EFFECTS OF SIGNALING ON CELLULAR SHAPE CHANGE AND NUCLEOID DYNAMICS DURING MYXOCOCCUS XANTHUS DEVELOPMENT By Y Ha My Hoang A DISSERTATION Submitted to Michigan State University in partial fulfilment of the requirements for the degree of Microbiology and Molecular Genetics–Doctor of Philosophy 2020 EFFECTS OF SIGNALING ON CELLULAR SHAPE CHANGE AND NUCLEOID DYNAMICS DURING MYXOCOCCUS XANTHUS DEVELOPMENT ABSTRACT By Y Ha My Hoang Myxococcus xanthus provides an attractive model to study how cells build multicellular structures and adopt alternative fates. Upon starvation, cells send signals to each other and coordinate their movement to construct uniform mounds. Mounds mature into fruiting bodies as rods inside mounds differentiate into round spores. Other cells undergo lysis or remain outside of fruiting bodies as peripheral rods. To investigate the response of developing rods to nutrients, I added a two-fold dilution series of nutrients at 18 h post-starvation (PS), when cells were not yet committed to form spores (Chapter 2). I discovered an ultrasensitive response to 12.5% vs. 25% nutrient medium addition. This two-fold difference in nutrient addition led to a 30-fold difference in the number of sonication-resistant spores. By systematically measuring transcript and protein levels after nutrient addition, I found that the transcript and protein levels of a key transcription factor, MrpC, correlated best with the sporulation response. The MrpC protein level decreased in the first three hours after nutrient addition, then recovered better after 12.5% vs. 25% nutrient add-back, suggesting that a threshold level of MrpC must be achieved for mounds to persist and spores to form. To visualize cellular shape change in situ, I stained cells with the membrane dye FM 4- 64 and visualized them by confocal microscopy (Chapter 3). At 18 h PS, cells were still rods. Rods began transitioning in shape at 24 h PS and many became spores by 42 h PS. Transitioning cells were irregularly-shaped. A second method involving mixtures of fluorescently labeled and unlabeled cells showed consistent results with FM 4-64 staining. Transitioning cells and spores were more abundant close to the radial center of nascent fruiting bodies from the bottom to several cell layers up, and C-signal-dependent gene expression was ! ! also greatest near the center. Interestingly, when developing cells were stained with the DNA- binding dye DAPI, the nucleoids condensed and segregated to form two loci as cells were changing shape within nascent fruiting bodies (Chapter 5). Nucleoid segregation was also supported by two other methods, both involving binding of a fluorescent protein to DNA in situ. Development requires not only nutrient scarcity, but also high cell density. C-signal, a short-range signal mediated by the CsgA protein, provides a mechanism for cells to communicate their spatial information. To investigate the effects of C-signaling on the developmental process, I co-developed wild-type and csgA mutant cells that were differentially labeled with constitutively-produced fluorescent proteins (Chapter 4). I found that wild-type cells rescued mound formation and sporulation of the cgsA mutant over a narrow range of ratios. A ratio of 1 wild-type cell to 4 csgA mutant cells (1:4) showed no mound formation or sporulation, at 1:2 mounds formed but not spores, and at 1:1 both mounds and spores formed. These observations indicate that threshold levels of C-signaling are required for mound formation and sporulation. Results from my dissertation suggest that the developmental process of M. xanthus responds ultrasensitively to both nutrients and C-signaling, leading to alternative cell fates. My work has advanced understanding of how cells can integrate multiple signals and respond appropriately to changes in their abiotic and biotic environments. ! ! ! Copyright by Y HA MY HOANG 2020 ! ! ! This dissertation is dedicated to my Parents, Hoang Duc and Nguyen Thi Ngoc Anh, and my little Sister, Hoang Ha My A, for constant support since the very beginning of this journey. v! ! ACKNOWLEDGEMENTS I would like to thank my PhD advisor, Dr. Lee Kroos, for supporting me during this work. I am very grateful for his scientific advice and many meaningful discussions and suggestions (with a record long 3.5-hour-meeting). Lee has given me the freedom to pursue various projects and introduced me to experts for collaboration and thoughtful insight. These opportunities expanded my network and made my PhD journey very exciting. Lee cares about his students as individuals, and is a great academic mentor, as well as life mentor. I learned so much about how to plan for future career and maintain a work life balance from him. Lee is my role model for how I want to be a scientist, a mentor, and a teacher in the future. I would like to thank my past mentors and peers, who have guided and supported me to join this program. I graduated from Vietnam National University and was awarded the Vietnam Foundation Fellowship to pursuit higher education in the U.S. I am thankful for the opportunity to become a student at Michigan State University in the Department of Microbiology and Molecular genetics. It was a fantastic experience with great colleagues and friends! I wish to thank my family and friends for their support. My parents and sister have always encouraged me to explore my interest, even when it means I will need to move to another city or a different country. I am grateful to have my Vietnamese friends, who have shared with me moments of being homesick, failure, and success. Finally, I want to express my unfailing gratitude and love to my husband, Curt Priest. He has supported and encouraged me to become a better version of myself every day, both in scientific research and personal life. ! vi! ! TABLE OF CONTENTS LIST OF TABLES........................................................................................................................ix LIST OF FIGURES ......................................................................................................................x KEY TO ABBREVIATIONS.........................................................................................................xiii CHAPTER 1: Myxococcus xanthus provides an attractive model to study bacterial interactions and cell fate determination............................................................................................................1 Introduction........................................................................................................................1 Conclusion………………………………………………………………………………………15 REFERENCES............................................................................................................................17 CHAPTER 2: Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC......................................................................25 Abstract………………………………………………………………………………………….26 Introduction......................................................................................................................27 Results.............................................................................................................................30 Discussion.......................................................................................................................42 Material and Methods......................................................................................................50 APPENDIX..................................................................................................................................55 REFERENCES ...........................................................................................................................63 CHAPTER 3: The spatiotemporal distribution of cellular shape and gene expression in nascent fruiting bodies of Myxococcus xanthus........................................................................................70 Abstract………………………………………………………………………………………….70 Introduction......................................................................................................................71 Results.............................................................................................................................73 Discussion.......................................................................................................................93 Materials and methods....................................................................................................98 APPENDIX................................................................................................................................103 REFERENCES .........................................................................................................................112 CHAPTER 4: Threshold effects of short-range C-signaling on emergent behaviors during Myxococcus xanthus development............................................................................................117 Abstract………………………………………………………………………………………...117 Introduction....................................................................................................................118 Results...........................................................................................................................120 Discussion.....................................................................................................................135 Materials and Methods..................................................................................................142 APPENDIX................................................................................................................................145 REFERENCES..........................................................................................................................147 CHAPTER 5: Segregation of sister chromosomes during the cellular shape change of developing Myxococcus xanthus.............................................................................................. 152 Abstract………………………………………………………………………………………...152 Introduction....................................................................................................................153 Results...........................................................................................................................155 Discussion.....................................................................................................................174 ! vii! ! Materials and Methods..................................................................................................180 APPENDIX................................................................................................................................183 REFERENCES..........................................................................................................................190 CHAPTER 6: Conclusions and Future Directions - An exciting opportunity to investigate bacterial interactions and cell fate determination inside a nascent fruiting body in situ………..195 Conclusions……………………………………………………………………………………195 Future directions………………………………………………………………………………198 REFERENCES..........................................................................................................................204 ! viii! ! LIST OF TABLES Table 1.1 Signaling protein numbers of M. xanthus and other bacteria…………………………...8 Table S2.1 Bacterial strains, plasmids, and primers used in this study…………………………..56 Table S3.1 Bacterial strains, plasmids, and primers used in this study…………………………104 Table S4.1 Bacterial strains and plasmids used in this study ………………………………..….146 Table S5.1 Bacterial strains and plasmids used in this study………………………………...…..184 ! ix! ! LIST OF FIGURES Figure 1.1 Starvation-induced development of M. xanthus………………….................................5 Figure 1.2 The gene regulatory network governing M. xanthus fruiting body formation……..…..9 Figure 1.3 Two models for the identity of C-signal: a proteolytic fragment of CgsA and/or lipids produced by CsgA phospholipase activity……………………………………………………………14 Figure 2.1 Simplified model of the gene regulatory network governing M. xanthus development……………………………………………………………………………………………..28 Figure 2.2 The ultrasensitive response of developing M. xanthus to nutrient medium addition..………………………………………................................................................................32 Figure 2.3 Effect of nutrient medium addition on formation of sonication-resistant spores and mature spores..………………..………………………………………………………….…….............33 Figure 2.4 Effect of nutrient medium addition on the mrpC transcript and MrpC protein levels...........................................................................................................................................35 Figure 2.5 Effect of nutrient medium addition on the mrpB transcript level……………………...37 Figure 2.6 MrpC protein stability after nutrient medium addition................................................38 Figure 2.7 Effect of nutrient medium addition on the fruA transcript and FruA protein levels….39 Figure 2.8 Effect of nutrient medium addition on late gene transcript levels……………......…..40 Figure 2.9 Effect of nutrient medium addition on exo-lacZ expression…………………...………42 Figure S2.1 Effect of nutrient medium addition on the formation of sonication-resistant spores…………………………………………………………………………………………………….57 Figure S2.2 Effect of nutrient medium addition on the MrpC protein level……………………….58 Figure S2.3 FruA protein stability after nutrient medium addition…………………………………59 Figure S2.4 Effect of nutrient medium addition on late gene transcript levels……………..……60 Figure S2.5 Inducible expression of MrpC in an mrpC mutant……………………………………61 Figure S2.6 Development after exposure to ≥25% CTTYE.…………………………….…………62 Figure 3.1 Cellular shape during development visualized by the membrane dye FM 4- 64……………………………………………………………………………………………………….…74 Figure 3.2 Cellular shape during development visualized by labeling cells with tdTomato…………………………………………………………………………………………………76 ! x! ! Figure 3.3 Shape classification of developing cells…………………………………………………78 Figure 3.4 Proportion of rods, transitioning cells, and spores in nascent fruiting bodies……………………………………………………………………………………………...……..80 Figure 3.5 Spatial distribution of rods, transitioning cells, and spores in nascent fruiting bodies…………………………………………………………………………………………….………83 Figure 3.6 Visualization of gene expression in nascent fruiting bodies……………………..……85 Figure 3.7 Quantification of gene expression in individual cells………………………….………..87 Figure 3.8 Spatial distribution of gene expression in nascent fruiting bodies……………………92 Figure S3.1 Effects of the membrane dye FM 4-64 on M. xanthus development………..……105 Figure S3.2 Spatial distribution of rods, TCs, and spores in individual nascent fruiting bodies. ……………………………………………………………………………………………………….…..106 Figure S3.3 Quantification of gene expression in individual cells without normalization………………………………………………………………………………………..….108 Figure S3.4 Quantification and spatial distribution of mNeonGreen expression from a vanillate- inducible promoter in strains YH14 and YH15……………………………………………………110 Figure S3.5 Quantification of gene expression from a vanillate-inducible promoter in strains YH8 and YH7…………………………………………………………………………………..………111 Figure 4.1 Development of wild-type and csgA mutant cells separately and co-development at different ratios............................................................................................................................123 Figure 4.2 Cell density upon co-development of WT and csgA mutant cells at different ratios..........................................................................................................................................126 Figure 4.3 Quantification of cellular shape change upon co-development of WT and csgA mutant cells at different ratios ………………………………………............................................128 Figure 4.4 Proportions of rods, transitioning cells, and spores upon co-development of WT and csgA mutant cells at different ratios. …………..................................……………….……………129 Figure 4.5 Radial distribution of rods, transitioning cells, and spores upon co-development of WT and csgA mutant cells at different ratios …………………………………......................……133 Figure 4.6 Vertical distribution of rods, transitioning cells, and spores upon co-development of WT and csgA mutant cells at different ratios..……………………….………...…......…......…….134 Figure 4.7 Model for co-development of wild-type and csgA mutant cells...……..............……139 Figure 5.1 Visualization of the chromosomal copy number during starvation-induced development ………………………….…………………………...............….................................156 ! xi! ! Figure 5.2 Visualization of DNA by DAPI staining during development……………………...…161 Figure 5.3 Visualization of nucleoids using mNeonGreen-FruA during development……..… 166 Figure 5.4 Visualization of fluorescence from mNeonGreen-FruA, DAPI, and FM 4-64 in the same cells during development………………………………………………………………………169 Figure 5.5 Quantification of segregated nucleoids………………………………………………..171 Figure 5.6 Visualization of the chromosomal copy number, DAPI-stained DNA loci, and cellular shape during glycerol-induced sporulation………………………………………………………….174 Figure S5.1 Visualization of the same nascent fruiting body during starvation-induced development……………………………………………………………………………………………186 Figure S5.2 Visualization of nucleoids using mNeonGreen-FruA during development…….…187 Figure S5.3 Visualization of the nucleoids by DAPI and mNeonGreen-FruA…………………..189 Figure 6.1 The multicellular starvation-induce development of M. xanthus……………………196 ! xii! KEY TO ABBREVIATIONS ! TFP Type IV pilus EPS exopolysaccharide HK Histidine kinase RR response regulator TCS two-component system ECF extracytoplasmic funcation STPKs Serine/threonine protein kinases GRN gene regulatory network Nla NtrC-like activator EBP enhancer-binding protein SCAD short-chain alcohol dehydrogenase DAG diacylglycerols PS post-starvation Pdev the dev promoter PfmgE the fmgE promoter WT wild-type TCs transitioning cells NFB nascent fruiting body MRFI median relative fluorescence intensity ! xiii! CHAPTER 1: Myxococcus xanthus provides an attractive model to study bacterial interactions and cell fate determination. Introduction Bacteria are ubiquitous and can interact with each other to respond to changes in the environment as a community. Microbial communities affect ecosystems and all the organisms that inhabit them by impacting global processes including the cycling of elements in soil, water, and air, and primary productivity of the oceans. They are involved in many aspects of human life, such as agriculture, medicine, and bioenergy production. Advancing knowledge of bacterial interactions would benefit efforts to manipulate microbial communities for human purposes. Bacteria respond to environmental cues via within-cell signaling, and they can also release molecules to signal each other. The first evidence of bacterial communication, known as quorum sensing, was observed in the symbiotic relationship between Vibrio fischeri and the Hawaiian bobtail squid (1). The squid provides the bacteria with food, and in return, the bacteria bioluminesce. The glow of the bacteria prevents the squid from casting a shadow, hiding it from predators swimming beneath. Light emission is tightly correlated with the population density of V. fischeri in the light organ. As the V. fischeri grow, they produce and release an autoinducer into the extracellular environment, and the autoinducer is trapped inside the light organ with the bacteria. When the autoinducer concentration reaches a certain threshold, it elicits a signaling cascade that culminates in the emission of light (2). Other examples of community behavior in bacteria involve virulence in Pseudomonas aeruginosa (3), competence in Streptococcus pneumonia (4), conjugation in Enterococcus faecalis (5), sporulation in Bacillus subtilis (6), and fruiting body development in Myxococcus xanthus (7). Among them, the system for quorum sensing in M. xanthus is different than in V. fischeri (8). The quorum sensing behavior in M. xanthus is not driven by a homoserine lactone autoinducer. It involves A-signal, a mixture of extracellular proteases, peptides, and amino acids, which provides a mechanism to sense cell ! 1! density early during the developmental process (9). Starving cells at a sufficiently high cell density coordinate their movements to form fruiting bodies of approximately 100,000 cells each. During this process, M. xanthus uses a second rather unusual signal. Positional information among cells is communicated via short-range C-signaling, which requires cell alignment and perhaps cell-cell contacts (10). M. xanthus is a model organism that exhibits not only social behavior, but also cellular differentiation. When starved, M. xanthus rod-shaped cells move in streams and build mounds which become fruiting bodies as some cells differentiate into round spores. Other cells undergo lysis or remain outside of fruiting bodies as peripheral rods. The rapid developmental process is completed in about two days. M. xanthus development is an amenable system with both bacterial genetics and ease of scaling up to perform molecular assays. The M. xanthus genome codes for a large number of signaling systems. By focusing on how M. xanthus cells integrate multiple signals to coordinate streaming and mound formation with spore differentiation, we may gain deep insight into emergent properties found commonly in development, such as assembly of multicellular structures and cell fate determination. M. xanthus development offers an attractive model to study bacterial signaling and cell fate determination. The research in this dissertation is designed to improve our understanding of interactions among bacteria, and with the environment, using M. xanthus development as the model system. The knowledge gained from this study is expected to transform thinking about how microbial communities integrate multiple signals and coordinate their behaviors to make complex decisions. Several important questions about the multicellular structure of fruiting bodies and the responses of developing M. xanthus cells to nutrients and C-signal will be discussed in the introduction of this dissertation. The ecology and lifecycle of M. xanthus. M. xanthus is a gram-negative species of myxobacteria, which belong to the delta subgroup of the proteobacteria. Myxobacteria typically live in the topmost layers of the soil. They can colonize decaying plant material and certain ! 2! types of herbivorous mammal dung such as rabbit, sheep, and goat (11). They can also be found in aquatic or special habitats such as deep caves, bogs, and deserts. All myxobacteria are capable of degrading biomacromolecules and can attack other microorganisms by secreting enzymes. The enzymatic attack is more efficient when the entire population works together because it increases enzyme concentration and minimizes diffusion. Indeed, the predatory behavior of M. xanthus has been compared to a multicellular “wolfpack” (12). Individual M. xanthus cells are also competent predators as one cell can lyse a micro-colony of about 20 Escherichia coli cells (13). A study in which living 13C-labelled E. coli were added to agricultural soil to identify microbes that are active in mineralization and carbon sequestration has shown that myxobacteria are close to the top of the microbial food chain (14). The predatory behavior of M. xanthus is facilitated by its motility. M. xanthus cells do not have flagella and are unable to swim in liquid environments. They move on solid surfaces via two genetically distinct motility systems: social (S)-motility and adventurous (A)-motility (15). A mutation in any gene of system A inactivates A-motility so that only S-motility remains, and vice versa. When cells are in a group, a type IV pilus (TFP) assembles at the leading cell pole and acts as grappling hooks to pull the cell forward by retraction (16). Retraction is stimulated by exopolysaccharide (EPS) on the surface of other cells or deposited on the substratum by moving cells (17). This is called S-motility as it is a cooperative form of group movement. At the colony’s edges, single cells move by A-motility. The Agl-Glt motor drives A-motility, which involves transient adhesion complexes that remain at fixed positions relative to the substratum as cells moved forward (18). Complexes assemble at leading cell poles and disperse at the rear of the cells. When cells reverse direction, the A-motility clusters re-localize to the new leading poles together with S-motility proteins. Hodgkin and Kaiser were able to identify a master regulator of cell movement, mglA (mutual gliding A), which is required for both motility systems (19). MglA is a 22-kDa protein that belongs to the Ras superfamily of small monomeric GTPases. MglA interacts with AglZ (20), an A-motility protein, and FrzS (21), an S-motility ! 3! protein. AglZ and FrzS appear to be paralogs that emerged following a gene duplication event in an ancestor of the Cystobacterineae (22). The velocity of wild-type cells under laboratory condition is 1-4 µm/min. M. xanthus cells periodically invert their polarity and reverse their direction of movement (23). Reversal rates impact the ability of cells to spread on a solid surface and form normal mounds and fruiting bodies (24). The social lifestyle of M. xanthus depends strongly on their motility. In the presence of nutrients, motile rod-shaped cells divide and form spreading colonies. Cells at the edges of the colony spread over the surface to form a thin, film-like structure. Upon nutrient depletion, cells undergo a phase transition from exploratory flocking to one-dimensional growing streams that are inherently stable in space (25). This transition is followed by formation of uniform structures called mounds. Rod-shaped cells inside mounds differentiate into spherical spores that can withstand harsh environments. Sporulation results in the maturation of mounds into fruiting bodies. The majority of the population undergoes lysis during the developmental process, while there are a few cells that remain outside of fruiting bodies as peripheral rods. When nutrients are again available, spores germinate to produce rod-shaped cells that enter vegetative growth. This dissertation focuses specifically on the part of the developmental process in which rod- shaped cells are converted into spherical spores. The developmental process of M. xanthus. Starvation-induced multicellular development: In response to nutrient scarcity, M. xanthus cells form multicellular structures filled with spores called fruiting bodies, which is one of the first examples of social, multicellular behavior in bacteria (26) (Fig.1.1). A fruiting body is comprised of approximately 50,000 to 100,000 cells and is dome-shaped. The diameter of a fruiting body ranges from 100 µm to a few millimeters. There are two domains inside a fruiting body: the hemispherical outer domain is densely packed with cells that move in a circular pattern, and the inner domain of less ordered cells is at 3-fold lower cell density (27). ! 4! Figure 1.1 Starvation-induced development of M. xanthus. Upon starvation, rod- shaped cells coordinate their movement to form uniform structures called mounds. The majority of the population undergoes lysis during this process. Rod-shaped cells inside mounds differentiate into spherical myxospores, which results in the maturation of mounds into fruiting bodies. Some rod-shaped cells still persist outside the fruiting bodies as peripheral rods. Adapted from (28). Microcinematography (time-lapse microscopy) of M. xanthus development has shown that wild-type cells progress through three distinct phases: a quiescent phase with some motility but little aggregation, then vigorous cell motility as growing streams, followed by formation of mounds, and finally maturation of the fruiting body coupled with sporulation (29) (25). In the first phase, there is little change in the pattern of cell layers. In the second phase, cell swarms coalesce resulting in the formation of many small cell towers of about three layers thick. Only a proportion of cell towers will become fruiting bodies and a small aggregate is more likely to disperse (30). Mounds extend vertically in a series of tiers, which involves the addition of a cell monolayer on top of the uppermost layer. However, the tiering can only be observed for a very short time because the materials that encapsulate mounds block visualization. During the last phase, the mounds darken as rod-shaped cells differentiate into spherical spores. Although it is known that mounds can be extended vertically by the addition of cells to the uppermost layer, it is unknown whether this is the only means of mound growth. Neither is it known when, and in what cell layer, cellular shape change from rods to spores happens inside a nascent fruiting body. Establishing the temporal and spatial distribution of cellular shape change is a crucial step toward a better understanding of the cell-cell interactions that lead to ! 5! sporulation during M. xanthus development. The ability to visualize individual cells within a nascent fruiting body would also offer the opportunity to measure gene expression in cells as they become committed to different fates in situ. Glycerol-induced development: Sporulation of M. xanthus can be chemically induced with 0.5 M glycerol (31), 0.7 M dimethyl sulfoxide (32), and beta-lactam antibiotics and D-amino- acids (33). Within 8 hours, all of the rod-shaped cells from an exponentially growing culture will differentiate into spherical spores. Compared with starvation-induced spores, which take about two days to mature, glycerol-induced spores have a considerably thinner spore protective layer (34). The spore coat of glycerol-induced spores also lack some of the proteins that have been identified in starvation-induced spores, such as the surface spore coat proteins S and C (35) (36), and the internal polyphosphate storage protein W (37). However, both starvation- and glycerol-induced spores are spherical, phase-bright microscopically, resistant to heat and sonication treatment, and able to geminate into rod-shaped vegetative cells. Under laboratory conditions, glycerol-induced sporulation is a simple and rapid process, in which the vast majority of cells synchronously bypass the complex signaling pathway of starvation-induced multicellular development. Therefore, the artificial induction of sporulation with glycerol can serve as a model system for investigating the core mechanisms of cellular shape change, albeit with the caveat that even the core mechanisms may differ from those used during starvation-induced sporulation. Nucleoid dynamics during the developmental process: It is crucial that the genetic material is packed and stored properly when spores are formed. DNA replication is tightly regulated during M. xanthus development. Vegetatively growing cells contain one or two copies of the genome, but the chromosome copy number drops to one when cells enter stationary phase (38). When starving cells aggregate to form mound structures, DNA replication is required for development to progress (39). Inhibition of DNA replication delays the developmental program and development can resume when the inhibitor is removed (40). If the ! 6! DNA replication inhibitor is added after the aggregation phase, it does not have any effect on developmental progression. A fruiting body-derived myxospore contains two copies of the chromosome with the origin and terminus regions localized to the periphery of the myxospore (38). During sporulation of M. xanthus, the rod-shaped cell is converted into a spherical spore by peptidoglycan remodeling (in contrast to endospore formation in B. subtilis). Since there is no compartmentalization following DNA replication during M. xanthus development, it may be interesting to investigate the arrangement of the two chromosomes when cells change shape and become spores inside a nascent fruiting body. Previous studies involve dispersion of myxospores from fruiting bodies (38). In contrast with starvation-induced myxospores, the chromosome copy number varies in glycerol-induced spores (38). It appears that DNA replication is less tightly regulated during the response to glycerol, since cells were able to bypass the checkpoint of DNA replication and proceed to cellular shape change. The M. xanthus genome codes for a large number of signaling systems. The relatively large 9.1 Mbp genome of M. xanthus codes for over 7,500 proteins (41) (42). Compared with other bacteria, the duplication of genes in M. xanthus results in a considerably higher number of signaling components (Table 1.1). Two-component His/Asp systems (TCS) are the most common bacterial signal transduction systems. M. xanthus has 135 histidine kinases (HK) and 127 response regulators (RR) (43). The HK can sense a signal via its N- terminal input domain, which results in autophosphorylation of a His residue in the C-terminal transmitter domain. The phosphoryl group from this His residue is then transferred to a conserved Asp residue in the receiver domain of the RR. The RR then becomes activated. Surprisingly, 58% of M. xanthus TCS are not traditional HK-RR pairs (43). M. xanthus employs TCS in a number of physiological processes such as motility, development, pili biogenesis, heat shock, and osmotic tolerance. While TCS have been positively selected for, the M. xanthus genome codes for only 220 one-component regulators (Table 1.1), a similar number as other species with half its genome size (43). This suggests that some of the functions that are usually ! 7! S. au C. cre B. su M. tb E. co P. ae S. co M. xa Category Two-component systems 33 17 Histidine Kinase 16 Response Regulator One-component 108 0 ECF Sigma Factor STPKb 1 GGDEFc 1 Polyketide Synthase 1 2.9 Genome Size (Mbp) 102 58 44 203 15 0 11 0 4.0 Bacterial Speciesa 69 37 32 243 8 3 4 11 4.1 29 16 13 220 10 11 1 17 4.4 60 29 31 298 2 0 15 3 5.4 117 50 67 435 19 4 33 9 6.3 237 150 87 776 51 36 9 16 9.0 262 135 127 220 41 98 17 48 9.1 regulated by one-component transcription factors in other species may instead be controlled by TCS and/or sigma factors in M. xanthus. Nevertheless, a key transcriptional regulator of the developmental process is MrpC, a one-component transcription factor. Table 1.1 Signaling protein numbers of M. xanthus and other bacteria. Adapted from (43). a S. au (Staphylococcus aureus USA300_FPR3757), C. cre (Caulobacter crescentus CB15), B. su (Bacillus subtilis BSn5), M. tb (Mycobacterium tuberculosis H37Rv), E. co (Escherichia coli 042), P. ae (Pseudomonas aeruginosa PAO1), S. co (Streptomyces coelicolor A3(2)), M. xa (Myxococcus xanthus DK1622). cyclic-di-GMP. b Serine-Threonine Protein Kinase (STPK) defined by the pfam domain Pkinase. c GGDEF domain defines di-guanylate cyclases and is an indication of the production of Extracytoplasmic function (ECF) sigma factors were also duplicated in M. xanthus. Environmental cues are sensed and responded to by 41 ECF sigma factors (43). However, only a few of the ECF sigma factors have been well-characterized. In addition to ECF sigma factors, non-ECF sigma factors also play important roles in the biology of M. xanthus. SigA is essential for growth (44). SigB, SigC, and SigE are important for multicellular development (45). SigD is important for viability in stationary phase and for stress responses (46). SigF plays a role in S- motility and development (47). SigG is not essential for growth or development; it directs flagellar gene expression (48). The M. xanthus genome codes for nearly 100 serine/threonine protein kinases (STPKs) that play important roles in the post-translational modification of proteins (43). M. xanthus also has the capacity for signaling via the production of c-di-GMP. There are 17 proteins containing a GGDEF-domain, which often produces c-di-GMP in response to a specific cellular cue (43). ! 8! C-di-GMP regulates motility in M. xanthus (49). ActA, which is a protein with a GGDEF-domain, plays an important role in development (50). M. xanthus cells are capable of producing specialized or secondary metabolites such as DKxanthene, a yellow pigment that is necessary for fruiting body formation and sporulation (51). Secondary metabolites may be critical in predation and competition in the soil environment. M. xanthus and B. subtilis can form “megastructrures” when they interact with each other (52). The gene regulatory network (GRN) governing M. xanthus fruiting body formation. The developmental GRN of M. xanthus is comprised of four modules: NtrC-like activator (Nla)24, an enhancer-binding protein (EBP) cascade, Mrp, and FruA (Fig. 1.2). Figure 1.2 The gene regulatory network governing M. xanthus fruiting body formation. Starvation increases the levels of guanosine penta- and tetraphosphate [(p)ppGpp] and cyclic diguanylate (c-di-GMP). (p)ppGpp accumulation produces extracellular A- and C-signals. c-di- GMP binds to Nla24 and activates the Nla24 module. Starvation also triggers the EBP cascade and Mrp module. The product of the Mrp module is the MrpC protein, which activates transcription of the fruA gene in the FruA module. Adapted from (28). The stringent response and the effects of (p)ppGpp: Fruiting body development of M. xanthus is induced when there is a shortage of amino acids. The stringent response that initiates development shares many features with that of E. coli. The lack of t-RNA charged with an amino acid causes ribosomes to stall. The ribosome-associated protein RelA synthesizes guanosine-5’-triphosphate-3’-diphosphate (pppGpp), which can be hydrolyzed to guanosine-5’- diphosphate-3’-diphosphate (ppGpp) (53) (54). The production of (p)ppGpp is necessary and ! 9! sufficient for early development gene expression (55). Accumulation of (p)ppGpp triggers a transcriptional response via RNA polymerase and the transcription factor DskA. In M. xanthus, there are four homologs of DskA, but their specific functions during development are still unknown. The stringent response triggers the production of extracellular A- and C-signal. A- signal is a mixture of extracellular proteases, peptides, and amino acids, which provide a mechanism to sense cell density early in the developmental process (9). C-signal is a short- range signal that requires cell alignment and possibly end-to-end contact between cells (56). C- signaling serves as a mechanism for cells to communicate positional information. The Nla24 module: Starvation induces transcription of a diguanylate cyclase, DmxB, which increases the level of c-di-GMP early in development (57). DmxB is essential for development, and functions downstream of the Dif chemosensory system to trigger the production of EPS. A c-di-GMP receptor, Nla24, is likely to be the target for DmxB-generated c- di-GMP. The accumulation of EPS enhances aggregation and sporulation. Many bacteria employ the second messenger c-di-GMP to regulate the production of EPS, which can also involve other second messengers or quorum sensors (58). The EBP cascade module: Starvation triggers a cascade of EBPs, which bind to sites located about 100 base pairs upstream of promoters and activate transcription by σ54 RNA polymerase (59). Most of the EBPs in the cascade contain a domain that can be phosphorylated by a protein kinase. However, some of the kinases have not been identified and none of the signals to which kinases respond have been identified. Signal integration by combinatorial regulation by more than one EBP, and positive autoregulation, are recurring themes in the EBP cascade. For example, the phosphorylated forms of Nla4 and Nla18 (Nla4~P and Nla18~P) likely activate transcription of nla6. Nla6~P and Nla28~P positively autoregulate, regulate each other, and both likely activate the downstream EBP, ActB~P, which positively autoregulates and likely activates transcription of the EBP, MXAN4899. ! 10! The Mrp module: The Mrp module consists of three proteins; MrpA, MrpB, and MrpC. MrpA is a phosphatase of MrpB~P (60). MrpB~P positively autoregulates transcription of the mrpAB operon (60). MrpB~P also activates transcription of mrpC, which codes for a member of the CRP/Fnr superfamily of transcriptional regulators. MrpC negatively autoregulates by competing with MrpB (61). MrpC is the key output of the Mrp module and regulates a large number of genes important for development. An mrpC mutant fails to aggregate and sporulate (60) (61). A specific cluster of threonines and serines (the TTSS motif) in the amino-terminal region is essential for MrpC activity in vivo (62). MrpC binds the promoter regions of about 300 genes known to be up- or down-regulated during development (63). Several post-translational regulatory mechanisms have been reported to impact the MrpC level. The EspAC signaling system regulates MrpC accumulation early during development (64) (65) (66). The addition of nutrients to developing cells activates the degradation of MrpC via metalloproteases, which blocks sporulation (67). Regulation of the MrpC level is a mechanism to control the timing of development. The FruA module: MrpC activates transcription of the fruA gene. It is proposed that C- signal regulates FruA activity, but the mechanism is unknown (68). FruA is similar to response regulators of TCS (69). However, among the 5 residues typically important for phosphorylation (the quintet) (70), FruA has two matches, two conservative substitutions, and one non- conservative substitution. There are eight additional residues downstream from the putative phosphorylation site (D59), which are not present in other receiver domains. A fruA mutant is defective in aggregation. Two positive feedback loops are proposed to affect the level of active FruA (designated FruA*). Aggregation bring cells closer, which enhances C-signaling, so that more FruA* will be made. The second feedback loop involves the act operon (71), of which ActB belongs to the EBP cascade module. MrpC and FruA cooperatively bind the promoter regions of many genes important for normal aggregation and sporulation (72) (63) (73) (74) (75) (76). Development of M. xanthus requires both starvation and cells to be in close proximity. ! 11! This dissertation focuses on the response of M. xanthus when the starvation signal is perturbed by replacing the starvation buffer with nutrients, and on how C-signaling affects emergent behaviors of the population during development. Response of M. xanthus to nutrient addition during development. During starvation-induced development, M. xanthus cells need to decide between three alternative fates: undergo lysis, differentiate into a myxospore, or persist as a peripheral rod. Commitment to form spores was investigated in an early study by collecting cells from developmental plates and subjecting them to nutrient medium (77). If cells were collected from the starvation agar before 18 h of incubation, they were unable to form myxospores. Cells harvested after 24 to 32 h on starvation agar failed to form spores if placed into nutrient medium directly. However, they were able to form spores if incubated in magnesium phosphate buffer 6 h prior to incubation in nutrient medium. Cells harvested after 36 h on starvation agar formed spores if placed directly into nutrient medium. These results suggest that commitment to spore formation occurs during development of M. xanthus. A later study of cell fate commitment in M. xanthus was conducted using submerged culture conditions (67). In this method, cells from a growing culture are sedimented and the supernatant is discarded. The cell pellet is resuspended in starvation buffer, placed in a plastic container, and incubated at 32°C. Cells adhere to the bottom of the plate and form a biofilm. Therefore, the overlay can be replaced with either fresh starvation buffer or nutrients to perturb development. The study showed that many M. xanthus cells commit to shape change and sonication resistance between 24 h and 30 h poststarvation (PS) (67). If the overlay is replaced with nutrient medium at 18 h, mounds fail to become compact and no spore-filled fruiting bodies are formed. On the other hand, nutrient addition at 30 h or 36 h PS does not stop the formation of fruiting bodies. It was shown that MrpC could provide a checkpoint for persistent starvation as the MrpC level rapidly decreased upon the addition of nutrient medium (67). This response involves ATP- ! 