figure 16 adding ADP to CckA phosphatase assays showed no difference in activity from assays done in absence of ADP. Adding both c-di-GMP and ADP had the same effect as adding c-di-GMP alone (Compare Figure 16A with Figure 14B). Adding ADP alone did not switch CckA from a kinase to a to phosphatase, as the CtrA band continued to decrease in intensity after the addition of ADP (Figure 16B). ADP was not used in any of the previous assays and, because phosphatase activity was observed in phosphorylation assays without the addition of ADP, it appears that the R. capsulatus CckA differs from the C. crescentus CckA in its response to ADP. It could be argued that ADP was bound by CckA during the induction process in E. coli cells and remained bound during purification. However, it has been noted that in vivo CckA would be predominantly complexed with ATP, because of a greater affinity and higher concentration relative to ADP (Dubey et al., 2016). An alternative interpretation is that it is conceivable that ADP produced by kinase activity was bound to CckA because of a localized increase in concentration; further testing of this latter possibility is presented in section 4.7.1 and addressed in section 5.3.1 of my discussion. 4.7 Modification of specific CckA amino acid residues results in altered kinase and phosphatase activities. The amino acid residues actively involved in the kinase and phosphatase activities of CckA have been studied in the C. crescentus protein (section 1.9, (Dubey et al., 2016)). Previously, in vivo assays were done of R. capsulatus mutants that express forms of full-length CckA hypothesized to have altered kinase and phosphatase activities, based on the results observed from the C. crescentus studies (Wiesmann, 2016). The results obtained from these experiments have so far been inconclusive, as some of them do not entirely fit the proposed model for the regulation of RcGTA gene expression, as discussed in section 1.9 (Fogg, 2019). To reassess the role of these CckA mutants and analyse their activity in vitro, the altered proteins were purified and assayed using the same methods used to analyse the “WT” (His-tagged) R. capsulatus CckA, ChpT and CtrA phosphorelay proteins. Purified CckA mutant proteins were incubated together with ChpT, CtrA and ATP in some experiments, whereas in others the proteins were with acetyl phosphate. Kinase activity was assayed as the phosphorylation of both CckA and CtrA in the ATP reaction, while phosphatase activity was assayed as the dephosphorylation of CtrA~P in both reactions after the addition of c-di-GMP. 55  Autophosphorylation of CtrA with acetyl phosphate allowed for determination of phosphatase activity even in the absence of kinase activity, by observing the dephosphorylation of pre-phosphorylated CtrA. 4.7.1 CckA H399A lacks kinase activity and has normal phosphatase activity. CckA H399A was expected to lack kinase activity but present normal phosphatase activity. If it were true that, as in the C. crescentus CckA, ADP is required for c-di-GMP stimulation of CckA phosphatase activity, then the CckA H399A mutant lacking kinase activity would not be able to bind ADP resulting from ATP-mediated autophosphorylation in the producing E. coli cells (section 4.6, Figure 16), and thus would not present phosphatase activity. To test this, ADP was added to one of set of samples for the phosphorylation assays of the H399A mutant (Figure 17). 56   Figure 17. CckA H399A mutant lacks kinase activity (A) and presents normal phosphatase activity in the presence of c-di-GMP and ADP (B). The absence of ADP has little to no effect on CckA H399A phosphatase activity in response to c-di-GMP (C). 57  These results show that CckA H399A lacks kinase activity, as no formation of bands corresponding to either CckA~P or CtrA~P were observed when incubating in the presence of ATP (Figure 17A). It was also observed that CckA H399A has phosphatase activity, as the bands corresponding to CtrA~P diminished in intensity after addition of c-di-GMP (Figure 17B), in a similar manner to that observed in the WT CckA (Figure 16A).  Additionally, it seems that ADP has little if any effect on the level of phosphatase activity presented by CckA H399A (Figure 16C), as the rate at which CtrA~P is dephosphorylated (shown by the reduction in band intensity) is very similar between the assay done with both c-di-GMP and ADP and the one done with c-di-GMP only. This result is consistent with the observations done in previous assays (section 4.6, Figure 16) and confirms that the presence of ADP has little or no effect on CckA phosphatase activity. 4.7.2 CckA V443P lacks both kinase and phosphatase activities. CckA V443P was expected to present normal kinase activity but lack phosphatase activity, as per the results observed in the C. crescentus mutants (section 1.9, (Y. E. Chen et al., 2009)). The V443P mutant was assayed the same way as the H399A mutant, with ADP being added along with c-di-GMP for the phosphatase reaction, which was evaluated with both CckA/ChpT-mediated phosphorylation of CtrA and acetyl phosphate-mediated CtrA autophosphorylation.  58   Figure 18. CckA V443P mutant lacks both kinase and phosphatase activities. In (A) the No ATP control was incubated for 30 minutes. In (B) the no acetyl phosphate control was incubated for 120 minutes. An absence of phosphorylation of CtrA as seen in Figure 18A shows that the CckA V443P mutant is, unexpectedly, lacking in kinase activity in its entirety. The respective phosphatase activity was absent as expected (Figure 18B). The implications of these results are further discussed in section 5.3.2 (Discussion). 4.7.3 CckA Y589D presents normal kinase activity and lacks phosphatase activity. The CckA Y589D mutant was expected to present normal kinase activity accompanied by low baseline levels of phosphatase activity, because it was expected to be unable to bind c-di-GMP, 59  like the C. crescentus and A. tumefaciens homologues (Kim et al., 2013; Lori et al., 2015). This mutant protein was assayed in the same manner as the previous ones.  Figure 19. CckA Y589D mutant presents normal kinase activity (A) but lacks phosphatase activity (A and B). No ATP control was incubated for 30 minutes, No Ac~P control was incubated for 120 minutes. The formation of a band corresponding to CtrA~P observed in Figure 19A shows that CckA Y589D has kinase activity. When c-di-GMP was added to the reaction no loss of CtrA~P was observed in either of the conditions assayed, thus indicating the expected loss of phosphatase activity.   