12! independent metalloprotease activity. The proteolysis of MrpC when nutrient medium was added to developing cells before and during the commitment period presumably halts the expression of genes required for commitment to sporulation. The proteolysis of MrpC is proposed to be a rapid mechanism to escape commitment and resume growth if nutrients reappear. Sporulation is a costly decision that consumes a lot of time and energy. The ability to resume growth if nutrients are available again could be an advantage for the survival of the population in response to environmental changes. Previous studies of cell fate commitment during M. xanthus development have established that cells do not commit to sporulation until they have been starving for about 18 h. The commitment to sporulation is a delayed and stepwise process. Under submerged culture conditions, the critical period of commitment to form a spore is from 24 h to 30 h PS. The addition of nutrient medium at 18 h PS blocks development. In this dissertation, I describe experiments aimed at determining whether the response of starving rods to nutrient medium is ultrasensitive. This involved the addition of a twofold dilution series of nutrient medium at 18 h PS. The sonication-resistant spore numbers were assessed as well as protein and transcript levels of GRN components. The response of M. xanthus to C-signaling. The identity of C-signal: C-signaling is mediated by the product of the csgA gene (78) (79). There are two models for the identity of the C-signal, which are not mutually exclusive. The C-signal may be a proteolytic fragment of CgsA and/or lipids produced by CsgA phospholipase activity (Fig. 1.3). Initially, C-signal was purified by detergent extraction and biochemical fractionation of starving cells, which identified a 17 kDa protein (p17) (80). It was subsequently shown that the full-length 25 kDa protein encoded by the csgA gene is cleaved by the subtilisin-like protease PopC (81). PopC accumulates in the cytoplasm of vegetative cells but is secreted during starvation. PopC secretion depends on the (p)ppGpp synthase RelA and the stringent response (82). Starvation increases (p)ppGpp accumulation, which triggers the production of FtsHD. FtsHD is able to degrade PopD, which ! 13! forms a soluble complex with PopC, resulting in release of PopC from inhibition of secretion. The p17 form of CsgA is the signal in this model. Figure 1.3 Two models for the identity of C-signal: a proteolytic fragment of CgsA and/or lipids produced by CsgA phospholipase activity. The first model involves the stringent response. Starvation increases the production of (p)ppGpp via RelA. The accumulation of (p)ppGpp raises the level of FtsHD, which can degrade a protein called PopD. In vegetative cells, PopD forms a soluble complex with PopC. The degradation of PopD releases PopC from the complex so that PopC is secreted from starving cells. Protease activity of PopC cleaves full- length CgsA to form p17. The second model proposes that CsgA has phopholipase activity, which oxidizes cardiolipin to form two diacylglycerols. p17 and/or diacylglycerols are the signal, but neither the receptor for p17 (?) nor the mechanism of diacylglycerol transmission is known. Adapted from (28). In the second model, CsgA is a phospholipase that oxidizes the 2’-OH glycerol moiety on cardiolipin to produce diacylglycerols (DAGs) (83). A lipid extract with a high concentration of DAGs from developing wild-type cells can rescue development of a csgA mutant. In the recipient cells, C-signal is proposed to activate FruA (68). However, the modes of signal transmission in the two models and the nature of FruA activation remain to be elucidated. Threshold effects of C-signaling: C-signaling begins at about 6 h PS. Mutants defective in C-signal synthesis are unable to aggregate or sporulate. The developmental defects of a csgA mutant can be rescued by codevelopment with wild-type cells or addition of purified C- signal (84) (85). A low level of C-signal (0.8 unit) rescues aggregation and expression of an early C-signal-dependent gene, whereas a slightly higher level (1.0 unit) can additionally trigger late C-signal-dependent gene expression and spore formation (86), indicative of distinct thresholds for different responses. C-signal transduction requires cell alignment and possibly ! 14! end-to-end contact between cells (56), which normally occurs as cells move into nascent fruiting bodies. Therefore, it is proposed that C-signaling communicates positional information to cells, telling them when they are close-packed in a nascent fruiting body and triggering sporulation. Since different threshold levels of C-signal rescue mound formation (0.8 unit) and sporulation (1.0 unit), we hypothesize that wild type would rescue different emergent behaviors of a csgA mutant over a narrow range of ratios. This dissertation describes experiments in which gene expression and cellular behavior of mixtures of wild-type and csgA mutant cells were examined over a narrow range of ratios. Conclusion M. xanthus development offers an exciting opportunity to study bacterial interactions and cell fate determination. Understanding how bacteria integrate signals to respond appropriately to environmental cues will facilitate efforts to manipulate microbial communities for human purposes. M. xanthus development also provides a model system to study how cells build multicellular structures of controlled size and shape, and adopt alternative fates, which are fundamental questions in biology. This dissertation focuses on improving the knowledge of M. xanthus response to nutrients and C-signaling during development. I have discovered and characterized the ultrasensitive response of starving cells to nutrient addition before the period of commitment, which provides a better understanding of how M. xanthus can avoid the high energy cost of the decision to sporulate. I have developed methods involving confocal fluorescence microscopy to visualize cellular shape change inside a nascent fruiting body. The nucleoid dynamics were also investigated as cells underwent morphological changes. It is exciting that I was able to measure single-cell gene expression in situ using confocal fluorescence microscopy and a custom image analysis pipeline. Results from my dissertation show that M. xanthus responds ultrasensitively to both nutrients and C-signaling during development, leading to bistability and alternative cell fates. My work has advanced ! 15! understanding of how cells can integrate multiple signals and respond appropriately to changes in their biotic and abiotic environments. ! 16! REFERENCES 17! ! REFERENCES Nealson KH, Platt T, Hastings JW. 1970. Cellular control of the synthesis and activity of the bacterial luminescent system. J Bacteriol 104(1):313–322 Engebrecht J, Nealson K, Silverman M. 1983. Bacterial bioluminescence: Isolation and genetic analysis of functions from Vibrio fischeri Cell!32:773-781 Passador L, Cook JM, Gambello MJ, Rust L, Iglewski BH. 1993. Expression of Pseudomonas aeruginosa virulence genes requires cell-to-cell communication. Science 260(5111):1127-30. 10.1126/science.8493556 Janzon L, Arvidson S. 1990. The role of the delta-lysin gene (hld) in the regulation of virulence genes by the accessory gene regulator (agr) in Staphylococcus aureus. EMBO J. 9(5):1391–99 Dunny GM, Brown BL, Clewell DB. 1978. Induced cell aggregation and mating in Streptococcus faecalis: evidence for a bacterial sex pheromone. Proc Natl Acad Sci U S A!75(7):3479–3483. doi:10.1073/pnas.75.7.3479 Hoch JA. 2014. Control of Cellular Development in Sporulating Bacteria by the Phosphorelay Two-Component Signal Transduction SystemTwo-Component Signal Transduction. See Hoch & Silhavy 1995, 129–44 1. 2. 3. 4. 5. 6. 7. 9. Dworkin M, Kaiser D. 1985. Cell interactions in myxobacterial growth and development. Science 230(4721):18-24. 10.1126/science.3929384 8. Miller MB, Bassler BL. 2001. Quorum Sensing in Bacteria. Annu Rev Microbiol 55:165- 99. 10.1146/annurev.micro.55.1.165 Kuspa A, Plamann L, Kaiser D. 1992. A-signalling and the cell density requirement for Myxococcus xanthus development. J Bacteriol 174(22):7360-69. 10.1128/jb.174.22.7360- 7369.1992 10. Lobedanz S, Søgaard-Andersen L. 2003. Identification of the C-signal, a contact- dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes Dev 17(17):2151-61. 10.1101/gad.274203 11. Reichenbach H. 1999. The ecology of the myxobacteria. Environ Microbiol 1(1):15-21. 10.1046/j.1462-2920.1999.00016.x 12. Keane R, Berleman J. 2016. The predatory life cycle of Myxococcus xanthus. Microbiol (United Kingdom) 162(1):1-11. 10.1099/mic.0.000208 13. McBride MJ, Zusman DR. 1996. Behavioral analysis of single cells of Myxococcus xanthus in response to prey cells of Escherichia coli. FEMS Microbiol Lett 137(2-3):227- 31. 10.1111/j.1574-6968.1996.tb08110.x 14. Lueders T, Kindler R, Miltner A, Friedrich MW, Kaestner M. 2006. Identification of ! 18! bacterial micropredators distinctively active in a soil microbial food web. Appl Environ Microbiol 72(8): 5342–5348. 10.1128/AEM.00400-06 15. Hodgkin J, Kaiser D. 1979. Genetics of gliding motility in Myxococcus xanthus (Myxobacterales): Two gene systems control movement. MGG Mol Gen Genet 171:167– 176. https://doi.org/10.1007/BF00270004 16. Sun H, Zusman DR, Shi W. 2000. Type IV pilus of Myxococcus xanthus is a motility apparatus controlled by the frz chemosensory system. Curr Biol 70(1):1-4. 10.1016/s0960-9822(00)00705-3 17. Li Y, Sun H, Ma X, Lu A, Lux R, Zusman D, Shi W. 2003. Extracellular polysaccharides mediate pilus retraction during social motility of Myxococcus xanthus. Proc Natl Acad Sci U S A 100(9):5443-8. 10.1073/pnas.0836639100 18. Mignot T, Shaevitz JW, Hartzell PL, Zusman DR. 2007. Evidence that focal adhesion complexes power bacterial gliding motility. Science. 10.1126/science.1137223 19. Hodgkin J, Kaiser D. 1977. Cell to cell stimulation of movement in nonmotile mutants of Myxococcus. Proc Natl Acad Sci U S A. 10.1073/pnas.74.7.2938 20. Yang R, Bartle S, Otto R, Stassinopoulos A, Rogers M, Plamann L, Hartzell P. 2004. AglZ is a filament-forming coiled-coil protein required for adventurous gliding motility of Myxococcus xanthus. J Bacteriol 186(18):6168-78. 10.1128/JB.186.18.6168-6178.2004 21. Ward MJ, Lew H, Zusman DR. 2000. Social motility in Myxococcus xanthus requires FrzS, a protein with an extensive coiled-coil domain. Mol Microbiol 37(6):1357-71. 10.1046/j.1365-2958.2000.02079.x 22. Guzzo M, Agrebi R, Espinosa L, Baronian G, Molle V, Mauriello EMF, Brochier-Armanet C, Mignot T. 2015. Evolution and Design Governing Signal Precision and Amplification in a Bacterial Chemosensory Pathway. PLoS Genet. 10.1371/journal.pgen.1005460 23. Kaimer C, Berleman JE, Zusman DR. 2012. Chemosensory signaling controls motility and subcellular polarity in Myxococcus xanthus. Curr Opin Microbiol 15(6):751-7. 10.1016/j.mib.2012.10.005 24. Bustamante VH, Martínez-Flores I, Vlamakis HC, Zusman DR. 2004. Analysis of the Frz signal transduction system of Myxococcus xanthus shows the importance of the conserved C-terminal region of the cytoplasmic chemoreceptor FrzCD in sensing signals. Mol Microbiol 53(5):1501-13. 10.1111/j.1365-2958.2004.04221.x 25. Thutupalli S, Sun M, Bunyak F, Palaniappan K, Shaevitz JW. 2015. Directional reversals enable Myxococcus xanthus cells to produce collective one-dimensional streams during fruiting-body formation. J R Soc Interface. 10.1098/rsif.2015.0049 26. Thaxter R. 1892. On the Myxobacteriaceæ, a New Order of Schizomycetes. Bot Gaz. 14:389–406 27. Sager B, Kaiser D. 1993. Two cell-density domains within the Myxococcus xanthus fruiting body. Proc Natl Acad Sci U S A 90(8): 3690–94. 10.1073/pnas.90.8.3690 ! 19! 28. Kroos L. 2017. Highly Signal-Responsive Gene Regulatory Network Governing Myxococcus Development. Trends Genet 33(1):3-15. 10.1016/j.tig.2016.10.006 29. Curtis PD, Taylor RG, Welch RD, Shimkets LJ. 2007. Spatial organization of Myxococcus xanthus during fruiting body formation. J Bacteriol 189(24):9126-30. 10.1128/JB.01008- 07 30. Xie C, Zhang H, Shimkets LJ, Igoshin OA. 2011. Statistical image analysis reveals features affecting fates of Myxococcus xanthus developmental aggregates. Proc Natl Acad Sci U S A 108(14):5915-5920. 10.1073/pnas.1018383108 31. Dworkin M, Gibson SM. 1964. A system for studying microbial morphogenesis: Rapid formation of microcysts in Myxococcus xanthus. Science 146(3641):243-4. 10.1126/science.146.3641.243 32. Sadler W, Dworkin M. 1966. Induction of cellular morphogenesis in Myxococcus xanthus. II. Macromolecular synthesis and mechanism of inducer action. J Bacteriol 91(4):1520-5. 33. O’Connor KA, Zusman DR. 1997. Starvation-independent sporulation in Myxococcus xanthus involves the pathway for β-lactamase induction and provides a mechanism for competitive cell survival. Mol Microbiol 24(4):839-50. 10.1046/j.1365- 2958.1997.3931757.x 34. Kottel RH, Bacon K, Clutter D, White D. 1975. Coats from Myxococcus xanthus: characterization and synthesis during myxospore differentiation. J Bacteriol 124(1):550-7. Inouye M, Inouye S, Zusman DR. 1979. Biosynthesis and self-assembly of protein S, a development-specific protein of Myxococcus xanthus. Proc Natl Acad Sci U S A 76(1):209-13. 10.1073/pnas.76.1.209 36. McCleary WR, Esmon B, Zusman DR. 1991. Myxococcus xanthus protein C is a major spore surface protein. J Bacteriol McCleary WR, Esmon B, Zusman DR. Myxococcus xanthus protein C is a major spore surface protein. J Bacteriol. 1991;173(6):2141–2145. doi:10.1128/jb.173.6.2141-2145.1991. 10.1128/jb.173.6.2141-2145.1991 37. Otani M, Kozuka S, Xu C, Umezawa C, Sano K, Inouye S. 1998. Protein W, a spore- specific protein in Myxococcus xanthus, formation of a large electron-dense particle in a spore. Mol Microbiol 30(1):57-66. 10.1046/j.1365-2958.1998.01037.x 38. Tzeng L, Singer M. 2005. DNA replication during sporulation in Myxococcus xanthus fruiting bodies. Proc Natl Acad Sci U S A. 10.1073/pnas.0506969102 39. Tzeng L, Ellis TN, Singer M. 2006. DNA replication during aggregation phase is essential for Myxococcus xanthus development. J Bacteriol 102(40):14428-33. 10.1128/JB.188.8.2774-2779.2006 40. Rosario CJ, Singer M. 2007. The Myxococcus xanthus developmental program can be delayed by inhibition of DNA replication. J Bacteriol 189(24):8793-8800. 10.1128/JB.01361-07 35. ! 20! 41. Goldman BS, Nierman WC, Kaiser D, Slater SC, Durkin AS, Eisen J, Ronning CM, Barbazuk WB, Blanchard M, Field C, Halling C, Hinkle G, Iartchuk O, Kim HS, Mackenzie C, Madupu R, Miller N, Shvartsbeyn A, Sullivan SA, Vaudin M, Wiegand R, Kaplan HB. 2006. Evolution of sensory complexity recorded in a myxobacterial genome. Proc Natl Acad Sci U S A 103(41):15200-5. 10.1073/pnas.0607335103 42. Müller S, Willett JW, Bahr SM, Darnell CL, Hummels KR, Dong CK, Vlamakis HC, Kirby JR. 2013. Draft genome sequence of Myxococcus xanthus wild-type strain DZ2, a model organism for predation and development. Genome Announc 1(3):e00217-13. 10.1128/genomeA.00217-13 43. Bretl DJ, Kirby JR. 2016. Molecular Mechanisms of Signaling in Myxococcus xanthus Development. J Mol Biol 428(19):3805-30. 10.1016/j.jmb.2016.07.008 44. Biran D, Kroos L. 1997. In vitro transcription of Myxococcus xanthus genes with RNA polymerase containing σA , the major sigma factor in growing cells 25:463-72. Mol Microbiol. 10.1046/j.1365-2958.1997.4751843.x 45. Ueki T, Inouye S. 2001. SigB, SigC, and SigE from Myxococcus xanthus homologous to σ32are not required for heat shock response but for multicellular differentiation. J Mol Microbiol Biotechnol 3(2):287-293. 10.1128/JB.183.21.6282-6287.2001 46. Viswanathan P, Singer M, Kroos L. 2006. Role of#sigmaD in regulating genes and signals during Myxococcus xanthus development. J Bacteriol 188:3246–3256. 10.1128/JB.188.9.3246-3256.2006 47. Ueki T, Xu CY, Inouye S. 2005. SigF, a new sigma factor required for a motility system of Myxococcus xanthus. J Bacteriol 187(24):8537-41. 10.1128/JB.187.24.8537-8541.2005 48. Whitworth DE. 2014. Transcriptional Regulatory Mechanisms during Myxococcus xanthus Development. Myxobacteria:!Multicellularity and Differentiation. 149–168 49. Skotnicka D, Petters T, Heering J, Hoppert M, Kaever V, Søgaard-Andersen L. 2016. Cyclic dI-GMP regulates type IV pilus-dependent motility in Myxococcus xanthus. J Bacteriol 198(1):77-90. 10.1128/JB.00281-15 50. Gronewold TMA, Kaiser D. 2001. The act operon controls the level and time of C-signal production for Myxococcus xanthus development. Mol Microbiol 40(3):744-56. 10.1046/j.1365-2958.2001.02428.x 51. Meiser P, Bode HB, Müller R. 2006. The unique DKxanthene secondary metabolite family from the myxobacterium Myxococcus xanthus is required for developmental sporulation. Proc Natl Acad Sci U S A 103(50):19128-33. 10.1073/pnas.0606039103 52. Müller S, Strack SN, Ryan SE, Kearns DB, Kirby JR. 2015. Predation by Myxococcus xanthus induces Bacillus subtilis to form spore-filled megastructures. Appl Environ Microbiol 81(1):203-10. 10.1128/AEM.02448-14 53. Harris BZ, Kaiser D, Singer M. 1998. The guanosine nucleotide (p)ppGpp initiates development and A-factor production in Myxococcus xanthus. Genes Dev 12(7):1022-35. 10.1101/gad.12.7.1022 ! 21! 54. Manoil C, Kaiser D. 1980. Guanosine pentaphosphate and guanosine tetraphosphate accumulation and induction of Myxococcus xanthus fruiting body development. J Bacteriol 141(1):305-15 55. Singer M, Kaiser D. 1995. Ectopic production of guanosine penta- and tetraphosphate can initiate early developmental gene expression in Myxococcus xanthus. Genes Dev 9:1633-44 56. Kim SK, Kaiser D. 1990. Cell aligmnent required in differentiation of Myxococcus xanthus. Science 249(4971):926-28. 10.1126/science.2118274 57. Skotnicka D, Smaldone GT, Petters T, Trampari E, Liang J, Kaever V, Malone JG, Singer M, Søgaard-Andersen L. 2016. A Minimal Threshold of c-di-GMP Is Essential for Fruiting Body Formation and Sporulation in Myxococcus xanthus. PLoS Genet 12(5): e1006080. 10.1371/journal.pgen.1006080 58. Romling U, Galperin MY, Gomelsky M. 2013. Cyclic di-GMP: the First 25 Years of a Universal Bacterial Second Messenger. Microbiol Mol Biol Rev 77(1):1-52. 10.1128/MMBR.00043-12 59. Bush M, Dixon R. 2012. The Role of Bacterial Enhancer Binding Proteins as Specialized Activators of 54-Dependent Transcription. Microbiol Mol Biol Rev 76(3):497–529. 10.1128/MMBR.00006-12 60. Sun H, Shi W. 2001. Genetic studies of mrp, a locus essential for cellular aggregation and sporulation of Myxococcus xanthus. J Bacteriol 183(16):4786–95. 183:4786–4795 61. McLaughlin PT, Bhardwaj V, Feeley BE, Higgs PI. 2018. MrpC, a CRP/Fnr homolog, functions as a negative autoregulator during the Myxococcus xanthus multicellular developmental program. Mol Microbiol 109:245-261. 10.1111/mmi.13982 62. Feeley BE, Bhardwaj V, McLaughlin PT, Diggs S, Blaha GM, Higgs PI. 2019. An amino- terminal threonine/serine motif is necessary for activity of the Crp/Fnr homolog, MrpC and for Myxococcus xanthus developmental robustness. Mol Microbiol 112(5):1531-51. 10.1111/mmi.14378 63. Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of hundreds of developmentally-regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. 10.1186/1471-2164-15-1123 64. Cho K, Zusman DR. 1999. Sporulation timing in Myxococcus xanthus is controlled by the espAB locus. Mol Microbiol 34(4):714-25. 10.1046/j.1365-2958.1999.01633.x 65. Higgs PI, Jagadeesan S, Mann P, Zusman DR. 2008. EspA, an orphan hybrid histidine protein kinase, regulates the timing of expression of key developmental proteins of Myxococcus xanthus. J Bacteriol 190:4416–4426. 10.1128/JB.00265-08 66. Schramm A, Lee B, Higgs PI. 2012. Intra- and interprotein phosphorylation between two- hybrid histidine kinases controls Myxococcus xanthus developmental progression. J Biol Chem 287:25060-72. 10.1074/jbc.M112.387241 ! 22! 67. Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736-47. 10.1128/JB.01692-14 68. Saha S, Patra P, Igoshin O, Kroos L. 2019. Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C-signaling. Mol Microbiol 111:1732-52. 10.1111/mmi.14249 69. Ellehauge E, Norregaard M, Søgaard-Andersen L. 1998. The FruA signal transduction protein provides a checkpoint for the temporal co-ordination of intercellular signals in Myxococcus xanthus development. Mol Microbiol 30:807–17. 10.1046/j.1365- 2958.1998.01113.x 70. Bourret RB. 2010. Receiver domain structure and function in response regulator proteins. Curr Opin Microbiol 13(2):142-9. 10.1016/j.mib.2010.01.015 71. Gronewold TMA, Kaiser D. 2007. Mutations of the act promoter in Myxococcus xanthus. J Bacteriol 189(5):1836-1844. 10.1128/JB.01618-06 72. Mittal S, Kroos L. 2009. A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proc Natl Acad Sci U S A 106:1965-70. 10.1073/pnas.0808516106 73. Lee JS, Son B, Viswanathan P, Luethy PM, Kroos L. 2011. Combinatorial regulation of fmgD by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 193(7):1681-89. 10.1128/JB.01541-10 74. Mittal S, Kroos L. 2009. Combinatorial regulation by a novel arrangement of frua and MrpC2 transcription factors during Myxococcus xanthus development. J Bacteriol 191:2753–2763. 10.1128/JB.01818-08 75. Son B, Liu Y, Kroos L. 2011. Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 193:2756–2766. 10.1128/JB.00205-11 76. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L. 2015. Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 197:240–51. 10.1128/JB.02310-14 77. Wireman JW, Dworkin M. 1977. Developmentally induced autolysis during fruiting body formation by Myxococcus xanthus. J Bacteriol 129(2):798–802 78. Hagen TJ, Shimkets LJ. 1990. Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. J Bacteriol 172(1):15-23. 10.1128/jb.172.1.15-23.1990 79. Shimkets LJ, Rafiee H. 1990. CsgA, an extracellular protein essential for Myxococcus xanthus development. J Bacteriol 172(9):5299-306. 10.1128/jb.172.9.5299-5306.1990 80. Kim SK, Kaiser D. 1990. Purification and properties of Myxococcus xanthus C-factor, an intercellular signaling protein. Proc Natl Acad Sci U S A 87(10):3635-39. ! 23! 81. Rolbetzki A, Ammon M, Jakovljevic V, Konovalova A, Søgaard-Andersen L. 2008. Regulated Secretion of a Protease Activates Intercellular Signaling during Fruiting Body Formation in M. xanthus. Dev Cell 15(4):627-34. 10.1016/j.devcel.2008.08.002 82. Konovalova A, Löbach S, Søgaard-Andersen L. 2012. A RelA-dependent two-tiered regulated proteolysis cascade controls synthesis of a contact-dependent intercellular signal in Myxococcus xanthus. Mol Microbiol 84(2):260-75. 10.1111/j.1365- 2958.2012.08020.x 83. Boynton TO, Shimkets LJ. 2015. Myxococcus CsgA, Drosophila sniffer, and human HSD10 are cardiolipin phospholipases. Genes Dev 29(18):1903–14. 10.1101/gad.268482.115 10.1073/pnas.87.10.3635 84. Hagen DC, Bretscher AP, Kaiser D. 1978. Synergism between morphogenetic mutants of Myxococcus xanthus. Dev Biol 64(2):284-96. 10.1016/0012-1606(78)90079-9 85. Kim SK, Kaiser D. 1990. C-factor: A cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 61(1):19-26. 10.1016/0092-8674(90)90211-v 86. Kim SK, Kaiser D. 1991. C-factor has distinct aggregation and sporulation thresholds during Myxococcus development. J Bacteriol 173:1722-28. 10.1128/jb.173.5.1722- 1728.1991 ! 24! CHAPTER 2: Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC This chapter was published in the Journal of Bacteriology and can be found at the following citation: Hoang Y, Kroos L. 2018. Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC. J Bacteriol. 200(22): e00456-18. 10.1128/JB.00456-18 All AMS journals provide authors the right to use and reprint published work without written permission as part of a dissertation, given the work is properly cited (see above). The publication reprint and supplemental data are provided on the following pages. ! ! 25! CHAPTER 2: Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC Abstract Upon depletion of nutrients, Myxococcus xanthus forms mounds on a solid surface. The differentiation of rod-shaped cells into stress-resistant spores within mounds creates mature fruiting bodies. The developmental process can be perturbed by the addition of nutrient medium before the critical period of commitment to spore formation. The response was investigated by adding a 2-fold dilution series of nutrient medium to starving cells. An ultrasensitive response was observed, as indicated by a steep increase in the spore number after the addition of 12.5% versus 25% nutrient medium. The level of MrpC, which is a key transcription factor in the gene regulatory network, correlated with the spore number after nutrient medium addition. The MrpC level decreased markedly by 3 h after adding nutrient medium but recovered more after the addition of 12.5% than after 25% nutrient medium addition. The difference in MrpC levels was greatest midway during the period of commitment to sporulation, and mound formation was restored after 12.5% nutrient medium addition but not after adding 25% nutrient medium. Although the number of spores formed after 12.5% nutrient medium addition was almost normal, the transcript levels of “late” genes in the regulatory network failed to rise normally during the commitment period. However, at later times, expression from a reporter gene fused to a late promoter was higher after adding 12.5% than after adding 25% nutrient medium, consistent with the spore numbers. The results suggest that a threshold level of MrpC must be achieved in order for mounds to persist and for cells within to differentiate into spores. ! ! 26! Introduction Myxococcus xanthus provides an excellent model system to study bacterial interactions, as it can undergo a multicellular developmental process which involves the coordinated movement and differentiation of thousands of cells (reviewed in reference 1). Upon depletion of nutrients, cells alter their movements to form mounds (reviewed in reference 2). The differentiation of rod-shaped cells inside mounds into round stress-resistant spores results in the formation of fruiting bodies, while some cells persist outside as peripheral rods (3). During the developmental process, the majority of the population lyses, and a minority of the cells remain as spores or peripheral rods (4) (5) (6). How cells adopt three different fates is unknown. Understanding cell fate determination in this model system has broad potential implications since microbes form communities (microbiomes and biofilms) in which cells differentiate and profoundly impact their environment, including other living organisms (7) (8) (9) (10) (11). The developmental process of M. xanthus is governed by a signal-responsive gene regulatory network (reviewed in reference 12). Starvation triggers the production of intracellular second messenger signals. The intracellular signal guanosine pentaphosphate and guanosine tetraphosphate [(p)ppGpp] leads to production of the extracellular A- and C-signals (13–16). The C-signal is a short-range signal that appears to be a proteolytic fragment of CsgA (17) (18) and/or lipids produced by phospholipase activity of intact CsgA (19). C-signaling appears to posttranslationally activate FruA (5) (20) (21), a transcription factor whose production is controlled by a cascade of transcription factors governed by starvation (reviewed in reference 12) (Fig. 2.1). The cascade appears to include phosphorylated MrpB (MrpB-P) and MrpC (22) (23) (Fig. 2.1). MrpC binds to the promoter regions of hundreds of developmentally regulated genes (24), including the promoter region of fruA, where MrpC appears to activate transcription (25) (Fig. 2.1). FruA and MrpC bind cooperatively to the promoter regions of many genes important for mound formation and/or sporulation (24) (26) (27) (28) (29) (30). Among the genes under combinatorial control of FruA and MrpC is the dev operon (26), whose products, DevTRS, ! ! 27! negatively autoregulate transcription (31) (32) (33) to prevent the overexpression of DevI, an inhibitor of the cellular shape change that leads to spore formation (31) (34) (Fig. 2.1). Figure 2.1 Simplified model of the gene regulatory network governing M. xanthus development. Starvation increases the level of phosphorylated MrpB (MrpB-P), which activates mrpC transcription. MrpC negatively autoregulates by competing with MrpB-P for binding to the mrpC promoter region. The addition of nutrient medium can halt development by lowering the mrpC transcript level (as described in this study but not depicted in the model) and by inducing proteolysis of MrpC, as depicted here. MrpC activates transcription of the gene for FruA and causes an increase in C-signaling, which posttransla- tionally activates FruA to FruA*. FruA* promotes mound formation, which enhances short-range C-signaling by bringing cells into proximity, creating a positive feedback loop. FruA* and MrpC bind cooperatively to the promoter region of the dev operon and activate transcription. The resulting DevTRS proteins negatively autoregulate. DevI normally delays cellular shape change leading to spore matura- tion, but if overproduced, DevI blocks sporulation. A rising level of FruA* in cells within mounds, together with MrpC, activates the transcription of genes involved in commitment to cellular shape change and spore maturation, resulting in spore-filled fruiting bodies. See the text for details and references. ! ! 28! Combinatorial control by FruA and MrpC has been proposed to provide checkpoints for mound formation and persistent starvation, respectively, prior to commitment to sporulation (5) (28) (35). FruA could provide a checkpoint for mound formation if C-signaling activates FruA and activated FruA (designated FruA*) stimulates mound formation, forming a positive feedback loop (Fig. 2.1). Efficient C-signaling requires cells to move into alignment (36) (37) (38), as they do during mound formation (39) (40) (41). C-signaling appears to activate FruA (5) (20), but the mechanism of activation is unknown. FruA* may cause methylation of the FrzCD component (42) (43) of the Frz chemosensory system that controls the gliding movements of cells (44), since C-signaling alters cell movements to promote mound formation (45) (46). As mounds form, cells move into alignment, enhancing short-range C-signaling, which would cause FruA* to rise further. The rising level of FruA* would eventually activate genes that commit a cell to form a spore within a mound, provided that MrpC is available to bind cooperatively to promoter regions (Fig. 2.1). MrpC could provide a checkpoint for persistent starvation, because MrpC in developing cells is susceptible to proteolysis upon the addition of nutrient medium (35) (Fig. 2.1). Under submerged culture conditions, in which growing cells are resuspended in starvation buffer and allowed to adhere to the bottom of a plastic container, the cells in the biofilm undergo development (47). Under these conditions, wild-type M. xanthus strain DK1622 forms mounds between 15 and 18 h poststarvation (PS) (35). Sonication-resistant spores form by 27 h PS, and the number of spores increases greatly by 36 to 48 h PS (5) (35). The critical period of commitment to sporulation was found to be between 24 and 30 h PS, since adding nutrient medium before or at 24 h blocked sporulation, whereas adding nutrient medium at 30 h or later did not block sporulation (35). Before and during the commitment period, adding nutrient medium caused the MrpC level to decline rapidly due to proteolysis. Proteolysis of MrpC was proposed to halt the expression of genes required for commitment to sporulation if nutrients reappear. ! ! 29! Sporulation involves cellular shape change and spore maturation (Fig. 2.1). During maturation, the spore becomes resistant to stresses, such as sonication and heat. One step during maturation is deposition of a polysaccharide coat on the spore surface, and this step is carried out by products of the exo and nfs operons (48) (49). These operons are expressed late during the developmental process under the control of C-signaling (50) (51), but their regulation is not well understood, although the expression of exo (previously called fdgA) depends in part on FruA, and the DNA-binding domain of FruA binds upstream of the three exo promoters (52). In this study, we investigated the response of developing M. xanthus to the addition of different nutrient medium concentrations. We added a 2-fold dilution series of nutrient medium to developing cells and measured phenotypic and molecular markers of the developmental process. We discovered that cells respond ultrasensitively to the addition of nutrient medium in terms of mound and spore formation, and recovery of the ability to form mounds and spores after nutrient medium addition correlated with the MrpC level during the period of commitment to sporulation. Our results suggest that a threshold level of MrpC must be reached in order to restore mound formation, perhaps by meeting a threshold requirement for C-signaling and FruA* to initiate mound formation and sustain a positive feedback loop that further elevates the FruA* level enough to efficiently activate genes that commit cells to form spores (Fig. 2.1). Results M. xanthus responds ultrasensitively to nutrient medium addition during development . Previous work has shown that 24 to 30 h PS is a critical period of commitment to spore formation during the development of M. xanthus under sub- merged culture conditions (35). Replacement of the starvation buffer overlaying the cells adhered to the bottom of the culture vessel with CTTYE nutrient medium (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH2PO4·K2HPO4, 8 mM MgSO4, [final pH 7.6], 0.2% yeast extract) prior to commitment halts development. To further investigate the effect of CTTYE addition, the starvation buffer at 18 h ! ! 30! PS was replaced with a 2-fold dilution series of CTTYE in starvation buffer, and sonication- resistant spores were measured at 30, 36, and 42 h of total incubation. At all three times, the number of sonication-resistant spores was much higher after the addition of ≤12.5% CTTYE than after the addition of ≥25% CTTYE (see Fig. S2.1 in the supplemental material). Therefore, the response to 12.5% versus 25% CTTYE was examined more thoroughly, including biological replicates and spore quantification at later times (42, 48, or 72 h of total incubation) to determine whether cells eventually formed spores after adding 25% CTTYE at 18 h PS. When the starvation buffer overlay was replaced with 25% or 100% CTTYE, cells dispersed from the mounds and appeared to grow as a biofilm of nearly uniform thickness (Fig. 2.2A). In contrast, after replacement with 12.5% CTTYE, the initial reversal of mound formation was overcome, and mounds were restored by 27 h of total incubation. These mounds began to darken at 42 h and became darker at 72 h, suggestive of ongoing sporulation. Consistent with the observation of darkened mounds, a steep increase in the number of sonication-resistant spores was observed with 12.5% CTTYE compared with that with 25% CTTYE at 42, 48, or 72 h of total incubation (Fig. 2.2B). A 2-fold decrease in nutrient concentration led to about a 100-fold increase in the number of sonication-resistant spores. Hence, the sporulation response is nonlinear over this range of nutrient medium addition to developing cells. The values of the slopes between 12.5% versus 25% CTTYE in the log-log plots shown in Fig. 2.2B range between -6 and -7, indicative of an ultrasensitive response (53). ! ! 31! Figure 2.2 The ultrasensitive response of developing M. xanthus to nutrient medium addition. Wild-type strain DK1622 was subjected to starvation under submerged culture conditions. The culture supernatant was replaced with the indicated percentage of CTTYE nutrient medium at 18 h poststarvation (PS). A. Effect of CTTYE addition on fruiting body formation. Images were obtained at 18 h PS and at the indicated times of total incubation. Mounds formed by 18 h PS (an arrow points to one) and darkened by 27 h of total incubation when the culture supernatant was replaced with fresh starvation buffer (0% CTTYE). Cells dispersed from mounds when the culture supernatant was replaced with CTTYE, but mounds reformed by 27 h of total incubation with 12.5% CTTYE, whereas cells remained dispersed with 25% or 100% CTTYE. Bar, 100 µm. Similar results were observed in three biological replicates. B. Effect of CTTYE addition on sonication-resistant spore formation. Cultures were harvested at the indicated times of total incubation for the measurement of sonication-resistant spores. Log-log plot shows the average of three biological replicates (y axis; error bars show one standard deviation) at each nutrient concentration (x axis). Mature spores form despite 12.5% CTTYE addition. Since a considerable number of sonication-resistant spores were observed despite the addition of 12.5% CTTYE at 18 h PS, we investigated whether these spores were “mature” (i.e., resistant to heat and sonication and able to germinate in the presence of nutrients and grow to produce a colony). The number of mature spores was lower than the number of sonication- resistant spores from the same treatment, but ! ! 32! the striking difference between 12.5% versus 25% CTTYE remained at 42 or 48 h of total incubation and was even greater on average at 72 h (Fig. 2.3). Clearly, mature spores form much more efficiently after adding 12.5% CTTYE than after adding 25% CTTYE. Figure 2.3 Effect of nutrient medium addition on formation of sonication-resistant spores and mature spores. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. The culture supernatant was replaced with fresh starvation buffer (0% CTTYE) or the indicated percentage of CTTYE nutrient medium at 18 h poststarvation, and cultures were harvested at the indicated times of total incubation for the measurement of sonication-resistant spores and mature spores. Values are the average of three biological replicates, and error bars show one standard deviation. MrpC protein level correlates with sporulation after CTTYE addition. Since sporulation exhibited an ultrasensitive response to CTTYE addition, the molecular response to the addition of 12.5% versus 25% CTTYE was compared by systematically measuring transcript and protein levels. MrpC is a key transcription factor in the gene regulatory network governing sporulation (reviewed in reference 12). Upon the addition of 100% CTTYE, MrpC has been reported to be rapidly degraded, halting the expression of genes important for sporulation (35). Therefore, the mrpC transcript and MrpC protein levels were measured after the addition of 0%, 12.5%, 25%, or 100% CTTYE at 18 h PS, until 30 h of total incubation (i.e., leading up to and including the critical period of commitment to spore formation). ! ! 33! The addition of CTTYE caused a 5-fold decline in mrpC transcripts within 1 h (Fig. 2.4A). While the mrpC transcript level remained low after the addition of 100% CTTYE, a partial recovery of the transcript level occurred after the addition of 12.5% or 25% CTTYE. On average, the mrpC transcript level recovered more after adding 12.5% CTTYE than after adding 25% CTTYE at 24 to 30 h of total incubation (i.e., during the commitment period). While the average mrpC transcript level was higher after the addition of 12.5% than after the addition of 25% CTTYE, especially at 24 and 27 h, the error of the measurements was large, and the differences were not statistically significant at the 95% confidence level (P = 0.15 at 24 h and P = 0.16 at 27 h in Student’s unpaired two-tailed t tests). Nevertheless, the results clearly show significant recovery of the mrpC transcript level after 12.5% CTTYE addition (e.g., P = 0.02 comparing 19 and 27 h). The MrpC protein level was consistent with the mrpC transcript level, but the initial decline in the MrpC level after CTTYE addition was not as large, and the difference after 12.5% versus 25% CTTYE addition was less pronounced at 24 h (Fig. 2.4B). The difference in MrpC levels was greatest at 27 h (P = 0.06 comparing 12.5% versus 25% CTTYE addition). In previous work, our group observed a rapid decline in MrpC protein levels and a slower decline of mrpC transcript levels after the addition of 100% CTTYE to developing cells (35). However, the samples for the two measurements were collected in different experiments under slightly different conditions (i.e., 6-well plates for protein measurements and petri plates for transcript measurements). Here, the samples for the two measurements were collected from petri plates in the same experiment, and by 1 h after adding CTTYE at 18 h PS, the mrpC transcript level had declined more than the MrpC protein level (Fig. 2.4). To test whether development in 6-well plates would yield a different outcome than development in petri plates, samples for protein measurements were collected from 6-well plates. The results were similar to those shown in Fig. 2.4B (i.e., petri plate samples), except the difference in MrpC level after 12.5% versus 25% CTTYE addition was more pronounced at 24 h (P = 0.05), similar to the ! ! 34! difference at 27 h (P = 0.03) (see Fig. S2.2 in the supplemental material). Clearly, a difference between development in 6-well plates versus petri plates does not explain why the MrpC level declined more slowly here (Fig. 2.4B and S2.2) than previously (35). We do not understand the discrepancy. Figure 2.4 Effect of nutrient medium addition on the mrpC transcript and MrpC protein levels. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. Cultures were harvested at the indicated times of total incubation. A. mrpC transcript levels. RNA was isolated from the cultures and subjected to quantitative reverse transcription-PCR (RT-qPCR) analysis. B. MrpC protein levels. Protein samples from the cultures were analyzed by immunoblotting using anti-MrpC antibodies. In both panels, values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. The mrpC transcript level declined slowly in previous work, but CTTYE was added to developing cells at 24 h PS (35). Here, CTTYE was added at 18 h PS, and the mrpC transcript level declined rapidly (Fig. 2.4A). We did not investigate this difference. Rather, we focused on possible explanations of the recovery of the mrpC transcript and MrpC protein levels after CTTYE addition (Fig. 2.4), because recovery of the MrpC level correlated with restoration of mound (Fig. 2.2A) and spore (Fig. 2.2B and 2.3) formation after 12.5% CTTYE addition at 18 h PS. ! ! 35! Recovery of the MrpC level after CTTYE addition is regulated by synthesis rather than stability. Our results showed recovery of both the mrpC transcript and MrpC protein levels after CTTYE addition (Fig. 2.4). A simple explanation of the recovery would be resumption of mrpC transcription, which is positively regulated by MrpB, an NtrC-like response regulator likely activated by phosphorylation (22). However, an assay for MrpB-P has not been developed. To determine whether CTTYE addition regulates an event necessary for MrpB-P production, we measured the mrpB transcript level after the addition of 0%, 12.5%, 25%, or 100% CTTYE at 18 h PS, until 30 h of total incubation. The mrpB transcript level rose at 19 h on average and then returned to the starting value at 21 h for all treatments (Fig. 2.5). Later, the level stayed about the same after adding 12.5% or 25% CTTYE, but oddly, the level was elevated on average and highly variable after the 0% or 100% treatments. Although we do not understand the results with the 0% and 100% treatments, clearly there was no correlation between the mrpB transcript level and the mrpC transcript level (Fig. 2.4A) after adding 12.5% versus 25% CTTYE. Since the response to CTTYE addition is not due to regulation of the mrpB transcript level, we infer that the response is due to regulation of the MrpB-P level and/or another factor that affects the mrpC transcript level (see Discussion). While recovery of the mrpC transcript level (Fig. 2.4A) may account for recovery of the MrpC protein level (Fig. 2.4B and S2.2) after CTTYE addition, we also considered the possibility that a difference in nutrient-regulated proteolysis of MrpC could play a role (35). To determine the stability of MrpC after CTTYE addition at 18 h PS, the protein synthesis inhibitor chloramphenicol was added at 27 h, and samples were collected every 15 min for 1 h. As expected, the MrpC levels in all samples declined after chloramphenicol addition, and the decline was on average more rapid in samples to which CTTYE had been added at 18 h PS (Fig. 2.6). However, the decline was more rapid on average in samples to which 12.5% CTTYE had been added than with 25% CTTYE. This would not account for the observed higher steady- state level of MrpC at 27 h after adding 12.5% than after adding 25% CTTYE (Fig. 2.4B and ! ! 