60  5 Discussion The expression of the R. capsulatus RcGTA genes is controlled by a network of bacterial regulatory mechanisms, including the proteins CckA, ChpT and CtrA (Lang & Beatty, 2002; Mercer et al., 2012). Homologues of these proteins form a phosphorelay in C. crescentus where they regulate the bacterial cell cycle and the expression of motility genes (Quon et al., 1996). Although ample experimental evidence exists for the activity and mechanisms of this phosphorelay in C. crescentus (Y. E. Chen et al., 2009), and continued research into RcGTA has led to the proposal of a regulatory model of RcGTA gene expression with the CckA/ChpT/CtrA phosphorelay at its center (Fogg, 2019), no solid biochemical proof existed that these proteins indeed form a phosphorelay in R. capsulatus. In this thesis I use phosphotransfer profiling assays coupled with a PhosTag PAGE method to show that the R. capsulatus CckA, ChpT and CtrA proteins form a phosphorelay in vitro. I also present data showing that c-di-GMP acts as a modulator of CckA activity, switching it from a kinase to a phosphatase and reversing the flow of phosphate within the phosphorelay. Additionally, I show that site-directed mutations on specific amino acid residues of CckA have effects on its kinase and phosphatase activities, and I analyze how the activity of these mutants is altered in comparison to their corresponding homologues in C. crescentus. 5.1 The R. capsulatus CckA, ChpT and CtrA proteins form a phosphorelay in vitro. To determine whether CckA, ChpT and CtrA form a phosphorelay in R. capsulatus I purified a set of recombinant R. capsulatus 6 His-tagged versions of CtrA, ChpT and the cytoplasmic domain of CckA after overexpression in E. coli. I then did a set of phosphotransfer profiling experiments on these proteins and analyzed the results from these experiments through phosphorylation-dependent gel mobility changes in PhosTag polyacrylamide gel electrophoresis (PhosTag PAGE).  The phosphotransfer profiling technique was developed by Laub, Biondi and Skerker for the systematic study of TCS pathways (Laub et al., 2007). Phosphotransfer profiling consists of the purification of the cytoplasmic domain of histidine kinases which are then autophosphorylated with ATP and tested for phosphotransfer against each possible receiver protein. This method was quickly applied for the study of the CckA/ChpT/CtrA phosphorelay in C. crescentus, but a similar approach was lacking in regard to the R. capsulatus homologues. 61  Most commonly, phosphotransfer profiling experiments have been performed in proteins that have been radiolabeled with radioactive phosphate (32PO4), but this method can provide inaccurate results as only a fraction of the phosphoryl groups are labeled and this fraction has proved to be difficult to quantify, as the non-phosphorylated proteins are not seen in autoradiography. An alternative has been provided in the form the PhosTag system, a method that allows for the separation of proteins in a polyacrylamide gel containing a phosphoryl group chelator (PhosTag) which results in differential protein migration depending on their phosphorylation status (Gao & Stock, 2013). In this method, both the phosphorylated and non-phosphorylated forms of proteins are visualized in the same gel. The results from my initial phosphotransfer profiling experiments are presented in Figure 10A. The data presented show that CckA autophosphorylates in the presence of ATP to form CckA~P, while neither CtrA nor ChpT can autophosphorylate using ATP. This CckA autophosphorylation is almost total, although an important observation to be made is that CckA incubated without ATP still presents a small degree of phosphorylation, perhaps the result of phosphate remaining bonded to the CckA protein during the purification process from E. coli cells. Incubating the different protein pairs together (CtrA+ChpT, CtrA+CckA and ChpT+CckA) shows that phosphate transfer occurs from CckA~P to ChpT but not directly from CckA to CtrA. Phosphorylation of CtrA only occurs when all three proteins are incubated together, resulting in the almost total phosphorylation of the CtrA sample, with partial phosphorylation being observed on CckA and ChpT. Although my results resemble those observed in similar experiments done on the C. crescentus phosphorelay (Biondi et al., 2006), additional experiments were done to corroborate the phosphorylation state of my proteins. First, I repeated the phosphotransfer profiling experiment but ran the samples in a standard SDS-PAGE. This was done to compare against the PhosTag PAGE results to determine if the observed changes in band mobility were due to phosphorylation changes and not due to any additional factors. These results are presented in Figure 10B and show that all the band changes that are present in the PhosTag gels are absent in the standard SDSPAGE, which is supportive of phosphorylation being the cause of the band shifts. 62  To further corroborate this, I phosphorylated a set of CckA and CtrA and samples through autophosphorylation: CckA with ATP (Laub et al., 2007), and CtrA with acetyl phosphate (Figure 11) (Lukat et al., 1992). The PhosTag PAGE analysis of these proteins indicated that the samples were a mix of phosphorylated and unphosphorylated forms of the respective protein, with CckA being mostly phosphorylated (Figure 10A) and CtrA being mostly unphosphorylated (Figure 11). First, these results support the model for the behaviors of CckA as a HK and CtrA as a RR as they undergo the predicted autophosphorylation using these phosphate donors (Laub et al., 2007; Lukat et al., 1992). Additionally, these proteins were used for ToF mass spectrometry to determine if the phosphorylated samples presented changes in mass corresponding to phosphorylation. The results of mass spectrometry analysis are presented in Figure 12 and are consistent with phosphorylation of the samples, with changes in mass observed to correspond to the addition of an 80 Da phosphate moiety to the proteins. The respective peak heights also follow the phosphorylated to unphosphorylated ratios approximated from Figures 10 and 11, with CckA being mostly phosphorylated and CtrA being mostly unphosphorylated. Taken together, these results give conclusive evidence that the R. capsulatus CckA/ChpT/CtrA proteins form a phosphorelay in vitro. 5.2 CckA changes its activity from a kinase to a phosphatase when incubated with c-di-GMP. The study of the C. crescentus phosphorelay model showed that c-di-GMP plays a key role in regulating CckA activity, as CckA responds to c-di-GMP presence by switching its enzymatic activity from kinase to phosphatase (Lori et al., 2015). A recent study from the Lang lab also showed that altering the expression of R. capsulatus diguanylate cyclases and c-di-GMP phosphodiesterases (enzymes responsible for the regulation of c-di-GMP levels) results in altered RcGTA gene expression and particle production (Pallegar et al., 2020). Having established a method to evaluate the activity in the CckA/ChpT/CtrA phosphorelay I proceeded to analyse how c-di-GMP affected CckA activity in my in vitro model. To test this, I used CtrA~P produced by acetyl phosphate-mediated phosphorylation of CtrA and did a phosphotransfer profiling assay by mixing it with CckA, ChpT and c-di-GMP. The results are 63  presented in Figure 13, and show a clear, rapid and almost complete dephosphorylation of CtrA~P to CtrA over the course of the first 15 minutes of the reaction. To further confirm that c-di-GMP switches CckA activity I designed a modified phosphotransfer assay in which the phosphorylation and dephosphorylation occurred in the same reaction mix. All three proteins (CckA, ChpT, CtrA) are incubated together and the kinase reaction is started with the addition of ATP and samples are taken at regular intervals. After 3 timepoints are taken to establish a trend, c-di-GMP is added to the mix and additional samples are taken for another 3 timepoints. The reaction was also done with acetyl phosphate as a phosphate donor, to evaluate if the phosphatase reaction occurred independently of previous CckA kinase activity. The results are presented in Figure 14, and show clearly how c-di-GMP affects CckA activity. Prior to the addition of c-di-GMP both CckA and CtrA become phosphorylated after the addition of the phosphate donor (ATP or acetyl phosphate, respectively). It is important to note that CckA becomes phosphorylated independently of the phosphate donor used (Figure 14A 7.5 min timepoint & B 30 min timepoint), this is in line with CckA having low baseline phosphatase activity even in absence of c-di-GMP, so it can be interpreted as CckA obtaining phosphate from CtrA~P. An alternative way for CckA to become phosphorylated is through acetyl phosphate-mediated autophosphorylation of the CckA RD, which would resemble the autophosphorylation of the CtrA RD. However, additional control experiments show that this phosphorylation of CckA is not observed when ChpT is absent from the reaction (Figure 15B), thus confirming that the low levels of CckA~P are a result of phosphate transfer from CtrA~P to CckA. Once c-di-GMP is added CtrA~P is rapidly lost, regardless of whether the phosphate donor is ATP or acetyl phosphate. It should be noted that at the end point of the assays all CtrA~P has been dephosphorylated when CtrA phosphorylation occurred through the phosphorelay, but there is still a fraction of CtrA~P present when CtrA is autophosphorylated using acetyl phosphate. I suggest that this happens because CtrA autophosphorylation activity is not inhibited by c-di-GMP, whereas CckA kinase activity is almost entirely suppressed by c-di-GMP. An additional phosphotransfer profile was done to evaluate if the phosphatase reaction needs all components of the phosphorelay to be present to occur. The results are presented in figure 15 and show how the absence of either CckA, ChpT or both results in CtrA~P not being dephosphorylated when c-di-GMP is added to the reaction. 64  5.2.1 Effect of ADP on CckA phosphatase activity. Previous research suggested that in C. crescentus ADP has a phosphatase-stimulating effect on CckA (Dubey et al., 2016). To test if a similar effect could be observed in my in vitro assays, I repeated the phosphatase experiments with the addition of ADP in absence of c-di-GMP. Adding ADP alone did not switch the activity from kinase to phosphatase (Figure 16B), although it may have slightly increased the low baseline CckA phosphatase activity as an increase in ChpT~P can be observed after ADP is added. This would occur as a result of phosphate being transferred from CtrA~P to ChpT and then from ChpT~P to CckA. However, as CckA is acting as a kinase, it mediates the flow of phosphate back to ChpT and CtrA, and as most of the CtrA becomes CtrA~P due to acetyl phosphate-mediated autophosphorylation, the flow of phosphate stops at ChpT. It has also been reported that c-di-GMP and ADP act synergistically to promote each other’s binding to CckA, thereby stimulating phosphatase activity more than the presence of each one alone (Dubey et al., 2016). Therefore, I repeated the phosphatase experiment with the addition of ADP + c-di-GMP. The results of this experiment are presented in Figure 16, where the presence of ADP did not significantly increase the phosphatase activity of CckA when c-di-GMP was present (Figure 16A, compare with Figure 14B).  As such, these analyses do not rule out ADP having an effect that results in a slight increase of CckA phosphatase activity by itself, or in the presence of c-di-GMP, but they show that c-di-GMP is capable of regulating CckA phosphatase activity without the addition of ADP. It should be noted most other phosphotransfer profiling experiments on the C. crescentus CckA did not require ADP to observe the CckA phosphatase activity, and only c-di-GMP was required (Mann et al., 2016; Mann & Shapiro, 2018). Therefore, there appears to be some disagreement among the C. crescentus research community about the role of ADP. An alternative explanation exists for the observation that c-di-GMP affects CckA phosphatase activity without adding ADP. It is possible that ADP produced as a result of CckA kinase activity was bound to the protein during induction in E. coli cells, and it remained throughout the purification process. This was addressed by the phosphorylation assays done on CckA mutants and will be further discussed in the next section. 65  Altogether, my results here confirm that CckA responds to c-di-GMP by switching its activity from kinase to phosphate and reversing the flow of phosphate in the CckA/ChpT/CtrA phosphorelay in vitro. Additionally, this activity results from c-di-GMP alone and does not require the addition of ADP. 5.3 Site specific-mutations of CckA alter its kinase and phosphatase activities. Structural analyses of the C. crescentus CckA have shown multiple amino acid residues that are vital for the enzymatic activities of CckA; His322 is essential for kinase activity, V366 is essential for phosphatase activity (Y. E. Chen et al., 2009), and Tyr514 is required for c-di-GMP binding (Lori et al., 2015). The C. crescentus CckA mutants H322A, V366P and Y514P have been used to study CckA regulatory activity in light of the known roles of these residues on CckA enzymatic activity. R. capsulatus strains with the homologous CckA site-specific mutations H399A, V443P and Y589D were constructed both at the Beatty and Lang labs to study their effects on RcGTA gene