36! S2.2). Given the short half-life of MrpC in both cases (Fig. 2.6), it is unlikely that a difference in MrpC stability earlier during the period between 18 h PS (when CTTYE was added) and 27 h of total incubation (when chloramphenicol was added) accounts for the observed difference in the MrpC level. Taken together, our results suggest that recovery of the mrpC transcript level after CTTYE addition involves regulation of transcript synthesis and/or stability, and this accounts for the recovery of the steady-state MrpC level due to protein synthesis (i.e., translation of mrpC transcripts) rather than regulation of protein stability. Figure 2.5 Effect of nutrient medium addition on the mrpB transcript level. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. Cultures were harvested at the indicated times of total incubation. RNA was isolated from the cultures and subjected to RT- qPCR analysis. Values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. ! ! 37! Figure 2.6 MrpC protein stability after nutrient medium addition. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions in 6-well plates. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. At 27 h of total incubation, culture supernatants were supplemented with 200 µg/ml chloramphenicol, and a culture was harvested immediately (t0) and at each indicated time (tx) after chloramphenicol addition, for measurement of the MrpC level by immunoblot using anti-MrpC antibodies. MrpC levels at tx were normalized to that at t0 for each of three biological replicates and used to determine the MrpC half-life for each replicate. The average half-life (average t1/2) and one standard deviation are shown in the legend. The graph shows the average ln(tx/t0) and one standard deviation for the three biological replicates. FruA protein level remains high after CTTYE addition, despite a drop in the fruA transcript level. MrpC activates transcription of the fruA gene (25), which encodes another important transcription factor for sporulation (reviewed in reference 12). Therefore, we measured the fruA transcript and FruA protein levels after the addition of 0%, 12.5%, 25%, or 100% CTTYE at 18 h PS, until 30 h of total incubation. The addition of CTTYE caused the fruA transcript level to drop quickly, and surprisingly, there was no recovery (Fig. 2.7A), unlike the mrpC transcript (Fig. 2.4A) and MrpC protein (Fig. 2.4B and S2.2) levels. In spite of the very low level of fruA transcript after CTTYE addition, the FruA protein level remained high at 19 and 21 h and only declined about 2-fold later, with the exception that the FruA level recovered on average at 27 h after adding 12.5% CTTYE (Fig. 2.7B); however, the error of the measurements ! ! 38! is large, and the difference from the 25% CTTYE addition is not statistically significant (P = 0.18). Figure 2.7 Effect of nutrient medium addition on the fruA transcript and FruA protein levels. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. Cultures were harvested at the indicated times of total incubation. A. fruA transcript levels. RNA was isolated from the cultures and subjected to RT-qPCR analysis. B. FruA protein levels. Protein samples from the cultures were analyzed by immunoblotting using anti-FruA antibodies. In both panels, values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. Transcript levels of “late” genes fail to rise normally after CTTYE addition. To test whether CTTYE addition affects the expression of operons induced later than the mrpC and fruA genes during development, we measured transcript levels after the addition of 0%, 12.5%, 25%, or 100% CTTYE at 18 h PS, until 30 h of total incubation, for operons we refer to as “late” genes. Among the late genes we chose (dev, exo, nfs, fadIJ, and MXAN_3259), the dev operon is the best characterized and requires cooperative binding of MrpC and FruA to the promoter region (26) in order to produce proteins that regulate the timing of sporulation (31). FruA has also been shown to bind in the promoter region of the exo operon (52), whose products together with those of the nfs operon are involved in forming the polysaccharide layer of the spore coat (48) (49). Developmental expression of fadIJ depends on MrpC and FruA and is important ! ! 39! for ︎oxidation of fatty acids stored in lipid bodies and both mound formation and spore maturation (54). The promoter region of the predicted MXAN_3259–MXAN_3263 operon is bound by the DNA-binding domain of Nla6, a developmental transcription factor, and the disruption of MXAN_3259 impairs spore formation but not mound formation (55). Figure 2.8 Effect of nutrient medium addition on late gene transcript levels. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. Cultures were harvested at the indicated times of total incubation. RNA was isolated from the cultures and subjected to RT- qPCR analysis. Values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. dev (A) and exo (B) transcript levels. Strikingly, all the late gene transcript levels decreased by 1 or 3 h after CTTYE addition (i.e., at 19 or 21 h of total incubation) and failed to recover by 30 h of total incubation, with the ! ! 40! notable exception that all five exhibited extremely high and varied transcript levels at 30 h after the 100% nutrient treatment (Fig. 2.8 and S2.4). The high transcript levels at 30 h after the addition of 100% CTTYE at 18 h PS were unexpected, since mound formation was reversed (Fig. 2.2A), very few spores formed at 30 h (Fig. S2.1) or later (Fig. 2.2B and 2.3), and the MrpC level remained low (Fig. 2.4B and S2.2). The dev transcript level was different from those of the other four since a decrease was not observed until 3 h after CTTYE addition and a large increase was not observed at 30 h in the 0% CTTYE addition control (Fig. 2.8A). The other four late gene transcript levels decreased on average by 1 h after CTTYE addition and increased on average from about 4- to 30-fold at 30 h in the 0% CTTYE addition control (Fig. 2.8B and S2.4). During the normal period of commitment to sporulation (24 to 30 h PS), the levels of all the late gene transcripts are considerably higher on average in the 0% CTTYE addition control than after the addition of 12.5% or 25% CTTYE at 18 h PS (Fig. 2.8 and S2.4). How then do spores form after the addition of 12.5% CTTYE at 18 h PS? We hypothesized that late genes are expressed more after adding 12.5% than after adding 25% CTTYE at 18 h PS but not until after 30 h of total incubation, because the number of sonication-resistant spores rises between 30 and 36 h (Fig. S2.1), and the number of mature spores rises even later, between 48 and 72 h (Fig. 2.3). To test our hypothesis, we measured the expression of a lacZ reporter fused to the exo promoter. Because the fusion was created by the insertion of Tn5 lac Ω7536 into exoC (48) (50), sonication-resistant spores fail to form, and β-galactosidase activity could easily be measured in sonic extracts of cells late during development without the need to break spores. We added 0%, 12.5%, or 25% CTTYE at 18 h PS and measured β-galactosidase activity at 42, 48, and 72 h of total incubation. Activity rose about 3-fold between 42 and 72 h in the 0% CTTYE addition control (Fig. 2.9). In agreement with our hypothesis, activity was higher after adding 12.5% CTTYE than 25% CTTYE at all times tested and rose about 2-fold between 42 and 72 h after the 12.5% nutrient treatment. ! ! 41! Higher expression of exo and other late genes may account for the higher number of spores observed after adding 12.5% versus 25% CTTYE (Fig. 2.2B and 2.3). Figure 2.9 Effect of nutrient medium addition on exo-lacZ expression. M. xanthus strain DK10524 was subjected to starvation under submerged culture conditions in 6-well plates. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the key. Cultures were harvested at the indicated times of total incubation for the measurement of β-galactosidase activity. Values are the average of three biological replicates, and error bars show one standard deviation. Discussion Our experiments revealed an ultrasensitive response of developing M. xanthus to the addition of nutrient medium. A 2-fold difference in the concentration of CTTYE determined whether mounds reformed and resulted in about a 100-fold difference in the number of spores formed. Systematic measurements of the molecular response showed that recovery of the MrpC level after the addition of CTTYE correlated with the ultrasensitive developmental response, suggesting a threshold requirement for MrpC. Recovery of the MrpC level was regulated at the level of synthesis and/or stability of the mrpC transcript, not at the level of MrpC stability. In contrast to MrpC, FruA was quite stable after CTTYE addition. We propose that recovery of the MrpC level is necessary to allow C-signaling and activation of FruA, which promotes mound formation and sporulation (Fig. 2.1). Although the formation of mounds and ! ! 42! spores is restored, the transcript levels of late genes are not fully restored, indicating that the developmental process is robust to this perturbation. A threshold level of MrpC may be necessary to meet a threshold requirement for C-signaling. MrpC is important for C-signaling, since an mrpC null mutant failed to rescue the development of a csgA mutant upon mixing (23). Also, the mrpC mutant accumulated less CsgA protein during development. Hence, MrpC positively regulates C-signaling (Fig. 2.1). Interestingly, distinct threshold levels of C-signaling are required for mound formation, sporulation, and expression of particular genes during development (56) (57) (58). In one study, small incremental differences (e.g., 0.2 unit) in the amount of CsgA added to a csgA mutant determined whether or not mounds formed, affected the sporulation efficiency 100-fold, and differentially influenced the expression of an early versus a late developmental gene (56). These observations suggest that the developmental process responds ultrasensitively to C- signaling. Since C-signaling is positively regulated by MrpC (23), we propose that after CTTYE addition, the MrpC level must recover to a certain threshold in order to meet a threshold requirement of C-signaling for mound formation. In order to test this hypothesis, it will be crucial to elucidate the mechanism by which C-signaling activates FruA and how FruA* initiates mound formation. Once initiated, mound formation would bring cells into proximity, enhancing short- range C-signaling (reviewed in reference 12) and resulting in more FruA*. This positive- feedback loop may explain the ultrasensitivity of the developmental process to C-signaling. Positive feedback is a common mechanism to generate ultrasensitivity leading to bistability and alternative cell fates (53) (59) (60). A rising level of FruA* produced in response to C-signaling during development would not only promote mound formation, it would trigger gene expression and lead to sporulation (56) (57) (58). FruA* and MrpC have been proposed to bind cooperatively to activate the transcription of genes that commit rod-shaped cells to form spores (5) (28) (35) (Fig. 2.1). The ! ! 43! dev operon is regulated in this way (26), and the products of the dev operon include the negative autoregulators DevTRS and the sporulation inhibitor DevI, forming a timer of sporulation (31) (34) (Fig. 2.1). The dev promoter may have a relatively low threshold for FruA*, whereas genes essential for commitment to sporulation may have a higher threshold (5), since some genes depend more strongly on C-signaling and are expressed later during development than dev (27) (30) (61). Neither the dev transcript level nor that of the other late genes we tested recovered significantly more after adding 12.5% CTTYE at 18 h PS than after adding 25% CTTYE by 30 h of total incubation (Fig. 2.8 and S2.4). However, none of these genes are essential for the initial cellular shape change during sporulation, but rather, dev controls the timing of sporulation (31), and the other late genes we tested are involved in spore maturation (48) (54) (S. Saha and L. Kroos, unpublished data). We propose that other late genes remain to be identified which have a relatively high threshold for FruA* and therefore depend strongly on C-signaling, and these are essential for the initial cellular shape change during sporulation. We predict that the transcript level of such genes would be higher during the critical period of commitment to sporulation (i.e., at about 27 h of total incubation, when the MrpC level is higher; Fig. 2.4B and S2.2) after adding 12.5% CTTYE at 18 h PS than after adding 25% CTTYE. We are searching for such genes. CTTYE addition regulates both mrpC transcript and MrpC protein levels. The rapid decline of the mrpC transcript level observed here (Fig. 2.4A) suggested a more important role of transcript synthesis and/or stability in the response to CTTYE addition than previously suspected (35). Interestingly, the mrpC transcript declined to a similar level by 1 h after the addition of different concentrations of CTTYE at 18 h PS (Fig. 2.4A). This observation could be explained if the addition of ≤ 12.5% CTTYE nearly halts the initiation of mrpC transcription, since the half-life of mrpC transcripts in developing cells at 18 h PS is about 10 min (5). MrpB-P appears to be an activator of mrpC transcription (22) (62). We showed that the mrpB transcript level (Fig. 2.5) did not respond to CTTYE addition in a way that would explain the observed ! ! 44! changes in the mrpC transcript level (Fig. 2.4A). Several other explanations are possible for the rapid decline and differential recovery of the mrpC transcript level after the addition of different concentrations of CTTYE. Nutrient-regulated dephosphorylation of MrpB-P would provide an attractive mechanism to control the activation of mrpC transcription. Such a mechanism could involve nutrient sensing by MrpA, which has been proposed to act as a phosphatase of MrpB rather than a kinase, based on mutant phenotypes (22). Alternatively, or in addition, MrpC itself could be a nutrient-sensing autorepressor. In two recent studies, MrpC was shown to negatively autoregulate (5) (62), and it may do so by competing with MrpB-P for overlapping binding sites in the mrpC promoter region (62). It is also possible that CTTYE addition destabilizes mrpC transcripts. The mechanism of mrpC transcript turnover has not been investigated. The MrpC protein level declined more slowly and recovered more slowly than the mrpC transcript level (Fig. 2.4). This pattern is consistent with regulation primarily of the transcript level. We expected the decline of the MrpC level after 100% CTTYE addition to be more rapid based on previous work by our group (35). Although we do not understand this difference, we did observe shortening of the MrpC half-life after the addition of CTTYE (Fig. 2.6), consistent with nutrient-regulated proteolysis of MrpC (35). Furthermore, our results indicate that lower concentrations of CTTYE (i.e., 12.5% or 25% rather than 100%) induce proteolysis that persists for at least 9 h, since CTTYE was added at 18 PS and MrpC stability was measured at 27 h of total incubation (Fig. 2.6). However, the average half-life of MrpC did not differ significantly after the addition of 12.5% versus 25% CTTYE, so a difference in nutrient-regulated proteolysis does not account for the observed differential recovery of the steady-state MrpC level at 27 h (Fig. 2.4B and S2.2). Rather, differential recovery of the mrpC transcript level (Fig. 2.4A) and the resulting synthesis of MrpC appear to account for differential recovery of the MrpC protein level (Fig. 2.4B and S2.2). ! ! 45! Recovery of the MrpC level after the addition of 12.5% CTTYE at 18 h PS reached a maximum at 24 to 27 h of total incubation (Fig. 2.4B and S2.2). According to our model, a threshold level of MrpC was reached so that C-signaling could activate FruA, restoring mound formation (Fig. 2.2A) and efficient sporulation (Fig. 2.2B and 2.3). We cannot rule out the possibility that the activity of MrpC is regulated (e.g., by phosphorylation or nucleotide binding), as well as the level of MrpC. It was reported previously that two serine-threonine protein kinases, Pkn8 and Pkn14, form a cascade in which Pkn8 phosphorylates Pkn14 and Pkn14 phosphorylates MrpC (63) (64). Deletion of pkn8 or pkn14 in the DZF1 strain background accelerated development (63). However, recently, it was found that the deletion of pkn14 in the DZ2 or DK1622 background (B. Feeley and P. Higgs, personal communication) or a plasmid insertion mutation in pkn14 in the DK1622 background has no effect on development (S. Saha, Y. Hoang, R. Rajagopalan, and L. Kroos, unpublished data). Therefore, it is unlikely that phosphorylation of MrpC by Pkn14 affects MrpC activity in a way that is important for the ultrasensitive response. MrpC has a potential nucleotide-binding domain (22), but nucleotide binding has not been shown to affect MrpC activity, despite testing many different nucleotides for an effect on DNA binding by MrpC (28). On the other hand, accelerated accumulation of MrpC or CsgA due to reduced proteolysis (65) (66) or overexpression (57), respectively, has been correlated with accelerated development, consistent with our model that a threshold level of MrpC must be reached to meet a threshold requirement for C-signaling. In an effort to test our model that MrpC must reach a threshold level in order for development to proceed, we fused mrpC to a vanillate-inducible promoter (67) and integrated the fusion ectopically in an mrpC mutant. Maximum induction with vanillate restored the MrpC level in the mrpC mutant to about 60% of the wild-type level at 18 h PS but was insufficient to restore mound formation or sporulation (see Fig. S2.5 in the supplemental material). These results suggest that the MrpC level must reach more than about 60% of the wild-type level in order to restore development. We note that this condition was met when 12.5% CTTYE was ! ! 46! added at 18 h PS, after 27 h of total incubation, but not when 25% CTTYE was added (Fig. 2.4B and S2.2). Unfortunately, the vanillate-inducible construct was unable to fully restore the MrpC level in the mrpC mutant, which our model predicts would restore development. An inducible system capable of higher expression during development will need to be engineered. The mrpC promoter may be a good starting point, since the mrpC transcript is highly abundant, about 10% of the 16S rRNA level in cells at 24 h PS (5). After the addition of 25% CTTYE, the MrpC level recovered slightly at 24 h of total incubation but did not increase thereafter by 30 h of incubation (Fig. 2.4B and S2.2). According to our model, the level of MrpC did not reach the threshold to allow C-signaling to activate FruA, so the positive-feedback loop that normally sustains mound formation was interrupted. Cells in the mounds that had formed by 18 h PS dispersed and appeared to grow as a biofilm (Fig. 2.2A). The spore number did not increase significantly between 36 and 72 h (Fig. S2.1, 2.2B, and 2.3). There was no indication that the developmental process had restarted by 72 h. The phenotypic effects of adding 100% CTTYE were similar to those of adding 25% CTTYE (Fig. 2.2 and 2.3), and there was no recovery of the MrpC level after the addition of 100% CTTYE (Fig. 2.4B and S2.2). These observations raise the question of why development fails to restart, rather than restarting later than after 12.5% CTTYE addition (e.g., after the nutrients in ≥25% CTTYE are exhausted). We considered the possibility that cells exposed to ≥25% CTTYE lose the ability to undergo development. Therefore, we added 25% or 100% CTTYE at 18 h PS; then, at 27 h of total incubation, we replaced the overlay with fresh starvation buffer. We found that cells retain the ability to form mounds by 42 h of total incubation, and the mounds darken by 96 h of total incubation (Fig. S2.6). Interestingly, the darkened mounds are abnormally large at 96 h under two conditions, either the addition of 12.5% CTTYE at 18 h PS or 25% CTTYE at 18 h PS followed by replacement of the overlay with fresh starvation buffer at 27 h of total incubation. These conditions appear to favor the merging of smaller mounds, a behavior observed for some mounds in close proximity without the addition of nutrient medium to ! ! 47! developing cells (68). The total number of spores formed was similar in all cases where darkened mounds were observed (Fig. S2.6). We conclude that exposure to ≥25% CTTYE (for 9 h) does not irreversibly damage cells so they lose the ability to form mounds and spores. Perhaps, exposure to ≥25% CTTYE for more than 9 h irreversibly damages cells for development. Alternatively, CTTYE may contain, or cause cells to produce, an inhibitor of development that can neither be utilized nor overcome when ≥25% CTTYE is added at 18 h PS. The impact on C-signal-dependent production of FruA* may be quite large when ≥25% CTTYE is added at 18 h PS. In a recent study conducted under the same conditions of submerged culture development used here, it was estimated that by 18 h PS, mound formation had progressed sufficiently for C-signaling to activate FruA at least 9-fold (5). This estimate is a minimum based on measurements of the dev transcript levels in the wild type and mutants, and on mathematical modeling assuming low occupancy of the dev promoter by active FruA and MrpC. If the dev promoter approaches saturation binding of active FruA and MrpC, C-signaling may activate FruA as much as 30-fold by 18 h PS. Even the addition of 12.5% CTTYE at 18 h PS likely lowers C-signal-dependent production of FruA* temporarily, accounting for the initial reversal of mound formation, which was restored by 27 h of total incubation (Fig. 2.2A). Sporulation despite low levels of fruA and late gene transcripts after CTTYE addition. It was surprising that the fruA transcript level after CTTYE addition (Fig. 2.7A) did not correlate with the MrpC level (Fig. 2.4B and S2.2), since MrpC binds to the fruA promoter region and appears to activate transcription (25). One possible explanation is that recovery of the MrpC level after CTTYE addition does not meet a threshold for binding to the fruA promoter region, even though, according to our model, recovery of the MrpC level after 12.5% CTTYE addition meets a threshold for C-signaling (i.e., perhaps the csgA promoter region or that of another gene important for C-signaling has a higher affinity for MrpC). Alternatively, CTTYE addition may induce a factor that prevents MrpC from activating fruA transcription or that destabilizes fruA transcripts. ! ! 48! Despite the low level of fruA transcripts after CTTYE addition, the FruA protein level remained fairly high, decreasing at most about 2-fold at 24 to 30 h of total incubation (Fig. 2.7B). At 27 h of total incubation, the half-life of FruA was ≥1 h whether or not CTTYE had been added at 18 h PS (Fig. S2.3), consistent with prior determinations of the FruA half-life with or without 100% CTTYE addition at 18 and 24 h PS (35). The stability of FruA ensures its presence despite the drop in fruA transcripts. Hence, FruA is available for activation by C-signaling under conditions permissive for the recovery of MrpC and C-signaling. Like the fruA transcript level, the late gene transcript levels declined after CTTYE addition and failed to recover, except all the late genes exhibited extremely high and varied transcript levels at 30 h after adding 100% CTTYE at 18 h PS (Fig. 2.8 and S2.4). The exception is very surprising, because there was no other indication that development had restarted after 100% CTTYE addition, either in terms of the MrpC level (Fig. 2.4B and S2.2) or phenotypically (Fig. 2.2 and 2.3). Rather, cells appear to grow as a biofilm after either 100% or 25% CTTYE addition (Fig. 2.2A). Yet, late gene transcripts accumulate abundantly only after adding 100% CTTYE, and not until 12 h later (i.e., 30 h of total incubation) (Fig. 2.8 and S2.4). Under this particular condition, at least some cells in the population appear to make one or more factors that increase synthesis and/or stability of late gene transcripts. Identification of such factors may provide insight into the potential biological significance of this unexpected response and could uncover novel regulatory mechanisms that normally control late gene expression. How do mature spores form after 12.5% CTTYE addition at 18 h PS, considering that four of the late genes we tested (exo, nfs, fadIJ, and MXAN_3259) are involved in spore maturation (48) (54) (S. Saha and L. Kroos, unpublished data) but their transcript levels failed to rise normally at 30 h of total incubation (Fig. 2.8B and S2.4)? We showed that exo is expressed more highly after adding 12.5% CTTYE than 25% CTTYE at 18 h PS, when examined between 42 and 72 h of total incubation (Fig. 2.9). Perhaps, the other three late genes also recover slowly after 12.5% CTTYE addition. However, recovery of exo expression was incomplete after ! ! 49! 12.5% CTTYE addition compared to that with 0% CTTYE (Fig. 2.9). Related to this, it was concluded recently that very little exo expression is required for sporulation, based on finding a low level of exo transcripts in a devI devS double mutant that nevertheless forms a normal number of mature spores (31). Perhaps, transcripts of the late genes involved in spore maturation are normally made in excess. On the other hand, the standard method of measuring mature spores, which we employ, may not reflect the selective pressures encountered by spores in natural environments (i.e., the spores made after 12.5% CTTYE addition may differ from spores made normally, in terms of resistance to insults harsher than what we employed). Of course, like FruA, the products of the late genes may be more stable than their transcripts after CTTYE addition. However, the fruA transcript and FruA protein levels were already quite high at 18 h PS (Fig. 2.7), whereas the transcript levels of the late genes involved in spore maturation were low at 18 h PS and increased on average from about 4- to 30-fold at 30 h in the 0% CTTYE addition control (Fig. 2.8B and S2.4). Some combination of the possibilities mentioned above likely explains how mature spores form after 12.5% CTTYE addition. Materials and Methods Bacterial strains, plasmids, and primers. The strains, plasmids, and primers used in this study are listed in Table S2.1 in the supplemental material. To construct pYH1, which was used to induce production of MrpC in an mrpC mutant background, primers Pvan-mrpC F and Pvan-mrpC R were used to generate a PCR product using chromosomal DNA from M. xanthus strain DK1622 as the template. The product was digested with NdeI and EcoRI and mixed with pMR3692 (also digested with the same restriction enzymes) in a Gibson assembly reaction to enzymatically join the overlapping DNA fragments (69). The reaction mixture was transformed into Escherichia coli strain DH5α with outgrowth in Luria-Bertani (LB) liquid prior to plating on LB agar supplemented with 50 µg/ml kanamycin sulfate for selection at 37°C. The DNA sequence of the joined fragments was verified, and the plasmid was transformed into M. xanthus strain ! ! 50! SW2808 using electroporation (70), with outgrowth in Casitone-Tris (CTT) liquid (see below for description of medium) prior to plating on CTT agar supplemented with 40 µg/ml kanamycin sulfate to select transformants. Ectopic integration of pYH1 was verified by colony PCR using primers pMR3691 MCS G-F and mrpC qPCR R. The resulting strain was named YH1. Growth and development. M. xanthus was grown at 32°C on CTT agar (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH2PO4·K2HPO4, 8 mM MgSO4, [final pH 7.6], 1.5% agar) or in CTTYE liquid medium (CTT with 0.2% yeast extract) with shaking at 350 rpm. Development was performed under submerged culture conditions in 8.5-cm-diameter plastic petri dishes or in 3.5-cm-diameter wells of plastic 6-well plates with starvation buffer MC7 (10 mM morpholinepropanesulfonic acid [MOPS] [pH 7.0], 1 mM CaCl2) (47). Briefly, cells from log- phase cultures in CTTYE liquid were collected by centrifugation and resuspended in MC7 to 1,000 Klett units. Either 1.5 ml of the cell suspension was added to 10.5 ml of MC7 in a petri dish or 240 µl of the cell suspension was added to 1.6 ml MC7 in one well of a 6-well plate. Upon incubation at 32°C, cells formed a biofilm at the bottom of the plate. At 18 h PS, the overlay was replaced with fresh MC7 as a control or with different concentrations of CTTYE to observe the effect of adding nutrient medium. To induce the production of MrpC from Pvan- mrpC, 0.5 mM vanillate was added during growth when the culture reached 50 Klett units, and 0.5 mM vanillate was added after cells were collected and resuspended in MC7 at the beginning of submerged culture development. Microscopy. The images of developing fruiting bodies were taken with a Leica Wild M8 microscope equipped with an Olympus E-620 digital camera. Sample collection. Samples were collected from petri dish submerged cultures unless indicated in the figure legends that 6-well plates were used. To collect samples from petri dishes, at the indicated times, the overlay was replaced with 5 ml of MC7. Cells were scraped from the plate, and the entire contents were aspirated into a 15-ml plastic tube. Samples were mixed thoroughly by pipetting and vortexing as described previously (35). For immunoblot ! ! 51! analysis, a 0.1-ml aliquot was mixed with an equal volume of 2X sample buffer (0.125 M Tris- HCl [pH 6.8], 20% glycerol, 4% sodium dodecyl sulfate [SDS], 0.2% bromophenol blue, 0.2 M dithiothreitol), boiled for 5 min, and stored at -20°C. For total protein analysis and/or spore measurement, a 0.4-ml aliquot was stored at -20°C. The rest of the sample (4.5 ml) was treated with 0.5 ml of RNase stop solution (5% phenol [pH = 7] in ethanol) and subjected to flash- freezing in liquid nitrogen before storing at -80°C. To collect samples from submerged cultures in 6-well plates, cells were scraped from one well, and the entire contents were aspirated in to a 15-ml plastic tube and thoroughly mixed as described previously (35). For immunoblot analysis, a 0.1-ml aliquot was prepared and stored as described above. For total protein analysis and/or spore measurement, a 0.4-ml aliquot was stored at -20°C. For the β-galactosidase assay, the overlay was replaced with 1 ml of MC7, cells were scraped and mixed thoroughly, and one 0.4-ml aliquot was collected and stored at -20°C. Spore measurement. Sonication-resistant spores and mature spores were measured as described previously (35). Briefly, a 0.4-ml aliquot collected as described above was sonicated for 10-s intervals three times with cooling on ice between. A Neubauer chamber was used to count sonication-resistant spores. To measure mature spores, samples were incubated at 50°C for 2 h before sonication, serial dilutions were made in MC7, aliquots (0.1 ml) were mixed with 2.5 ml CTT soft (0.7%) agar and plated on CTT agar, and colonies were counted after incubation at 32°C for 10 days. RNA extraction and analysis. RNA was extracted from samples collected as described above and treated with RNase stop solution. RNA was extracted using the hot- phenol method, and the RNA was digested with DNase I (Roche) as described previously (65). Total RNA (1 µg) was used to synthesize cDNA with Superscript III reverse transcriptase (Lifetech) and random primers (Promega), according to the manufacturer’s instructions. Control reactions were not subjected to cDNA synthesis. Quantitative PCR (qPCR) was performed as ! ! 52! described previously (5). Briefly, cDNA was subjected to qPCR in quadruplicate using a LightCycler 480 system (Roche). Gene expression was quantified using the relative standard curve method (user bulletin 2; Applied Biosystems, MA, USA), with a standard curve generated with M. xanthus DK1622 chromosomal DNA for each set of qPCRs. The transcript level in each sample was normalized to the amount of 16S rRNA, which was also determined by qPCR and served as an internal control. The resulting values were further normalized to the average of the 18-h-PS samples, which was set as 1. Total protein and immunoblot analyses. For total protein analysis, a 0.4-ml aliquot was sonicated as described above for spore measurement and centrifuged at 10,000 x g for 1 min, and the protein content in the supernatant was measured using a Bradford (71) assay kit (catalog number 5000006; Bio-Rad Laboratories). For immunoblot analysis, we used a semiquantitative method to measure the relative levels of MrpC and FruA, as described previously (5). Briefly, an equal volume of each sample was subject to SDS-PAGE and immunoblotting, as described previously (72), with anti-MrpC (1:20,000 dilution) or anti-FruA antibodies (1:1,000 dilution). On each immunoblot, a sample of the wild-type strain DK1622 at 18 h PS served as an internal control for normalization of signal intensities across immunoblots. Signals were detected using a ChemiDoc MP imaging system (Bio-Rad), with exposure times short enough to ensure that signals were not saturated, and signal intensities were quantified using Image Lab 5.1 software (Bio-Rad) with the global analysis setting. After normalization to the internal control, each signal intensity was divided by the total protein concentration of the corresponding sample. The resulting values were further normalized to the average of the 18-h- PS samples, which was set as 1. Protein stability. The culture supernatant was replaced with fresh starvation buffer or different percentages of CTTYE at 18 h PS and then supplemented with 200 µg/ml chloramphenicol at 27 h of total incubation. Samples were collected immediately and at 15-min intervals for 1 h for total protein and immunoblot analyses, as described above, except without ! ! 53! an internal control, since only signals on the same blot were compared. Each signal intensity was divided by the total protein concentration of the corresponding sample and then normalized to the value for the relevant sample collected immediately after chloramphenicol addition, which was set as 1. The natural log of the resulting values was plotted against the time after chloramphenicol addition, and the slope of a linear fit of the data was used to calculate the protein half-life, assuming a first-order kinetic degradation reaction, as described previously (66). β-Galactosidase assay. Samples collected and stored as described above for the β- galactosidase assay were thawed on ice and then sonicated and centrifuged as described above for total protein analysis. The β-galactosidase-specific activity of the supernatant was determined as described previously (73) using o-nitrophenyl- β -D-galactoside (ONPG) as the substrate. Optical densities were read using a microplate reader SpectraMax M2 (Molecular Devices), and the specific activity was calculated using the following equation: (213 x OD420)/[(milliliter of supernatant assayed)(milligrams of protein/milliliter of supernatant)(minutes of incubation)] (OD420, optical density at 420 nm). Acknowledgments We thank Oleg Igoshin for discussions about the ultrasensitive response, Mitch Singer for providing bacterial strains, and Shreya Saha for suggesting primers for RNA analysis. This research was supported by National Science Foundation grant MCB-1411272 and by salary support for L.K. from Michigan State University AgBioResearch. ! ! 54! APPENDIX ! ! 55! Table S2.1. Bacterial strains, plasmids, and primers used in this study Strain, plasmid, or primer Strain E. coli DH5a M. xanthus DK1622 DK10524 SW2808 YH1 Plasmid pMR3692 pYH1 Primer 16S rRNA fwd 16S rRNA rev mrpC qPCR F mrpC qPCR R mrpB qPCR F1 mrpB qPCR R1 fruA oPH252 fruA oPH253 cas6-F cas6-R exoA-NF4 exoA-NR4 nfsA-NF nfsA-NR Mxan_5372 F1 Mxan_5372 R1 Mxan_3259 F3 Mxan_3259 R3 Pvan-mrpC F Pvan-mrpC R pMR3691 MCS G-F ! Description Source or reference l- f80dlacZDM15 D(lacZYA-argF)U169 recA1 endA1 hsdR17(rK- mK-) supE44 thi-1 gyrA relA1 (74) Laboratory strain Tn5 lac Ω7536 Dm_r_p_C_ _ DmrpC MXAN_0018-MXAN_0019::pYH1 MXAN_0018-MXAN_0019-PR3-4::vanR- MXAN_0018-MXAN_0019-PR3-4::vanR- Pvan::lacZ, Kmr Pvan::mrpC, Kmr CAAGGGAACTGAGAGACAGG CTCTAGAGATCCACTACTTGCG GGAGGCCATCGACTTCAAGG GGCCGGACTTCAGCAGGTAG CCTGCCAGATCCTTGGAAT TTGAGGGCTCTGGCTCTGT CGTCACGGAAGGCATCAATC CGAGATGATTTCCGGTGTGC TGGGGAAATCTAATGGTGTTTG GAGAACAGCAGATAGGCATGGT CAGCAAGGGCGGACAGAT CGGAGCATGACCTCGTGT TTCTTCATCCTGGACAAGCAC TCCAGGTTGACGCGGTAG CTGGAGTCTTCACGGACGAT TCTGTTCGACAACGAGGTCA TCCTCTCCGGGCAGAAGAC GCATCGATGATCTCCGTCA GATGCGAGGAAACGCATATGCACGGTTTCA ACCGCCCCCTC GTACGCGTAACGTTCGAATTCCTACTTCTC CTTGCCGGCGATCTC CACGATGCGAGGAAACGCA (75) (50) (22) This study (67) This study (77) (77) (35) (35) This study This study (31) (31) (5) (5) This study This study This study This study This study This study This study This study This study This study (5) ! 56! Figure S2.1 Effect of nutrient medium addition on the formation of sonication-resistant spores. Wild-type M. xanthus strain DK1622 was subject to starvation under submerged culture conditions in 6-well plates. Culture supernatants were replaced with a twofold dilution series of CTTYE nutrient medium in starvation buffer at 18h poststarvation. Cultures were harvested at the indicated times of total incubation for the measurement of sonication-resistant spores. Vertical dashed lines highlight the values observed after 12.5% versus 25% CTTYE addition. ! ! 57! Figure S2.2 Effect of nutrient medium addition on the MrpC protein level. The experiment was performed as described in the Figure 2.4 legend, except submerged culture was done in 6- well plates instead of petri plates. Protein samples from the cultures were analyzed by immunoblot using anti-MrpC antibodies. Values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. ! ! 58! Figure S2.3 FruA protein stability after nutrient medium addition. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions in 6-well plates. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the legend. At 27 h of total incubation, culture supernatants were supplemented with 200 µg/ml of chloramphenicol and a culture was harvested immediately (t0) and at each indicated time (tX) after chloramphenicol addition, for measurement of the FruA level by immunoblot using anti-FruA antibodies. FruA levels at tX were normalized to that at t0 for each of three biological replicates. In most cases, the time course did not extend long enough to determine the half-life. Since the error was large, we concluded the half-life was ≥1 h whether or not CTTYE was added. The graph shows the average ln(tX/t0) and one standard deviation for the three biological replicates. ! ! 59! Figure S2.4 Effect of nutrient medium addition on late gene transcript levels. Wild-type M. xanthus strain DK1622 was subjected to starvation under submerged culture conditions. At 18 h poststarvation, culture supernatants were replaced with fresh starvation buffer (0%) or the percentage of CTTYE nutrient medium indicated in the legend. Cultures were harvested at the indicated times of total incubation. RNA was isolated from the cultures and subjected to RT- qPCR analysis. Values are the average of three biological replicates, relative to the sample at 18 h, and error bars show one standard deviation. (A) nfs, (B) fadIJ, and (C) MXAN_3259 transcript levels. ! ! 60! Figure S2.5 Inducible expression of MrpC in an mrpC mutant. Wild-type M. xanthus strain DK1622, mrpC mutant strain SW2808, and the mrpC mutant bearing Pvan-mrpC were subjected to starvation under submerged culture conditions. A. MrpC protein levels at different times of development. Protein samples from the cultures were analyzed by immunoblot using anti- MrpC antibodies. Values for wild type (black circles) and mrpC Pvan-mrpC (gray triangles) are shown, relative to the wild-type sample at 0 h. No signal was detected for the mrpC mutant (data not shown). B. MrpC protein levels at 18 h poststarvation. Protein samples from the cultures were analyzed by immunoblot using anti-MrpC antibodies. Values are the average of three biological replicates, relative to the average of the wild-type samples at 18 h, and error bars show one standard deviation. C. Fruiting body formation. Wild type had formed darkened mounds (an arrow points to one) by 72 h poststarvation, but neither the mrpC mutant nor mrpC Pvan-mrpC formed mounds. Bar, 100 µm. ! ! 61! Figure S2.6 Development after exposure to ≥25% CTTYE. Wild-type strain DK1622 was subjected to starvation under submerged culture conditions. The culture supernatant was replaced with the indicated percentage of CTTYE nutrient medium at 18 h poststarvation (PS) and in some cases the overlay was replaced with fresh starvation buffer (0% CTTYE) at 27 h of total incubation. Images were obtained at the indicated times of total incubation. Darkened mounds formed by 42 h PS (a white arrow points to one in the control). Bar, 100 µm. Abnormally large darkened mounds formed by 96 h of total incubation under two conditions (black arrows). The number of sonication-resistant spores at 96 h of total incubation is expressed as a percentage of the number observed for the control, which was 7.5 x 107 spores/mL. ! ! 62! REFERENCES ! ! 63! REFERENCES 1. Yang Z, Higgs P. 2014. Myxobacteria: genomics, cellular and molecular biology. Caister Academic Press, Norfolk, UK. 2. Kaiser D., Robinson M., Kroos L. 2010. Myxobacteria, polarity, and multicellular morphogenesis. Cold Spring Harb Perspect Biol 2:a000380. doi:10.1101/cshperspect.a000380 3. O'Connor KA, Zusman DR. 1991. Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. J Bacteriol 173:3318–3333. doi:10.1128/jb.173.11.3318-3333.1991 4. Lee B, Holkenbrink C, Treuner-Lange A, Higgs PI. 2012. Myxococcus xanthus developmental cell fate production: heterogeneous accumulation of developmental regulatory proteins and reexamination of the role of MazF in developmental lysis. J Bacteriol 194:3058–3068. doi:10.1128/JB.06756-11 5. Saha S, Patra P, Igoshin OA, Kroos L. 2018. Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C-signaling. bioRxiv http://biorxiv.org/cgi/content/short/415331v1 6. Wireman JW, Dworkin M. 1977. Developmentally induced autolysis during fruiting body formation by Myxococcus xanthus. J Bacteriol 129:796–802 7. Flemming HC, Wingender J, Szewzyk U, Steinberg P, Rice SA, Kjelleberg S. 2016. Biofilms: an emergent form of bacterial life. Nat Rev Microbiol 14:563–575. doi:10.1038/nrmicro.2016.94 8. Nadell CD, Drescher K, Foster KR. 2016. Spatial structure, cooperation and competition in biofilms. Nat Rev Microbiol 14:589–600. doi:10.1038/nrmicro.2016.84 9. Shapiro J, Dworkin M. 1997. Bacteria as multicellular organisms. Oxford University Press, Oxford. 10. Wintermute EH, Silver PA. 2010. Dynamics in the mixed microbial concourse. Genes Dev 24:2603–2614. doi:10.1101/gad.1985210 11. Barratt MJ, Lebrilla C, Shapiro HY, Gordon JI. 2017. The gut microbiota, food science, and human nutrition: a timely marriage. Cell Host Microbe 22:134–141. doi:10.1016/j.chom.2017.07.006 12. Kroos L. 2017. Highly signal-responsive gene regulatory network governing Myxococcus development. Trends Genet 33:3–15. doi:10.1016/j.tig.2016.10.006 13. Crawford EW, Jr, Shimkets LJ. 2000. The Myxococcus xanthus socE and csgA genes are regulated by the stringent response. Mol Microbiol 37:788–799. doi:10.1046/j.1365- 2958.2000.02039.x ! ! 64! 14. Harris BZ, Kaiser D, Singer M. 1998. The guanosine nucleotide (p)ppGpp initiates development and A-factor production in Myxococcus xanthus. Genes Dev 12:1022–1035. doi:10.1101/gad.12.7.1022 15. Konovalova A, Lobach S, Sogaard-Andersen L. 2012. A RelA-dependent two-tiered regulated proteolysis cascade controls synthesis of a contact-dependent intercellular signal in Myxococcus xanthus. Mol Microbiol 84:260–275. doi:10.1111/j.1365- 2958.2012.08020 16. Konovalova A, Wegener-Feldbrugge S, Sogaard-Andersen L. 2012. Two intercellular signals required for fruiting body formation in Myxococcus xanthus act sequentially but non-hierarchically. Mol Microbiol 86:65–81. doi:10.1111/j.1365-2958.2012.08173.x 17. Kim SK, Kaiser D. 1990. C-factor: a cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 61:19–26. doi:10.1016/0092-8674(90)90211-V 18. Lobedanz S, Sogaard-Andersen L. 2003. Identification of the C-signal, a contact- dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes Dev 17:2151–2161. doi:10.1101/gad.274203 19. Boynton TO, Shimkets LJ. 2015. Myxococcus CsgA, Drosophila Sniffer, and human HSD17B10 are cardiolipin phospholipases. Genes Dev 29:1903–1914. doi:10.1101/gad.268482.115 20. Ellehauge E, Norregaard-Madsen M, Sogaard-Andersen L. 1998. The FruA signal transduction protein provides a checkpoint for the temporal co-ordination of intercellular signals in Myxococcus xanthus development. Mol Microbiol 30:807–817. doi:10.1046/j.1365-2958.1998.01113.x 21. Ogawa M, Fujitani S, Mao X, Inouye S, Komano T. 1996. FruA, a putative transcription factor essential for the development of Myxococcus xanthus. Mol Microbiol 22:757–767. doi:10.1046/j.1365-2958.1996.d01-1725.x 22. Sun H, Shi W. 2001. Genetic studies of mrp, a locus essential for cellular aggregation and sporulation of Myxococcus xanthus. J Bacteriol 183:4786–4795. doi:10.1128/JB.183.16.4786-4795.2001 23. Sun H, Shi W. 2001. Analyses of mrp genes during Myxococcus xanthus development. J Bacteriol 183:6733–6739. doi:10.1128/JB.183.23.6733-6739.2001 24. Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of many developmentally-regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. doi:10.1186/1471-2164-15-1123 25. Ueki T, Inouye S. 2003. Identification of an activator protein required for the induction of fruA, a gene essential for fruiting body development in Myxococcus xanthus. Proc Natl Acad Sci U S A 100:8782–8787. doi:10.1073/pnas.1533026100 26. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L. 2015. Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 197:240–251. doi:10.1128/JB.02310-14 ! ! 65! 27. Lee J, Son B, Viswanathan P, Luethy P, Kroos L. 2011. Combinatorial regulation of fmgD by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 193:1681– 1689. doi:10.1128/JB.01541-10 28. Mittal S, Kroos L. 2009. A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proc Natl Acad Sci U S A 106:1965–1970. doi:10.1073/pnas.0808516106 29. Mittal S, Kroos L. 2009. Combinatorial regulation by a novel arrangement of FruA and MrpC2 transcription factors during Myxococcus xanthus development. J Bacteriol 191:2753–2763. doi:10.1128/JB.01818-08 30. Son B, Liu Y, Kroos L. 2011. Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 193:2756–2766. doi:10.1128/JB.00205-11 31. Rajagopalan R, Kroos L. 2017. The dev operon regulates the timing of sporulation during Myxococcus xanthus development. J Bacteriol 199:e0078816. doi:10.1128/JB.00788-16 32. Thöny-Meyer L, Kaiser D. 1993. devRS, an autoregulated and essential genetic locus for fruiting body development in Myxococcus xanthus. J Bacteriol 175:7450–7462. doi:10.1128/jb.175.22.7450-7462.1993 33. Viswanathan P, Murphy K, Julien B, Garza AG, Kroos L. 2007. Regulation of dev, an operon that includes genes essential for Myxococcus xanthus development and CRISPR- associated genes and repeats. J Bacteriol 189:3738–3750. doi:10.1128/JB.00187-07 34. Rajagopalan R, Wielgoss S, Lippert G, Velicer GJ, Kroos L. 2015. devI is an evolutionarily young negative regulator of Myxococcus xanthus development. J Bacteriol 197:1249–1262. doi:10.1128/JB.02542-14 35. Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736–2747. doi:10.1128/JB.01692-14 36. Kim SK, Kaiser D. 1990. Cell motility is required for the transmission of C-factor, an intercellular signal that coordinates fruiting body morphogenesis of Myxococcus xanthus. Genes Dev 4:896–905. doi:10.1101/gad.4.6.896 37. Kim SK, Kaiser D. 1990. Cell alignment required in differentiation of Myxococcus xanthus. Science 249:926–928. doi:10.1126/science.2118274 38. Kroos L, Hartzell P, Stephens K, Kaiser D. 1988. A link between cell movement and gene expression argues that motility is required for cell-cell signaling during fruiting body development. Genes Dev 2:1677–1685. doi:10.1101/gad.2.12a.1677 39. O'Connor KA, Zusman DR. 1989. Patterns of cellular interactions during fruiting-body formation in Myxococcus xanthus. J Bacteriol 171:6013–6042. doi:10.1128/jb.171.11.6013-6024.1989 ! ! 66! 40. Sager B, Kaiser D. 1993. Two cell-density domains within the Myxococcus xanthus fruiting body. Proc Natl Acad Sci U S A 90:3690–3694 41. Shi W, Ngok FK, Zusman DR. 1996. Cell density regulates cellular reversal frequency in Myxococcus xanthus. Proc Natl Acad Sci U S A 93:4142–4146 42. Søgaard-Andersen L, Kaiser D. 1996. C factor, a cell-surface-associated intercellular signaling protein, stimulates the cytoplasmic Frz signal transduction system in Myxococcus xanthus. Proc Natl Acad Sci U S A 93:2675–2679 43. Søgaard-Andersen L, Slack F, Kimsey H, Kaiser D. 1996. Intercellular C-signaling in Myxococcus xanthus involves a branched signal transduction pathway. Genes Dev 10:740–754. doi:10.1101/gad.10.6.740 44. Blackhart BD, Zusman DR. 1985. “Frizzy” genes of Myxococcus xanthus are involved in control of frequency of reversal of gliding motility. Proc Natl Acad Sci U S A 82:8767– 8770. 45. Jelsbak L, Sogaard-Andersen L. 1999. The cell surface-associated intercellular C-signal induces behavioral changes in individual Myxococcus xanthus cells during fruiting body morphogenesis. Proc Natl Acad Sci U S A 96:5031–5036. 46. Jelsbak L, Sogaard-Andersen L. 2002. Pattern formation by a cell surface-associated morphogen in Myxococcus xanthus. Proc Natl Acad Sci U S A 99:2032–2037. doi:10.1073/pnas.042535699 47. Kuner JM, Kaiser D. 1982. Fruiting body morphogenesis in submerged cultures of Myxococcus xanthus. J Bacteriol 151:458–461 48. Müller FD, Schink CW, Hoiczyk E, Cserti E, Higgs PI. 2012. Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Mol Microbiol 83:486–505. doi:10.1111/j.1365-2958.2011.07944.x 49. Wartel M, Ducret A, Thutupalli S, Czerwinski F, Le Gall AV, Mauriello EM, Bergam P, Brun YV, Shaevitz J, Mignot T. 2013. A versatile class of cell surface directional motors gives rise to gliding motility and sporulation in Myxococcus xanthus. PLoS Biol. 11:e1001728. doi:10.1371/journal.pbio.1001728 50. Licking E, Gorski L, Kaiser D. 2000. A common step for changing cell shape in fruiting body and starvation-independent sporulation of Myxococcus xanthus. J Bacteriol 182:3553–3558. doi:10.1128/JB.182.12.3553-3558.2000 51. Müller FD, Treuner-Lange A, Heider J, Huntley SM, Higgs PI. 2010. Global transcriptome analysis of spore formation in Myxococcus xanthus reveals a locus necessary for cell differentiation. BMC Genomics 11:264. doi:10.1186/1471-2164-11-264 52. Ueki T, Inouye S. 2005. Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development. J Biol Chem 280:32279–32284. doi:10.1074/jbc.M507191200 ! ! 67! 53. Ray JC, Tabor JJ, Igoshin OA. 2011. Non-transcriptional regulatory processes shape transcriptional network dynamics. Nat Rev Microbiol 9:817–828. doi:10.1038/nrmicro2667 54. Bullock HA, Shen H, Boynton TO, Shimkets LJ. 2018. Fatty acid oxidation is required for Myxococcus xanthus development. J Bacteriol 200:e00572-17. doi:10.1128/JB.00572-17 55. Giglio KM, Zhu C, Klunder C, Kummer S, Garza AG. 2015. The enhancer binding protein Nla6 regulates developmental genes that are important for Myxococcus xanthus sporulation. J Bacteriol 197:1276–1287. doi:10.1128/JB.02408-14 56. Kim SK, Kaiser D. 1991. C-factor has distinct aggregation and sporulation thresholds during Myxococcus development. J Bacteriol 173:1722–1728. doi:10.1128/jb.173.5.1722- 1728.1991 57. Kruse T, Lobedanz S, Berthelsen NM, Sogaard-Andersen L. 2001. C-signal: a cell surface-associated morphogen that induces and co-ordinates multicellular fruiting body morphogenesis and sporulation in Myxococcus xanthus. Mol Microbiol 40:156–168. doi:10.1046/j.1365-2958.2001.02365.x ! ! 68! 58. Li S-F, Lee B, Shimkets LJ. 1992. csgA expression entrains Myxococcus xanthus development. Genes Dev 6:401–410. doi:10.1101/gad.6.3.401 59. Dubnau D, Losick R. 2006. Bistability in bacteria. Mol Microbiol 61:564–572. doi:10.1111/j.1365-2958.2006.05249.x 60. Smits WK, Kuipers OP, Veening JW. 2006. Phenotypic variation in bacteria: the role of feedback regulation. Nat Rev Microbiol 4:259–271. doi:10.1038/nrmicro1381 61. Kroos L, Kaiser D. 1987. Expression of many developmentally regulated genes in Myxococcus depends on a sequence of cell interactions. Genes Dev 1:840–854. doi:10.1101/gad.1.8.840 62. McLaughlin PT, Bhardwaj V, Feeley BE, Higgs PI. 2018. MrpC, a CRP/Fnr homolog, functions as a negative autoregulator during the Myxococcus xanthus multicellular developmental program. Mol Microbiol 109:245–261. doi:10.1111/mmi.13982 63. Nariya H, Inouye S. 2005. Identification of a protein Ser/Thr kinase cascade that regulates essential transcriptional activators in Myxococcus xanthus development. Mol Microbiol 58:367–379. doi:10.1111/j.1365-2958.2005.04826.x 64. Nariya H, Inouye S. 2006. A protein Ser/Thr kinase cascade negatively regulates the DNA-binding activity of MrpC, a smaller form of which may be necessary for the Myxococcus xanthus development. Mol Microbiol 60:1205–1217. doi:10.1111/j.1365- 2958.2006.05178.x 65. Higgs PI, Jagadeesan S, Mann P, Zusman DR. 2008. EspA, an orphan hybrid histidine protein kinase, regulates the timing of expression of key developmental proteins of Myxococcus xanthus. J Bacteriol 190:4416–4426. doi:10.1128/JB.00265-08 66. Schramm A, Lee B, Higgs PI. 2012. Intra- and inter-protein phosphorylation between two hybrid histidine kinases controls Myxococcus xanthus developmental progression. J Biol Chem 287:25060–25072. doi:10.1074/jbc.M112.387241 67. Iniesta AA, Garcia-Heras F, Abellon-Ruiz J, Gallego-Garcia A, Elias-Arnanz M. 2012. Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl-β-d-thiogalactopyranoside or vanillate. J Bacteriol 194:5875–5885. doi:10.1128/JB.01110-12 68. Xie C, Zhang H, Shimkets LJ, Igoshin OA. 2011. Statistical image analysis reveals features affecting fates of Myxococcus xanthus developmental aggregates. Proc Natl Acad Sci U S A 108:5915–5920. doi:10.1073/pnas.1018383108 69. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, III, Smith HO. 2009. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6:343–345. doi:10.1038/nmeth.1318 70. Kashefi K, Hartzell P. 1995. Genetic suppression and phenotypic masking of a Myxococcus xanthus frzF− defect. Mol Microbiol 15:483–494. doi:10.1111/j.1365- 2958.1995.tb02262.x 71. Bradford M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248– 254. doi:10.1016/0003-2697(76)90527-3 72. Yoder-Himes D, Kroos L. 2006. Regulation of the Myxococcus xanthus C-signal- dependent Ω4400 promoter by the essential developmental protein FruA. J Bacteriol 188:5167–5176. doi:10.1128/JB.00318-06 73. Kroos L, Kuspa A, Kaiser D. 1986. A global analysis of developmentally regulated genes in Myxococcus xanthus. Dev Biol 117:252–266. doi:10.1016/0012-1606(86)90368-4 74. Hanahan D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557-580. 75. Kaiser D. 1979. Social gliding is correlated with the presence of pili in Myxococcus xanthus. Proc. Natl. Acad. Sci. USA 76:5952-5956. 76. Licking E, Gorski L, Kaiser D. 2000. A common step for changing cell shape in fruiting body and starvation-independent sporulation of Myxococcus xanthus. J. Bacteriol. 182:3553-3558. 77. Ossa F, Diodati ME, Caberoy NB, Giglio KM, Edmonds M, Singer M, Garza AG. 2007. The Myxococcus xanthus Nla4 protein is important for expression of stringent response- associated genes, ppGpp accumulation, and fruiting body development. J. Bacteriol. 189:8474-8483. ! ! 69! CHAPTER 3: The spatiotemporal distribution of cellular shape and gene expression in nascent fruiting bodies of Myxococcus xanthus Abstract Myxococcus xanthus provides an attractive model system to study formation of multicellular structures and adoption of alternative cell fates. Upon starvation, rod-shaped cells send signals to each other and coordinate their movements to construct uniform mounds. Mounds mature into fruiting bodies as rods differentiate into round spores. Other cells undergo lysis or remain outside of fruiting bodies as peripheral rods. To determine when and where cells change shape, their membrane was stained and then visualized using confocal fluorescence microscopy. Mounds formed by 18 h poststarvation (PS), some rods began to change shape by 24 h, and many cells completed the transition to spores by 42 h. Transitioning cells (TCs) and spores were more abundant closer to the center of nascent fruiting bodies, in optical sections near the bottom. To allow quantification of cellular shape in three dimensions, mixtures of fluorescently labeled and unlabeled cells were co-developed. The proportion of TCs and spores increased by 36 to 48 h PS, especially near the radial center of nascent fruiting bodies, and near the bottom for spores at 48 h. Expression of a gene coding for a fluorescent protein was fused to two promoters with differential dependence on C-signaling. Transcription from the dev promoter appeared to be five-fold greater than from the fmgE promoter at 27 h PS, but similar at 48 h. In TCs and spores at 36 h, transcription from the fmgE promoter appeared to be 10-fold greater than in rods, while the difference was smaller from the dev promoter. The results suggest that transcription from the dev promoter is more sensitive to C-signaling than that from the fmgE promoter, and transcription from both promoters is greater in TCs and spores than in rods. Transcription from both promoters also appeared to be greater near the radial center of nascent fruiting bodies, consistent with the location of TCs and spores, and with a model that C- signaling triggers gene expression important for sporulation. ! 70! Introduction How cells build multicellular structures and adopt alternative fates are fundamental questions in developmental biology. Pattern formation is widely observed from prokaryotes to eukaryotes. Some examples are the one-dimensional pattern of heterocysts in filaments of the cyanobacterium Anabaena (1) and the pattern of cell fates along the anterior–posterior and dorsal–ventral axes of Drosophila (2) (3). Compared with other bacteria, Myxococcus xanthus forms a relatively complex, three-dimensional, multicellular structure called a “fruiting body.” The M. xanthus genome encodes a large number of signaling systems, which rival those of lower eukaryotes in complexity and include some eukaryotic-like features, such as the use of short-range signaling to coordinate emergent behaviors (4) (5). The starvation-induced development of M. xanthus offers a unique model system to study pattern formation during a process with eukaryotic-like features, using the power of bacterial genetics. Upon starvation, M. xanthus cells send signals to each other and coordinate their movements to construct uniform mounds (6). Mounds mature into fruiting bodies as rod-shaped cells differentiate into round spores. Other cells undergo lysis or remain outside of fruiting bodies as peripheral rods (7) (8). How cells adopt these alternative fates is unknown. A dome- shaped nascent fruiting body (NFB) consists of approximately 50,000 to 100,000 cells, which are organized into two domains with differential cell densities, arrangements, and movements (9). The outer domain contains densely packed rods that move in bidirectional streams, whereas cells in the inner domain are at lower density, less ordered, and less motile. It has been proposed that rods transitioning to spores accumulate in the inner domain (10). Microcinematography (time-lapse microscopy) of M. xanthus development has shown that mounds extend vertically in a series of tiers, which involves the addition of a cell monolayer on top of the uppermost layer (11). However, the spatial and temporal distribution of cellular shape change inside NFBs is unknown. Establishing the spatiotemporal distribution of cellular shape ! 71! change would be an important step toward a better understanding of the events that lead to sporulation during M. xanthus development. Spatiotemporal coordination of fruiting body formation is regulated by short-range C- signaling (12) (13) (14) (15). C-signaling is mediated by the product of the csgA gene (16) (17). There are two models for the identity of the C-signal, which are not mutually exclusive. The C- signal may be a proteolytic fragment of CgsA (18) (19) and/or lipids produced by CsgA phospholipase activity (20). Initially, C-signal was purified by detergent extraction and biochemical fractionation of starving cells, which identified a 17 kDa protein (p17) that could rescue development of a csgA mutant (21). It was subsequently shown that the full-length 25 kDa CsgA protein is cleaved to p17 by the subtilisin-like protease PopC (22). In the second model, CsgA is a phospholipase that oxidizes the 2’-OH glycerol moiety on cardiolipin to produce diacylglycerols (DAGs) (20). A lipid extract with a high concentration of DAGs from developing wild-type M. xanthus can rescue development of a csgA mutant. However, the modes of signal transmission in the two models remain to be elucidated. In the recipient cells, C-signaling is proposed to activate FruA by an unknown mechanism (23). Activated FruA (designated FruA*) and another transcription factor, MrpC, are proposed to bind cooperatively to activate transcription of developmentally-regulated genes, such as the dev operon (24) and the fmgE gene (25). The dev operon regulates the timing of sporulation (26) and disruption of the fmgE gene delays aggregation and reduces sporulation by about 8-fold (25). The architecture of the dev and fmgE promoter regions shares some similarity, but also has a notable difference. Both dev and fmgE employ a cooperative binding site located just upstream of the promoter, which activates transcription, and distal binding sites. The proximal upstream site in the dev promoter (Pdev) region exhibits a higher affinity for cooperative binding of FruA and MrpC than the distal upstream site (24) or several other sites (25) (27). On the other hand, the distal upstream site in the fmgE promoter (PfmgE) region has higher affinity for cooperative binding of MrpC and FruA, which acts negatively by competing for ! 72! binding with the lower affinity site just upstream of the promoter (25). Therefore, it has been hypothesized that Pdev has a relatively low threshold for FruA*, whereas PfmgE has a higher threshold (23) (24) (25). In other words, Pdev is predicted to be more sensitive than PfmgE to FruA* produced by C-signaling. Previous studies of M. xanthus development have involved sample collection and low- resolution imaging that hindered in situ examination of the shape and gene expression of individual cells (9) (10) (28). No study has shown the morphology of rods transitioning to spores and the spatial distribution of this process inside unperturbed NFBs. Here, we describe methods using confocal laser scanning microscopy and a custom image analysis pipeline to visualize and quantify the shape and gene expression of individual cells in situ. We found that TCs and spores are more abundant in the inner domain, providing evidence for a previous proposal (10). A new insight was that initially spores were more abundant farther from the bottom of NFBs, and later TCs and spores were more abundant closer to the bottom. For the first time, we were able to quantify gene expression of individual cells inside unperturbed NFBs. Pdev appeared to be more sensitive than PfmgE to C-signaling, as predicted, and PfmgE activity appeared to correlate closely with initiation of cellular shape change near the radial center of NFBs. Results Visualization of cellular shape during development with the membrane stain FM 4- 64. During the starvation-induced development of M. xanthus, rod-shaped cells transition into round spores. However, TCs have not been visualized inside NFBs in situ. To observe cellular shape during development, wild-type strain DK1622 was starved under submerged culture conditions. The lipophilic dye FM 4-64 was added at the beginning of starvation. Effects of the membrane stain on development are shown in Figure S3.1. Bright-field images of the same NFBs from 18 to 42 h PS showed that the timing of development was not affected by the ! 73! addition of FM 4-64 (Fig. S3.1A). Mounds formed by 18 h, and NFBs darkened by 42 h with or without FM 4-64 addition. The number of sonication-resistant spores at 42 h was greater with FM 4-64 addition than without (Fig. S3.1B) and the difference of ~2,500 spores/mL appeared to be credible (Fig. S3.1C), suggesting FM 4-64 somehow benefits sporulation. A 18 h 24 h B Hours PS 18 24 30 42 30 h 42 h C Hours PS 30 31 32 34 36 38 40 42 Figure 3.1 Cellular shape during development visualized by the membrane dye FM 4-64. M. xanthus wild-type strain DK1622 was starved under submerged culture conditions. FM 4-64 (2 µg/mL) was added at the start of starvation. Confocal images were acquired at the indicated times PS to show FM 4-64 staining of cell membranes (gray). Images are representative of three biological replicates. A. Confocal images of the same nascent fruiting body. Images show an optical section near the bottom of a NFB stained with FM 4-64. The white arrow indicates a rod-shaped cell, yellow arrows indicate TCs, and the red arrow indicates a spore. Bar, 20 µm. B. Close-up images of representative cells. Higher magnification images from panel A show cells with the predominant shape (rods at 18 h, TCs at 24 and 30 h, and spores at 42 h). Colored arrows point to the same cell types as in panel A. Bar, 2 µm. C. Representative images of a cell transitioning into a spore. Images were extracted from a time-lapse movie of three TCs indicated by magenta, blue, and orange arrows at 30 h. At later times, the arrows indicate the same three positions. Bar, 1 µm. At low magnification, optical sections near the bottom of the same NFB showed mostly rods aligned circularly in the outer domain at 18 h PS (Fig. 3.1A). Regions stained brightly with the FM 4-64 dye may be membranes from lysed cells. At 24 h, some TCs were observed, mostly in the inner domain, while rods were still aligned circularly in the outer domain. At 30 h, ! 74! there were fewer rods in the outer domain and more TCs in the inner domain compared to 24 h. By 42 h, many round spores had formed in both domains, especially in the inner domain. The close-up images of different cell shapes are shown in Figure 3.1B. Rod-shaped cells (white arrow) at 18 h were densely packed together in a mound. Some rods shorten and thicken by 24 h (yellow arrow). TCs at 24 h were lanceolate in shape. TCs underwent further morphological change by 30 h, becoming shorter and elliptical in shape. By 42 h, round spores (red arrow) had formed, which appeared to be smaller and more uniformly stained with FM 4-64 compared to TCs in the optical sections. To further investigate the TCs, we recorded a time-lapse movie of developing cells stained with FM4-64 from 30 to 42 h PS (images were taken every 15 minutes). Although TCs did not show motility, they could be moved by rods. Therefore, it is challenging to track individual TCs for 12 h since many are moved out of the focal plane rapidly. Figure 3.1C shows representative frames from the time-lapse movie. Three apparent TCs can be seen at 30 h. The cell indicated by the blue arrow could not be seen clearly at 34 h or thereafter. The cell indicated by the magenta arrow appeared to change shape by 31 h, but could not be seen clearly at 40 h or thereafter. These two cells may have been moved out of the focal plan or may have undergone lysis. The cell indicated by the orange arrow did not appear to be moved in the movie and eventually had the appearance of a spore (i.e., small, round, and more uniformly stained with FM 4-64). Visualization of cellular shape during development using a mixture of fluorescently labeled and unlabeled cells. The membrane stain FM 4-64 offered a method to visualize all the cells in optical sections in two dimensions. However, quantification of cellular shape was not possible because cells in NFBs were not sufficiently well-separated. To visualize and quantify cellular shape in three dimensions, we co-developed a strain producing tdTomato under control of a vanillate-inducible promoter (Pvan) with unlabeled cells at a ratio of 1 to 5. The mixture was delayed about 6 to 9 h compared with development of unlabeled wild-type cells ! 75! alone. Figure 3.2A shows optical sections near the bottom of the same nascent fruiting body at 27, 30, 36, and 48 h PS. At 27 h, rod-shaped cells were aligned circularly in the outer domain of the mound. TCs were observed at 30 h and became more abundant at 36 h. The fruiting body was filled with many spores at 48 h. Figure 3.2B shows close-up images of different cell shapes. Rod-shaped cells (white arrow) were ~5 µm long. TCs (yellow arrows) were shorter and thicker than rods. Spores (red arrows) were smaller and round, with a diameter of approximately 1 µm. FIG 2 A 27 h 30 h 36 h 48 h B Hours PS 27 30 36 48 Figure 3.2 Cellular shape during development visualized by labeling cells with tdTomato. Strain YH8 cells, which produce tdTomato under control of a vanillate-inducible promoter (Pvan), were mixed with unlabeled wild-type strain DK1622 cells at a ratio of 1 to 5 and starved under submerged culture conditions. Vanillate (0.5 mM) was added at the start of starvation. Confocal images of the same mound were acquired at the indicated times PS to show fluorescence from tdTomato (shown as gray for clarity). Images are representative of three biological replicates. A. Confocal images of the same nascent fruiting body. Images show an optical section near the bottom of a NFB. The white arrow indicates a rod-shaped cell, yellow arrows indicate TCs, and the red arrows indicates spores. Bar, 20 µm. B. Close-up images of representative cells. Higher magnification images from panel A show cells with the predominant shape (rods at 27 h, TCs at 30 and 36 h, and spores at 48 h). Colored arrows point to the same cell types as in panel A. Bar, 2 µm. ! 76! Cellular volume remains the same as rods transition into spores. To quantify cellular shape in three dimensions, we developed a computational pipeline involving segmentation and shape classification of individual cells (see Materials and Methods). In the experiment described in the preceding section, z-stacks of optical sections (~0.5 µm in thickness and a step size of 0.2 µm) were collected from near the bottom (Fig. 3.2) to 10 µm up into NFBs (one from each of three biological replicates at 27, 30, 36, and 48 h PS). Figure 3.3A shows cartoon depictions of representative cell shapes reconstructed from the segmentation. Rod-shaped cells appeared to be either straight or bent inside mounds. Some TCs were shortened, thickened rods, while others were more elliptical or nearly spherical but irregularly- shaped. Spores were round and very uniform in shape. The first column of graphs in Figure 3.3B shows the distributions of morphological parameters for cells classified as rods, TCs, and spores; the second column shows the pairwise differences between the posterior probability distributions of the three cell types for each parameter. The length of rods was 4 µm on average and cell length decreased as cells changed shape and became spores. TCs were on average 2 µm shorter than rods and 1.5 µm longer than spores. In contrast to cell length, mean width was the lowest for rods (0.3 µm) and the highest for spores (0.5 µm). TCs were on average 0.05 µm wider than rods and 0.1 µm narrower than spores. Sphericity, ranging from 0 to 1, is a measure of how closely the shape of an object resembles that of a perfect sphere. Spores had the highest sphericity (0.9). TCs had a sphericity of 0.8, which was higher than that of rods. The difference in sphericity between rods and spores was nearly 0.25. The measurements of cell length, mean width, and sphericity were consistent with our observations in two dimensions (Fig. 3.1 and 3.2) and with a priori expectations for rods, TCs, and spores, supporting our cellular shape classification method. ! 77! A Representative cell shapes B rods transitioning spores Figure 3.3 Shape classification of developing cells. In the experiment described in the legend of Figure 3.2 and in the text, z-stacks of optical sections were acquired at the indicated four times PS from the same nascent fruiting body for each of three biological replicates. ! 78! Figure 3.3 (cont’d) Segmented cells from the 12 z-stacks were classified as rods, TCs, or spores. The total number of cells classified from each biological replicate was similar (1246, 1175, 1555). A. Representative cell shapes. Three-dimensional reconstructions are shown for three cells in each classification. B. Morphological parameter distributions of cells classified as rods, transitioning cells, or spores. The first column of graphs shows violin plots of the distributions of cell length, mean width, sphericity, and equivalent diameter of cells classified as indicated. The line indicates the median. The second column shows the posterior probability distributions of the difference of each parameter of the indicated pair of cell types. The thick and thin vertical lines show the 50% and 95% highest density intervals, respectively. It was unknown whether the cellular volume changes as rods transition into spores. Therefore, we calculated the equivalent diameter of cells, which is the diameter of a sphere with the same volume as the object (Fig. 3.3B, bottom row of graphs). The equivalent diameter of rods, TCs, and spores ranged only from 1.6 to 1.7 µm, and the shapes of their distributions were similar. Since the differences ranged only 0.1 µm, which was only 6% of the measurements, these results suggest that the cellular volume changes very little as rods transition into spores. Temporal distribution of cellular shapes in nascent fruiting bodies. Having classified cellular shape in z-stacks of three NFBs as described above, we calculated the proportion of each cell type for each z-stack at each time PS. At 27 and 30 h, more than 80% of the cells were rods and small percentages of TCs and spores were observed (Fig. 3.4). By 36 h, the proportion of rods had decreased almost twofold on average, that of TCs had increased slightly, and that of spores had increased considerably to nearly 40%. At 48 h, the proportion of rods was only ~15%, that of TCs was ~30%, and that of spores was ~55%. The proportions of each cell type were consistent for the three NFBs examined, with the exception that one had a greater proportion of rods and a smaller proportion of spores at 36 h, suggesting it was less mature than the other two, but by 48 h it had similar proportions of cell types as the other two. The proportions agree qualitatively with visual inspection of images of optical sections in the z- stacks (Fig. 4.2 and data not shown), supporting our cellular shape classification method. ! 79! Figure 3.4 Proportion of rods, transitioning cells, and spores in nascent fruiting bodies. In the experiment described in the legend of Figure 3.2 and in the text, z-stacks of optical sections were acquired at the indicated times PS. Segmented cells from the z-stacks were classified as rods, TCs, or spores. Each dot represents one biological replicate (some dots overlap). Lines connect means of the three replicates. Spatiotemporal distribution of cellular shapes in nascent fruiting bodies. It has been proposed that cells transitioning to spores accumulate in the inner domain of NFBs (10). This appeared to be the case in optical sections near the bottom of NFBs (Fig. 4.1 and 4.2). The z-stacks described above provide an opportunity to quantify cellular shapes from near the bottom to near 10 µm up into NFBs at different times PS. We computed the proportions of rods, TCs, and spores as functions of radial distance from the center and of vertical distance from the bottom. Since the radius of all three NFBs was ~60 µm, we examined the radial distribution of cellular shapes over that distance (Fig. 3.5A). The z-stacks were 10 µm thick, but we examined the vertical distribution of cellular shapes from 1 to 9.5 µm up in order to eliminate edge effects (Fig. 3.5B). At 27 and 30 h PS, as noted above, most cells were rods (Fig. 3.4), and our analysis of their spatial distribution revealed them to be distributed uniformly both radially and vertically, judging from the means represented by the lines in Figure 3.5. Uncertainty in the proportions of rods farther from the radial center and farther from the bottom increased between 27 and 30 h, as indicated by broadening of the red shaded regions that show the 95% credible intervals. At ! 80! those locations, variations between the three NFBs increased between 27 and 30 h, accounting for the increased uncertainty (Fig. S3.2). In particular, the proportion of rods decreased and that of spores increased distal from the center and from the bottom in one biological replicate at 30 h (top row in both panels). The increased proportion of spores at those locations in just one replicate accounts for broadening of the blue shaded regions between 27 and 30 h (Fig. 3.5), even though in general for the three NFBs the proportions of TCs and spores were small at 27 and 30 h, as noted above (Fig. 3.4), and both cell types were distributed uniformly, judging from the lines (Fig. 3.5). By 36 h PS, not only had the proportions of all three cell types changed (Fig. 3.4), their spatial distributions were nonuniform. On average, rods accounted for only ~25% of the cells at the radial center, but their proportion was still ~65% at the edge (Fig. 3.5A). TCs had increased to ~25% at the center. Spores, interestingly, showed a maximum proportion of ~50% at ~20 µm from the center and decreased to ~25% at the edge. In comparison to 30 h, by 36 h the proportion of rods had declined dramatically near the center, where the proportions of TCs and spores had increased, providing evidence for their accumulation in the inner domain, as proposed (10). The 95% credible intervals indicated by the shaded regions overlapped extensively for the three cell types, due primarily to one nascent fruiting body with a greater proportion of rods and a smaller proportion of spores (Fig. 3.4) distal from the center (Fig. S3.2A, middle row). That replicate also had a greater proportion of rods and a smaller proportion of spores distal from the bottom at 36 h (Fig. S3.2B, middle row), which broadened the corresponding shaded regions in Figure 3.5B. On average, though, the proportion of rods declined and that of spores rose farther from the bottom, while the vertical distribution of TCs was uniform. In comparison to 30 h, by 36 h the most dramatic change in the vertical distribution was the increased proportion of spores higher up in NFBs, which has not been reported previously. ! 81! By 48 h PS the spatial distribution of cellular shapes had become less variable among the three NFBs than at 36 h (Fig. S3.2), resulting in narrower shaded regions showing the 95% credible intervals (Fig. 3.5). Consistent with spores being the majority cell type at 48 h (Fig. 3.4), spores were on average the greatest proportion of cells across most of both the radial and vertical distributions (Fig. 3.5). Radially, spores gradually decreased from ~75% at the center to <40% at the edge, TCs were uniformly ~25%, and rods increased sharply from very few in the inner domain (i.e., <30 µm from the center) to nearly 40% at the edge (Fig. 3.5A). Vertically, there were very few rods between 1 and 5 µm from the bottom, and their proportion increased sharply to nearly 40% at 9.5 µm up (Fig. 3.5B). TCs declined sharply from >50% near the bottom to ~25% from 4 to 9.5 µm up. Spores rose sharply between 1 and 4 µm from the bottom, reaching nearly 70%, then decreasing gradually to ~40% at 9.5 µm up. In comparison to 36 h, by 48 h the most dramatic changes in the radial distribution were the increased proportions of spores in the inner domain and of TCs in the outer domain (Fig. 3.5A), suggesting that spores may increase in the outer domain after 48 h. The vertical distributions of all three cell types changed most dramatically near the bottom, with the proportion of rods decreasing and the proportions of TCs and spores increasing (Fig. 3.5B). At 36 h two of the three NFBs had a greater proportion of spores higher up, and by 48 h all three had a maximum at ~4 µm up, but all three had a large proportion of TCs between 1 and 4 µm from the bottom (Fig. S3.2B). Taken together, these observations suggest that spores may increase near the bottom after 48 h, although their proportion initially increases farther up. ! 82! A B Figure 3.5 Spatial distribution of rods, transitioning cells, and spores in nascent fruiting bodies. In the experiment described in the legend of Figure 3.2 and in the text, z-stacks of optical sections were acquired at the indicated times PS. Segmented cells from the z-stacks were classified as rods, TCs, or spores. A. Radial distribution of cellular shapes. The proportion of each cell type radially from the center of NFBs to 60 µm away is shown. Lines represent the means and shaded regions represent 95% credible intervals. B. Vertical distribution of cellular shapes. The proportion of each cell type vertically from 1 to 9.5 µm up from the bottom of NFBs is shown. Lines represent the means and shaded regions represent 95% credible intervals. Visualization of gene expression in individual cells during fruiting body formation. Having investigated the temporal and spatial distribution of cellular shapes inside NFBs, we wanted to begin investigating molecular events that may govern cellular shape change during development. Therefore, we devised a method to measure gene expression in individual cells in situ. M. xanthus strains with transcriptional fusions to genes coding for two different fluorescent proteins were created. Developmentally-regulated promoters were fused to tdTomato, which codes for a red fluorescent protein. The constructs were then transformed into strain YH7, which has Pvan fused to mNeonGreen (encoding a green fluorescent protein) integrated in its genome. The activity of each developmental promoter was normalized by dividing the fluorescence intensity from tdTomato by that from mNeonGreen in the same cell. Fluorescence from mNeonGreen serves as an internal control. In this work, we investigated ! 83! activity of the dev and fmgE developmental promoters, which have been predicted to exhibit differential sensitivity to C-signaling. Figure 3.6 shows low magnification and higher magnification of optical sections near the bottom of one nascent fruiting body at different times PS for strains YH14 and YH15 bearing tdTomato fused to Pdev (panel A) and PfmgE (panel B), respectively, each co-developed with unlabeled wild-type cells at a ratio of 1 to 5. The mixtures were delayed about 6 to 9 h compared with development of unlabeled wild-type cells alone, as described above for mixtures (Fig. 3.2), and all the mixtures formed NFBs with rods (white arrows), TCs (yellow arrows), and spores (red arrows) in similar arrangements near the bottom at the same times PS (Fig. 3.2 and 3.6). As expected, all cell types of both strains YH14 and YH15 exhibited green fluorescence (Fig. 3.6), since both strains have Pvan-mNeonGreen and vanillate was added at the start of starvation. Strikingly though, rods of the strain bearing Pdev-tdTomato showed red fluorescence, but rods of the strain bearing PfmgE-tdTomato did not show red fluorescence. TCs and spores of both strains exhibited red fluorescence. The results show that Pdev is active at 27 h, earlier than PfmgE, and that PfmgE activity differs more than Pdev activity between rods and TCs or spores. These observations are consistent with the prediction that Pdev is more sensitive to C-signaling than PfmgE (23) (24) (25), since C-signal has been shown to increase between the beginning of mound formation and the beginning of sporulation (13) (15). PfmgE activity appeared to correlate closely with initiation of cellular shape change. ! 84! A 27 h 30 h 36 h 48 h B 27 h 30 h 36 h 48 h Hours PS 27 30 36 48 mNeonGreen tdTomato Merged Hours PS 27 30 36 48 mNeonGreen tdTomato Merged Figure 3.6 Visualization of gene expression in nascent fruiting bodies. Strain YH14 (bearing Pdev-tdTomato) or YH15 (bearing PfmgE-tdTomato) was mixed with unlabeled wild-type strain DK1622 at a ratio of 1 to 5 and co-developed under submerged culture conditions. Both YH14 and YH15 contain Pvan-mNeonGreen. Vanillate (0.5 mM) was added at the start of starvation. Confocal images near the bottom of the same nascent fruiting body were acquired at the indicated times PS. Images show tdTomato fluorescence (red) under control of the dev or fmgE promoter and mNeonGreen fluorescence (green) under control of the vanillate- inducible promoter. White arrows indicate rod-shaped cells, yellow arrows indicate TCs, and red arrows indicate spores. Images are representative of three biological replicates. A. Fluorescence from the dev promoter. The panels on the left show the mixture with strain YH14 at low magnification with the red and green channels merged (bar, 20 µm). The panels on the right show close-up images of representative cells in the green (top), red (middle), and merged (bottom) channels. Bar, 2 µm. B. Fluorescence from the fmgE promoter. The panels on the left show the mixture with strain YH15 at low magnification with the red and green channels merged (bar, 20 µm). The panels on the right show close-up images of representative cells in the green (top), red (middle), and merged (bottom) channels. Bar, 2 µm. ! 85! Quantification of gene expression in individual cells supports the prediction that the dev promoter is more sensitive to C-signaling than the fmgE promoter. In the experiment described in the preceding section, z-stacks of optical sections (~0.5 µm in thickness and a step size of 0.2 µm) were collected from near the bottom (Fig. 3.6) to 5 µm up into NFBs (one from each of three biological replicates at 27, 30, 36, and 48 h PS). The z- stacks collected were thinner than in the experiment described in the legend of Figure 3.2, because in this experiment we imaged in two channels (red and green), and too much laser exposure interferes with development and causes photobleaching. We used the same computational pipeline involving segmentation and shape classification of individual cells (see Materials and Methods). In addition, for each segmented cell, we determined the relative fluorescence intensity (i.e., red fluorescence from tdTomato produced under the control of Pdev or PfmgE was normalized to green fluorescence from mNeonGreen produced under the control of Pvan). At 27 h PS, the median relative fluorescence intensity (MRFI) from Pdev in rods, the majority cell type, was slightly less than 0 on a log10 scale, meaning that red fluorescence from tdTomato under Pdev control was slightly less than green fluorescence from mNeonGreen under Pvan control (Fig. 3.7). Far fewer TCs and spores were observed at 27 h, and had similar MRFI as rods. By 30 h, the number of rods had decreased and the numbers of TCs and spores had increased. The MRFI of spores had increased to ~three-fold greater than that of rods and TCs. By 36 h, spores were the majority cell type and their MRFI had increased to ~five-fold greater than that of rods, while that of TCs increased slightly. By 48 h, the MRFI of TCs was similar to that of spores. Hence, between 27 and 48 h, the median ratio of Pdev activity to Pvan activity changed very little in rods, but it increased ~fivefold in cells that changed shape. ! 86! Pdev-tdTomato! PfmgE-tdTomato! Figure 3.7 Quantification of gene expression in individual cells. In the experiment described in the legend of Figure 3.6 and in the text, z-stacks of optical sections were acquired at the indicated times PS, without changing the microsope settings. Segmented cells from the z-stacks were classified as rods, TCs, or spores. For each cell type, the distribution of log10 relative fluorescence intensity (i.e., red/green) at each time PS is plotted with strain YH14 bearing Pdev-tdTomato on the left and strain YH15 bearing PfmgE-tdTomato on the right. The results from three biological replicates are shown. Each dot represents one cell. Bars represent the median of the total population from the three biological replicates. The MRFI from PfmgE exhibited a similar pattern as that from Pdev, but there were some notable differences (Fig. 3.7). At 27 h PS, the MRFI from PfmgE was ~four-fold less than from Pdev, and both were similar in all cell types. By 30 h, the MRFI from PfmgE of spores was ~eight- fold greater than that of rods and TCs. By 36 h, the PfmgE MRFI of TCs had increased to nearly that of spores, whereas the Pdev MRFI of TCs had only increased slightly. By 48 h, the PfmgE MRFI of rods had increased considerably to nearly the Pdev MRFI of rods, and the PfmgE MRFI of TCs and spores was similar to that of Pdev. The increased PfmgE MRFI of rods was unexpected, based on observing very little red fluorescence from rods near the bottom of NFBs at 48 h (Fig. 3.6B), and we investigated this further as described below. For cells that changed shape, the median ratio of PfmgE activity to Pvan activity increased >10-fold for TCs at 36 h and for spores at ! 