expression, with previous results in this regard having been reported by Christina Wiesmann (Wiesmann, 2016). Here, I analyzed the effect that these mutants have on phosphotransfer within the phosphorelay in vitro and how that may affect their overall regulatory capabilities. I worked on recombinant mutants of the 6 His-tagged cytoplasmic domain CckA with the indicated amino acid changes, after purifying them from E. coli cells. I analyzed the respective kinase and phosphatase activities of these mutants with the previously designed phosphotransfer profiling assays and compared the activities to the WT (His-tagged, soluble) CckA. 5.3.1 H399A The results presented in Figure 17 show the observed activities of the CckA H399A mutant. In addition to the normal kinase (Figure 17A) and phosphatase assays (Figure 17C), I did an additional phosphatase assay with added ADP (Figure 17B). This was done to address the issue raised near the end of section 5.2, regarding the possibility that ADP was being bound to CckA during induction in E. coli cells as a result of its autokinase activity, thereby creating ADP in close proximity for binding. CckA H399A lacks kinase activity, as shown by the lack of CtrA~P bands in Figure 17A, and therefore would not be able to produce ADP during induction. If ADP produced in this way explained the active phosphatase activity of the WT CckA in the absence 66  of added ADP in vitro, then CckA H399A would be unable to do the same and would lack phosphatase activity in the absence of ADP. However, the results here shown indicate that CckA H399A presents normal phosphatase activity when incubated with c-di-GMP, regardless of whether ADP is added (compare  Figures 17B and 17C with Figure 16A). Therefore, the absence of a great stimulation of WT CckA phosphatase activity by ADP is not due to binding of ADP produced by kinase activity in vivo, and retained by the protein during purification. As discussed above for the results presented in Figure 16, there is still a possibility that ADP slightly increases CckA phosphatase activity, but if it does the change is too small to be noticeable by this type of assay. The observation that the R. capsulatus CckA responds to ADP and c-di-GMP differently than what has been reported in the C. crescentus CckA may just be a normal difference between the two proteins. This is supported by previous studies noting that different HKs can present different activation mechanisms for their phosphatase reactions (Casino et al., 2014). These results show that His399 in R. capsulatus CckA is essential for CckA kinase activity but not required for CckA phosphatase activity. This mutant has the same enzymatic activities as the homologous C. crescentus mutant, confirming that His399 is the key histidine residue that becomes phosphorylated in R. capsulatus CckA autophosphorylation. 5.3.2 V443P Figure 18 shows the results I obtained for the CckA V443P phosphotransfer profiling experiments. V443P showed an absence of both kinase and phosphatase activities. This is because no CtrA~P band was formed by incubating the phosphorelay with ATP (Figure 18A), and no loss of CtrA~P was observed after addition of c-di-GMP (Figure 18B). The lack of both activities raised the question of whether the specific mutation affected both activities, or if this particular preparation of the purified protein lacked all enzymatic activity. However, these assays were repeated with three independently purified samples of the CckA V443P mutant using the same protocol used for the purification of the other mutants, which follows the same principle as the one used for the purification of the C. crescentus homologue (Y. E. Chen et al., 2009); yet all assays showed the same results, with CckA V443P presenting neither kinase nor phosphatase activities. The sequence of the plasmid containing the mutant 67  protein was verified via Sanger sequencing and shown to contain only the specific V443P mutation (Table 2). These results point towards the R. capsulatus CckA V443P mutant as lacking both kinase and phosphatase activities, which does not match the results observed in the homologous V366P C. crescentus mutant (Y. E. Chen et al., 2009). This indicates that either the R. capsulatus Val443 residue has a different function than the C. crescentus Val366, or that this amino acid change has a different effect on the overall R. capsulatus CckA structure and function than that of the C. crescentus protein. The C. crescentus V366P CckA mutant has been reported in several publications to present increased kinase activity and it is likely that the lack of phosphatase activity is a result of this increased kinase activity. Chen et al. (2009) suggested that the properties of this mutant are a result of altered domain-domain interactions, as Val366 lies near a linker that connects the DHp and CA domains; as such it is possible that the V443P mutation also affects the larger 3D structure of R. capsulatus CckA, but in a different way. This is somewhat supported by the in vivo results reported by Wiesmann, where the V443P mutant appeared to retained some in vivo phosphatase activity, and that this mutant increased WT CckA kinase activity levels when expressed in trans, which would indicate that it is capable of forming functional heterodimers with the WT protein (Wiesmann, 2016).  Because the V443P CckA mutant presents neither kinase nor phosphatase activities in vitro, it is likely that this mutation has a different effect than its C. crescentus homologue. There exists a possibility that it retains some level of activity but so low as to not be observable in this type of phosphotransfer profiling assay. Alternatively, the mutation alters the protein in such a way that it becomes functionally inactive in vitro, such as undergoing irreversible oligomerization during the purification process. Further study needs to be done to properly assess the activity of this mutant. 5.3.3 Y589D The Tyr514 in C. crescentus CckA is one of the few amino acid residues that mediates the binding of c-di-GMP to the protein (Dubey et al., 2016). A Y514D mutant presents normal kinase activity and baseline phosphatase activity but does not switch its activity to a phosphatase when in the presence of c-di-GMP, as the protein is impaired in binding to this cofactor (Lori et al., 2015). 