87! 30 and 36 h. Hence, PfmgE activity appeared to increase more than Pdev activity as cells underwent shape change. To investigate normalization of fluorescence from tdTomato to that of mNeonGreen as a possible source of the PfmgE MRFI increase of rods at 48 h PS (Fig. 3.7), the red fluorescence intensity from tdTomato was graphed without normalization. Importantly, the median red fluorescence intensity of rods bearing PfmgE-tdTomato did not increase at 48 h (Fig. S3.3). Aside from this difference, the pattern of intensity changes was similar in Figures 3.7 and S3.3, indicating that red fluorescence intensity correlates well with relative fluorescence intensity for both Pdev and PfmgE in all three cell types at most times. Since the median red fluorescence intensity of rods bearing PfmgE-tdTomato did not increase at 48 h PS (Fig. S3.3), we reasoned that their median green fluorescence intensity from Pvan-mNeonGreen must have decreased in order to account for their PfmgE MRFI increase (Fig. 3.7). Indeed, we found that the median green fluorescence intensity of all cell types decreased ~sixfold between 27 and 48 h PS (Fig. S3.4A). The decrease was gradual from 27 to 36 h for cells bearing Pdev-tdTomato and abrupt between 36 and 48 h for cells bearing PfmgE- tdTomato. These decreases elevated the Pdev MRFI slightly at 30 h and considerably at 36 h compared to the PfmgE MRFI, but equally for all cell types, so our observation that PfmgE activity appeared to increase more than Pdev activity as cells underwent shape change, based on MRFIs, is unaffected by the decreases. Although Pvan activity appeared to change temporally during development, it did so consistently for all three cells types (Fig. S3.4A), making it potentially valuable as an internal control for developmental promoter activity that is cell type-specific. Pvan activity could also be valuable as an internal control for developmental promoter activity that is spatially regulated, unless the spatial distribution of Pvan activity is nonrandom. To determine whether green fluorescence intensity from Pvan-mNeonGreen was random in the z-stacks, the mean intensity of each cell was determined and plotted as a blue shaded dot at its radial and vertical location, for ! 88! each NFB that had cells bearing Pdev-tdTomato (Fig. S3.4B) or PfmgE-tdTomato (Fig. S3.4C). The mean intensities indicated by the blue shading were distributed randomly in each z-stack (i.e., blue dots of each shade were randomly located), indicating that green fluorescence intensity from Pvan-mNeonGreen can serve as an internal control for developmental promoter activity that is spatially regulated. We note the greater heterogeneity between NFBs that had cells bearing PfmgE-tdTomato at 36 h (Fig. S3.4C). In particular, the NFB in the bottom row, and to a lesser extent the NFB in the top row, had fewer cells with low green fluorescence intensity (i.e., darker blue dots) than the NFB in the middle row at 36 h. Perhaps the NFB in the middle row is closer in maturity to the NFBs that had cells bearing Pdev-tdTomato at 36 h (Fig. S3.4B). In any case, by 48 h all the NFBs had many cells with low green fluorescence intensity (Fig. S3.4B and S3.4C) and the median green fluorescence intensity of both strains had decreased in all cell types (Fig. S3.4A), including rods, thus accounting for the PfmgE MRFI increase of rods at 48 h (Fig. 3.7). To determine whether fluorescence intensity from Pvan fused to a gene encoding a different fluorescent protein would also decrease at later times, we compared red fluorescence intensity from Pvan-tdTomato. We co-developed a strain bearing Pvan-tdTomato with unlabeled cells at a ratio of 1 to 5, just as we had in the experiment described in the Figure 3.2 legend. In parallel, we co-developed a strain bearing Pvan-mNeonGreen with unlabeled cells at the same ratio. The strain bearing Pvan-mNeonGreen in this experiment was the parent of the two strains used in the experiment described in the Figure 3.6 legend. For this strain, as expected, the median green fluorescence intensity decreased consistently for all three cell types. The decrease was abrupt between 30 and 36 h PS (Fig. S3.5), which was somewhat different from either derived strain (Fig. S3.4A), and the magnitude of the decrease was somewhat greater for the parent (Fig. S3.5) than for either derivative (Fig. S3.4A). These differences could be due to differential maturity of the NFBs examined, as mentioned above (Fig. S3.4C), and/or due to variation between the two experiments that we do not understand. Importantly, the median red ! 89! fluorescence intensity from Pvan-tdTomato decreased at later times PS, but not as much as the median green fluorescence intensity from Pvan-mNeonGreen (Fig. S3.5). The smaller decrease could make Pvan-tdTomato a better internal control, but the decrease in spores was greater than in rods and TCs at 30 h, which would be undesirable if it is reproducible. The smaller decrease may indicate that tdTomato is more stable than mNeonGreen in cells at later times PS. This would have elevated the relative fluorescence intensity at later times PS in Figure 3.7, since red fluorescence from Pdev-tdTomato and PfmgE-tdTomato was normalized to green fluorescence from Pvan-mNeonGreen, but the comparisons we made above between different cell types and between the two developmental promoters would not be affected. Since the median fluorescence intensity from both Pvan-mNeonGreen and Pvan-tdTomato decreased at later times PS, perhaps Pvan activity decreases. Presumably, the major sigma factor, σA, directs transcription from Pvan, and decrease in the level of σA during development has been reported (29). In summary, we conclude that normalization to green fluorescence from Pvan- mNeonGreen is a valuable internal control for developmental promoter activity that is cell type- specific and/or spatially regulated. The median Pdev/Pvan and PfmgE/Pvan activities increased ~fivefold and >10-fold, respectively, in cells that changed shape, and Pdev was active earlier than PfmgE. Pdev activity appeared to be more sensitive to C-signaling and PfmgE activity appeared to correlate better with initiation of cellular shape change. The spatiotemporal distribution of gene expression implies that C-signaling is greater near the center of nascent fruiting bodies, where transitioning cells and spores are most abundant. If short-range C-signaling coordinates mound formation with sporulation in NFBs as has been proposed (12) (13) (14) (15), the spatiotemporal distribution of relative fluorescence intensity from C-signal-dependent promoters such as Pdev and PfmgE may correlate with the spatiotemporal distribution of TCs and spores in NFBs (Fig. 3.5). In the preceding section, we reported temporal quantification of gene expression from Pdev and PfmgE fused to ! 90! tdTomato using Pvan-mNeonGreen as an internal control (Fig. 3.7), and we reported that the spatial distribution of Pvan activity was random (Fig. S3.4B and S3.4C), so we used the z-stacks collected in the experiment described in the legend of Figure 3.6 and in the text to investigate the spatiotemporal distribution of relative fluorescence intensity from Pdev and PfmgE. At 27 h PS, the mean Pdev intensity for all three cell types was distributed uniformly both radially and vertically (Fig. 3.8). As expected, the mean PfmgE intensity was less than that of Pdev, and the PfmgE radial distribution was uniform, but in the few spores present, it was ~10-fold greater 1 µm from the bottom than 4.5 µm up. By 30 h, both promoters exhibited ~10-fold greater mean intensity in spores near the radial center than near the edge, and in spores 1 µm from the bottom than 4.5 µm up. By 36 h, all three cell types showed similar patterns. Pdev mean intensity was greater near the center than near the edge, while that of PfmgE was greatest 20-30 µm from the center. Although Pdev mean intensity in TCs fluctuated vertically, the shaded regions indicating the 95% credible intervals overlapped with those of rods and spores. PfmgE mean intensity in TCs fluctuated vertically in a similar fashion at 48 h, but the 95% credible intervals likewise overlapped for all three cell types. Strikingly though, at 48 h, for all three cell types and both promoters, the mean intensity was greater near the center than near the edge by >10-fold. ! 91! A B d e v f m g E d e v f m g E Figure 3.8 Spatial distribution of gene expression in nascent fruiting bodies. In the experiment described in the legend of Figure 3.6 and in the text, z-stacks of optical sections were acquired at the indicated times PS, without changing the microsope settings. Segmented cells from the z-stacks were classified as rods, TCs, or spores. For each cell, the log10 relative fluorescence intensity (i.e., red/green) at each time PS was determined. The results from three biological replicates were combined. A. Radial distribution of relative fluorescence intensity. The intensity for each cell type radially from the center of NFBs to 60 µm away is shown. Lines represent the means and shaded regions represent 95% credible intervals for promoters of genes indicated on the right. B. Vertical distribution of relative fluorescence intensity. The intensity for each cell type vertically from 1 to 4.5 µm up from the bottom of NFBs is shown. Lines represent the means and shaded regions represent 95% credible intervals for promoters of genes indicated on the right. We conclude that C-signal-dependent promoter activity appeared to be greater near the radial center of NFBs, consistent with the location of TCs and spores (Fig. 3.5). Moreover, PfmgE mean intensity in all three cell types was greatest 20-30 µm from the center at 36 h PS (Fig. ! 92! 3.8A), and spores were most abundant there at 36 h (Fig. 3.5A). Although few spores were present at 27 and 30 h (Fig. 3.7), PfmgE mean intensity was greatest near the bottom of NFBs (Fig. 3.8B), where TCs were abundant at 27 h and both TCs and spores were abundant at 48 h (Fig. 3.5B). These spatiotemporal correlations between PfmgE mean intensity and cellular shape change, as well as the >10-fold increase in PfmgE MRFI of cells that changed shape (Fig. 3.7), indicate that PfmgE activity is a good marker for initiation of cellular shape change. Discussion In this work, we have devised a methods to visualize and quantify cellular shape and gene expression of individual cells in situ. We found that TCs and spores increased over time and were more abundant closer to the center near the bottom of NFBs. Normalized activity of the C-signal-dependent dev promoter was five-fold greater than that of PfmgE at 27 h PS, but at 48 h their activity was similar. At 36 h, PfmgE activity in cells that already changed shape was ~10-fold greater than in rods, while the difference in Pdev activity was ~fivefold. Inside NFBs at 36 and 48 h, both Pdev and PfmgE were more active in the inner domain, consistent with the location of TCs and spores. Morphology of transitioning cells. We have developed methods to observe the shape of TCs in nascent fruiting bodies in situ. Staining cells with the membrane dye FM 4-64 showed that the shape of TCs varied between 24 and 30 h PS. TCs were lanceolate-shaped at 24 h and then became more oval at 30 h (Fig. 3.1B). Three-dimensional images of TCs were reconstructed for the first time (Fig. 3.3A) using a computational pipeline involving cell segmentation and classification. Some TCs were shortened, thickened rods and others were irregularly-shaped and resembled spores more than rods. Cellular shape of bacteria is determined by the shape of the peptidoglycan sacculus, which in gram-negative bacteria like M. xanthus is an exoskeleton located in the periplasm that protects cells from internal osmotic pressure (30) (31). White et al. suggested that the cell ! 93! envelope of M. xanthus vegetative cells contains patches of discontinuous peptidoglycan (32). During glycerol-induced sporulation, there was a decrease in peptide cross-linkage of the peptidoglycan (32) and it was degraded (33). Although glycerol-induced and starvation-induced spores differ in some aspects (34), their morphology and resistance properties are similar. Therefore, we propose that when M. xanthus cells change shape during starvation, the process of peptidoglycan remodeling is non-uniform, resulting in TCs of various shapes. It is also possible that during shape change, the peptidoglycan becomes flexible so that cells adopt shapes that depend on their surroundings. Later during starvation-induced spore formation, upon completion of peptidoglycan remodeling, it may become more round and rigid, or it may be degraded, as during glycerol-induced sporulation (33), and rely on deposition of the polysaccharide spore coat for shape and rigidity (35). We also found that the volume of cells remained the same during shape change (Fig. 3.3B, bottom row of graphs). Previously, it was reported that preliminary measurements of thin sections of rods and glycerol-induced spores suggested that their surface areas are similar (32). Pattern of cellular shape change inside nascent fruiting bodies. Visualization near the bottom of the same NFBs during development with FM 4-64 (Fig. 3.1A) supported previous observations of two domains with different cellular orientation and density (9) (10). While the majority of rods aligned circularly in the outer domain, TCs and spores localized mostly in the inner domain of NFBs. In a second method, mixtures of fluorescently labeled and unlabeled cells were examined. Confocal microscopy and development of a computational pipeline to quantify the shape of individual cells in situ facilitated the investigation of the boundary between the two domains. Analysis of the collected z-stacks showed that the average radius of NFBs under our submerged culture conditions is ~60 µm. At 48 h PS, the proportion of rods was very low at the center and increased greatly from ~30 to 60 µm radially from the center (Fig. 3.5A). On the other hand, the proportions of both TCs and spores decreased considerably beyond 30 ! 94! to 40 µm from the center at 36 and 48 h. Therefore, the boundary between the two domains is ~30-40 µm from the center. Where does cellular shape change happen inside NFBs? Rods were the dominant cell type at 27 and 30 h. At 36 h, a considerable change in the proportions of TCs and spores occured. The maximum proportion of spores at 36 h was ~20 µm from the radial center (Fig. 3.5A and S3.2A). At 48 h, the peak of the spore distribution shifted to the center of NFBs. Since our time-lapse movies showed that TCs and spores could not move, it is possible that the circular movements of rods in the outer domain push spores toward the center as has been proposed (10). TCs have not investigated in situ before, so their spatial distribution may provide new insights into the location of cellular shape change. The maximum proportion of TCs was at the radial center of NFBs at 36 h and their radial distribution was uniform at 48 h (Fig. 3.5A and S3.2A), suggesting that cellular shape change occured first near the center and later expanded outward. This is not consistent with the inward pushing model (10), but the shapes of TCs and spores may affect their movement by rods, and a definitive answer will require time-lapse movies. The vertical distribution of cellular shapes inside NFBs was examined for the first time. At 36 h PS, the proportion of spores increased from 1 to 9.5 µm up (Fig. 3.5B). However, the peak of the spore distribution shifted to ~4 µm up at 48 h. TCs were distributed uniformly at 36 h and became very abundant near the bottom of NFBs at 48 h. Rods virtually disappeared near the bottom between 36 and 48 h, but whether they became TCs and spores, moved, or lysed is unknown. Again, a definitive answer will require time-lapse movies. Even so, the initial appearance of spores farther from the bottom at 36 h is a novel observation worthy of further investigation. In this work, we collected 10-µm z-stacks in the experiment described in the Figure 3.2 legend since resolution decreased farther up, and 5-µm z-stacks were collected in the experiment described in the Figure 3.6 legend since we imaged in two channels and too much ! 95! laser exposure interferes with development and causes photobleaching. Light-sheet microscopy is a method to obtain three-dimensional, live images of large volumes much faster than the point-by-point scanning of a typical confocal microscope (36). Light-sheet microscopy illuminates an entire plane of the sample, captures a wide-field image, then moves the plane (36). Therefore, individual cells are exposed to less light. In the future, this approach may enable us to collect thicker z-stacks to examine cell layers farther up from the bottom, and record time-lapse movies to track individual cells inside mounds. The spatial distribution of C-signal-dependent gene expression correlates with the pattern of cellular shape change. C-signal is a short-range morphogen that regulates M. xanthus cell movements to facilitate fruiting body formation (37). The level of C-signal during growth is very low, but it increases during development via two positive feedback loops. In the first one, C-signaling increases expression of the csgA gene early during development (13) (38) (39). The second positive feedback loop involves movements of rods. As cells move in streams, cell-to-cell contacts or proximity is increased (7) (10), resulting in a boost in C- signaling. Cells experience a gradual increase of C-signal during starvation and reach a threshold for expression of genes important for mound formation (13) (28). Inside the mounds, cells move in circular tracks with extensive end-to-end contacts (40), which may increase C- signaling further to meet a threshold for expression of genes important for sporulation (13). Interestingly, genes whose expression is limited to the fruiting body are dependent on C- signaling either directly or indirectly (28). In this work, we examined the spatial distribution of expression of C-signal-dependent genes, including the dev operon and the fmgE gene. We found that Pdev activity was always higher in the inner domain than in the outer domain of NFBs from 27 to 48 h PS (Fig. 3.8A). PfmgE, however, showed a peak of activity at ~20-30 µm from the radial center at 36 h. At 48 h, the distribution of PfmgE activity was similar to Pdev, highest at the center and decreasing radially. Interestingly, cells of all three types showed similar patterns at 36 at 48 h. Our quantitative ! 96! results are consistent with prior qualitative studies that showed some C-signal-dependent genes are expressed more highly in the inner domain of fruiting bodies (10) (28). In contrast, one C- signal-dependent gene showed initial expression in the outer domain and eventually was expressed in both domains (10). Perhaps the promoter of that gene is more sensitive than Pdev or PfmgE to a low level of C-signaling that begins in the outer domain. In any case, Pdev and PfmgE appeared to be greater near the radial center of NFBs (Fig. 3.8A), consistent with the location of TCs and spores (Fig. 3.5), suggesting that C-signaling becomes abundant in the inner domain and triggers gene expression important for sporulation. The vertical distribution of the activities of Pdev and PfmgE in spores was interesting at 27 and 30 h PS. At 30 h, Pdev activity in spores decreased ~10-fold from 1 to 4.5 µm up in NFBs. PfmgE activity in spores likewise decreased ~10-fold farther up at 27 and 30 h. These results suggest that at early times, when very few spores have formed, C-signaling is greater near the bottom of NFBs. At 36 and 48 h, we did not see differential activites of Pdev and PfmgE from 1 to 4.5 µm, suggesting there may no longer a vertical gradient of C-signaling. The ability to examine more cell layers with light-sheet microscopy promises to advance our knowledge about the distribution of C-signaling inside NFBs. C-signaling differentially regulates the temporal and cell type-specific expression of the dev operon and the fmgE gene. Although both Pdev and PfmgE are C-signal-dependent, their architectures are notably different. Both promoters employ a cooperative binding site for MrpC and FruA located just upstream of the promoter, which activates transcription (24) (25). However, the PfmgE region has a distal upstream site with higher affinity for cooperative binding of MrpC and FruA, which acts negatively by competing for binding with the lower-affinity site just upstream of the promoter (25). We found that Pdev activity was five-fold greater than PfmgE activity at 27 h PS (Fig. 3.7), consistent with the prediction based on promoter architecture (24) (25) and transcript levels (23) that Pdev would be more sensitive to C-signaling than PfmgE. Conversely, PfmgE exhibited stronger cell type-specific regulation than Pdev. PfmgE and Pdev ! 97! activities increased >10-fold and ~fivefold, respectively, in cells that changed shape. Pdev appears to be activated during streaming and mound formation, and the products of the dev operon delay initiation of cellular shape change (26), whereas PfmgE activity correlates closely with initiation of cellular shape change and the product of fmgE enhances sporulation. Materials and Methods Bacterial strains, plasmids, and primers. Table S3.1 lists the strains, plasmids, and primers used in this study. To construct pYH8, which was used to induce production of tdTomato by addition of vanillate, primers Van-tdTom F and Van-tdTom R were used to generate a PCR product of the open reading frame coding tdTomato sequence, using plasmid tdTomato-N1 (Addgene, item ID 54642). The product was digested with NdeI and EcoRI and mixed with pMR3691 (also digested with the same restriction enzymes) in a Gibson assembly reaction to enzymatically join the overlapping DNA fragments (41). The reaction mixture was transformed into Escherichia coli strain DH5α with outgrowth in Luria-Bertani (LB) liquid prior to plating on LB agar (1.5%) supplemented with 15 µg/mL tetracycline for selection at 37°C. The DNA sequence of the joined fragments was verified, and the plasmid was transformed into M. xanthus strain DK1622 using electroporation (42), with outgrowth in Casitone-Tris (CTT) (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH2PO4-K2HPO4, 8 mM MgSO4, [final pH 7.6]) liquid medium prior to plating on CTT agar (1.5%) supplemented with 15 µg/mL tetracycline to select transformants. Ectopic integration of pYH8 was verified by colony PCR using primers pMR3691 MCS G-F and pMR3691 MCS G-R. The resulting strain was named YH8. A similar method was used to generate pYH7, which bears Pvan-mNeonGreen (with primers Van-mNeonGreen F and Van-mNeonGreen R; the template is pNSI-ptre-mNeonGreen (Danny Ducat, personal communication). pYH7 was transformed into strain DK1622 to create strain YH7. To construct pYH14, a DNA fragment containing the dev promoter region was amplified from genomic DNA with Pdev F and Pdev R primers, which contain EcoRI and BamHI sites, ! 98! respectively. The tdTomato sequence was amplified from plasmid tdTomato-N1 (Addgene) using primers TomatoF and TomatoR, which contain BamHI and HindIII sites, respectively. Plasmid pSWU19 (43) was digested with BamHI and EcoRI. The three fragments were joined in a ligation reaction. The reaction mixture was transformed into E. coli strain DH5α with outgrowth in LB liquid prior to plating on LB agar supplemented with 5 µg/mL kanamycin for selection at 37°C. The DNA sequence of the joined fragments was verified, and the pYH14 was transformed into strain YH7 using electroporation (42), with outgrowth in CTT liquid prior to plating on CTT agar supplemented with 5 µg/mL kanamycin to select transformants. Integration of pYH14 was verified by colony PCR and the resulting strain was named YH14. The dev promoter region fragment of pYH14 was replaced by an fmgE promoter region fragment to construct pYH15. The fmgE promoter region was amplified from genomic DNA with primers PfmgE F and PfmgE R. pYH15 was transformed into YH7, resulting in strain YH15. Growth and development. M. xanthus was grown at 32°C on CTT agar or in CTTYE liquid medium (CTT with 0.2% yeast extract) shaking at 350 rpm. Development was performed under submerged culture conditions in 8-well µ-slides from Ibidi with starvation buffer MC7 (10 mM morpholinepropanesulfonic acid [MOPS, pH 7.0], 1 mM CaCl2). Cells from log-phase cultures in CTTYE liquid were resuspended in MC7 to 1,000 Klett units as described previously (44). The cell suspension (26 µL) was added to 174 µL of MC7 in each well. To collect samples for quantification of sonication-resistant spores, cells were scraped from the slide and the entire contents were aspirated into an 1.5 mL microcentrifuge tube. MC7 (300 µL) was added and samples were sonicated. Sonication-resistant spores were measured as described previously (45). Microscopy. Images of NFBs were acquired with a Nikon A1 Laser Scanning Confocal Microscope, which was configured on a Nikon Ti inverted platform with an XY automated stage and a 100X objective. Fluorescence from FM 4-64 and tdTomato were examined using a 560- nm laser for excitation and a 595/50 band pass emission filter. Fluorescence from ! 99! mNeonGreen was examined using a 488 nm laser for excitation and a 525/50 band pass emission filter. Time-lapse confocal images were taken every 15 min. Images “near the bottom of NFBs” were the first optical section above the bottom of the well in which cells could be clearly visualized. Cell segmentation. Segmentation was done using a custom MATLAB (Mathworks Inc.) script. The pixel intensity of each image in each image stack was normalized by subtracting the background intensity (mode of the intensity distribution) and dividing by the intensity at the 99.99th percentile. Then, the second derivative of the local voxel intensity variation (Hessian matrix) was calculated and decomposed into eigenvectors to estimate the local curvature of fluorescent objects. Voxels with large negative values for three or two eigenvectors are part of bright blob-like or tubular structures, respectively. The two most negative eigenvectors were summed and inverted. The resulting image was eroded using a 3x3x3 structure to separate close objects, binarized using a threshold at the 99th percentile, and opened with a 2x2x2 structure to filter stray voxels. Small objects of less than 100 voxels (0.308 µm3) were filtered. To separate merged cells, objects that were larger than the median volume of the segmented objects were processed further. First, the skeleton of the object was extracted to calculate the number of endpoints (more than two indicates merged objects). Then branches with the highest angle from the trunk were separated into independent objects until objects had only two endpoints. Finally, the morphological properties of each object were calculated (volume, centroid coordinates, equivalent diameter, surface area). The medial axis length and mean cell diameter along the cell axis were calculated from the cell skeleton transformation. The mean voxel intensities in the different fluorescent channels were calculated using the unprocessed images and subtracting the background intensity for each stack. To filter cells that were truncated at the image boundaries, cells with centroids outside the z-heights range of 1 to 9.5 µm (for Fig. 3.3, 3.4, 3.5, and S3.2) or 1 to 4.5 µm (for Fig. 3.7, 3.8, S3.3, S3.4, and S3.5) were discarded. ! 100! Cell shape classification. Cells were classified into different types based on their morphologies by fitting a Gaussian mixture model (GMM) with four components using the Mclust package (46). The mixture was fitted once using equivalent diameter, square root of cell surface area, length of the cell medial axis, and mean cell width on the ensemble of segmented objects across fruiting bodies and time points. All variables were log-transformed before fitting to produce variables distributions that were approximately Gaussian. The cells were assigned to one of four groups (rods, TCs, spores, and debris) based on maximum likelihood. Cell classification was assessed by visual inspection of three-dimensional images of cells randomly sampled from each cluster across fruiting bodies. Comparisons of morphological parameters. Equivalent diameter, skeleton size (referred as length in Fig.3.3B), and mean width were modeled using an intercept-only log- normal model and Bayesian sampling using brms (47), with a separate intercept for each cell type (Morph_param ~ 0 + cell_type). Sphericity was modeled using an intercept-only beta model with a separate intercept for each morphological classification. Statistical significance was calculated for the difference between the posterior probability distributions of each parameter. For a broad overview of this kind of technique, see (48). Bayesian sampling was done with 4 chains with 4,000 iterations each with default parameters. Plots were generated in the R statistical environment using the ggplot2 (49) and tidybayes packages (50). Estimating the proportions of cell types and fluorescence intensity in space. The proportion of each cell type as a function of distance within the fruiting body was fitted with a multinomial logistic regression with a generalized additive model (GAM) of tensor products to smooth the spatial dependency using the mcgv package (51). Variability between fruiting bodies was treated as a random effect and each time point was fitted independently. The model was fitted using Bayesian sampling using the brms package (47) to determine the mean and 95% credible intervals of P_celltype, the probability of a given cell type at a given location (cell_type|trials(1) ~ 1, P_celltype ~ t2(distance) + t2(distance| fruiting_body_id)). To ! 101! model the mean relative fluorescence intensity of each cell as a function of space, we used the same approach but replaced the multinomial model with a log-normal model. Bayesian sampling was done with 8 chains with 5,000 iterations each with default parameters except for ‘adapt_delta’=0.99 and ‘inits’=0. Plots were generated in the R statistical environment using the ggplot2 (49) and tidybayes (50) packages. Acknowledgments We thank Montse Elias-Arnanz for providing pMR3691. We are grateful to Melinda Frame and the Center for Advanced Microscopy at Michigan State University for assistance with microscopy. This research was supported by National Science Foundation grants MCB-1411272 and IOS-1951025, and by salary support for L.K. from Michigan State University AgBioResearch. ! 102! APPENDIX 103! ! Table S3.1 Bacterial strains, plasmids, and primers used in this study Strain, plasmid, or primer Strain E. coli DH5α M. xanthus DK1622 YH7 YH8 YH14 YH15 Plasmid pMR3691 pSWU19 pYH7 pYH8 pYH14 pYH15 Primers Van-tdTom F Van-tdTom R Description λ - φ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17(rK -mK - ) supE44 thi-1 gyrA relA1 Laboratory wild-type strain DK1622 with MXAN_0018-MXAN_0019::pYH7 DK1622 with MXAN_0018-MXAN_0019::pYH8 YH7 with attB::pYH14 YH7 with attB::pYH15 MXAN_0018-MXAN_0019-PR3-4::vanR-Pvan::lacZ, Tetr Cloning vector with Mx8 attP locus for integration at attB in the M. xanthus genome, Kmr MXAN_0018-MXAN_0019-PR3-4::vanR- Pvan::mNeonGreen, Tetr MXAN_0018-MXAN_0019-PR3-4::vanR-Pvan::tdTomato, Tetr pSWU19 with Pdev-tdTomato, Kmr pSWU19 with PfmgE-tdTomato, Kmr GATGCGAGGAAACGCAT ATG GTG AGC AAG GGC GAG GTACGCGTAACGTTCGAATTCTTACTTGTACAGCTCG TCCATG Van-mNeonGreen F GATGCGAGGAAACGCATATGGTCAGCAAAGGTGAA GAAGAC Van-mNeonGreen R GTACGCGTAACGTTCGAATTCCTTGTACAGTTCGTCC Pdev F Pdev R PfmgE F PfmgE R TomatoF ATACCCATC GCTAGAATTC CGCCGTTGGCTCGGATGC GCTAGGATCCGACCTGAGCTGATACCAAC GCTAGAATTCCATTAACGGGCGATCTTCCTC GCTAGGATCCCTAAAGGCGATTGCGTTCCTGC GCTAGGATCCGAAAGGAGGCGCATATGGTGAGCAA GGGCGAG GCTA AAGCTTTTACTTGTACAGCTCGTCCATG TomatoR pMR3691 MCS G-F CACGATGCGAGGAAACGCA pMR3691 MCS G-R CACCGGTACGCGTAACGTTC ! 104! Source or reference (52) (53) This study This study This study This study (54) (55) This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study A FM 4-64 (µg/mL) 0 2 B B Hours PS 18 24 30 42 C FM 4-64 concentration (µg/mL) Figure S3.1 Effects of the membrane dye FM 4-64 on M. xanthus development. Wild-type strain DK1622 was subjected to starvation under submerged culture conditions. FM 4-64 (2 µg/mL) was added at the start of starvation. A. Bright-field images of the same fruiting body from 18 to 42 h PS. Darkening inside mounds typically reflects the formation of spores. Images are representative of three biological replicates. B. Quantification of sonication-resistant spores. Each dot represents one biological replicate. C. Effect of the membrane dye FM 4-64 on sporulation. Posterior probability distributions of the differences in the number of sonication-resistant spores in response to 0 and 2 µg/mL of FM 4-64 evaluated from a linear mixed-effect model (3 biological replicates). The filled circle indicates the mean, the thick line indicates the 50% highest density interval, and the thin line indicates 95% highest density interval. ! 105! A B Figure S3.2 Spatial distribution of rods, TCs, and spores in individual nascent fruiting bodies. In the experiment described in the legend of Figure 3.2 and in the text, z-stacks of optical sections were acquired at the indicated times PS. Segmented cells from the z-stacks were classified as rods, TCs, or spores. ! 106! Figure S3.2 (cont’d) A. Radial distribution of cellular shapes of individual mounds. Each row shows the proportion of each cell type radially from the center of nascent fruiting bodies to 60 µm of the same mound overtime. Lines represent the means and shaded regions represent 95% credible intervals. B. Vertical distribution of cellular shapes of individual mounds. Each row shows the proportion of each cell type vertically from 1 to 9.5 µm up from the bottom of nascent fruiting bodies of the same mound overtime. Lines represent the means and shaded regions represent 95% credible intervals. ! 107! YH14! YH15! Figure S3.3 Quantification of gene expression in individual cells without normalization. In the experiment described in the legend of Figure 3.6 and in the text, z-stacks of optical sections were acquired at the indicated times PS, without changing the microsope settings. Segmented cells from the z-stacks were classified as rods, TCs, or spores. For each cell type, the distribution of log10 red fluorescence intensity at each time PS is plotted with strain YH14 bearing Pdev-tdTomato on the left and strain YH15 bearing PfmgE-tdTomato on the right. The results from three biological replicates are shown. Each dot represents one cell. Bars represent the median of the total population from the three biological replicates. ! 108! YH14! YH15! A B ! 109! Figure S3.4 (cont’d) Figure S3.4 Quantification and spatial distribution of mNeonGreen expression from a vanillate-inducible promoter in strains YH14 and YH15. In the experiment described in the legend of Figure 3.6 and in the text, z-stacks of optical sections were acquired at the indicated times PS, without changing the microsope settings. Segmented cells from the z-stacks were classified as rods, TCs, or spores. A. Quantification of gene expression in individual cells. For each cell type, the distribution of log10 green fluorescence intensity at each time PS is plotted with strain YH14 on the left and strain YH15 on the right. The results from three biological replicates are shown. Each dot represents one cell. Bars represent the median of the total population from the three biological replicates. B. Spatial distribution of gene expression in nascent fruiting bodies of strain YH14. Each row shows the spatial distribution of mNeonGreen expression inside the same mound over time. Each dot represents one cell. C. Spatial distribution of gene expression in nascent fruiting bodies of strain YH15. Each row shows the spatial distribution of mNeonGreen expression inside the same mound over time. Each dot represents one cell. ! 110! Pvan-mNeonGreen! Pvan-tdTomato! Figure S3.5 Quantification of gene expression from a vanillate-inducible promoter in strains YH8 and YH7. Strain YH7 (bearing Pvan-mNeonGreen) or YH8 (bearing Pvan-tdTomato) was co-developed with unlabeled wild-type strain DK1622 cells (1 to 5 ratio) under submerged culture conditions. Vanillate (0.5 mM) was added at the start of starvation. Z-stacks of optical sections were acquired at the indicated times PS, without changing the microsope settings. Segmented cells from the z-stacks were classified as rods, TCs, or spores. For each cell type, the distribution of log10 fluorescence intensity (i.e., green or red) at each time PS is plotted with strain YH7 bearing Pvan-mNeonGreen on the left and strain YH8 bearing Pvan-tdTomato on the right. The results from three biological replicates are shown. Each dot represents one cell. Bars represent the median of the total population from the three biological replicates. ! 111! REFERENCES 112! ! REFERENCES Kumar K, Mella-Herrera RA, Golden JW. 2010. Cyanobacterial heterocysts. Cold Spring Harb Perspect Biol 2(4): a000315. 10.1101/cshperspect.a000315 Driever W, Nüsslein-Volhard C. 1988. The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner. Cell 54:95-104. Roth S, Stein D, Nüsslein-Volhard C. 1989. A gradient of nuclear localization of the dorsal protein determines dorsoventral pattern in the Drosophila embryo. Cell 59(6):1189-202. 10.1016/0092-8674(89)90774-5 Bretl DJ, Kirby JR. 2016. Molecular Mechanisms of Signaling in Myxococcus xanthus Development. J Mol Biol 428(19):3805-30. 10.1016/j.jmb.2016.07.008 Kroos L. 2017. Highly Signal-Responsive Gene Regulatory Network Governing Myxococcus Development. Trends Genet 33(1):3-15. 10.1016/j.tig.2016.10.006 Kaiser D, Robinson M, Kroos L. 2010. Myxobacteria, polarity, and multicellular morphogenesis. Cold Spring Harb Perspect Biol 2(8):a000380. 10.1101/cshperspect.a000380 7. O’Connor KA, Zusman DR. 1991. Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. J Bacteriol 173:3318–33. 10.1128/jb.173.11.3318-3333.1991 8. Wireman JW, Dworkin M. 1977. Developmentally induced autolysis during fruiting body formation by Myxococcus xanthus. J Bacteriol 129(2):798-802. Sager B, Kaiser D. 1993. Two cell-density domains within the Myxococcus xanthus fruiting body. Proc Natl Acad Sci U S A 90(8): 3690–4. 10.1073/pnas.90.8.3690 10. Sager B, Kaiser D. 1993. Spatial restriction of cellular differentiation. Genes Dev 7:1645- 53. 10.1101/gad.7.9.1645 1. 2. 3. 4. 5. 6. 9. 11. Curtis PD, Taylor RG, Welch RD, Shimkets LJ. 2007. Spatial organization of Myxococcus xanthus during fruiting body formation. J Bacteriol 189(24):9126-30. 10.1128/JB.01008- 07 12. Kim SK, Kaiser D. 1990. Cell aligmnent required in differentiation of Myxococcus xanthus. Science 249(4971):926-8. 10.1126/science.2118274 13. Kim SK, Kaiser D. 1991. C-factor has distinct aggregation and sporulation thresholds during Myxococcus development. J Bacteriol 173:1722–8. 10.1128/jb.173.5.1722- 1728.1991 14. Li S, Lee BU, Shimkets LJ. 1992. csgA expression entrains Myxococcus xanthus development. Genes Dev 6:401–10. 10.1101/gad.6.3.401 15. Kruse T, Lobedanz S, Berthelsen NMS, Søgaard-Andersen L. 2001. C-signal: A cell ! 113! surface-associated morphogen that induces and co-ordinates multicellular fruiting body morphogenesis and sporulation in Myxococcus xanthus. Mol Microbiol 40:156–68. 10.1046/j.1365-2958.2001.02365.x 16. Hagen TJ, Shimkets LJ. 1990. Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. J Bacteriol 172(1):15-23. 10.1128/jb.172.1.15-23.1990 17. Shimkets LJ, Rafiee H. 1990. CsgA, an extracellular protein essential for Myxococcus xanthus development. J Bacteriol 172(9):5299-306. 10.1128/jb.172.9.5299-5306.1990 18. Kim SK, Kaiser D. 1990. C-factor: A cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 61(1):19-26. 10.1016/0092-8674(90)90211-V 19. Lobedanz S, Søgaard-Andersen L. 2003. Identification of the C-signal, a contact- dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes Dev 17:2151–61. 10.1101/gad.274203 20. Boynton TO, Shimkets LJ. 2015. Myxococcus CsgA, Drosophila sniffer, and human HSD10 are cardiolipin phospholipases. Genes Dev 29(18):1903-14. 10.1101/gad.268482 21. Kim SK, Kaiser D. 1990. Purification and properties of Myxococcus xanthus C-factor, an intercellular signaling protein. Proc Natl Acad Sci U S A 87(10):3635-9. 10.1073/pnas.87.10.3635 22. Rolbetzki A, Ammon M, Jakovljevic V, Konovalova A, Søgaard-Andersen L. 2008. Regulated Secretion of a Protease Activates Intercellular Signaling during Fruiting Body Formation in M. xanthus. Dev Cell 15(4):627-34. 10.1016/j.devcel.2008.08.002 23. Saha S, Patra P, Igoshin O, Kroos L. 2019. Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C-signaling. Mol Microbiol 111: 1732-52. 10.1111/mmi.14249 24. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L. 2015. Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 197:240–251. 10.1128/JB.02310-14 25. Son B, Liu Y, Kroos L. 2011. Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 193:2756–66. 10.1128/JB.00205-11 26. Rajagopalan R, Kroos L. 2017. The dev operon regulates the timing of sporulation during Myxococcus xanthus development. J Bacteriol 199 (10) e00788-16. 10.1128/JB.00788- 16 27. Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of hundreds of developmentally-regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. 10.1186/1471-2164-15-1123 28. Julien B, Kaiser a D, Garza a. 2000. Spatial control of cell differentiation in Myxococcus xanthus. Proc Natl Acad Sci U S A 97:9098–103. 10.1073/pnas.97.16.9098 ! 114! 29. Hao T, Biran D, Velicer GJ, Kroos L. 2002. Identification of the Ω4514 regulatory region, a developmental promoter of Myxococcus xanthus that is transcribed in vitro by the major vegetative RNA polymerase. J Bacteriol 184(12): 3348–59. 10.1128/JB.184.12.3348- 3359.2002 30. Vollmer W, Blanot D, De Pedro MA. 2008. Peptidoglycan structure and architecture. FEMS Microbiol Rev 32(2):149-67. 10.1111/j.1574-6976.2007.00094.x 31. Weidel W, Pelzer H. 1964. Bagshaped macromolecules - a new outlook on bacterial cell walls. Adv Enzymol Relat Areas Mol Biol 26:193-232. 10.1002/9780470122716.ch5 32. White D, Dworkin M, Tipper DJ. 1968. Peptidoglycan of Myxococcus xanthus: structure and relation to morphogenesis. J Bacteriol 95(6): 2186–97. 33. Bui NK, Gray J, Schwarz H, Schumann P, Blanot D, Vollmer W. 2009. The peptidoglycan sacculus of Myxococcus xanthus has unusual structural features and is degraded during glycerol-induced myxospore development. J Bacteriol 191(2):494-505. 10.1128/JB.00608-08 Inouye M, Inouye S, Zusman DR. 1979. Biosynthesis and self-assembly of protein S, a development-specific protein of Myxococcus xanthus. Proc Natl Acad Sci U S A 76(1):209-13. 10.1073/pnas.76.1.209 35. M??ller FD, Schink CW, Hoiczyk E, Cserti E, Higgs PI. 2012. Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Mol Microbiol 83:486–505. 10.1111/j.1365-2958.2011.07944.x 36. Santi PA. 2011. Light sheet fluorescence microscopy: A review. J Histochem Cytochem 59(2):129-38. 10.1369/0022155410394857 37. Jelsbak L, Søgaard-Andersen L. 2000. Pattern formation: Fruiting body morphogenesis in Myxococcus xanthus. Curr Opin Microbiol 3(6):637-42. 10.1016/S1369-5274(00)00153-3 38. Gronewold TMA, Kaiser D. 2001. The act operon controls the level and time of C-signal production for Myxococcus xanthus development. Mol Microbiol 40(3):744-56. 10.1046/j.1365-2958.2001.02428.x 39. Kaiser D. 2015. Signaling in Swarming and Aggregating Myxobacteria. Springer. 10.1007/978-94-017-9642-2_22 40. Sager B, Kaiser D. 1994. Intercellular C-signaling and the traveling waves of Myxococcus. Genes Dev 8: 2793-804. 10.1101/gad.8.23.2793 41. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, Smith HO. 2009. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 343–345. 10.1038/nmeth.1318 42. Kashefi K, Hartzell PL. 1995. Genetic suppression and phenotypic masking of a Myxococcus xanthus frzF- defect. Mol Microbiol 15:483–494. 10.1111/j.1365- 2958.1995.tb02262.x 34. ! 115! 43. Samuel SWU, Kaiser D. 1997. Regulation of expression of the pilA gene in Myxococcus xanthus. J Bacteriol 179(24):7748-58. 10.1128/jb.179.24.7748-7758.1997 44. Hoang Y, Kroos L. 2018. Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC. J Bacteriol 200(22):e00456-18. 10.1128/JB.00456-18 45. Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736–47. 