68  The homologous Y589D R. capsulatus CckA mutant behaves in a similar manner, as indicated by the results in Figure 19. CckA Y589D presents normal kinase activity (Figure 19A) but does not show an increase in its phosphatase activity when incubated with c-di-GMP (Figure 19B). It still presents a baseline level of phosphatase activity, which is shown by the formation of weak ChpT~P bands in a manner similar as the one observed in Figure 16B, and discussed in section 5.2: CtrA is autophosphorylating using acetyl phosphate, and because no ATP is present CckA does not autophosphorylate and lacks kinase activity, and so it instead acts as a phosphatase and mediates the transfer of phosphate from CtrA~P to ChpT and from ChpT~P to CckA. However, since CckA is not responding to c-di-GMP, once it acquires phosphate on its RD, it transfers it back to ChpT to phosphorylate CtrA. As CtrA keeps being phosphorylated by acetyl phosphate, phosphate starts accumulating on ChpT~P and therefore the band increases in intensity. This result also supports the slight increase of CckA phosphatase activity that results from addition of ADP. Thus, this confirms that the R. capsulatus CckA Y589D mutant behaves in a similar manner to the homologous C. crescentus Y514D CckA, by presenting normal kinase activity and basal phosphatase activity, but not being affected by the presence of c-di-GMP.      69  6 Conclusions and Future Directions In conclusion, my research shows the following: The R. capsulatus CckA/ChpT/CtrA proteins form a phosphorelay in vitro. In this system CckA becomes autophosphorylated in the presence of ATP to form CckA~P. CckA~P and ChpT interact to mediate phosphate transfer and form ChpT~P. ChpT~P then interacts with CtrA to form CtrA~P. This confirms the latest model of regulation of RcGTA gene expression. R. capsulatus CckA switches its activity from a kinase to a phosphatase in the presence of c-di-GMP and reverses the phosphate flow within the CckA/ChpT/CtrA phosphorelay, from CtrA~P back to CckA. This effect occurs in the absence of ADP, and ADP did not appear to amplify the c-di-GMP effect, although it is possible that ADP minimally increases the baseline level of CckA phosphatase activity without affecting its kinase activity. This result differs from what was expected based on results reported for the C. crescentus CckA by Dubey et al. (2016). However, it has been reported by Mann and Shapiro (2018) that in their hands the C. crescentus CckA binding of c-di-GMP and ADP was additive rather than cooperative. Furthermore, other studies have suggested that “the specific mechanism of activation of the phosphatase reaction may differ from protein to protein” (Casino et al., 2014). The CckA His399 residue appears to be the phosphate-receiving residue in CckA, and it is essential for CckA kinase activity but not required for CckA phosphatase activity. The CckA V443P mutant presents neither kinase nor phosphatase activities. It is unclear whether this is due to overall protein misfolding occurring as a result of the mutation, or if the mutant protein folds correctly but has greatly reduced activities that are not observable under the conditions tested. It is clear that the mutant behaves differently from its C. crescentus homologue, and further study should be done to properly identify the mechanism through which the mutation causes its effect.  The CckA Tyr589 residue is essential for the interaction between CckA and c-di-GMP, based on the similarity with the homologous C. crescentus mutant. The CckA Y589D mutant presents normal kinase activity and basal levels of phosphatase activity. It does not present increased phosphatase activity when in the presence of c-di-GMP.  70  Although very similar to their C. crescentus homologues, the R. capsulatus phosphorelay proteins present some differences in structure and activity and more studies should be done in regards to these aspects to increase our understanding of their role in regulating the expression of RcGTA genes. Further study should be done in regards to the possibility of an additive or synergistic effect between ADP and c-di-GMP in activating CckA phosphatase activity.  A more focused structural analysis of the R. capsulatus CckA would help address some of these questions more directly. A high-resolution crystal structure should help identify the amino acid residues essential for binding of c-di-GMP, as well as giving more information on the effect certain mutations, such as V443P, have on the overall structure of the protein. Further phosphotransfer profiling assays should be able to solve some of the questions regarding the role of ADP and the low basal phosphatase activity of CckA: use of phosphodiesterase enzymes to degrade any present ADP/ATP would allow CckA basal phosphatase activity to be more easily observed and eliminate the possibility of residual ADP being bound to the purified protein. It also would be interesting to couple the PhosTag assay with detailed kinetic measurements. Although I took timepoints to show changes over time, more timepoints over shorter and longer timescales might reveals subtle changes in response to ADP. Finally, to more properly connect these in vitro results to in vivo observations, the following approach can be explored: coupling the PhosTag system to western blot assays to evaluate the phosphorylation ratios of the phosphorelay proteins in vivo. A system has been reported in which RcGTA-producing cells can be separated from recipient cells by fluorescence activated cell sorting (FACS), in an R. capsulatus strain where the RcGTA structural gene cluster has been replaced with an mCherry reporter gene (Fogg et al., 2012). A western blot of a PhosTag gel separating proteins in sorted subpopulations of donor and recipient cells would show differential levels of phosphorylation, if these exist in the two populations. I attempted this approach but was unsuccessful, due to technical limitations regarding cell sorting and non-specific binding of anti-CtrA antibodies. If these issues can be solved, this method will allow to obtain for a large amount of information to be obtained regarding the balance of phosphorylated and non-phosphorylated proteins in vivo  ¸ results that could then be integrated into the overall model for regulation of RcGTA gene expression. 71  Finally, if this approach were successful then it could be coupled to the study of the role that the extracellular Rc280 protein has in the overproduction of RcGTA, to aid in discovering the nature of the Rc280 regulatory pathway.  72  References Anderson, B., Goldsmith, C., Johnson, A., Padmalayam, I., & Baumstark, B. (1994). Bacteriophage‐like particle of Rochalimaea henselae. Molecular Microbiology, 13(1), 67–73. Bachhawat, P., Swapna, G. V. T., Montelione, G. T., & Stock, A. M. (2005). Mechanism of activation for transcription factor PhoB suggested by different modes of dimerization in the inactive and active states. Structure (London, England : 1993), 13(9), 1353–1363. Barbieri, C. M., & Stock, A. M. (2008). Universally applicable methods for monitoring response regulator aspartate phosphorylation both in vitro and in vivo using Phos-tag-based reagents. Analytical Biochemistry, 376(1), 73–82. Beatty, J. T., & Gest, H. (1981). Biosynthetic and bioenergetic functions of citric acid cycle reactions in Rhodopseudomonas capsulata. Journal of Bacteriology, 148(2), 584–593. Bertani, G. (1999). Transduction-like gene transfer in the methanogen Methanococcus voltae. Journal of Bacteriology, 181(10), 2992–3002. Biers, E. J., Wang, K., Pennington, C., Belas, R., Chen, F., & Moran, M. A. (2008). Occurrence and expression of gene transfer agent genes in marine bacterioplankton. Applied and Environmental Microbiology, 74(10), 2933–2939. Biondi, E. G., Reisinger, S. J., Skerker, J. M., Arif, M., Perchuk, B. S., Ryan, K. R., & Laub, M. T. (2006). Regulation of the bacterial cell cycle by an integrated genetic circuit. Nature, 444(7121), 899–904. Blair, J. A., Xu, Q., Childers, W. S., Mathews, I. I., Kern, J. W., Eckart, M., Deacon, A. M., & Shapiro, L. (2013). Branched Signal Wiring of an Essential Bacterial Cell-Cycle Phosphotransfer Protein. Structure, 21(9), 1590–1601. Blokesch, M. (2016). Natural competence for transformation. Current Biology : CB, 26(21), R1126–R1130. Brilli, M., Fondi, M., Fani, R., Mengoni, A., Ferri, L., Bazzicalupo, M., & Biondi, E. G. (2010). The diversity and evolution of cell cycle regulation in alpha-proteobacteria: a comparative genomic analysis. BMC Systems Biology, 4, 52. Brimacombe, C. A., Ding, H., & Beatty, J. T. (2014). Rhodobacter capsulatus DprA is essential for RecA-mediated gene transfer agent (RcGTA) recipient capability regulated by quorum-sensing and the CtrA response regulator. Molecular Microbiology, 92(6), 1260–1278. Brimacombe, C. A., Ding, H., Johnson, J. A., & Thomas Beatty, J. (2015). Homologues of genetic transformation DNA import genes are required for Rhodobacter capsulatus gene transfer agent recipient capability regulated by the response regulator CtrA. Journal of Bacteriology, 197(16), 2653–2663. 73  Brimacombe, C. A., Stevens, A., Jun, D., Mercer, R. G., Lang, A. S., & Beatty, J. T. (2013). Quorum-sensing regulation of a capsular polysaccharide receptor for the Rhodobacter capsulatus gene transfer agent (RcGTA). Molecular Microbiology, 87(4), 802–817. Burbulys, D., Trach, K. A., & Hoch, J. A. (1991). Initiation of sporulation in B. subtilis is controlled by a multicomponent phosphorelay. Cell, 64(3), 545–552. Cabezón, E., Ripoll-Rozada, J., Peña, A., de la Cruz, F., & Arechaga, I. (2015). Towards an integrated model of bacterial conjugation. FEMS Microbiology Reviews, 39(1), 81–95. Casino, P., Miguel-Romero, L., & Marina, A. (2014). Visualizing autophosphorylation in histidine kinases. Nature Communications, 5, 3258. Casino, P., Rubio, V., & Marina, A. (2009). Structural insight into partner specificity and phosphoryl transfer in two-component signal transduction. Cell, 139(2), 325–336. Casino, P., Rubio, V., & Marina, A. (2010). The mechanism of signal transduction by two-component systems. Current Opinion in Structural Biology, 20(6), 763–771. Chen, F., Spano, A., Goodman, B. E., Blasier, K. R., Sabat, A., Jeffery, E., Norris, A., Shabanowitz, J., Hunt, D. F., & Lebedev, N. (2009). Proteomic analysis and identification of the structural and regulatory proteins of the Rhodobacter capsulatus gene transfer agent. Journal of Proteome Research, 8(2), 967–973. Chen, I., & Dubnau, D. (2004). DNA uptake during bacterial transformation. Nature Reviews. Microbiology, 2(3), 241–249. Chen, Y. E., Tsokos, C. G., Biondi, E. G., Perchuk, B. S., & Laub, M. T. (2009). Dynamics of two Phosphorelays controlling cell cycle progression in Caulobacter crescentus. Journal of Bacteriology, 191(24), 7417–7429. Cottrell, J. C., & Green, B. N. (1998). MaxEnt: An Essential Maximum Entropy Based Tool for Interpreting Multiply-Charged Electrospray Data. Curtis, P. D., & Brun, Y. V. (2010). Getting in the loop: regulation of development in Caulobacter crescentus. Microbiology and Molecular Biology Reviews : MMBR, 74(1), 13–41. Curtiss, R. (1969). Bacterial conjugation. Annual Review of Microbiology, 23, 69–136. Daubin, V., & Szöllősi, G. J. (2016). Horizontal Gene Transfer and the History of Life. Cold Spring Harbor Perspectives in Biology, 8(4), a018036. Delavat, F., Miyazaki, R., Carraro, N., Pradervand, N., & van der Meer, J. R. (2017). The hidden 74  life of integrative and conjugative elements. FEMS Microbiology Reviews, 41(4), 512–537. Ding, H., Grüll, M. P., Mulligan, M. E., Lang, A. S., & Beatty, J. T. (2019). Induction of Rhodobacter capsulatus Gene Transfer Agent Gene Expression Is a Bistable Stochastic Process Repressed by an Extracellular Calcium-Binding RTX Protein Homologue. Journal of Bacteriology, 201(23). Domingues, S., & Nielsen, K. M. (2017). Membrane vesicles and horizontal gene transfer in prokaryotes. Current Opinion in Microbiology, 38, 16–21. Dubey, B. N., Lori, C., Ozaki, S., Fucile, G., Plaza-Menacho, I., Jenal, U., & Schirmer, T. (2016). Cyclic di-GMP mediates a histidine kinase/phosphatase switch by noncovalent domain cross-linking. Science Advances, 2(9), e1600823. Fioravanti, A., Clantin, B., Dewitte, F., Lens, Z., Verger, A., Biondi, E. G., & Villeret, V. (2012). Structural insights into ChpT, an essential dimeric histidine phosphotransferase regulating the cell cycle in Caulobacter crescentus. Acta Crystallographica. Section F, Structural Biology and Crystallization Communications, 68(Pt 9), 1025–1029. Fogg, P. C. M. (2019). Identification and characterization of a direct activator of a gene transfer agent. Nature Communications, 10(1), 595. Fogg, P. C. M., Westbye, A. B., & Beatty, J. T. (2012). One for all or all for one: heterogeneous expression and host cell lysis are key to gene transfer agent activity in Rhodobacter capsulatus. PloS One, 7(8), e43772. Gao, R., & Stock, A. M. (2013). Probing kinase and phosphatase activities of two-component systems in vivo with concentration-dependent phosphorylation profiling. Proceedings of the National Academy of Sciences of the United States of America, 110(2), 672–677. Griffith, F. (1928). The Significance of Pneumococcal Types. The Journal of Hygiene, 27(2), 113–159. Heinrich, K., Sobetzko, P., & Jonas, K. (2016). A Kinase-Phosphatase Switch Transduces Environmental Information into a Bacterial Cell Cycle Circuit. PLoS Genetics, 12(12), e1006522. Humphrey, S. B., Stanton, T. B., Jensen, N. S., & Zuerner, R. L. (1997). Purification and characterization of VSH-1, a generalized transducing bacteriophage of Serpulina hyodysenteriae. Journal of Bacteriology, 179(2), 323–329. Hynes, A. P., Mercer, R. G., Watton, D. E., Buckley, C. B., & Lang, A. S. (2012). DNA packaging bias and differential expression of gene transfer agent genes within a population during production and release of the Rhodobacter capsulatus gene transfer agent, RcGTA. Molecular Microbiology, 85(2), 314–325.  