10.1128/JB.01692-14 46. Scrucca L, Fop M, Murphy TB, Raftery AE. 2016. Mclust 5: Clustering, classification and density estimation using Gaussian finite mixture models. R J 8(1):289-317. https://cran.r- project.org/web/packages/mclust/citation.html 47. Bürkner PC. 2018. Advanced Bayesian multilevel modeling with the R package brms. R J. 10(1):395–411. 10.32614/RJ-2018-017. 48. Kruschke JK. 2013. Bayesian estimation supersedes the t test. J Exp Psychol Gen! 142(2):573–603. 10.1037/a0029146. 49. Wickham H. 2016. ggplot2: elegant graphics for data analysisJournal of the Royal Statistical Society: Series A (Statistics in Society) 978-3-319-24277-4. https://ggplot2.tidyverse.org 50. M K. 2020. tidybayes: Tidy Data and Geoms for Bayesian Models. 10.5281/zenodo.1308151 51. Wood SN. 2011. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J R Stat Soc Ser B Stat Methodol 73(1):3-36. https://cran.r-project.org/web/packages/mgcv/citation.html 52. Hanahan D. 1983. Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166(4):557-80. 10.1016/s0022-2836(83)80284-8 53. Kaiser D. 1979. Social gliding is correlated with the presence of pili in Myxococcus xanthus. Proc Natl Acad Sci U S A 76(11): 5952–6. 10.1073/pnas.76.11.5952 54. ! Iniesta AA, Garcia-Heras F, Abellon-Ruiz J, Gallego-Garcia A, Ellias-Arnanz M. 2012. Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl- β -d-thiogalactopyranoside or vanillate. J Bacteriol 194:5875–5885. 10.1128/JB.01110-12 55. Wu SS, Kaiser D. 1995. Genetic and functional evidence that Type IV pili are required for social gliding motility in Myxococcus xanthus. Mol Microbiol 18:547-58. 10.1111/j.1365- 2958.1995.mmi_18030547.x 116! CHAPTER 4: Threshold effects of short-range C-signaling on emergent behaviors during Myxococcus xanthus development Abstract Bacterial interactions involving diffusible signals are relatively well-understood in comparison with interactions involving short-range signals. Short-range signaling is an important mechanism for development of multicellular structures and cell fate determination. In this study, the M. xanthus developmental process was used as a model to investigate short- range C-signaling, which coordinates cell movements to build uniform mounds upon starvation. Mounds mature into fruiting bodies as some of the rod-shaped cells differentiate into round spores. Other cells lyse or remain outside of fruiting bodies as peripheral rods. Genes important for spore formation and maturation are regulated by C-signaling. Here, wild-type (WT) and csgA mutant (defective in C-signaling) cells were co-developed at slightly different ratios and examined using methods described in Chapter 3 to visualize and quantify cellular shape. At a ratio of 1 WT to 4 csgA mutant cells (1:4), neither mounds nor spores formed. At a 1:2 ratio, broad mounds formed with twofold less cell density than normal, and very few spores formed. At a 1:1 ratio, normal mounds formed and csgA spores were slightly more abundant than WT spores. Apparently, a threshold for C-signaling was reached in the 1:1 mixture. At ratios of 2:1 and 4:1, csgA mutant cells were disproportionately overrepresented in mounds and sporulated with similar efficiency as WT cells. Hence, cheating by the csgA mutant occurred during mound formation. Cells in transition from rods to spores were observed near the radial center of nascent fruiting bodies, and csgA mutant spores were typically observed closer to the center than WT spores. We conclude that short-ranging C-signaling between cells is threshold- dependent and susceptible to cheating during mound formation, and we propose it is most abundant near the radial center of nascent fruiting bodies. ! 117! Introduction Bacterial interactions affect ecosystems and all organisms inhabiting them. Bacterial interactions involving diffusible signals are relatively well-understood (1) in comparison with interactions involving short-range signals (2). Examples of bacteria utilizing short-range signaling are fruiting body formation in Myxococcus xanthus, where a signal protein is displayed on the surface and is proposed to interact with a receptor on an adjacent cell (3), and the criss- cross signaling between the mother cell and the forespore during sporulation in Bacillus subtilis, where communication occurs between cells that are in intimate contact (4). In eukaryotes, short-range (paracrine) signaling controls a number of processes such as neurotransmission, blood clotting, angiogenesis, and embryonic differentiation. Short-range signaling is an important mechanism for development of multicellular structures and cell fate determination. General principles learned from studying bacterial short-range signaling will be broadly applicable and provide insight into the evolution of multicellularity (5). In this study, the M. xanthus developmental process was used as a model to investigate short-range signaling. The M. xanthus genome sequence reveals prodigious capacity for signaling and sensory capability, which rivals those of lower eukaryotes in complexity (6). Its multicellular development occurs rapidly (2 days) and synchronously under laboratory conditions. Growing cells adhere to the bottom of a plastic surface to form a biofilm. Upon starvation, cells coordinate their movements to align in streams and build mounds. The mounds mature into fruiting bodies as some of the rod-shaped cells differentiate into round spores. Other cells lyse or remain outside fruiting bodies as peripheral rods (7) (8). During the developmental process, short-range C-signaling coordinates streaming and mound formation with spore differentiation (9) (10) (11) (12), but the mechanisms driving these emergent behaviors are not fully known. C-signaling is mediated by the product of the csgA gene (13) (14). There are two models for the identity of the C-signal, which are not mutually exclusive. The C-signal may be a ! 118! proteolytic fragment of CgsA and/or lipids produced by CsgA phospholipase activity. Initially, C- signal was purified by detergent extraction and biochemical fractionation of starving cells, which identified a 17 kDa protein (p17) fragment of CsgA that could rescue development of a csgA mutant (15) (16). It was subsequently shown that the full-length 25 kDa protein encoded by the csgA gene is cleaved by a secreted subtilisin-like protease, PopC, at the cell surface (17). In the second model, CsgA is a phospholipase that oxidizes the 2’-OH glycerol moiety on cardiolipin to produce diacylglycerols (DAGs) (18). A lipid extract with a high concentration of DAGs from developing wild-type cells rescued development of a csgA mutant. The receptor of the C‐signal has yet to be identified. The DNA‐binding response regulator protein FruA is the first recognized component in the pathway (19). However, the nature of the effect of C-signaling on FruA is still unknown (20). Short-range C-signaling temporally and spatially regulates mound formation and sporulation (21). A csgA mutant is unable to form mounds or spores (14) (22). Overexpression of the csgA gene early during development results in premature mound formation and sporulation (12). The developmental defects of a csgA mutant can be rescued by co- development with wild-type (WT) cells (23) or addition of purified C-signal as mentioned above. A low level of C-signal (0.8 unit) rescues mound formation and expression of an early C-signal- dependent gene, whereas a slightly higher level (1.0 unit) can additionally trigger late C-signal- dependent gene expression and spore formation (10). The isolation of C-signal from wild-type cells at different developmental stages showed that the C-signal concentration increased only about fourfold between the beginning of mound formation and the beginning of sporulation. During this period, β-galactosidase production from a csgA-lacZ transcriptional fusion also increased about fourfold. These results suggested that slightly different threshold levels of C- signaling control mound formation and sporulation. Therefore, we hypothesized that WT cells would rescue different emergent behaviors of a csgA mutant over a narrow range of ratios in cell mixtures. ! 119! A prior study has shown that a csgA mutant exhibits cheating behavior when mixed with WT cells (24). The csgA mutant was overrepresented among resulting spores relative to their initial frequency in the mixture. When csgA mutant and WT cells were mixed with a ratio of 1:99, sporulation of the csgA mutant was rescued and its efficiency was even slightly better than WT, relatively to their initial proportions in the mixture. In a 1:1 mixture, csgA mutant spores outnumbered WT spores and the total spore count fell below the WT level (i.e., when 100% of the developing cells were WT). Although a csgA mutant fails to contribute C-signaling to the mixture, it performs better than WT cells in terms of the spore numbers, but how the csgA mutant cheats on the community is unknown. Neither is it known when cheating occurs during development. For example, cheating behavior could occur during mound formation and/or during sporulation. In this work, we used methods described in Chapter 3 to visualize and quantify cellular shape change during development of WT and csgA mutant cells in mixtures over a narrow range of ratios. Our results demonstrate threshold dependence of emergent behaviors on short-range C-signaling between cells and reveal that cheating by a csgA mutant occurs during mound formation. Our results also provide evidence that C-signaling is most abundant near the radial center of nascent fruiting bodies. Results Wild-type cells rescue mound formation and sporulation of a csgA mutant over a narrow range of cell ratios upon co-development. In order to visualize and quantify both WT and csgA mutant cells during co-development, we labeled the cells by engineering them to produce different fluorescent proteins. We transformed a construct bearing Pvan-mNeonGreen into WT strain DK1622 and a construct bearing Pvan-tdTomato into csgA mutant strain MRR33, resulting in strains YH7 and YH11, respectively. Pvan is a vanillate-inducible promoter and the plasmids integrate into the M. xanthus chromosome at a site that does not affect development ! 120! (25). Under submerged culture conditions with vanillate added at the start of starvation, labeled YH7 mixed with its unlabeled WT parent at a ratio of 1:5 in order to allow visualization of individual YH7 cells, formed mounds at 24 h PS and many spores were observed at 42 h (Fig. 4.1A), similar to both unmixed strains (data not shown). On the other hand, labeled YH11 mixed with its unlabeled cgsA mutant parent at 1:5 failed to form mounds and spores (Fig. 4.1A), similar to both unmixed strains (data not shown). For YH11, many small dots of red fluorescence were observed (Fig. 4.1A), which based on their size, we interpret to be cross- sectional views of rods oriented vertically in the biofilm. In this work, half of each cell type was labeled and mixtures at ratios close to 1:1 were co-developed and examined by confocal laser scanning microscopy. Only half of the cells were labeled in order to allow visualization of individual cells and computational analysis of their shape. In the images, the ratio of labeled WT cells exhibiting green fluorescence to labeled csgA mutant cells exhibiting red fluorescence is expected to reflect the overall ratio of WT to csgA mutant cells in the mixture. Figure 4.1B shows confocal images near the bottom of the submerged culture well for the same field of view at low magnification from 24 to 42 h PS. Figure 4.1C shows images at higher magnification so that representative cellular shapes can be seen. The co-developed mixtures were also examined at 72 and 96 h (data not shown), and looked similar as at 42 h. ! 121! Hours PS 24 42 Hours PS 24 42 A WT csgA Hours PS 24 30 36 42 B WT : csgA 1:4 1:2 1:1 2:1 4:1 ! 122! C Hours PS 24 30 36 42 WT : csgA 1:4 1:2 1:1 2:1 4:1 Figure 4.1 Development of wild-type and csgA mutant cells separately and co- development at different ratios. Cell mixtures were starved under submerged culture conditions. Vanillate (0.5 mM) was added at the start of starvation. Confocal images of the same field of view were acquired near the bottom of the biofilm or a mound at the indicated times PS. Images are representative of five biological replicates. A. Development of wild-type and csgA mutant cells separately. The top row of panels show labeled WT strain YH7 (green fluorescence) mixed with unlabeled WT strain DK1622 at a ratio of 1:5. The bottom row of panels show labeled csgA mutant strain YH11 (red fluorescence) mixed with unlabeled csgA mutant strain MRR33 at a ratio of 1:5. Arrows indicate small dots of red fluorescence, perhaps due to cross-sectional views of rods oriented vertically. Panels on the left and right are at lower (bar, 20 µm) and higher (bar, 5 µm) magnification, respectively. B. Confocal images of the same field of view during co-development at different ratios. WT and csgA mutant cells were mixed at ratios indicated on the left. Half of the WT cells were the labeled strain YH7 (green fluorescence) and half were the unlabeled strain DK1622. Likewise, half of the csgA mutant cells were the labeled strain YH11 (red fluorescence) and half were the unlabeled strain MRR33. Images show the green and red channels merged. Bar, 20 µm. C. Close-up images of representative cells during co-development at different ratios. Higher magnification images from panel B show representative cellular shapes. The white arrow indicates a rod-shaped cell, the yellow arrow indicates a transitioning cell, and the red arrow indicates a spore. Bar, 5 µm. When WT and csgA mutant cells were mixed at a ratio of 1:4, mounds and spores failed to form (Fig. 4.1B and 4.1C). At a ratio of 1:2, mounds formed by 24 h, but were less compact than normal. The broad mounds persisted from 24 to 42 h, and neither WT nor csgA mutant cells appeared to undergo cellular shape change. Co-developing equal amounts of WT and ! 123! csgA mutant cells (i.e., a 1:1 ratio) allowed formation of normal-looking mounds at 24 h, and rods that appeared to be transitioning to spores (called “transitioning cells” (TCs) in Chapter 3 and hereafter), as well as spores, were observed near the bottom of nascent fruiting bodies at 36 h. Strikingly, there were more csgA mutant spores (red) than WT spores (green) at both 36 and 42 h, consistent with cheating by the csgA mutant (24). Mixtures with ratios of 2:1 and 4:1 also appeared to have more csgA mutant spores than expected, based on the ratios in the initial mixtures. In addition, a few csgA mutant TCs and spores (red) were seen for these mixtures at 30 h, but WT cells did not appear to be undergoing cellular shape change, suggesting that cgsA mutant cells were more advanced in development. Our qualitative visual assessment of co-development in the mixtures suggests that short- range C-signaling has reached a threshold for the emergent behaviors of compact mound formation and sporulation when WT cells comprise 50% of the population (i.e., the 1:1 ratio), but not when WT cells comprise 33% of the population (i.e., the 1:2 ratio). Above the threshold, csgA mutant cells appear to exhibit cheating behavior, forming a disproportionately large number of the spores in the population, while not contributing C-signal, a public good. Cell density is two-fold lower in broad mounds than in compact mounds, and csgA mutant cells are disproportionately overrepresented in compact mounds formed when wild-type cells are initially in excess. Development requires starvation and high cell density to proceed (26). When WT and csgA mutant cells were mixed at different ratios, no mounds were formed at 1:4 and broad mounds were observed at 1:2 (Fig. 4.1B). To determine the cell density achieved at different ratios, z-stacks of optical sections were collected from near the bottom of the biofilm or mound, to 5 µm up. As described in Chapter 3, computational analysis of the z-stacks led to segmentation and shape classification of individual cells. Since we did not observe any mounds or cellular shape change at the 1:4 ratio from 24 to 42 h PS, the density of rods at 42 h was measured, and was found to be ~0.005 rod/µm3 (Fig. 4.2A, blue dots). For the other ratios, very little cellular shape change had occurred by 24 h, so the density ! 124! of the population of primarily rods in the broad or compact mounds that had formed was measured. Interestingly, the density in broad mounds at the 1:2 ratio was ~0.006 rod/µm3, which was very close to the density at 42 h in the biofilm formed at the 1:4 ratio. Presumably, a low level of C-signaling in the 1:4 mixture accounts for the lack of mound formation, and it may also account for the small difference in cell density compared to the 1:2 mixture at the earlier developmental time, since csgA mutants have been reported to undergo less lysis during development (27) (28) (20). The density in compact mounds formed by the other mixtures (1:1, 2:1, 4:1) was ~two-fold greater than in broad mounds formed by the 1:2 mixture (Fig. 4.2A, blue dots). Compact mounds formed by the 1:1, 2:1, and 4:1 mixtures had similar cell density, which was ~0.005 rod/µm3 greater than for broad mounds formed by the 1:2 mixture (Fig. 4.2B). We conclude that when WT cells comprise 50% or more of the population, short-range C-signaling has reached a threshold that allows compact mounds with ~two-fold greater cell density to form at 24 h, which is correlated with subsequent ability to form spores (Fig. 4.1B and 4.1C). The density of WT and csgA mutant cells in each mixture is also shown in Figure 4.2A. Strikingly, the density of WT and csgA mutant cells in mounds did not match their initial ratios in the mixtures. For the 1:1, 2:1, and 4:1 mixtures that formed compact mounds, the density of WT and csgA mutant cells was similar at ~0.005 rod/µm3 of each. These results indicate that C- signaling by WT cells efficiently rescued mound formation by csgA mutant cells in the 1:1 mixture, and that csgA mutant cells were disproportionately overrepresented in mounds formed by the 2:1 and 4:1 mixtures. In other words, the csgA mutant appears to be cheating during mound formation. If those csgA mutant cells subsequently form spores with similar efficiency as WT cells, it would explain the disproportionately large number of csgA mutant spores observed near the bottom of nascent fruiting bodies formed by the 2:1 and 4:1 mixtures (Fig. 4.1B and 4.1C). Since the 1:1 mixture also exhibited a disproportionately large number of csgA mutant spores, but csgA mutant cells were not overrepresented in mounds (Fig. 4.2A), the csgA mutant also appears to be cheating during sporulation, at least near the bottom of nascent fruiting ! 125! bodies. In contrast, below the threshold for C-signaling that allows compact mound formation, in the 1:2 mixture that formed broad mounds, a slightly higher density of WT cells than csgA mutant cells was observed (Fig. 4.2A). Hence, WT cells were disproportionately overrepresented in the broad mounds, suggesting that C-signaling by WT cells allows them to form mounds inefficiently, and rescues mound formation by csgA mutant cells even less efficiently. In the 1:4 mixture that did not form mounds, the ratio of WT to csgA mutant cell densities was ~1:3 at 42 h, so there was no evidence that WT cells had undergone more lysis than csgA mutant cells in the mixture, as has been reported for comparisons between unmixed WT and csgA mutant strains (27) (28) (20). Perhaps csgA mutant cells in the mixture interfere with lysis of WT cells, in addition to interfering with mound formation and sporulation. A B ) 3 l l / m µ d o r ( y t i s n e d e C Cell density (rod/µm3) Figure 4.2 Cell density upon co-development of WT and csgA mutant cells at different ratios. In the experiment described in the legend of Figure 4.1 and in the text, z-stacks of optical sections from near the bottom of the biofilm or mound, to 5 µm up, were acquired at 24 h PS, except for the 1:4 mixture the z-stack was acquired at 42 h. Segmented cells from the z- stacks were classified as rods, TCs, or spores, but very few of the latter two cell types were observed, so only rods were considered. A. Density of WT, csgA mutant, and total cells after co-development of mixtures with different initial ratios of cells. Each dot represents one biological replicate (five replicates, except one replicate for the 1:4 mixture; some dots overlap for other mixtures). B. Comparison of total cell density after co-development of mixtures with different initial ratios of cells. For the indicated mixtures, the posterior probability distributions of the differences in total cell density were shown. The filled circle indicates the mean, and the thick and thin vertical lines show the 50% and 95% highest density intervals, respectively. Mixtures that form compact mounds exhibit similar numbers of transitioning cells and spores. To examine whether cell density correlates with cellular shape change, TCs and ! 126! spores were quantified in the z-stacks collected at 42 h PS as described above. Consistent with our observations from confocal images acquired near the bottom of the biofilm (Fig. 4.1B and 4.1C), when WT and csgA mutant cells were mixed at a 1:4 ratio, very few rods changed shape (Fig. 4.3A). Although cells in the 1:2 mixture were able to coordinate their movements to form broad mounds (Fig. 4.1B), very few cells changed shape (Fig. 4.3A). In the 1:1 mixture, typically more than 300 spores and nearly 100 TCs were observed, and similar numbers of spores and TCs were formed during co-development of the 2:1 and 4:1 mixtures. To compare cellular shape change for the 1:1, 2:1, and 4:1 mixtures that formed compact mounds, we generate posterior probability distributions of the differences. The differences in the number of TCs overlapped with 0 in the 95% highest density interval (Fig. 4.3B). Similarly, the differences in the number of spores overlapped with 0 in the 50% highest density interval (Fig. 4.3C). Taken together, our results imply that when WT cells comprise 50% or more of the population, short-range C-signaling has reached a threshold that not only allows compact mounds with similar cell density to form at 24 h (Fig. 4.2), it also allows similar numbers of TCs and spores to form at 42 h (Fig. 4.3). ! 127! A! B! C! Figure 4.3 Quantification of cellular shape change upon co-development of WT and csgA mutant cells at different ratios. In the experiment described in the legend of Figure 4.1 and in the text, z-stacks of optical sections from near the bottom of the biofilm or mound, to 5 µm up, were acquired at 42 h PS. Segmented cells from the z-stacks were classified as rods, TCs, or spores, and here we report the number of the latter two cell types. A. The number of transitioning cells and spores after co-development of mixtures with different initial ratios of cells. Each dot represents one biological replicate (five replicates, except one replicate for the 1:4 mixture; some dots overlap for other mixtures). B. Comparison of the number of transitioning cells after co-development of mixtures with different initial ratios of cells. For the indicated mixtures, posterior probability distributions of the differences in the number of TCs were shown. The filled circle indicates the mean, and the thick and thin vertical lines show the 50% and 95% highest density intervals, respectively. C. Comparison of the number of spores after co-development of mixtures with different initial ratios of cells. Differences in the number of spores were analyzed as described for panel B. The csgA mutant cheats during sporulation in the 1:1 mixture, but during mound formation when wild-type cells are initially in excess. For the mixtures that formed compact mounds, csgA mutant cells were disproportionately overrepresented in mounds formed by the 2:1 and 4:1 mixtures at 24 h PS (Fig. 4.2A), suggesting that the csgA mutant cheats during mound formation. Since the total numbers of TCs and spores formed by the 1:1, 2:1, and 4:1 mixtures at 42 h was similar (Fig. 4.3), and given the apparent preponderance of csgA mutant spores near the bottom of nascent fruiting bodies formed by those mixtures (Fig. 4.1B and ! 128! 4.1C), we wondered whether the csgA mutant also cheats during sporulation. Therefore, we quantified csgA mutant and WT rods, TCs, and spores in the z-stacks collected as described above at each developmental time point, for the mixtures that formed compact mounds and for the 1:2 mixture that formed broad mounds. Figure 4.4 Proportions of rods, transitioning cells, and spores upon co-development of WT and csgA mutant cells at different ratios. In the experiment described in the legend of Figure 4.1 and in the text, z-stacks of optical sections from near the bottom of mounds to 5 µm up were acquired at the indicated times PS. Segmented cells from the z-stacks were classified as rods, TCs, or spores, and here we report the proportion of the population for each cell type. Each dot represents one biological replicate (five replicates; some dots overlap). The lines connect the means for a given cell type at different times PS. For the 1:2 mixture, as noted above based on cell density (Fig. 4.2A), WT rods were disproportionately overrepresented in broad mounds at 24 h PS, comprising ~60% of the population, while csgA mutant rods comprised ~35% (Fig. 4.4). Hence, the proportions of WT and csgA!mutant rods in broad mounds were nearly the inverse of their initial ratio in the 1:2 mixture. The proportions of WT and csgA mutant rods changed slightly over time, but WT rods ! 129! always outnumbered csgA mutant rods. As noted above (Fig. 4.3A), very few TCs and almost no spores were observed (Fig. 4.4). For the 1:1 mixture, WT and csgA mutant rods each comprised ~50% of the population in compact mounds at 24 h PS (Fig. 4.4), as expected from their equivalent cell densities (Fig. 4.2A). There was little change at 30 h, but at 36 h the proportion of spores rose considerably, with csgA mutant spores on average (~25%) slightly outnumbering WT spores (~20%), rods of both types declined to ~20% each, and TCs of both types comprised the remaining ~15% of cells (Fig. 4.4). The trends continued at 42 h, though changes were less dramatic. On average, the proportions of csgA mutant TCs and spores were slightly greater than for WT cells at 36 and 42 h, suggesting that csgA mutant cells form spores slightly more efficiently than WT cells in the mixture. Such cheating during sporulation may explain the apparent overrepresentation of csgA mutant spores near the bottom of nascent fruiting bodies formed by the 1:1 mixture (Fig. 4.1B and 4.1C) When the initial ratio of WT to csgA mutant cells was 2:1, similar proportions of WT and csgA mutant rods (~50% each) were observed at 24 h PS (Fig. 4.4), indicative of cheating by the csgA mutant during mound formation, as noted above based on cell density (Fig. 4.2A). At 30 h, WT rods still accounted for ~50% of the population, while csgA mutant rods decreased to ~35% on average (Fig. 4.4). Interestingly, the proportion of csgA mutant TCs was ~2.5% at 24 h and ~5% at 30 h, while the proportion of WT TCs was nearly 0% at 24 h and 2.5% at 30 h. In the 1:1 mixture, very few TCs were seen at 24 or 30 h. Hence, cellular shape change began earlier in the 2:1 mixture than in the 1:1 mixture, and began earlier for csgA mutant cells than for WT cells, consistent with our observations from confocal images acquired near the bottom of nascent fruiting bodies (Fig. 4.1B and 4.1C). Also in agreement with earlier developmental progression in the 2:1 mixture, at 36 h the proportion of spores rose greatly (~35% WT, ~25% csgA mutant) and rods of both types declined to ~10-15%, with little change in these proportions at 42 h (Fig. 4.4). Importantly, similar proportions of WT and csgA mutant rods in mounds at 24 ! 130! h resulted in roughly similar proportions of spores at 42 h, so in the 2:1 mixture there was no evidence of cheating during sporulation. Rather, cheating occurred during mound formation. For the 4:1 mixture, cheating also occurred during mound formation. As noted above, csgA mutant rods were disproportionately overrepresented in mounds at 24 h PS, reaching nearly an equivalent density as WT rods (Fig. 4.2A), despite the initial 4:1 ratio. The proportion WT rods was only ~1.5-fold greater than that of csgA mutant rods on average at 24 h, and csgA mutant TCs already comprised ~5% of the population (Fig. 4.4). At 30 h, while the WT rod proportion remained the same, csgA mutant rods decreased slightly and spores were observed. As in the 2:1 mixture, development progressed rapidly in the 4:1 mixture, with the proportions of spores rising and rods declining greatly at 36 h, and changing little at 42 h. Also as in the 2:1 mixture, the proportions of WT and csgA mutant rods in mounds at 24 h resulted in roughly similar proportions of spores at 42 h in the 4:1 mixture, so there was no evidence of cheating during sporulation. In summary, our data suggest that the csgA mutant cheats during sporulation in the 1:1 mixture. Both cell types are present equally in compact mounds at 24 and 30 h PS, but the csgA mutant forms spores slightly more efficiently than the WT strain. In contrast, there was no evidence that the csgA mutant cheats during sporulation in the 2:1 or 4:1 mixtures. Rather, cheating occurred during mound formation. Apparently, the minority of csgA mutant rods take advantage of abundant C-signaling by WT rods in the 2:1 and 4:1 mixtures during mound formation so that csgA mutant rods are disproportionately overrepresented in mounds by 24 h, then both cell types form spores with similar efficiency by 42 h, although small proportions of csgA mutant cells appear to form TCs and spores earlier than WT cells. In the 1:1 mixture, C- signaling by WT rods reaches a threshold for rescue of compact mound formation by csgA mutant rods, but not cheating, since csgA mutant rods are not overrepresented in the mounds. However, limited C-signaling in the mounds appears to result in cheating by the csgA mutant during sporulation. ! 131! csgA mutant spores accumulated closer to the center of mounds than wild-type spores. For the unmixed wild-type strain DK1622, TCs and spores were more abundant closer to the radial center of nascent fruiting bodies in z-stacks of optical sections from near the bottom to 10 µm up (Chapter 3). Here, we report the spatial distribution of WT and csgA mutant rods, TCs, and spores in the z-stacks collected as described above, for the mixtures that formed compact mounds and for the 1:2 mixture that formed broad mounds. Since the radius of all compact mounds from collected z-stacks was approximately 60 µm (data not shown), we examined the radial distribution of cell shapes from the center of mounds to 60 µm (Fig. 4.5). Collected z-stacks were 5 µm thick, but we only examined cells in the range of 1 to 4.5 µm up from the bottom of mounds to eliminate edge effects (Fig. 4.6). For the 1:2 mixture, WT rods were more abundant than csgA mutant rods farther from the center at all times (Fig. 4.5). Interestingly, a very small proportion of the csgA mutant formed spores far from the center, as if cheating on C-signaling by WT rods occurred there, albeit inefficiently. The distributions of WT and csgA mutant rods varied little vertically (Fig. 4.6). For the 1:1 mixture, WT and csgA mutant rods were distributed similarly both radially (Fig. 4.5) and vertically (Fig. 4.6) at 24 and 30 h PS. At 36 and 42 h, csgA mutant TCs and spores were closer to the center than WT spores, there were very few WT TCs, and both WT and csgA mutant rods were far from the center (Fig. 4.5). The vertical distributions at 36 and 42 h were overlapping in the shaded 95% credible intervals for WT and csgA mutant cells in the same shape classification, but on average csgA mutant spores were more abundant than WT spores, especially farther from the bottom (Fig. 4.6). ! 132! Figure 4.5 Radial distribution of rods, transitioning cells, and spores upon co- development of WT and csgA mutant cells at different ratios. In the experiment described in the legend of Figure 4.1 and in the text, z-stacks of optical sections from near the bottom of mounds to 5 µm up were acquired at the indicated times PS. Segmented cells from the z- stacks were classified as rods, TCs, or spores. The proportion of each cell type radially from the center of mounds to 60 µm away is shown. Lines represent the means and shaded regions represent 95% credible intervals (five biological replicates). ! 133! Figure 4.6 Vertical distribution of rods, transitioning cells, and spores upon co- development of WT and csgA mutant cells at different ratios. In the experiment described in the legend of Figure 4.1 and in the text, z-stacks of optical sections from near the bottom of mounds to 5 µm up were acquired at the indicated times PS. Segmented cells from the z- stacks were classified as rods, TCs, or spores. The proportion of each cell type vertically from 1 to 4.5 µm up from the bottom of mounds is shown. Lines represent the means and shaded regions represent 95% credible intervals (five biological replicates). The 2:1 and 4:1 mixtures exhibited similar spatial distributions. At 24 and 30 h PS, WT rods were more abundant than csgA mutant rods farther from the center, as for the 1:2 mixture (Fig. 4.5). Near the radial center of mounds formed by the 2:1 and 4:1 mixtures, there was a small proportion of csgA TCs at all times, as for the 1:1 mixture at 36 and 42 h. Also like the 1:1 mixture at 36 and 42 h, for mounds formed by the 2:1 and 4:1 mixtures at 36 and 42 h, csgA mutant spores were closer to the center than WT spores, and both WT and csgA mutant rods ! 134! were far from the center. A small proportion of WT TCs was also observed near the radial center of mounds formed by the 2:1 and 4:1 mixtures at 36 and 42 h. Vertically, the proportions of cells showed little variation for the 2:1 and 4:1 mixtures, with csgA mutant spores on average only slightly more abundant farther from the bottom (as for the 1:1 mixture), and WT spores on average slightly more abundant partway up (Fig. 4.6). Discussion We have characterized the threshold effects of short-range C-signaling on the emergent behaviors of mound formation, sporulation, and cheating during M. xanthus development. Differentially labeled WT and csgA mutant cells were co-developed at slightly different ratios and examined by confocal laser scanning microscopy. At a ratio of 1 WT to 4 csgA mutant cells (1:4), neither mounds nor spores were able to form. At a 1:2 ratio, rod-shaped cells formed broad mounds with twofold less cell density than normal. Although WT cells comprised only 33% of the initial mixture, they accounted for ~60% of the cells in mounds at 24 h PS, but neither strain went on to form many TCs or spores. Co-developing an equal amount of WT and csgA mutant cells (1:1 ratio) allowed compact mound formation and sporulation, indicating that a threshold for C-signaling had been reached. The mounds had similar numbers of WT and csgA mutant rods at 24 and 30 h, but at 36 and 42 h csgA spores slightly outnumbered WT spores on average, suggesting that the csgA mutant cheated during sporulation. In contrast, at initial ratios of 2:1 and 4:1, the csgA mutant cheated during mound formation rather than during sporulation. For the 1:1 and 2:1 mixtures, csgA mutant TCs were more abundant near the radial center of nascent fruiting bodies than WT TCs, suggesting that C-signaling is most abundant near the center (in z-stacks from the bottom to 5 µm up). Spores of the csgA mutant also were observed closer to the center than WT spores, and csgA mutant spores were on average slightly more abundant farther from the bottom. The implications of these findings for ! 135! C-signal-dependent threshold effects on emergent behaviors during fruiting body development are discussed below. Threshold effects of C-signaling on mound formation and sporulation. Kim and Kaiser showed that the C-signal concentration increases only about fourfold between the beginning of mound formation and the beginning of sporulation (10). They also showed that β- galactosidase production from csgA-lacZ increased about fourfold during that interval. Furthermore, 0.8 unit of C-signal, but not 0.6 unit, rescued mound formation by a csgA mutant, and 1.0 unit additionally rescued sporulation, suggesting that distinct thresholds of C-signaling are required for the emergent behaviors of mound formation and sporulation. Consistent with that suggestion, we found that mound formation and sporulation of a csgA mutant were rescued upon co-development with a WT strain over a narrow range of initial ratios in cell mixtures. Increasing WT cells from 20% in the 1:4 mixture to 33% in the 1:2 mixture partially restored mound formation (Fig. 4.1B), but csgA mutant rods were underrepresented in the broad mounds (Fig. 4.4), which had twofold lower cell density than normal compact mounds (Fig. 4.2). The underrepresentation of the csgA mutant indicates poor rescue by C-signaling from WT cells, as do the abnormal breadth and low cell density of the mounds. Further increasing WT cells to 50% in the 1:1 mixture profoundly improved both mound formation and sporulation. We tested mixtures with between 33% and 50% WT cells, but the appearance of the mounds was not reproducible, so we did not characterize them further. Hence, we were unable to identify distinct thresholds of C-signaling required for compact mound formation versus sporulation by using mixtures of WT and csgA mutant cells. Nevertheless, comparison of WT and csgA mutant rods during mound formation (i.e., before 24 h PS) in the 1:2 and 1:1 mixtures, e.g., with respect to their motility and cell-cell interactions, may yield additional insight into the role of C- signaling in compact mound formation. Cell motility and alignment are required for C-signaling (29) (9) (30), but whether cells must contact each other, and if so, the nature of those contacts (end-to-end, end-to-side, side-to-side), is unknown. ! 136! The mixtures with 67% (2:1) or 80% (4:1) WT cells formed compact mounds with similar appearance (Fig. 4.1B), cell density (Fig. 4.2), and numbers of TCs and spores at 42 h (Fig. 4.3) as the 1:1 mixture, supporting the conclusion that a threshold of C-signaling had been reached in the 1:1 mixture that allowed normal mound formation and sporulation. However, our data do not rule out the possibilities that mounds formed by the 2:1 and/or 4:1 mixtures are taller and/or that the nascent fruiting bodies contain more spores in total than those formed by the 1:1 mixture, since we only analyzed z-stacks from the bottom to 5 µm up. Light scattering limits the height at which cells in nascent fruiting bodies can be imaged at high-resolution. Close inspection of images near the bottom of nascent fruiting bodies formed by the 2:1 and 4:1 mixtures revealed subtle differences from those formed by the 1:1 mixture. TCs were evident at 30 h PS (Fig. 4.1B and 4.1C) and a few were seen at 24 h (Fig. 4.1B) only for the mixtures with >50% WT cells initially. Most of the TCs were csgA mutant (red). In agreement, the 2:1 and 4:1 mixtures had more csgA mutant TCs than the 1:1 mixture at 24 and 30 h (Fig. 4.4). These results show that csgA mutant rods begin transitioning to spores earlier than WT rods in mixtures with more WT rods capable of C-signaling. Short-range C-signaling increases during development both because cells move into closer proximity as mounds form (29) (9) (30) and because expression of the csgA gene and production of C-signal increase (10) (11). The increase in C-signal is controlled by the transcription factor MrpC in response to starvation (31) (31) (Fig. 4.7). MrpC may directly activate csgA transcription, since a ChIP-seq study revealed an MrpC binding site in the csgA promoter region (32). Addition of nutrient medium to developing cells causes rapid degradation of MrpC, halting commitment of cells to sporulation (33). My results reported in Chapter 2 show that developing cells respond ultrasensitively to nutrient medium addition and that the level of MrpC correlates with the ability to form mounds and spores. Therefore, we propose that a threshold level of MrpC is necessary to meet threshold requirements of C-signaling for mound formation and sporulation (Fig. 4.7). MrpC also appears to directly activate transcription of fruA ! 137! (34), which codes for another key transcription factor of the developmental gene regulatory network (35) (19) (20). C-signaling is proposed to activate FruA, forming FruA*, but the mechanism is unknown (19) (20) (Fig. 4.7). FruA* is proposed to promote mound formation and sporulation. As cells move into close proximity in mounds, C-signaling presumably increases, further elevating FruA* in a positive feedback loop. According to this model, FruA* is the downstream effector of C-signaling, raising the question of whether the threshold effects of C- signaling on mound formation and sporulation reflect different threshold requirements of FruA* for these emergent behaviors. Different threshold levels of FruA* have been proposed to be required for transcription from particular C-signal-dependent promoters (20) (36) (37). As explained in Chapter 3, Pdev was predicted to have a relatively low threshold for FruA* and a higher threshold was predicted for PfmgE. Consistent with the predictions and with the notion that the FruA* level rises during development (Fig. 4.7), Pdev was activated earlier during development than PfmgE (Fig. 3.6 and 3.7). Furthermore, the difference between Pdev activity in rods versus cells that had changed shape was less than for PfmgE (Fig. 3.7), so perhaps the threshold level of FruA* required to trigger sporulation is greater than that required for mound formation, as suggested by the model (Fig. 4.7). Future studies will aim to measure Pdev and PfmgE activities in a csgA mutant co- developed in different ratios with WT cells. We hypothesize that Pdev activity will be more sensitive to C-signaling, exhibiting some activity in a 1:2 mixture and greater activity in a 1:1 mixture, whereas PfmgE will exhibit no activity in a 1:2 mixture and may exhibit earlier and greater activity in mixtures as the proportion of WT cells is increased. ! 138! Starvation MrpC FruA C-signaling mound FruA* rod transitioning cell spore fruiting body Figure 4.7 Model for co-development of wild-type and csgA mutant cells. Similar proportions of WT (green) and csgA mutant (red) cells in the initial mixture are assumed for this illustration. Starvation triggers the production of MrpC, which drives FruA production. A threshold level of MrpC is proposed to be required to meet a threshold requirement of C- signaling for activation of FruA (denoted as FruA*). FruA* is proposed to promote mound formation, which increases cell proximity, resulting in more C-signaling. The positive feedback loop thus created would further elevate FruA*, which is proposed to bind cooperatively with MrpC and activate genes important for sporulation. The csgA mutant transitions to spores slightly more efficiently than the WT strain, especially near the radial center of the nascent fruiting body, and perhaps farther up from the bottom. A threshold effect of C-signaling on cheating. Velicer et al. reported that csgA mutant spores dominated WT spores upon co-development of a 1:1 mixture!(24). When WT cells outnumbered csgA mutant cells 99:1 in a mixture, csgA mutant spores accounted for more than 1% of the resulting spores. In both mixtures, the csgA mutant cheated, forming a disproportionately large number of spores without contributing to a public good, C-signal. The cell mixtures were spotted on Tris-buffered agar supplemented with phosphate and magnesium to initiate development, samples were collected at 68 h PS for heat treatment and sonication, spores were plated on nutrient agar with and without antibiotics, and colonies were counted after several days of incubation to infer spore numbers. In this work, we were able to quantify ! 139! WT and csgA mutant cells co-developed in ratios over a narrow range, at multiple times PS in situ. Interestingly, our data provide evidence for a threshold effect of C-signaling on cheating behavior, since a lower level of C-signaling in a 1:1 mixture appeared to result in cheating during sporulation (Fig. 4.4) and higher levels of C-signaling in the 2:1 and 4:1 mixtures resulted in cheating during mound formation (Fig. 4.2 and 4.4). We propose that abundant C-signaling by WT rods in the 2:1 and 4:1 mixtures changes the motility behavior of csgA mutant rods in a way that favors their entry into mounds and/or disfavors their exit from mounds. Both purified C-signal (38) (39) and WT cells (40) have been shown to alter motility of a csgA mutant in ways that could promote mound formation. In the 1:1 mixture, C-signaling by WT rods was sufficient to rescue mound formation by csgA mutant rods, but not disproportionately so (Fig. 4.2 and 4.4). We hypothesize that equal numbers of WT and csgA mutant rods then compete for limited C-signaling in the mounds, with the csgA mutant forming spores slightly more efficiently than the WT strain, presumably owing to the physiological state of the csgA mutant being more favorable for sporulation due to not having produced C-signal. In mounds formed by the 2:1 mixture, equal numbers of WT and csgA mutant rods were present in mounds at 24 h (as for the 1:1 mixture), yet the csgA mutant did not form spores more efficiently than the WT strain (Fig. 4.2 and 4.4). We propose that the different outcome in the two mixtures relates to the abundance of C-signaling during mound formation (i.e., prior to 24 h) and its effects on the csgA mutant, as reflected by the decrease in the proportion of csgA mutant rods and the increase in the proportion of csgA mutant TCs at 30 h in the 2:1 mixture, but not in the 1:1 mixture. As noted above, csgA mutant rods transitioning to spores earlier than WT rods was evident in mixtures with >50% WT cells initially (Fig. 4.1B and 4.1C). We speculate that earlier developmental progression of the csgA mutant in the 2:1 mixture changes C-signaling in mounds after 24 h so that WT rods form spores at least as efficiently as the csgA mutant by 36 h (Fig. 4.4). Likewise, earlier developmental progression of the csgA mutant in ! 140! mounds formed by the 4:1 mixture (Fig. 4.1B and 4.1C) may change C-signaling after 24 h so that WT spores form as efficiently as csgA mutant spores by 36 h (Fig. 4.4). In terms of a threshold effect of C-signaling on cheating by the csgA mutant, limited C- signaling in the 1:1 mixture does not reach a threshold for cheating during mound formation, but does so for cheating during sporulation. More abundant C-signaling in the 2:1 and 4:1 mixtures reaches a threshold for cheating during mound formation, but earlier developmental progression of the csgA mutant (perhaps related to the presumed physiological state favorable for sporulation speculated above to explain cheating during sporulation in the 1:1 mixture) changes C-signaling in mounds after 24 h so that a threshold for cheating during sporulation no longer exists (e.g., the csgA mutant begins transitioning to spores, no longer competing for C-signaling, allowing WT rods to engage in C-signaling and reach the threshold for sporulation). Spatial distributions of wild-type and csgA mutant cellular shape change differ. Transcription from C-signal-dependent promoters appeared to be greatest near the radial center of nascent fruiting bodies (Fig. 3.8A), suggesting that C-signaling is greatest near the center. TCs and spores were more abundant near the center (Fig. 3.5A), consistent with the possibility that abundant C-signaling near the center triggers cellular shape change there. However, TCs can be moved by rods (Fig. 3.1C), and rods moving circularly in the outer domain of nascent fruiting bodies have been proposed to push spores inward (41), as discussed in Chapter 3. Here, we found that csgA mutant TCs and spores were more abundant than those from the WT strain near the radial center of nascent fruiting bodies formed by the 1:1 and 2:1 mixtures (Fig. 4.5). For those mixtures, as well as the 4:1 mixture, the maximum proportion of csgA mutant spores was observed at or very close to the center, while that of WT spores was typically ~30 µm from the center. The greater proximity of csgA mutant spores than WT spores to the center could reflect greater sensitivity of the csgA mutant to C-signaling that is more abundant near the center. Alternatively or in addition, earlier sporulation of the csgA mutant could allow more time for inward pushing by rods. Although earlier sporulation of the csgA mutant was not observed ! 141! for the 1:1 mixture, it cannot be ruled out during the period between 30 and 36 h PS (Fig. 4.4). However, that would leave little time for inward pushing by rods to create the different radial distributions of csgA mutant and WT spores observed at 36 h. Moreover, csgA mutant TCs were observed near the radial center of nascent fruiting bodies formed by the 2:1 and 4:1 mixtures at 24 and 30 h, perhaps also limiting the time for inward pushing by rods. These considerations, together with the distribution of C-signal-dependent promoter activity (Fig. 3.8A), lead us to favor the possibility that C-signaling is more abundant near the radial center of nascent fruiting bodies and triggers cellular shape there, and that the csgA mutant is more sensitive than the WT strain to C-signaling. The maximum proportion of csgA mutant spores was typically ~4.5 µm from the bottom (i.e., near the top of the z-stack, corresponding to a few several cell layers up from the bottom), whereas that of WT spores was typically near the middle of the z-stack, but the differences were small (Fig. 4.6). As discussed above for differences in radial distributions, differences in vertical distributions of csgA mutant versus WT spores could reflect differential sensitivity to C-signaling and/or movement of spores by rods. We favor the latter possibility since the vertical distribution of C-signal-dependent promoter activity was uniform (Fig. 3.8B). Specifically, earlier sporulation by the csgA mutant could favor upward movement of those spores by rods pushing less mature WT spores inward. Interestingly, the maximum proportion of spores was found at ~4 µm from the bottom of 48-h nascent fruiting bodies formed by the unmixed WT strain, in z-stacks collected from the bottom to 10 µm up (Fig. 3.5B). Materials & Methods Bacterial strains and plasmids. Table S3.1 lists the strains and plasmids used in this study. To construct M. xanthus strain MRR33 with a plasmid insertion mutation in csgA, pRR028 was transformed into strain DK1622 using electroporation (42) with outgrowth in Casitone-Tris (CTT) liquid medium (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH2PO4- ! 142! K2HPO4, 8 mM MgSO4, [final pH 7.6]) prior to plating on CTT agar (1.5% agar) supplemented with 40 µg/mL kanamycin to select transformants. pYH8, which was constructed to induce production of tdTomato (chapter 3), was transformed into the csgA mutant strain MRR33 and outgrowth in Casitone-Tris (CTT) liquid prior to plating on CTT agar supplemented with 15 µg/ml tetracycline. Ectopic integration of pYH8 was verified by colony PCR using primers pMR3691 MCS G-F and pMR3691 MCS G-R. The resulting strain was named YH11. Growth and development. M. xanthus was grown at 32°C on CTT agar or in CTTYE liquid medium (CTT with 0.2% yeast extract) shaking at 350 rpm. Development was performed under submerged culture conditions in 8-well µ-slides (Ibidi) with starvation buffer MC7 (10 mM morpholinepropanesulfonic acid [MOPS, pH 7.0], 1 mM CaCl2). Cells from log-phase cultures in CTTYE liquid were resuspended in MC7 at 1,000 Klett units as described previously (43). Cell suspension (26 µL) of strains DK1622, YH7, MR33, and YH11 in different ratios was added to 174 µL of MC7 for each well. Microscopy. Images of nascent fruiting bodies were acquired with a Nikon A1 Laser Scanning Confocal Microscope configured on a Nikon Ti inverted platform with an XY automated stage and a 100X objective. Fluorescence from tdTomato was examined using a 560-nm laser for excitation and a 595/50 band pass emission filter. Fluorescence from mNeonGreen was examined using a 488-nm laser for excitation and a 525/50 band pass emission filter. Images “near the base of mounds” were the first optical section above the bottom of the well in which cells could be clearly visualized. Cell segmentation, cell shape classification, and estimating the proportions of cell types in space. See Chapter 3. ! 143! Acknowledgments We thank Ramya Rajagopalan for constructing M. xanthus strain MRR33. We are grateful to Melinda Frame and the Center for Advanced Microscopy at Michigan State University for assistance with microscopy. This research was supported by National Science Foundation grants MCB-1411272 and IOS-1951025, and by salary support for L.K. from Michigan State University AgBioResearch. ! 144! APPENDIX ! 145! Table S4.1. Bacterial strains and plasmids used in this study Strain or plasmid Description Strain E. coli DH5α M. xanthus DK1622 MRR33 YH7 YH11 Plasmid pRR028 pYH7 pYH8 λ - φ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17(rK -mK - ) supE44 thi-1 gyrA relA1 Laboratory strain csgA mutation in DK1622 using pRR028 Pvan-mNeonGreen in DK1622 using pYH7 Pvan-tdTomato in MRR33 using pYH8 pCR2.1-TOPO with internal DNA fragment from csgA MXAN_0018-MXAN_0019-PR3-4::vanR- Pvan::mNeonGreen, Tcr MXAN_0018-MXAN_0019-PR3-4::vanR-Pvan::tdTomato, Tcr Source or reference (44) (45) This study Chapter 3 This study (46) Chapter 3 Chapter 3 ! 146! REFERENCES 147! ! REFERENCES Papenfort K, Bassler BL. 2016. Quorum sensing signal-response systems in Gram- negative bacteria. Nat Rev Microbiol 14(9):576-88. 10.1038/nrmicro.2016.89 Troselj V, Cao P, Wall D. 2018. Cell-cell recognition and social networking in bacteria. Environ Microbiol 20(3):923–933. 10.1111/1462-2920.14005 Lobedanz S, Søgaard-Andersen L. 2003. Identification of the C-signal, a contact- dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes Dev 17:2151–61. 10.1101/gad.274203 Losick R, Stragier P. 1992. Crisscross regulation of cell-type-specific gene expression during development in B. subtilis. Nature 355:601–4. 10.1038/355601a0 Claessen D, Rozen DE, Kuipers OP, Søgaard-Andersen L, Van Wezel GP. 2014. Bacterial solutions to multicellularity: A tale of biofilms, filaments and fruiting bodies. Nat Rev Microbiol 12:115–24. 10.1038/nrmicro3178 6. Goldman BS, Nierman WC, Kaiser D, Slater SC, Durkin AS, Eisen J, Ronning CM, Barbazuk WB, Blanchard M, Field C, Halling C, Hinkle G, Iartchuk O, Kim HS, Mackenzie C, Madupu R, Miller N, Shvartsbeyn A, Sullivan SA, Vaudin M, Wiegand R, Kaplan HB. 2006. Evolution of sensory complexity recorded in a myxobacterial genome. Proc Natl Acad Sci U S A 103(41):15200-5. 10.1073/pnas.0607335103 7. Wireman JW, Dworkin M. 1977. Developmentally induced autolysis during fruiting body formation by Myxococcus xanthus. J Bacteriol 129(2):798–802. 8. O’Connor KA, Zusman DR. 1991. Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. J Bacteriol 173:3318–33. 10.1128/jb.173.11.3318-3333.1991 Kim SK, Kaiser D. 1990. Cell aligmnent required in differentiation of Myxococcus xanthus. Science 249(4971):926-8. 10.1126/science.2118274 10. Kim SK, Kaiser D. 1991. C-factor has distinct aggregation and sporulation thresholds during Myxococcus development. J Bacteriol 173:1722–8. 10.1128/jb.173.5.1722- 1728.1991 11. Li S, Lee BU, Shimkets LJ. 1992. csgA expression entrains Myxococcus xanthus development. Genes Dev 6:401–10. 10.1101/gad.6.3.401 12. Kruse T, Lobedanz S, Berthelsen NMS, Søgaard-Andersen L. 2001. C-signal: A cell surface-associated morphogen that induces and co-ordinates multicellular fruiting body morphogenesis and sporulation in Myxococcus xanthus. Mol Microbiol 40:156–8. 10.1046/j.1365-2958.2001.02365.x 13. Hagen TJ, Shimkets LJ. 1990. Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. J Bacteriol. 10.1128/jb.172.1.15-23.1990 1. 2. 3. 4. 5. 9. ! 148! 14. Shimkets LJ, Rafiee H. 1990. CsgA, an extracellular protein essential for Myxococcus xanthus development. J Bacteriol 172(9):5299-306. 10.1128/jb.172.9.5299-5306.1990 15. Kim SK, Kaiser D. 1990. C-factor: A cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 61(1):19-26. 10.1016/0092-8674(90)90211-V 16. Kim SK, Kaiser D. 1990. Purification and properties of Myxococcus xanthus C-factor, an intercellular signaling protein. Proc Natl Acad Sci U S A 87(10):3635-9. 10.1073/pnas.87.10.3635 17. Rolbetzki A, Ammon M, Jakovljevic V, Konovalova A, Søgaard-Andersen L. 2008. Regulated Secretion of a Protease Activates Intercellular Signaling during Fruiting Body Formation in M. xanthus. Dev Cell 15(4):627-34. 10.1016/j.devcel.2008.08.002 18. Boynton TO, Shimkets LJ. 2015. Myxococcus CsgA, Drosophila sniffer, and human HSD10 are cardiolipin phospholipases. Genes Dev 29(18):1903-14. 10.1101/gad.268482.115 19. Ellehauge E, Norregaard M, Søgaard-Andersen L. 1998. The FruA signal transduction protein provides a checkpoint for the temporal co-ordination of intercellular signals in Myxococcus xanthus development. Mol Microbiol 30:807–817. 10.1046/j.1365- 2958.1998.01113.x 20. Saha S, Patra P, Igoshin O, Kroos L. 2019. Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C-signaling. Mol Microbiol 111:1732-52. 10.1111/mmi.14249 21. Søgaard-Andersen L, Overgaard M, Lobedanz S, Ellehauge E, Jelsbak L, Rasmussen AA. 2003. Coupling gene expression and multicellular morphogenesis during fruiting body formation in Myxococcus xanthus. Mol Microbiol 48(1):1-8. 10.1046/j.1365- 2958.2003.03399.x 22. Shimkets LJ, Gill RE, Kaiser D. 1983. Developmental cell interactions in Myxococcus xanthus and the spoC locus. Proc Natl Acad Sci U S A 80(5):1406-10. 10.1073/pnas.80.5.1406 23. Hagen DC, Bretscher AP, Kaiser D. 1978. Synergism between morphogenetic mutants of Myxococcus xanthus. Dev Biol 64(2):284-96. 10.1016/0012-1606(78)90079-9 24. Velicer GJ, Kroos L, Lenski RE. 2000. Developmental cheating in the social bacterium Myxococcus xanthus. Nature 404:598–601. 10.1038/35007066 25. ! Iniesta AA, Garcia-Heras F, Abellon-Ruiz J, Gallego-Garcia A, Ellias-Arnanz M. 2012. Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl-!β -D-thiogalactopyranoside or vanillate. J Bacteriol 194:5875–5885. 10.1128/JB.01110-12 26. Kuspa A, Plamann L, Kaiser D. 1992. A-signalling and the cell density requirement for Myxococcus xanthus development. J Bacteriol 174(22):7360-9. 10.1128/jb.174.22.7360- 7369.1992 149! 27. Janssen GR, Dworkin M. 1985. Cell-cell interactions in developmental lysis of Myxococcus xanthus. Dev Biol 112(1):194-202. 10.1016/0012-1606(85)90133-2 28. Lee B, Holkenbrink C, Treuner-Lange A, Higgs PI. 2012. Myxococcus xanthus developmental cell fate production: Heterogeneous accumulation of developmental Regulatory proteins and reexamination of the role of MazF in developmental lysis. J Bacteriol 194:3058–68. 10.1128/JB.06756-11 29. Kroos L, Hartzell P, Stephens K, Kaiser D. 1988. A link between cell movement and gene expression argues that motility is required for cell-cell signaling during fruiting body development. Genes Dev 2:1677-85. 10.1101/gad.2.12a.1677 30. Kim SK, Kaiser D. 1990. Cell motility is required for the transmission of C-factor, an intercellular signal that coordinates fruiting body morphogenesis of Myxococcus xanthus. Genes Dev 4(6):896-904. 10.1101/gad.4.6.896 31. Sun H, Shi W. 2001. Analyses of mrp genes during Myxococcus xanthus development. J Bacteriol 183 (23) 6733-6739. 10.1128/JB.183.23.6733-6739.2001 32. Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of hundreds of developmentally-regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. 10.1186/1471-2164-15-1123 33. Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736–47. 10.1128/JB.01692-14 34. Ueki T, Inouye S. 2003. Identification of an activator protein required for the induction of fruA, a gene essential for fruiting body development in Myxococcus xanthus. Proc Natl Acad Sci 100:8782–7. 10.1073/pnas.1533026100 35. Ogawa M, Fujitani S, Mao X, Inouye S, Komano T. 1996. FruA, a putative transcription factor essential for the development of Myxococcus xanthus. Mol Microbiol 22(4):757-67. 10.1046/j.1365-2958.1996.d01-1725.x 36. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L. 2015. Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 197:240–51. 10.1128/JB.02310-14 37. Son B, Liu Y, Kroos L. 2011. Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 193:2756–66. 10.1128/JB.00205-11 38. Søgaard-Andersen L, Kaiser D. 1996. C factor, a cell-surface-associated intercellular signaling protein, stimulates the cytoplasmic Frz signal transduction system in Myxococcus xanthus. Proc Natl Acad Sci U S A 93(7):2675–9. 10.1073/pnas.93.7.2675 39. Jelsbak L, Søgaard-Andersen L. 1999. The cell surface-associated intercellular C-signal induces behavioral changes in individual Myxococcus xanthus cells during fruiting body morphogenesis. Proc Natl Acad Sci U S A 96(9):5031-6. 10.1073/pnas.96.9.5031 ! 150! 40. Jelsbak L, Søgaard-Andersen L. 2002. Pattern formation by a cell surface-associated morphogen in Myxococcus xanthus. Proc Natl Acad Sci U S A 99 (4) 2032- 7.10.1073/pnas.042535699. 41. Sager B, Kaiser D. 1993. Spatial restriction of cellular differentiation. Genes Dev 7:1645- 3. 10.1101/gad.7.9.1645 42. Kashefi K, Hartzell PL. 1995. Genetic suppression and phenotypic masking of a Myxococcus xanthus frzF- defect. Mol Microbiol 15:483–94. 10.1111/j.1365- 2958.1995.tb02262.x 43. Hoang Y, Kroos L. 2018. Ultrasensitive response of developing Myxococcus xanthus to the addition of nutrient medium correlates with the level of MrpC. J Bacteriol 200(22): e00456-18. 10.1128/JB.00456-18 44. Hanahan D. 1983. Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166(4):557-80. 10.1016/s0022-2836(83)80284-8 45. Kaiser D. 1979. Social gliding is correlated with the presence of pili in Myxococcus xanthus. Proc Natl Acad Sci U S A 76(11): 5952–6. 10.1073/pnas.76.11.5952 46. Rajagopalan R, Kroos L. 2017. The dev operon regulates the timing of sporulation during Myxococcus xanthus development. J Bacteriol 199(10):e00788-16. 10.1128/JB.00788-16 ! 151! CHAPTER 5: Segregation of sister chromosomes during the cellular shape change of developing Myxococcus xanthus Abstract Chromosomal organization is a critical feature in maintaining genetic integrity for the continuation of a species. Most studies of bacterial nucleoids have focused on rod-shaped or crescent-shaped bacteria (Escherichia coli, Bacillus subtilis and Caulobacter crescentus). Studying nucleoid dynamics during development of M. xanthus offers the unique opportunity to investigate localization of two sister chromosomes when a rod-shaped cell is converted into a round spore. During starvation-induced development, DNA replication is required for starving rods to progress, resulting in spores with two copies of the chromosome. In this study, we developed methods using confocal fluorescence microscopy to observe nucleoid localization in developing cells in situ. By staining cells with the membrane dye FM 4-64 and the DNA-binding dye DAPI, we found that the two sister chromosomes condensed and often segregated when rod-shaped cells transitioned to round spores. A similar result was observed when we tracked a fluorescent protein fused to a transcription factor that binds to numerous DNA sites during development. Segregated nucleoids were observed in some spores, but appeared to be decondensed in others. DNA replication is usually followed by chromosomal segregation and a compartmentalization event, such as cellular division or endospore formation. However, during M. xanthus development, DNA replication is followed by the entire rod-shaped cell converting into a round spore without obvious compartmentalization. Chromosomal condensation and segregation in a portion of the developing cells may be a bet-hedging strategy to facilitate rapid division after germination if nutrients become available prior to spore maturation and/or an evolutionary remnant of ancestral events that included cellular division to produce spores with one copy of the chromosome. ! 152! Introduction Spatial organization is a critical feature of all living systems. Although bacteria are the smallest form of cellular life, they still possess complex intracellular organization such as polar localization of chemoreceptors (1), protein oscillations (2), and nucleoid organization (3) (4) (5). In terms of nucleoid organization, the genomic DNA must be compacted several orders of magnitude since the bacterial chromosome is ~1,000 times longer than the micrometer-sized cell. The chromosomal organization is accomplished by both physical and biochemical factors such as DNA supercoiling (6) and nucleoid-associated proteins (7) (8). The organization of the two nucleoids after DNA replication facilitates segregation of chromosomes during cellular division. Proper chromosomal segregation is essential to preserve the integrity of the genetic material for the continuation of the species. With the advancement of fluorescence microscopy and live-cell imaging, our knowledge of nucleoid condensation and segregation has increased. However, most of the studies of bacterial nucleoids have focused on rod-shaped or crescent- shaped bacteria (E. coli (9), B. subtilis (10), and C. crescentus (11)) (12). In this work, we studied nucleoid dynamics during the multicellular development of M. xanthus, which offers the unique opportunity to investigate sister chromosome condensation and segregation as rod- shaped cells are converted into round spores. The starvation-induced development of M. xanthus requires the coordinated movements of ~105 cells to form a dome-shaped mound. Large populations can be induced to form many mounds of similar width and height. A majority of the population undergoes lysis during this process (13). Some rod-shaped cells persist outside of mounds as peripheral rods (14). The temporal and spatial distribution of cells transitioning from rods to spores inside mounds or so- called “nascent fruiting bodies” was described in Chapter 3. There are more transitioning cells and spores in the inner domain than in the outer domain of nascent fruiting bodies. The transitioning cells undergo peptidoglycan remodeling and adopt different shapes that may depend on their surroundings and/or non-uniformity in the remodeling process. During the ! 153! starvation-induced developmental process of M. xanthus, DNA replication is tightly regulated to ensure that all spores have two copies of the chromosome (15). During vegetative growth, cells contain one or two copies of the genome, but the chromosomal copy number drops to one when cells enter the stationary phase (15). As starving cells begin to form mounds, DNA replication is required for development to progress (16). Inhibition of DNA replication delays the developmental program, which can resume if the inhibitor is removed (17). If a DNA replication inhibitor is added after mounds have formed, there is no effect on developmental progression. Fruiting body-derived spores contain two copies of the chromosome with the origin and terminus regions localized to the periphery of the spore (15). However, it is unknown when and how nucleoid arrangement changes during the transition of rods to spores. M. xanthus can also form spores in response to certain chemicals. In contrast to starvation-induced spores, the chromosomal copy number varies in glycerol-induced spores (15). It appears that DNA replication is less tightly regulated during glycerol-induced sporulation. Cells appear to bypass the DNA replication checkpoint and proceed to cellular shape change. In this study, we investigated the nucleoid arrangement during both starvation- induced development and glycerol-induced sporulation. We devised three methods that rely on confocal fluorescence microscopy to investigate nucleoid dynamics of M. xanthus cells in situ. The chromosomal copy number was examined using a fluorescent repressor operator system (FROS) (18). Chromosomal condensation and segregation was tracked using the DNA-binding dye DAPI and using a fluorescent protein fused to a transcription factor that binds to numerous DNA sites. Our methods allowed us to simultaneously visualize cellular shape. We discovered that the two chromosomes condense and often segregate during the transition of rods to spores in nascent fruiting bodies, and during glycerol-induced sporulation of cells that had replicated the chromosome. ! 154! Results Visualization of two copies of the chromosome in developing cells inside nascent fruiting bodies. Starvation-induced spores contain two copies of the chromosome (15). In a previous study,!spores were dispersed from fruiting bodies and flow cytometry was used to investigate their DNA content (15). In this study, we used a FROS and confocal laser scanning microscopy to visualize the chromosomal copy number in developing cells inside nascent fruiting bodies. M. xanthus strains with ~240 tetO operator sequences inserted at specific chromosomal locations were described previously (18). In that study, TetR-YFP was produced under the control of a copper-inducible promoter and time-lapse fluorescence microscopy was used to investigate chromosomal arrangement and dynamics during vegetative growth. However, copper interferes with submerged culture development (Emily Titus and L. K., unpublished data). Therefore, we constructed strains to produce TetR-YFP under the control of an IPTG-inducible promoter in backgrounds with tetO sequences inserted at 33° and 270° from the origin of replication (oriC), resulting in M. xanthus strain YH3 and YH4, respectively. Upon IPTG addition, TetR-YFP would be produced and bind to tetO sequences, indicating the chromosomal copy number of cells. As a control, strain YH5 lacking tetO sequences but able to produce TetR-YFP upon IPTG addition, was also constructed. The strains were starved under submerged culture conditions that induce development. IPTG and the lipophilic dye FM 4-64 were added at the beginning of starvation. FM 4-64 stains the cellular membrane, allowing cellular shape to be visualized. Confocal images of the same nascent fruiting bodies were acquired at intervals poststarvation (PS). ! 155! Hours PS 24 30 42 Hours PS 24 30 42 YH3 (33o) YH4 (270o) YH3" YH4" D Hours PS 24 30 42 A B 1.2" C 3 1" 0.8" 0.6" 0.4" 0.2" ) m µ ( e c n a t s D 0" i 2 1 43" YH5 spores spores rods transitioning cells 0 rods transitioning cells Figure 5.1 Visualization of the chromosomal copy number during starvation-induced development. M. xanthus strains YH3, YH4, and YH5 were starved under submerged culture conditions. IPTG (1 mM) and FM 4-64 (5 µg/mL) were added at the start of starvation. Confocal images were acquired at the indicated times PS and merged to show both TetR-YFP (yellow) and FM 4-64 staining of the cellular membrane (red). A. Confocal images of the same nascent fruiting body during development. Images show an optical section near the base of a nascent fruiting body of strain YH3 at the indicated times. ! 156! Figure 5.1 (cont’d) Arrows indicate a rod-shaped cell at 24 h, a transitioning cell at 30 h, and a spore at 42 h. Bar, 20 µm. See Figure S5.1 for images of strains YH4 and YH5. B. Close-up images of representative cells. Higher magnification images from panel A and Figure S5.1A show cells with the predominant shape (rods at 24 h, transitioning cells at 30 h, spores at 42 h) at the indicated times. Arrows indicate TetR-YFP loci. Strains are indicated on the left with the position of the tetO sequences relative to oriC in parenthesis. Bar, 1 µm. C. Distances between TetR-YFP loci. For cells that exhibited two TetR-YFP loci, the distance between the loci was measured (rods at 24 h, transitioning cells at 30 h, spores at 42 h). The graph shows the data points and average for 20 cells from 10 nascent fruiting bodies, and error bars show one standard deviation. D. Close-up images of representative cells of the control strain lacking tetO sequences. Higher magnification images from Figure S5.1B show a spherical cell lacking yellow fluorescence and a rod-shaped cell exhibiting cytoplasmic yellow fluorescence at 24 h, and a transitioning cell and a spore with cytoplasmic yellow fluorescence at 30 and 42 h, respectively. Bar, 1 µm. Mounds formed by 24 h PS and cells were still rod-shaped in optical sections near the base of mounds for all three strains (Fig. 5.1A and S5.1). In the same mounds, cells in transition between rods and spores were observed at 30 h, primarily closer to the center of the nascent fruiting body. Farther from the center, rods and some transitioning cells are arranged in a circular pattern. The transitioning cells near the center and the rods farther from the center likely correspond to the inner and outer domains of nascent fruiting bodies described previously (19) (20). By 42 h, many cells have become round spores, in both domains. At low magnification, tiny yellow dots can be seen in the images of strains YH3 (Fig. 5.1A) and YH4 (Fig. S5.1A), as expected due to localized binding of TetR-YFP to tetO sequences. Also as expected, TetR-YFP loci were not observed in the images of strain YH5 (Fig. S5.1B), which lacks tetO sequences. Importantly, not all TetR-YFP loci are expected to be seen, since an optical section (0.5 µm) is as thin or thinner than cells (~0.5, 1, and 2 µm for rods, spores, and transitioning cells, respectively), and cells are in various orientations relative to the optical section. Figure 5.1B shows representative cells at higher magnification. Cells with the predominant shape at each time PS, and exhibiting one (left panels) or two (right panels) TetR- YFP loci are shown for strains YH3 and YH4 with tetO sequences at 33° and 270° from oriC. ! 157! For both strains at each time, most cells exhibited one TetR-YFP locus or none, presumably since not all TetR-YFP loci are seen, as explained above. Even so, by examining an optical section near the base of 10 different fruiting bodies for each strain at each time, at least 20 cells with the predominant shape and two TetR-YFP loci could be found. Although we realized that individual cells would have slightly different orientations relative to the optical section, we wanted to compare the distance between the two TetR-YFP loci in rods at 24 h, transitioning cells at 30 h, and spores at 42 h, for strains YH3 and YH4 with tetO sequences at different chromosomal locations. We reasoned that significant differences in the distance between the two TetR-YFP loci would suggest differences in the arrangement of sister chromosomes in developing cells. The distance between the two TetR-YFP loci in rods was greater than in transitioning cells for strain YH3 (p = 0.0187, 95% confidence level in paired two-tailed t-tests) (Fig. 5.1C). The distance in rods, on the other hand, was smaller than in transitioning cells for strain YH4 (p = 0.0027, 95% confidence level in paired two-tailed t-tests). For both strains, the distance in transitioning cells was greater than in spores (p = 0.02 for strain YH3 and p < 0.0001 for strain YH4, 95% confidence level in paired two-tailed t-tests with Welch’s correction). TetR- YFP distance in rods did not show a difference between two strains (p = 0.5411, 95% confidence level in two-tailed Student’s t-tests). Interestingly, the distance between TetR-YFP loci in transitioning cells spanned a wide range in both strains, suggesting that the loci move when cells are changing shape. On average, the distance between loci in transitioning cells was two-fold greater for strain YH4 with tetO sequences at 270° from oriC than for strain YH3 with tetO sequences at 33° (p = 0.0003, 95% confidence level in two-tailed Student’s t-tests). Similarly, the distance between loci in spores was significantly greater for strain YH4 than for strain YH3 (p = 0.0094, 95% confidence level in two-tailed Student’s t-tests). The differences in distance between loci for the two strains suggest that the sister chromosomes have a favored arrangement in developing cells. Taken together, the results suggest both that the arrangement ! 158! of sister chromosomes changes as developing cells undergo shape change and that the chromosomes have a favored arrangement. As expected, strain YH5 lacking tetO sequences failed to show TetR-YFP loci even at high magnification (Fig. 5.1D). Rather, fluorescence from TetR-YFP was evenly distributed in the cytoplasm of rods, transitioning cells, and spores. This control shows that the TetR-YFP loci observed in strains YH3 and YH4 are not an artifact caused by aggregation of the fluorescent protein. For all three strains, we noticed a small proportion of spherical cells with no fluorescence at 24 h PS (Fig. 5.1D shows a representative cell). Based on additional characterization presented below, the spherical cells appear to be in the process of lysing, which is known to be the fate of the majority of cells during development. We speculate that lysing cells cannot maintain the level or proper folding of YFP, hence the spherical cells exhibit no yellow fluorescence. Sister chromosomes appear to condense and segregate in transitioning cells. The FROS enabled us to visualize the location of the tetO sequences of sister chromosomes in some developing cells. To visualize the entire chromosomes, we used the DNA-binding dye DAPI. We found that DAPI interfered with M. xanthus development, so images had to be acquired from submerged cultures initiated in parallel but in separate wells of a microscope slide. FM 4-64 was added at the beginning of starvation to stain the cellular membrane, but DAPI was added 30 min prior to image acquisition. At low magnification, optical sections near the base of mounds showed mostly rods at 24 h PS, transitioning cells near the center and a mix of rods and transitioning cells farther from the center at 30 h, and mostly spores at 42 h (Fig. 5.2A), consistent with the results shown in Figure 5.1A. At higher magnification, images of representative cells show their cellular membrane (top panels), DNA (middle panels), and both (bottom panels) (Fig. 5.2B). At 24 h, the DNA occupied primarily the central portion of rods, consistent with localization of the DNA during vegetative growth (18). In spherical cells, which were much less numerous than rods, DNA appeared to be ! 159! evenly distributed in the cytoplasm (Fig. 5.2B), perhaps reflecting a loss of DNA condensation and/or integrity in lysing cells. At 30 h, transitioning cells exhibited DNA at one locus (left panel), in a crescent along one side of the cell (center panel), or at two loci (right panel). At 42 h, spores showed DNA at one locus (left panel), at two loci (center panel), or evenly distributed in the cytoplasm (right panel). As noted above for TetR-YFP loci, not all the DNA is expected to be seen, owing to the thinness of the optical sections. However, since entire chromosomes are much larger than the tetO sequence arrays, DNA loci were expected to be larger than TetR-YFP loci, and this was the case. Importantly, cells contain sufficient DNA to fill the cytoplasm of a suspected lysing cell or a spore, yet the DNA was localized in the majority of rods and transitioning cells, and in some spores, indicative of DNA condensation and/or exclusion. The two DNA loci observed in some transitioning cells and spores appeared to be well- separated, as if sister chromosomes were condensed and segregated. To quantify the separation, we examined the fluorescence intensity profiles of transitioning cells and spores that showed two DNA loci. For transitioning cells at 30 h PS, a line was drawn through the apparent centers of the two loci and was extended to the cellular membrane at each end (Fig. 5.2C). The graph shows the normalized intensity of fluorescence from the DNA and membrane stains plotted versus the normalized distance along the line drawn, for a total of 10 cells. On average, the DNA fluorescence intensity reached maxima of ~0.8 at distances of 0.3 and 0.7, indicating that the apparent centers of the two DNA loci were separated by ~40% of the distance between the cellular membrane at each end of the line. Importantly, the DNA fluorescence intensity reached ~0.5 at a distance of 0.5, midway between the maxima and similar in intensity to the minima of ~0.4 observed at each end of the line. We conclude that the two DNA loci are well- separated in some transitioning cells at 30 h, consistent with the notion that sister chromosomes are condensed and segregated. ! 160! Hours PS 24 30 42 A Hours PS 24 30 42 B C D 1.2" 1" 0.8" 0.6" 0.4" 0.2" 0" 1.2" 1" 0.8" 0.6" 0.4" 0.2" 0" y t i s n e t n i t n e c s e r o u l f d e z i l a m r o N y t i s n e t n i t n e c s e r o u l f d e z i l a m r o N DAPI FM 4-64 0" Normalized distance between membranes 0.8" 0.6" 0.2" 0.4" DAPI" FM"4/64" 0.2" 0" Normalized distance between membranes 0.8" 0.6" 0.4" 1" E 1" Figure 5.2 Visualization of DNA by DAPI staining during development. M. xanthus wild- type strain DK1622 was starved under submerged culture conditions. FM 4-64 (5 µg/mL) was added at the start of starvation. DAPI (10 µg/mL) was added 30 min before imaging. Confocal images were acquired at the indicated times PS to show FM 4-64 staining of the cellular membrane (red) and DAPI staining of DNA (blue). ! 161! Figure 5.2 (Cont”d) A. Confocal images of nascent fruiting bodies during development. Images show an optical section near the base of different nascent fruiting bodies, with the red and blue channels merged. Arrows indicate a rod-shaped cell at 24 h, a transitioning cell at 30 h, and a spore at 42 h. Bar, 20 µm. B. Close-up images of representative cells. Higher magnification images from panel A show cells with the predominant shape (rods at 24 h, transitioning cells at 30 h, spores at 42 h) and a spherical cell at 24 h (leftmost panels). Rows show red (top), blue (middle), and merged (bottom) channels. Arrows point to intense DNA fluorescence at one or two locations in an individual cell. Bar, 1 µm. C. Fluorescence intensity profile of transitioning cells with two DNA loci. For each of 10 transitioning cells at 30 h, a line was drawn through the apparent centers of the two DNA loci and was extended to the cellular membrane at each end, as shown in the image with red and blue channels merged. The line length was normalized to 1. The fluorescence intensity of each channel was measured from 0 to 1 in increments of 0.04 along the line and normalized to its maximum, which was assigned a value of 1. Lines show the average normalized fluorescence intensity of DAPI staining of DNA (blue) and FM 4-64 staining of the cell membrane (red). Error bars indicate one standard deviation. D. Fluorescence intensity profile of spores with two DNA loci. For each of 10 spores at 42 h, a line was drawn along the cellular periphery, starting at one DNA locus where it was closest to the other, as shown in the image with red and blue channels merged. The line length was normalized to 1 and the fluorescence intensities were measured, normalized, and graphed as in panel C. E. Close-up image of representative spores stained with FM 4-64 but not DAPI. A high magnification image was acquired at 42 h in the red channel. Arrows point to intense membrane fluorescence at two locations in an individual cell. Bar, 1 µm. To quantify the separation of DNA loci in spores at 42 h PS, a different approach was required, because the DNA loci were at the cellular periphery and there was little membrane staining there. Therefore, a line was drawn along the cellular periphery, starting at one DNA locus where it was closest to the other (Fig. 5.2D). On average, the DNA fluorescence intensity reached maxima of ~0.8 at distances of 0 and 0.7, indicating that the centers of the two DNA loci were separated by ~70% of the distance around the cellular periphery in the longer direction (~30% in the shorter direction). The two DNA loci were well-separated, as the fluorescence intensity reached minima of ~0.4 between the maxima. Interestingly, membrane fluorescence intensity was out-of-phase with DNA fluorescence intensity. This apparent localization of membrane staining was not due to the DNA stain, since localized membrane staining was observed in the majority of spores at 42 h without DAPI addition (Fig. 5.2E). The out-of-phase nature of the DNA and membrane fluorescence intensities (Fig. 5.2D) suggests that DNA ! 162! association with membrane interferes with FM 4-64 staining. Alternatively or in addition, areas of intense membrane staining may reflect excess accumulation of membrane material in areas where DNA is not associated with the membrane. In any case, our results confirm and extend prior work on fruiting body spores harvested at 5 days PS, which showed that the origin and terminus regions of both chromosomes are localized to the periphery (15). Nucleoids appear to condense and segregate in transitioning cells. DAPI staining of DNA provided a simple method to visualize the chromosomes in developing cells. However, because DAPI stained the DNA of all the cells, which were packed inside nascent fruiting bodies, it was challenging to examine individual cells, particularly in three dimensions (i.e. in z- stacks of optical sections). Moreover, DAPI interferes with development, so the same nascent fruiting bodies could not be observed over time. Therefore, we engineered strain YH6, which produces a translational fusion of a very bright fluorescent protein, mNeonGreen, to FruA, a transcription factor produced during development. FruA has been shown to bind cooperatively with MrpC to numerous DNA sites (21). We reasoned that mNeonGreen-FruA might provide a method to visualize nucleoids in developing cells. Strain YH6 was engineered to produce mNeonGreen-FruA under control of the native fruA promoter at the native chromosomal location, so native FruA would not be synthesized. We found that strain YH6 was delayed for development by ~6 h compared to wild-type strain DK1622 (Fig. S5.2A). Immunoblot analysis with FruA antibodies detected a protein of the expected size (51 kDa) for mNeonGreen-FruA at 30 h PS (Fig. 5.3A). A less abundant species matching native FruA in size was also detected in strain YH6. This species may be a proteolytic fragment of the fusion protein resulting from cleavage near the junction between mNeonGreen and FruA. Neither mNeonGreen-FruA nor the species matching FruA in size was as abundant as native FruA in strain DK1622. The lower abundance may account for the delayed development of strain YH6. Alternatively or in addition, mNeonGreen-FruA and/or the species matching FruA in size may not be fully functional. Nevertheless, considerable mNeonGreen- ! 163! FruA accumulates, so if it binds to numerous DNA sites, it would allow nucleoids to be visualized. To facilitate the potential observation of nucleoids in strain YH6, it was mixed with strain DK1622 at a ratio of approximately 1 YH6 cell to 3 DK1622 cells (1:3). The 1:3 ratio would provide separation between YH6 cells in which nucleoids might fluoresce and DK1622 cells in which nucleoids would not fluoresce. In this experiment, FM 4-64 was added at the beginning of starvation in order to stain the membrane of all cells. Like strain YH6 alone (Fig. S5.2A), development of the 1:3 mixture was delayed by ~6 h compared to strain DK1622 alone, so confocal images of the same nascent fruiting bodies were acquired at 30, 36, and 48 h in order to match the stages of development shown in Figures 5.1 and 5.2. At low magnification, optical sections near the base of mounds showed mostly rods at 30 h, transitioning cells near the center and a mix of rods and transitioning cells farther from the center at 36 h, and mostly spores at 48 h (Fig. 5.3B), consistent with the results shown in Figures 5.1A and 5.2A, except in those experiments a similar stage of development was reached 6 h earlier. Importantly, it appeared that mNeonGreen-FruA allowed nucleoids to be visualized (Fig. 5.3B), with similar localization as DAPI-stained chromosomes (Fig. 5.2A) At higher magnification, images of representative cells show their cellular membrane (top panels), mNeonGreen-FruA fluorescence (middle panels), and both (bottom panels) (Fig. 5.3C). At 30 h PS, mNeonGreen-FruA fluorescence occupied primarily the central portion of rods, consistent with localization of the DNA at 24 h (Fig. 5.2B) and during vegetative growth (18). In spherical cells at 30 h, which were much less numerous than rods, no mNeonGreen- FruA fluorescence was observed (Fig. 5.3C). The absence of mNeonGreen-FruA fluorescence was consistent with the absence of TetR-YFP fluorescence in spherical cells at 24 h (Fig. 5.1D). As noted above, spherical cells may be in the process of lysing and therefore unable maintain the level or proper folding of fluorescent proteins. To test whether mNeonGreen-FruA fluorescence would be observed in a larger number of spherical cells, strain YH6 alone was ! 