2958.2012.08113.x Hynes, A. P., Shakya, M., Mercer, R. G., Grüll, M. P., Bown, L., Davidson, F., Steffen, E., Matchem, H., Peach, M. E., Berger, T., Grebe, K., Zhaxybayeva, O., & Lang, A. S. (2016). Functional and Evolutionary Characterization of a Gene Transfer Agent’s Multilocus “Genome”. Molecular Biology and Evolution, 33(10), 2530–2543. Jacobs, C., Domian, I. J., Maddock, J. R., & Shapiro, L. (1999). Cell cycle-dependent polar localization of an essential bacterial histidine kinase that controls DNA replication and cell division. Cell, 97(1), 111–120. Jenal, U., Reinders, A., & Lori, C. (2017). Cyclic di-GMP: second messenger extraordinaire. Nature Reviews. Microbiology, 15(5), 271–284. Johnston, C., Martin, B., Fichant, G., Polard, P., & Claverys, J.-P. (2014). Bacterial transformation: distribution, shared mechanisms and divergent control. Nature Reviews. Microbiology, 12(3), 181–196. Kim, J., Heindl, J. E., & Fuqua, C. (2013). Coordination of division and development influences complex multicellular behavior in Agrobacterium tumefaciens. PloS One, 8(2), e56682. Kuchinski, K., Brimacombe, C. A., Westbye, A. B., Ding, H., & Beatty, J. T. (2016). The SOS Response Master Regulator LexA Regulates the Gene Transfer Agent of Rhodobacter capsulatus and Represses Transcription of the Signal Transduction Protein CckA. Journal of Bacteriology, 198(7), 1137–1148. Lang, A. S., & Beatty, J. T. (2000). Genetic analysis of a bacterial genetic exchange element: the gene transfer agent of Rhodobacter capsulatus. Proceedings of the National Academy of Sciences of the United States of America, 97(2), 859–864. Lang, A. S., & Beatty, J. T. (2001). The gene transfer agent of Rhodobacter capsulatus and “constitutive transduction” in prokaryotes. Archives of Microbiology, 175(4), 241–249. Lang, A. S., & Beatty, J. T. (2002). A bacterial signal transduction system controls genetic exchange and motility. Journal of Bacteriology, 184(4), 913–918. Lang, A. S., & Beatty, J. T. (2006). Importance of wide spread gene transfer agent genes in alpha-proteobacteria. Trends in Microbiology, 15(2), 54–62. Lang, A. S., Taylor, T. A., & Beatty, J. T. (2002). Evolutionary implications of phylogenetic analyses of the gene transfer agent (GTA) of Rhodobacter capsulatus. Journal of Molecular Evolution, 55(5), 534–543. Lang, A. S., Westbye, A. B., & Beatty, J. T. (2017). The Distribution, Evolution, and Roles of Gene Transfer Agents in Prokaryotic Genetic Exchange. Annual Review of Virology, 4(1), 87–104. 76  Lang, A. S., Zhaxybayeva, O., & Beatty, J. T. (2012). Gene transfer agents: phage-like elements of genetic exchange. Nature Reviews. Microbiology, 10(7), 472–482. Laub, M. T., Biondi, E. G., & Skerker, J. M. (2007). Phosphotransfer Profiling: Systematic Mapping of Two-Component Signal Transduction Pathways and Phosphorelays. Methods in Enzymology, 423(07), 531–548. Laub, M. T., Chen, S. L., Shapiro, L., & McAdams, H. H. (2002). Genes directly controlled by CtrA, a master regulator of the Caulobacter cell cycle. Proceedings of the National Academy of Sciences of the United States of America, 99(7), 4632–4637. Lederberg, J. (1947). Gene Recombination and Linked Segregations in Escherichia coli. Genetics, 32(5), 505–525. Leung, M. M., Brimacombe, C. A., & Beatty, J. T. (2013). Transcriptional regulation of the Rhodobacter capsulatus response regulator CtrA. Microbiology (United Kingdom), 159(1), 96–106. Leung, M. M., Brimacombe, C. A., Spiegelman, G. B., & Beatty, J. T. (2012). The GtaR protein negatively regulates transcription of the gtaRI operon and modulates gene transfer agent (RcGTA) expression in Rhodobacter capsulatus. Molecular Microbiology, 83(4), 759–774. Llosa, M., & de la Cruz, F. (2005). Bacterial conjugation: a potential tool for genomic engineering. Research in Microbiology, 156(1), 1–6. Lori, C., Ozaki, S., Steiner, S., Böhm, R., Abel, S., Dubey, B. N., Schirmer, T., Hiller, S., & Jenal, U. (2015). Cyclic di-GMP acts as a cell cycle oscillator to drive chromosome replication. Nature, 523(7559), 236–239. Lukat, G. S., McCleary, W. R., Stock, A. M., & Stock, J. B. (1992). Phosphorylation of bacterial response regulator proteins by low molecular weight phospho-donors. Proceedings of the National Academy of Sciences of the United States of America, 89(2), 718–722. Madigan, M. T., & Gest, H. (1979). Growth of the photosynthetic bacterium Rhodopseudomonas capsulata chemoautotrophically in darkness with H2 as the energy source. Journal of Bacteriology, 137(1), 524–530. Madigan, M. T., & Jung, D. O. (2009). The Purple Phototrophic Bacteria (C. N. Hunter, F. Daldal, M. C. Thurnauer, & J. T. Beatty (eds.); 1st ed., Vol. 28). Springer Netherlands. Mann, T. H., Seth Childers, W., Blair, J. A., Eckart, M. R., & Shapiro, L. (2016). A cell cycle kinase with tandem sensory PAS domains integrates cell fate cues. Nature Communications, 7(7), 11454. Mann, T. H., & Shapiro, L. (2018). Integration of cell cycle signals by multi-PAS domain 77  kinases. Proceedings of the National Academy of Sciences of the United States of America, 115(30), E7166–E7173. Marrs, B. (1974). Genetic recombination in Rhodopseudomonas capsulata. Proceedings of the National Academy of Sciences of the United States of America, 71(3), 971–973. Marrs, B. (2002). The early history of the genetics of photosynthetic bacteria: a personal account. In Govindjee, J. T. Beatty, H. Gest, & J. F. Allen (Eds.), Discoveries in Photosynthesis (Vol. 73, Issues 1–3, pp. 999–1002). Springer-Dordrecht. McDaniel, L. D., Young, E. C., Ritchie, K. B., & Paul, J. H. (2012). Environmental factors influencing gene transfer agent (GTA) mediated transduction in the subtropical ocean. PloS One, 7(8), e43506. Mell, J. C., & Redfield, R. J. (2014). Natural competence and the evolution of DNA uptake specificity. Journal of Bacteriology, 196(8), 1471–1483. Mercer, R. G., Quinlan, M., Rose, A. R., Noll, S., Beatty, J. T., & Lang, A. S. (2012). Regulatory systems controlling motility and gene transfer agent production and release in Rhodobacter capsulatus. FEMS Microbiology Letters, 331(1), 53–62. Miller, M. B., & Bassler, B. L. (2001). Quorum sensing in bacteria. Annual Review of Microbiology, 55, 165–199. Mónico, A., Martínez-Senra, E., Cañada, F. J., Zorrilla, S., & Pérez-Sala, D. (2017). Drawbacks of Dialysis Procedures for Removal of EDTA. PloS One, 12(1), e0169843. Narayanan, S., Kumar, L., & Radhakrishnan, S. K. (2018). Sensory domain of the cell cycle kinase CckA regulates the differential DNA binding of the master regulator CtrA in Caulobacter crescentus. Biochimica et Biophysica Acta. Gene Regulatory Mechanisms, 1861(10), 952–961. Pallegar, P., Peña-Castillo, L., Langille, E., Gomelsky, M., & Lang, A. S. (2020). Cyclic di-GMP-Mediated Regulation of Gene Transfer and Motility in Rhodobacter capsulatus. Journal of Bacteriology, 202(2), 554–573. Penadés, J. R., Chen, J., Quiles-Puchalt, N., Carpena, N., & Novick, R. P. (2015). Bacteriophage-mediated spread of bacterial virulence genes. Current Opinion in Microbiology, 23, 171–178. Pini, F., Frage, B., Ferri, L., De Nisco, N. J., Mohapatra, S. S., Taddei, L., Fioravanti, A., Dewitte, F., Galardini, M., Brilli, M., Villeret, V., Bazzicalupo, M., Mengoni, A., Walker, G. C., Becker, A., & Biondi, E. G. (2013). The DivJ, CbrA and PleC system controls DivK phosphorylation and symbiosis in Sinorhizobium meliloti. Molecular Microbiology, 90(1), 54–71. Poncin, K., Gillet, S., & De Bolle, X. (2018). Learning from the master: Targets and functions 78  of the CtrA response regulator in Brucella abortus and other alpha-proteobacteria. FEMS Microbiology Reviews. Quon, K. C., Marczynski, G. T., & Shapiro, L. (1996). Cell cycle control by an essential bacterial two-component signal transduction protein. Cell, 84(1), 83–93. Rapp, B. J., & Wall, J. D. (1987). Genetic transfer in Desulfovibrio desulfuricans. Proceedings of the National Academy of Sciences of the United States of America, 84(24), 9128–9130. Redfield, R. J. (2001). Do bacteria have sex? Nature Reviews. Genetics, 2(8), 634–639. Redfield, R. J., & Soucy, S. M. (2018). Evolution of Bacterial Gene Transfer Agents. Frontiers in Microbiology, 9, 2527. Richards, T. A., & Archibald, J. M. (2011). Cell evolution: gene transfer agents and the origin of mitochondria. Current Biology : CB, 21(3), R112-4. Sambrook, J. ; Fritsch, E. F. ; Maniatis, T. (1989). Molecular cloning: a laboratory manual. In University of Texas South Western Medical Center, USA. (Vol. 61, Issue 1). Schaefer, A. L., Taylor, T. A., Beatty, J. T., & Greenberg, E. P. (2002). Long-chain acyl-homoserine lactone quorum-sensing regulation of Rhodobacter capsulatus gene transfer agent production. Journal of Bacteriology, 184(23), 6515–6521. Shakya, M., Soucy, S. M., & Zhaxybayeva, O. (2017). Insights into origin and evolution of α-proteobacterial gene transfer agents. Virus Evolution, 3(2), vex036. Sherlock, D., Leong, J. X., & Fogg, P. C. M. (2019). Identification of the First Gene Transfer Agent (GTA) Small Terminase in Rhodobacter capsulatus and Its Role in GTA Production and Packaging of DNA. Journal of Virology, 93(23). Skerker, J. M., Prasol, M. S., Perchuk, B. S., Biondi, E. G., & Laub, M. T. (2005). Two-component signal transduction pathways regulating growth and cell cycle progression in a bacterium: A system-level analysis. PLoS Biology, 3(10). Solioz, M., & Marrs, B. (1977). The gene transfer agent of Rhodopseudomonas capsulata. Purification and characterization of its nucleic acid. Archives of Biochemistry and Biophysics, 181(1), 300–307. Solioz, M., Yen, H. C., & Marrs, B. (1975). Release and uptake of gene transfer agent by Rhodopseudomonas capsulata. Journal of Bacteriology, 123(2), 651–657. 79 Soucy, S. M., Huang, J., & Gogarten, J. P. (2015). Horizontal gene transfer: building the web of life. Nature Reviews. Genetics, 16(8), 472–482. Spencer, W., Siam, R., Ouimet, M.-C., Bastedo, D. P., & Marczynski, G. T. (2009). CtrA, a global response regulator, uses a distinct second category of weak DNA binding sites for cell cycle transcription control in Caulobacter crescentus. Journal of Bacteriology, 191(17), 5458–5470. Stock, A. M., Robinson, V. L., & Goudreau, P. N. (2000). Two-component signal transduction. Annual Review of Biochemistry, 69(1), 183–215. Thomas, C. M., & Nielsen, K. M. (2005). Mechanisms of, and barriers to, horizontal gene transfer between bacteria. Nature Reviews. Microbiology, 3(9), 711–721. Tsokos, C. G., & Laub, M. T. (2012). Polarity and cell fate asymmetry in Caulobacter crescentus. Current Opinion in Microbiology, 15(6), 744–750. Wang, H., Ziesche, L., Frank, O., Michael, V., Martin, M., Petersen, J., Schulz, S., Wagner-Döbler, I., & Tomasch, J. (2014). The CtrA phosphorelay integrates differentiation and communication in the marine alphaproteobacterium Dinoroseobacter shibae. BMC Genomics, 15(1), 130. Weaver, P. F., Wall, J. D., & Gest, H. (1975). Characterization of Rhodopseudomonas capsulata. Archives of Microbiology, 105(3), 207–216. West, A. H., & Stock, A. M. (2001). Histidine kinases and response regulator proteins in two-component signaling systems. Trends in Biochemical Sciences, 26(6), 369–376. Westbye, A. B. (2016). Regulation of production, maturation and lytic release of the Rhodobacter capsulatus gene transfer agent. (Issue April) [University of British Columbia]. Westbye, A. B., Beatty, J. T., & Lang, A. S. (2017). Guaranteeing a captive audience: coordinated regulation of gene transfer agent (GTA) production and recipient capability by cellular regulators. Current Opinion in Microbiology, 38, 122–129. Westbye, A. B., Kater, L., Wiesmann, C., Ding, H., Yip, C. K., & Beatty, J. T. (2018). The Protease ClpXP and the PAS Domain Protein DivL Regulate CtrA and Gene Transfer Agent Production in Rhodobacter capsulatus. Applied and Environmental Microbiology, 84(11), 1–17. Westbye, A. B., Kuchinski, K., Yip, C. K., & Beatty, J. T. (2016). The Gene Transfer Agent RcGTA Contains Head Spikes Needed for Binding to the Rhodobacter capsulatus Polysaccharide Cell Capsule. Journal of Molecular Biology, 428(2 Pt B), 477–491. 80 Westbye, A. B., Leung, M. M., Florizone, S. M., Taylor, T. A., Johnson, J. A., Fogg, P. C. M., & Beatty, J. T. (2013). Phosphate concentration and the putative sensor kinase protein CckA modulate cell lysis and release of the Rhodobacter capsulatus gene transfer agent. Journal of Bacteriology, 195(22), 5025–5040. Westbye, A. B., O’Neill, Z., Schellenberg-Beaver, T., & Beatty, J. T. (2017a). The Rhodobacter capsulatus gene transfer agent is induced by nutrient depletion and the RNAP omega subunit. Microbiology (Reading, England), 163(9), 1355–1363. Wiesmann, C. (2016). A site-directed mutagenesis approach to study the functions of the histidine kinase CckA in Rhodobacter capsulatus gene transfer agent production and recipient capability (Issue December) [University of British Columbia]. Yen, H. C., Hu, N. T., & Marrs, B. (1979). Characterization of the gene transfer agent made by an overproducer mutant of Rhodopseudomonas capsulata. Journal of Molecular Biology, 131(2), 157–168. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items