164! subjected to submerged culture. Still no mNeonGreen-FruA fluorescence was observed in any of the spherical cells (Fig. 5.3D). Transitioning cells at 36 h exhibited mNeonGreen-FruA fluorescence at one locus (left panel), in a crescent along one side of the cell (center panel), or at two loci (right panel) (Fig. 5.3C). At 48 h, spores showed mNeonGreen-FruA fluorescence at one locus (left panel), at two loci (center panel), or evenly distributed in the cytoplasm (right panel). The patterns of mNeonGreen-FruA fluorescence are very similar to the patterns of DAPI fluorescence at a similar stage of development ~6 h earlier (Fig. 5.2B), suggesting that both allow visualization of nucleoids, which appear to condense and segregate in transitioning cells. As a control, we engineered strain YH7 to ectopically produce mNeonGreen under the control of a vanillate-inducible promoter and subjected the strain to submerged culture development in the presence of vanillate. The timing of development was normal for strain YH7 (Fig. S5.2B), presumably because FruA is produced normally and mNeonGreen does not interfere with development. Fluorescence from mNeonGreen was evenly distributed in the cytoplasm of strain YH7 rods, transitioning cells, and spores (Fig. 5.3E), in contrast to the localized mNeonGreen-FruA fluorescence observed for strain YH6 (Fig. 5.3C). Hence, the FruA portion of the fusion protein is required for localization, which we infer reflects binding to numerous DNA sites, owing to the similar patterns of DAPI fluorescence observed (Fig. 5.2B). Spherical cells of strain YH7 at 24 h show no fluorescence (Fig. 5.3E), consistent with the notion that these cells may be lysing and unable to maintain the level or proper folding of fluorescent proteins. ! 165! Most spores of strains YH6 and YH7 showed two arcs of intense FM 4-64 fluorescence at opposite sides of their periphery (Fig. S5.2, 5.3B, 5.3C, and 5.3E), consistent with spores stained with FM 4-64 alone (Fig. 5.2E) or with FM 4-64 and DAPI (Fig. 5.2A, 5.2B, and 5.2D). The correlations between fluorescence from mNeonGreen-FruA, DAPI, and FM 4-64 in the same cells will be described below. Figure 5.3 Visualization of nucleoids using mNeonGreen-FruA during development. M. xanthus strains were starved under submerged culture conditions. FM 4-64 (5 µg/mL) was added at the start of starvation. Samples were collected for immunoblot analysis or confocal images were acquired at the indicated times PS. Confocal images show FM 4-64 staining of the cellular membrane (red) and mNeonGreen-FruA fluorescence (green). ! 166! Figure 5.3 (cont’d) A. Immunoblot analysis. Samples of wild-type strain DK1622 and strain YH6 were collected at 30 h and equal volumes analyzed by immunoblot with FruA antibodies. The blue arrow indicates a protein of the expected size for mNeonGreen-FruA and the black arrow indicates the expected size for native FruA. The position of migration of molecular weight markers of the indicated size in kDa is shown on the left. B. Confocal images of the same nascent fruiting body during development. Images show an optical section near the base of a nascent fruiting body of a 1:3 mixture of strains YH6 and DK1622, with the red and green channels merged. Arrows indicate a rod-shaped cell at 30 h, a transitioning cell at 36 h, and a spore at 48 h. Bar, 20 µm. C. Close-up images of representative cells. Higher magnification images from panel B show cells with the predominant shape (rods at 30 h, transitioning cells at 36 h, spores at 48 h) and a spherical cell at 30 h (leftmost panels). Rows show red (top), green (middle), and merged (bottom) channels. Arrows point to intense mNeonGreen-FruA fluorescence at one or two locations in an individual cell. Bar, 1 µm. D. Confocal image of a mound of strain YH6 alone. The image shows an optical section near the base of a mound of strain YH6 at 30 h, with the red and green channels merged. The arrow indicates a group of spherical cells, all lacking mNeonGreen-FruA fluorescence. Bar, 5 µm. Lower magnification images are shown in Fig. S5.2A. E. Close-up images of representative cells engineered to produce mNeonGreen. Strain YH7 was starved under submerged culture conditions. FM 4-64 (5 µg/mL) and vanillate (0.5 mM) were added at the start of starvation. Higher magnification images from Fig. S2B show cells with the predominant shape (rods at 24 h, transitioning cells at 30 h, spores at 42 h) and a spherical cell at 24 h (leftmost panel), with the red and green channels merged. Bar, 1 µm. Correlations between fluorescence in the same cells. To examine fluorescence from mNeonGreen-FruA, DAPI, and FM 4-64 in the same cells during development, strain YH6 was subjected to submerged culture conditions. FM 4-64 was added at the start of starvation and DAPI was added 30 min before imaging. At low magnification, optical sections near the base of mounds showed mostly rods at 30 h, transitioning cells near the center and a mix of rods and transitioning cells farther from the center at 36 h, and mostly spores at 48 h (Fig. S5.2C), consistent with the results shown in Figure S5.2A for strain YH6 without DAPI staining. At higher magnification, images of representative cells show fluorescence from FM 4-64 and DAPI (top panels), FM 4-64 and mNeonGreen-FruA (middle panels), and all three channels (bottom panels) (Fig. 5.4A). At 30 h PS, fluorescence from DAPI and mNeonGreen-FruA overlapped primarily in the central portion of rods. At 36 and 48 h, when cells were predominantly transitioning and spores, respectively, fluorescence from DAPI and mNeonGreen-FruA overlapped in patterns consistent with those described for the experiments ! 167! presented in Figures 5.2 (except 6 h earlier PS) and 5.3. Figure 5.4A shows a transitioning cell at 36 h and a spore at 48 h in which two loci of fluorescence from DAPI and mNeonGreen-FruA overlapped. The overlapping fluorescence strongly supports the interpretation that both DAPI staining of DNA and mNeonGreen-FruA binding to numerous sites in DNA allow visualization of nucleoids, which appear to condense and segregate in transitioning cells. As for the experiment presented in Figure 5.2, we focused on transitioning cells and spores in which two loci of fluorescence from DAPI and mNeonGreen-FruA were observed in the experiment presented in Figure 5.4, in order to quantify the separation between nucleoids and the correlations between fluorescence in all three channels in the same cells. For lines drawn through the apparent centers of the two loci in transitioning cells, the normalized fluorescence intensity profiles from DAPI and mNeonGreen-FruA were similar, with the latter showing greater maxima and lesser minima on average (Fig. 5.4B). We conclude that two nucleoids are well-separated in some transitioning cells, strengthening support for the notion that sister chromosomes are condensed and segregated. For lines drawn along the periphery of spores, starting at one locus where it was closest to the other, the normalized fluorescence intensity profiles from DAPI and mNeonGreen-FruA were similar, but in this case the DAPI signal showed greater maxima on average (Fig. 5.4C). Both signals were out-of-phase with the FM 4-64 signal, consistent with the out-of-phase nature of the DAPI and FM 4-64 signals in the experiment presented in Figure 5.2D, and suggesting that nucleoid association with membrane interferes with FM 4-64 staining and/or excess accumulation of membrane material. ! 168! A B C Hours PS 30 36 48 DAPI mNeonGreen-FruA Merged 1.2" 1" 0.8" 0.6" 0.4" 0.2" 0" 0" 1.2" 1" 0.8" 0.6" 0.4" 0.2" 0" 0" y t i s n e t n i t n e c s e r o u l f d e z i l a m r o N y t i s n e n t i t n e c s e r o u l f d e z i l a m r o N DAPI" mNeonGreen4FruA" FM"4464" 0.2" 0.8" Normalized distance between membranes 0.4" 0.6" 0.2" Normalized distance between membranes 0.4" 0.6" 0.8" 1" 1" Figure 5.4 Visualization of fluorescence from mNeonGreen-FruA, DAPI, and FM 4-64 in the same cells during development. M. xanthus strain YH6 was starved under submerged culture conditions. FM 4-64 (5 µg/mL) was added at the start of starvation. DAPI (10 µg/mL) was added 30 min before imaging. Confocal images were acquired at the indicated times PS to show FM 4-64 staining of the cellular membrane (red), DAPI staining of DNA (blue), and mNeonGreen-FruA fluorescence (green). ! 169! Figure 5.4 (cont’d) A. Close-up images of representative cells. Higher magnification images from Fig. S5.2C show cells with the predominant shape (rods at 30 h, transitioning cells at 36 h, spores at 48 h). Rows show red and blue channels merged (top), red and green channels merged (middle), and all three channels merged (bottom). Arrows point to intense fluorescence from DAPI and/or mNeonGreen-FruA at one or two locations in an individual cell. Bar, 1 µm. B. Fluorescence intensity profile of transitioning cells with two loci of fluorescence from DAPI and mNeonGreen-FruA. For each of 10 transitioning cells at 36 h, a line was drawn through the apparent centers of the two loci and was extended to the cell membrane at each end, as shown in the image with all three channels merged. The line length was normalized to 1. The fluorescence intensity of each channel was measured from 0 to 1 in increments of 0.04 along the line and normalized to its maximum, which was assigned a value of 1. Lines show the average normalized fluorescence intensity of DAPI staining of DNA (blue), mNeonGreen-FruA (green), and FM 4-64 staining of the cellular membrane (red). Error bars indicate one standard deviation. C. Fluorescence intensity profile of spores with two loci of fluorescence from DAPI and mNeonGreen-FruA. For each of 10 spores at 48 h, a line was drawn along the cellular periphery, starting at one locus where it was closest to the other, as shown in the image with all three channels merged. The line length was normalized to 1 and the fluorescence intensities were measured, normalized, and graphed as in panel B. Nucleoids condense and segregate in at least ~40% of transitioning cells. Because FM 4-64 stained the membrane of all the cells, which were packed inside nascent fruiting bodies, it was challenging to examine individual cells, particularly in three dimensions (i.e. in z- stacks of optical sections). Therefore, we engineered cells to ectopically produce tdTomato under the control of a vanillate-inducible promoter, reasoning that red fluorescence from tdTomato would be evenly distributed in the cytoplasm of cells and allow their shape to be examined. To visualize nucleoids in the same cells, a derivative of strain YH6 that produces mNeonGreen-FruA as described above, was engineered to produce tdTomato upon addition of vanillate. The resulting strain, YH9, was mixed with wild-type strain DK1622 in a 1:3 ratio and subjected to submerged culture development with vanillate added at the start of starvation. The 1:3 ratio would provide separation between YH9 cells from which fluorescence was expected and DK1622 cells, which would not fluoresce. Like strain YH9 alone (Fig. S5.2D), development of the 1:3 mixture was delayed by ~6 h compared to strain DK1622 alone, so confocal images of the same nascent fruiting bodies were acquired at 30, 36, and 48 h PS. At low magnification, optical sections near the base of mounds showed mostly rods at 30 h, transitioning cells near ! 170! the center and a mix of rods and transitioning cells farther from the center at 36 h, and mostly spores at 48 h (Fig. 5.5A). At 36 and 48 h, z-stacks of optical sections from near the base of nascent fruiting bodies to 5 µm up were collected. Resolution decreased in optical sections farther than 5 µm up. A preliminary analysis of YH9 cells in the z-stacks is presented below. Quantitative, three-dimensional analysis of nucleoids and cellular shape is ongoing. A Hours PS 30 36 48 Hours PS 36 48 B tdTomato mNeonGreen-FruA merged C i c o l o w t h t i w n o i t r o p o r P 0.6 0.8 0.6 0.5 0.4 0.4 0.2 0.3 0.0 transitioning cells spores transitioning cells spores Figure 5.5 Quantification of segregated nucleoids. M. xanthus strain YH9 was mixed with wild-type strain DK1622 at a ratio of 1:3 and the mixture was subjected to starvation under submerged culture conditions. Vanillate (0.5 mM) was added at the start of starvation. Confocal images were acquired at the indicated times PS to show tdTomato (red) and mNeonGreen-FruA (green) fluorescence of strain YH9 cells. ! 171! Figure 5.5 (cont’d) A. Confocal images of the same nascent fruiting body during development. Images show an optical section near the base of a nascent fruiting body, with the red and green channels merged. Arrows indicate a rod-shaped cell at 30 h, a transitioning cell at 36 h, and a spore at 48 h. Bar, 20 µm. B. Close-up images of representative cells in maximum intensity projections of z-stacks. Cells with the predominant shape (transitioning cells at 36 h, spores at 48 h) that did not appear to overlap with any other cell in the z-stack projection are shown at high magnification. Rows show red (top), green (middle), and merged (bottom) channels. Arrows point to intense mNeonGreen-FruA fluorescence at one or two locations in an individual cell. Bar, 1 µm. C. Proportion of cells with two loci of mNeonGreen-FruA fluorescence. Cells that did not appear to overlap with any other cell in the z-stack projection were classified as having one locus (condensed, cresent-shaped, or decondensed) or two loci of nucleoid-associated mNeonGreen-FruA fluorescence. Graphs show the data points from five nascent fruiting bodies at 36 h (total of 109 transitioning cells classified) or at 48 h (total of 82 spores classified) and error bars show one standard deviation. Each z-stack was projected onto a plane orthogonal to the z-axis to create a maximum intensity projection. Only the voxels with maximum intensity along the z-axis are projected onto the plane, resulting in a partial representation of the z-stack in two dimensions. Cells that appeared to overlap with other cells were ignored. Among the remaining cells, images of representative cells at high magnification show fluorescence from tdTomato (top panels), mNeonGreen-FruA (middle panels), and both channels (bottom panels) (Fig. 5.5B). As expected, tdTomato fluorescence appeared to be evenly distributed in the cytoplasm, allowing cellular shape to be determined. Transitioning cells at 36 h exhibited mNeonGreen-FruA fluorescence at one locus (left panel), in a crescent along one side of the cell (center panel), or at two loci (right panel), consistent with results observed for strain YH6 (Fig. 5.3C). At 48 h, spores showed mNeonGreen-FruA fluorescence at two loci (left panel) or evenly distributed in the cytoplasm (right panel) (Fig. 5.5B). Spores with mNeonGreen-FruA fluorescence at one locus were not observed, in contrast to the results for strain YH6 (Fig. 5.3C). However, in the case of strain YH6, only one optical section was examined, so a second locus might not have been seen, owing to the thinness of the optical section. Conceivably, a second locus could also be missed in a maximum intensity projection of a z-stack, but this should occur less often, especially if condensed and segregated nucleoids are well-separated, as our data suggest (Fig. ! 172! 5.2 and 5.4). Therefore, we determined the proportions of transitioning cells and spores with two loci of mNeonGreen-FruA fluorescence in the maximum intensity projections of z-stacks of five nascent fruiting bodies at 36 and 48 h, respectively. For both transitioning cells and spores, ~40% of the cells exhibited two loci (Fig. 5.5C). The proportion of cells exhibited two loci in spores is greater on average than in transitioning cells, though not statistically significant at the 95% confidence level (p = 0.08, paired two-tailed t-test). The result suggests that segregation occurs in transitioning cells and is preserved in spores. Most of the other transitioning cells had a crescent-shaped nucleoid (~65%) and the other ~60% of spores had decondensed nucleoids. This preliminary analysis is a minimum estimate of the percentage of cells with two loci for the reason mentioned above. Based on this analysis, it appears that nucleoids condense and segregate in at least ~40% of transitioning cells at 36 h, and are condensed and segregated in a similar percentage of spores at 48 h. Sister chromosomes appear to condense and segregate in some glycerol-induced spores. M. xanthus can form spores not only in fruiting bodies during starvation, but also without starvation or fruiting body formation, when growing cells are treated with certain chemicals, including glycerol. A previous study of glycerol-induced spores showed that their chromosomal copy number varies from one to two or even more (15). Here, strains YH3 and YH4 bearing the FROS to visualize chromosomal copy number, and the wild-type strain DK1622 stained with DAPI to visualize DNA, were stained with FM 4-64 to visualize the cellular membrane during glycerol-induced sporulation. Growing cells had one or two copies of the chromosome, as indicated by the number of TetR-YFP loci in cells of strains YH3 and YH4 prior to glycerol addition (Fig. 5.6). At 1 h after glycerol addition, cells in transition from rods to spores were observed. Transitioning cells with one or two TetR-YFP loci (strains YH3 and YH4) or DAPI-stained DNA loci (strain DK1622) were observed. Likewise, spores observed at 3 h after glycerol addition had one or two TetR- YFP loci or DAPI-stained DNA loci. These results show that some glycerol-induced ! 173! transitioning cells and spores contain two copies of the chromosome, which appear to condense and segregate as during starvation-induced development. We did not observe any cells with more than two TetR-YFP loci or DAPI-stained DNA loci. Time (h) 0 1 3 0 1 3 0 1 3 YH3 (33o) YH4 (270o) DK1622 Figure 5.6 Visualization of the chromosomal copy number, DAPI-stained DNA loci, and cellular shape during glycerol-induced sporulation. To M. xanthus strains YH3, YH4, and DK1622 growing in nutrient medium, FM 4-64 (5 µg/mL) was added. IPTG (1 mM) was also added to the cultures of strains YH3 and YH4 (with tetO sequences at the position relative to oriC indicated in parenthesis) to induce production of TetR-YFP. Samples were taken before the addition of glycerol (0.5 M) (designated 0 h) and at 1 and 3 h post-addition. DAPI (10 µg/mL) was added to the DK1622 samples 30 min before imaging. Confocal images were acquired to show FM 4-64 staining of the cellular membrane (red), TetR-YFP loci (yellow) in strains YH3 and YH4, and DAPI staining of DNA (blue) in strain DK1622. High magnification images show cells with the predominant shape (rods at 0 h, transitioning cells at 1 h, spores at 3 h), with the red channel merged with the yellow or blue channel. Rows show cells with one (top) or two (bottom) TetR-YFP or DNA loci, as indicated by arrows. Bar, 1 µm. ! Discussion In this study, we have developed methods to observe nucleoid localization in developing cells in situ using confocal microscopy. We found that sister chromosomes condense and often segregate when rod-shaped cells transition to round spores during both starvation- and glycerol- induced development. Staining cells with the DNA-binding dye DAPI and producing the DNA- binding transcription factor mNeonGreen-FruA showed consistent results for nucleoid localization. During starvation-induced development, our preliminary analysis suggests that nucleoids condense and segregate in at least ~40% of transitioning cells at 36 h PS. Most of the other transitioning cells had a crescent-shaped nucleoid that may be indicative of ongoing ! 174! segregation. At 48 h, the nucleoids appeared to have decondensed in the majority of spores, but remained condensed and segregated in at least ~40% of the spores. The potential mechanism of chromosome segregation during cellular shape change. M. xanthus uses a parABS system for chromosome segregation in vegetative cells (18). ParB proteins bind to parS sequences in chromosomes (18). ParA proteins are P-loop ATPases that interact with the ParB/parS complex (22). It is likely that chromosome segregation during M. xanthus developmental shape change is also mediated by its parABS system. To test this hypothesis, parA, parB, and parAB conditional mutant strains would need to be created, because ParA and ParB are essential for cell viability in M. xanthus (18) (22). Specifically, strains engineered to produce ParA and/or ParB under control of vanillate-inducible promoters would be subjected to vanillate withdrawal during development just prior to the time that many rods begin transitioning to spores. We predict that chromosome segregation would be inhibited upon depletion of ParA or ParB. If so, it would be interesting to determine whether spores are formed and to test their resistance and germination characteristics. The ability to monitor ParA and ParB depletion by immunoblot would be important for the interpretation of these experiments, and antibodies to these proteins have been reported (18). If ParA and ParB are important for chromosome segregation in transitioning cells, it would be of interest to investigate localization of these proteins during development. In vegetative cells, ParA clusters at the poles and this localization pattern depends on ParB (18) (22). ParB inhibits the nonspecific interaction of ParA with DNA, and ParA co-localizes with chromosomal DNA only when ParB is depleted. For these studies, fusions of ParA and ParB to mCherry and YFP, respectively, were created, but neither fusion protein was functional, so merodiploid strains that also produced the native proteins were used (18). We plan to request the merodiploid strains for initial studies of ParA-mCherry and ParB-YFP localization in developing cells. We predict that both fusion proteins will localize to the poles of rods and/or transitioning cells at about the time cellular shape change begins. ! 175! It is also of interest to investigate the orientation of sister chromosomes in developing cells. Our results suggest that the arrangement of the two chromosomes changes as cells undergo shape change and that the chromosomes have a favored arrangement (Fig. 5.1C). It has been shown that during DNA replication in vegetative cells, one ori region remains in the original subpolar region, while the second copy segregates unidirectionally to the opposite subpolar region, followed by the rest of the chromosomes (18). At the same time, the ter region of the mother chromosome relocates to midcell. After completion of replication and segregation, the two chromosomes show an ori-ter-ter-ori arrangement with mirror symmetry about a transverse axis at midcell. Many of these conclusions relied on interpretation of time-lapse movies of individual cells. Future work aimed at devising methods to track individual cells during development is essential to determine whether replication and segregation of chromosomes in developing cells is similar as in vegetative cells. We have found that it is challenging to make time-lapse movies and track individual rods as they move into mounds and transition into spores. Although transitioning cells are not motile, they can still be moved out of the focal plane by rods. To solve this problem, z-stacks need to be collected. However, our preliminary efforts have shown that collecting z-stacks too often using confocal laser scanning microscopy leads to photobleaching and developmental defects. Methods involving lower laser intensity or other approaches such as light-sheet microscopy (23) will need to be explored. Nucleoid segregation as a bet-hedging strategy or an evolutionary remnant. A previous study has shown that when cells starving on an agar surface begin to form mounds, DNA replication is required for development to progress (16). Our study employed submerged culture conditions, and we observed mounds containing a large number of rod-shaped cells with only one apparent TetR-YFP locus (Fig. 5.1), DAPI-stained DNA locus (Fig. 5.2), or mNeonGreen-FruA locus (Fig. 5.3). Although a second locus might have been missed owing to the thinness of the optical sections examined, it is also possible that the different developmental conditions impact DNA replication. For example, under submerged culture conditions, DNA ! 176! replication may occur later than on starvation agar, perhaps even in transitioning cells rather than in rods. It appears that a more definitive analysis of nucleoid localization and cellular shape change will be possible using the approach described for the experiment shown in Figure 5.5. A quantitative, three-dimensional analysis of nucleoids and cellular shape visualized with mNeonGreen-FruA and tdTomato, respectively, in the z-stacks collected during that experiment is ongoing in collaboration with Joshua Franklin and Yann Dufour here at Michigan State University. We cannot rule out the possibility that DNA replication fails to occur in some cells under submerged culture conditions; however, our preliminary analysis of maximum intensity projections of z-stacks suggests that ~60% of transitioning cells have a crescent-shaped nucleoid that may be indicative of ongoing segregation of chromosomes during or after replication, and at least ~40% of transitioning cells appear to have completed replication and segregation (Fig. 5.5C). Analysis of spores purified from 5-day-old fruiting bodies formed on starvation agar showed that at least ~70% have two copies of the chromosome (15). Flow- cytometric profiles of these spores were indistinguishable from profiles of spores formed under submerged culture conditions. It was proposed that two chromosomes in mature spores enhance their survival and ability to germinate in the soil environment. Our results raise the question of why nucleoids condense and segregate in developing cells. Chromosomal condensation typically hinders gene expression, which could be important for conserving cellular resources during sporulation. As mentioned above, it remains to be tested whether ParA and ParB mediate chromosomal segregation in transitioning cells and if so, how loss of segregation impacts spore resistance and germination. Chromosomal condensation and segregation is crucial during cellular growth and division, as a mechanism to ensure that each progeny cell has one copy of the chromosome. During Bacillus subtilis endospore formation in response to starvation, chromosomal translocation across an asymmetrically-positioned division septum distributes a copy of the chromosome to the forespore (24). In both of these cases, chromosomal segregation is ! 177! accompanied by a compartmentalization event. However, during M. xanthus development, the whole rod-shaped cell is converted into a round spore. There is no compartmentalization after chromosomal segregation. One possible explanation for nucleoid condensation and segregation during M. xanthus development is that it is a bet-hedging strategy to facilitate rapid division after germination if nutrients become available prior to spore maturation. Nutrient addition to developing cells has provided evidence for commitment to spore formation at about the time that cellular shape change begins (25) (26). In one of these studies, there were indications that some newly formed spores may germinate in response to nutrients (26). As spores mature, their nucleoids may decondense. The nucleoids appeared to have decondensed in the majority of spores by 48 h PS (Fig. 5.5). Under submerged culture conditions on a glass surface, very little or no nucleoid condensation was observed by 96 h (Fig. S5.3). Perhaps the quiescent state of spores makes it difficult to maintain condensation of nucleoids. Whether mature spores germinate and divide more slowly than newly formed spores is an open question. Another possible explanation for nucleoid condensation and segregation during M. xanthus development is that it is an evolutionary remnant of ancestral events that included cellular division to produce spores with one copy of the chromosome. Like M. xanthus, its putative ancestor may have coupled DNA replication with nucleoid condensation and segregation tightly upon starvation, but the ancestor may have differed from M. xanthus by undergoing cellular division before or during spore formation. To our knowledge, the chromosomal copy number in spores of myxobacteria other than M. xanthus has not been determined. Interestingly, many glycerol-induced spores of M. xanthus appear to contain one copy of the chromosome, based on flow-cytometric analysis (15) and on our analyses using the FROS and DNA visualization with DAPI (Fig. 5.6). However, cellular division after glycerol addition is unlikely to account for spores with one copy of the chromosome, since spores form quickly (e.g. transitioning cells are observed at 1 h post-addition in Fig. 5.6). Rather, it appears ! 178! that coupling between DNA replication and spore formation is weaker after glycerol addition than during starvation-induced development (15). If nucleoid condensation and segregation during M. xanthus development is an ancient trait, perhaps spores with two copies of the chromosome proved to be advantageous for survival and ability to germinate in the soil environment as has been proposed (15), resulting in evolutionary loss of cellular division during starvation-induced development. Spherical cells may be in the process of lysing. Several of our experiments revealed a small number of spherical cells that were distinct from transitioning cells or spores. Under our submerged culture conditions of development, mounds form at ~18 h PS, and by this time, the number of rods remaining is only ~30% of the initial number (i.e. 70% have undergone lysis) (26) (27). Although many cells undergo lysis, they may not persist in the spherical state for long. If spherical cells are short-lived, it would explain their small number compared to the number of rods remaining at a given time. The spherical cells are unlikely to be spore precursors, since sonication-resistant spores are not detected until 27 h (27), yet we have observed spherical cells at 18 h (data not shown). Moreover, we have shown that rods transitioning to spores are elliptical, not spherical, and that both transitioning cells and spores differ from spherical cells in terms of their ability to exhibit fluorescence from YFP (Fig. 5.1C) or mNeonGreen (Fig. 5.3). The lack of fluorescence from spherical cells suggests they are unable to maintain the level or proper folding of fluorescent proteins. Spherical cells also did not exhibit DNA localization, as did the majority of rods and transitioning cells, and many spores (Fig. 5.2). The lack of DNA localization in spherical cells suggests an inability to maintain DNA condensation and/or integrity. Altogether, the characteristics of spherical cells suggest they are undergoing lysis. Dye permeability provides another approach to assess cell viability. For example, the LIVE/DEAD BacLight staining method tests cell permeability to dyes (28), which is often interpreted as “dead” (red) or “live” (green) (29). This method has been used previously in studies of M. ! 179! xanthus and a large proportion of the cells stained red as development progressed (30) (31). If the spherical cells stain red, it would support our hypothesis that spherical cells are undergoing lysis. The dye permeability assay presumably interferes with development, whereas inducible expression of a gene coding for a fluorescent protein, or simply staining the cell membrane with FM 4-64, do not block development, so the methods we have described that detect spherical cells may provide a way to monitor ongoing cellular lysis during fruiting body formation. Materials and Methods Bacterial strains, plasmids, and primers. M. xanthus strains, plasmids, and primers used in this study are listed in Table S5.1. pYH3, which was used to induce production of TetR- YFP with IPTG, was constructed by replacing the lacZ gene of pMR3487 with a tetR-yfp fragment amplified from pLAU53 using tetR-YFP-F and tetR-YFP-R primers. pMR3487 was digested with XbaI and KpnI restriction enzymes. A Gibson assembly reaction was used to enzymatically join the overlapping DNA fragments (32) (33). The reaction mixture was transformed into E. coli strain DH5α with outgrowth in Luria-Bertani (LB) liquid medium prior to plating on LB agar (1.5%) supplemented with 15 µg/mL tetracycline for selection at 37°C. The construct was verified by sequencing with pMR3487 F and pMR3487 R primers. pYH3 was electroporated (34) into M. xanthus strains SA4212 and SA4118, resulting in strains YH3 and YH4, respectively. pYH6 was used to replace fruA with mNeonGreen-fruA in the M. xanthus chromosome. To construct pYH6, DNA fragments corresponding to regions flanking the fruA start codon were amplified from M. xanthus chromosomal DNA using 5 flank FOR and 5 flank REV primers, and using 3 flank FOR and 3 flank REV primers. An mNeonGreen fragment was amplified from pFM13 (Penelope Higgs, personal communication) using GFP F and GFP R primers. The three fragments were mixed with SmaI-digested pBJ114, in a Gibson assembly reaction (33). The reaction mixture was transformed into E. coli strain DH5α with outgrowth in LB liquid medium ! 180! prior to plating on LB agar supplemented with 5 µg/mL kanamycin for selection at 37°C. The DNA sequence of the joined fragments was verified, and the plasmid was transformed into M. xanthus strain DK1622 using electroporation (34), with outgrowth in Casitone-Tris (CTT) liquid medium (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH2PO4-K2HPO4, 8 mM MgSO4, [final pH 7.6]) prior to plating on CTT agar (1.5%). The positive negative screening was carried out as described previously with 40 µg/mL kanamycine and 5% galactose (35). Colonies that grew in the presence of galactose, but not when patched onto CTT agar supplemented with 40 µg/mL kanamycin, were tested by colony PCR using GFP FOR and GFP REV primers, and a strain that produced the expected PCR product was named YH6. Growth and development. M. xanthus was grown at 32°C on CTT agar or in CTTYE liquid medium (CTT with 0.2% yeast extract) shaking at 350 rpm. Development was performed under submerged culture conditions in 8-well µ-slides (Ibidi) with starvation buffer MC7 (10 mM morpholinepropanesulfonic acid [MOPS, pH 7.0], 1 mM CaCl2). Cells from log-phase cultures in CTTYE liquid were resuspended in MC7 at 1,000 Klett units as described previously (26). Cell suspension (26 µL) was added to MC7 (174 µL) in each well. Upon incubation at 32°C, cells formed a biofilm at the bottom of the well and underwent development. Microscopy. Images of nascent fruiting bodies were acquired with a Nikon A1 Laser Scanning Confocal Microscope configured on a Nikon Ti inverted platform with an XY automated stage and a 100X objective. Fluorescence from FM 4-64 and TdTomato was examined using a 560-nm laser for excitation and a 595/50 band pass emission filter. DAPI fluorescence was examined using a 402-nm laser for excitation and a 450/50 band pass emission filter. Fluorescence from mNeonGreen was examined using a 488-nm laser for excitation and a 525/50 band pass emission filter. Images “near the base of mounds” were the first optical section above the bottom of the well in which cells could be clearly visualized. Image analysis. The distance between two TetR-YFP loci was measured using ImageJ software (36). Fluorescence intensity profiles were measured using the software of the Nikon ! 181! A1 Laser Scanning Confocal Microscope. First, FM 4-64 fluorescence was used to visualize the cell membrane, then DAPI and/or mNeonGreen-FruA fluorescence was used to visualize DNA and nucleoids in cells selected for analysis. For each cell, a line was drawn across the cell (transitioning cells) or along the periphery (spores) and the fluorescence intensities were measured and analyzed as described in the text and the figure legends. Acknowledgements We thank Lotte Sogaard-Andersen, Montse Elias-Arnanz, and Bryan Julien for providing strains or plasmids. We are grateful to Melinda Frame and the Center for Advanced Microscopy at Michigan State University for assistance with microscopy. This research was supported by National Science Foundation grants MCB-1411272 and IOS- 1951025, and by salary support for L.K. from Michigan State University AgBioResearch. ! 182! APPENDIX ! 183! Table S5.1 Bacterial strains, plasmids, and primers used in this study Strain, plasmid, or primer Description Source or reference Strain E. coli DH5α λ- φ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17(rK ) supE44 thi-1 gyrA relA1 - - mK M. xanthus DK1622 SA4212 SA4118 YH3 YH4 YH5 YH6 YH7 YH9 Plasmid pMR3487 pLAU53 pYH3 pBJ114 pYH6 pYH7 pYH9 Primer Laboratory wild-type strain tetO-array in mxan0733 region (pNV3) tetO-array in mxan5499 region (pMAT18) PIPTG-tetR-yfp in SA4212 using pYH3 PIPTG-tetR-yfp in SA4118 using pYH3 PIPTG-tetR-yfp in DK1622 using pYH3 PfruA-mNeonGreen-FruA in DK1622 using pYH6 Pvan-mNeonGreen in DK1622 using pYH7 Pvan-tdTomato in YH6 using pYH9 1.38-kb-PIPTG-MCS_A-PR4::lacI, Tcr Para-tetR-yfp 1.38-kb-PIPTG-tetR-yfp_A-PR4::lacI, Tcr Backbone for in-frame deletions; galK Kmr PfruA-mNeonGreen-FruA, Kmr MXAN_0018-MXAN_0019-PR3-4::vanR- Pvan::mNeonGreen, Tcr MXAN_0018-MXAN_0019-PR3-4::vanR- Pvan::tdTomato, Tcr (37) (38) (18) (18) This study This study This study This study Chapter 3 This study (39) (32) This study (40) This study Chapter 3 Chapter 3 tetR-YFP-F GGAATCTAGAATGGTGTCTAGATTAGATAAAAG This study ! 184! pMR3487 F pMR3487 R 5 flank FOR 5 flank REV 3 flank FOR 3 flank REV GFP FOR GFP REV Van-tdTom F Van-tdTom R TCTTCACCTTTGCTGACCATGCGAAGGCCCCCC AG CCGT This study GCGGGCTCCGCGGCGGGCTCCGGCGAGTTCAT GGCA ACCAATCAAGCAGC This study GGTCGACTCTAGAGGATCCCCCAATCTTCAGGT TGTCCGCG This study ATGGTCAGCAAAGGTGAAGAAGAC GCCGGAGCCCGCCGCGGAGCCCGCGGAGCCC TTGTA CAGTTCGTCCATACCCATC This study This study GATGCGAGGAAACGCATATGGTGAGCAAGGGC GAG Chapter 3 GTACGCGTAACGTTCGAATTCTTACTTGTACAG CTCGTCCATG Chapter 3 Chapter 3 Chapter 3 Table S5.1 (cont’d) tetR-YFP-R GCCAGGTACC TTACTTGTACAGCTCGTC GTAAATGTGAGCACTCACAAT CCCGCACTCAGCTTGGAGGTG CGAATTCGAGCTCGGTACCCGTCGGCAGCATA CAC GTCTG This study This study This study This study pMR3691 MCS G- F pMR3691 MCS G- R CACGATGCGAGGAAACGCA CACCGGTACGCGTAACGTTC ! 185! Hours PS 24 30 42 A B Hours PS 24 30 42 Figure S5.1 Visualization of the same nascent fruiting body during starvation-induced development. M. xanthus strains YH4 and YH5 were starved under submerged culture conditions. IPTG (1 mM) and FM 4-64 (5 µg/mL) were added at the start of starvation. Confocal images were acquired at the indicated times PS and merged to show both TetR-YFP (yellow) and FM 4-64 staining of the cellular membrane (red). Arrows indicate a rod-shaped cell at 24 h, a transitioning cell at 30 h, and a spore at 42 h. Bar, 20 µm. A. Confocal images of the same nascent fruiting body of strain YH4. Images show an optical section near the base of a nascent fruiting body of strain YH4 at the indicated times. B. Confocal images of the same nascent fruiting body of strain YH5. Images show an optical section near the base of a nascent fruiting body of strain YH5 at the indicated times. ! 186! Hours PS 30 36 48 Hours PS 24 30 42 Hours PS 30 36 48 Hours PS 30 36 48 A! B! C! D! Figure S5.2 Visualization of nucleoids using mNeonGreen-FruA during development. M. xanthus strains YH6 or YH7 were starved under submerged culture conditions. FM 4-64 (5 µg/mL) was added at the start of starvation. Confocal images show FM 4-64 staining of the cellular membrane (red) and mNeonGreen-FruA fluorescence (green). A. Confocal images of the same nascent fruiting body of strain YH6. Images show an optical section near the base of a nascent fruiting body of strains YH6 with the red and green ! 187! channels merged. Arrows indicate a rod-shaped cell at 30 h, a transitioning cell at 36 h, and a spore at 48 h. Bar, 20 µm. B. Confocal images of the same nascent fruiting body of strain YH7. Images show an optical section near the base of a nascent fruiting body of strains YH7 with the red and green channels merged. Arrows indicate a rod-shaped cell at 24 h, a transitioning cell at 30 h, and a spore at 42 h. Bar, 20 µm. C. Confocal images of different fruiting bodies of strain YH6 with DAPI. DAPI (10 µg/mL) was added 30 min before imaging. Images show an optical section near the base of a nascent fruiting body of strains YH6 with the blue (DAPI), red and green channels merged. Arrows indicate a rod-shaped cell at 30 h, a transitioning cell at 36 h, and a spore at 48 h. Bar, 20 µm. D. Confocal images of the same nascent fruiting body of strain YH9. Images show an optical section near the base of a nascent fruiting body of strains YH9 with the red and green channels merged. Arrows indicate a rod-shaped cell at 30 h, a transitioning cell at 36 h, and a spore at 48 h. Bar, 20 µm. ! 188! 24 h 48 h 72 h 96 h Hours PS 48 72 96 FM4-64 & DAPI FM4-64 & mNeonGreen-FruA Figure S5.3 Visualization of the nucleoids by DAPI and mNeonGreen-FruA. M. xanthus strains DK1622 and YH6 were starved under submerged culture conditions in a 6-well, glass bottom plate. Glass surface interferes with M. xanthus motility, which results in a delay of about 6 h in development compared to on plastic. FM 4-64 (5 µg/mL) was added at the start of starvation. DAPI (10 µg/mL) was added 30 min before imaging strain DK1622. Confocal images were acquired at the indicated times PS to show FM 4-64 staining of the cellular membrane (red), DAPI staining of DNA (blue), and mNeonGreen-FruA fluorescence (green). Rows show strain DK1622 with red and blue channels merged (top), and strain YH6 with red and green channels merged (bottom). Arrows point to intense fluorescence from DAPI and/or mNeonGreen-FruA at one or two locations in an individual cell. Bar, 1 µm. ! 189! REFERENCES 190! ! REFERENCES Alley MRK, Maddock JR, Shapiro L. 1992. Polar localization of a bacterial chemoreceptor. Genes Dev 6:825-836. 10.1101/gad.6.5.825 Hu Z, Lutkenhaus J. 1999. Topological regulation of cell division in Escherichia coli involves rapid pole to pole oscillation of the division inhibitor MinC under the control of MinD and MinE. Mol Microbiol 34(1):82-90. 10.1046/j.1365-2958.1999.01575.x 3. Glaser P, Sharpe ME, Raether B, Perego M, Ohlsen K, Errington J. 1997. Dynamic, mitotic-like behavior of a bacterial protein required for accurate chromosome partitioning. Genes Dev 11(9):1160-8. 10.1101/gad.11.9.1160 Teleman AA, Graumann PL, Lin DCH, Grossman AD, Losick R. 1998. Chromosome arrangement within a bacterium. Curr Biol 8(20):1102-9. 10.1016/S0960-9822(98)70464- 6 5. Webb CD, Teleman A, Gordon S, Straight A, Belmont A, Lin DCH, Grossman AD, Wright A, Losick R. 1997. Bipolar localization of the replication origin regions of chromosomes in vegetative and sporulating cells of B. subtilis. Cell 88(5):667-674. 10.1016/s0092- 8674(00)81909-1 Luijsterburg MS, Noom MC, Wuite GJL, Dame RT. 2006. The architectural role of nucleoid-associated proteins in the organization of bacterial chromatin: A molecular perspective. J Struct Biol 156(2):262-272. 10.1016/j.jsb.2006.05.006 Toro E, Shapiro L. 2010. Bacterial chromosome organization and segregation. Cold Spring Harb Perspect Biol 31:171-99. 10.1146/annurev-cellbio-100814-125211 Dillon SC, Dorman CJ. 2010. Bacterial nucleoid-associated proteins, nucleoid structure and gene expression. Nat Rev Microbiol 185–195(2010). 10.1038/nrmicro2261 1. 2. 4. 6. 7. 8. 9. Verma SC, Qian Z, Adhya SL. 2019. Architecture of the Escherichia coli nucleoid. PLoS Genet. 10.1371/journal.pgen.1008456 10. Wu LJ, Errington J. 2004. Coordination of cell division and chromosome segregation by a nucleoid occlusion protein in Bacillus subtilis. Cell 117(7):915-925. 10.1016/j.cell.2004.06.002 11. Ricci DP, Melfi MD, Lasker K, Dill DL, McAdams HH, Shapiro L. 2016. Cell cycle progression in Caulobacter requires a nucleoid-associated protein with high AT sequence recognition. Proc Natl Acad Sci U S A 113 (40) E5952-E5961. 10.1371/journal.pgen.1008456 12. Kleckner N, Fisher JK, Stouf M, White MA, Bates D, Witz G. 2014. The bacterial nucleoid: Nature, dynamics and sister segregation. Curr Opin Microbiol 22:127-137. 10.1016/j.mib.2014.10.001 13. Lee B, Holkenbrink C, Treuner-Lange A, Higgs PI. 2012. Myxococcus xanthus ! 191! developmental cell fate production: Heterogeneous accumulation of developmental Regulatory proteins and reexamination of the role of MazF in developmental lysis. J Bacteriol 194:3058–3068. 0.1128/JB.06756-11 14. O’Connor KA, Zusman DR. 1991. Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. J Bacteriol. 173:3318–3333. 10.1128/jb.173.11.3318-3333.1991 15. Tzeng L, Singer M. 2005. DNA replication during sporulation in Myxococcus xanthus fruiting bodies. Proc Natl Acad Sci U S A. 102(40):14428-33. 10.1073/pnas.0506969102 16. Tzeng L, Ellis TN, Singer M. 2006. DNA replication during aggregation phase is essential for Myxococcus xanthus development. J Bacteriol 188(8):2774-79. 10.1128/JB.188.8.2774-2779.2006 17. Rosario CJ, Singer M. 2007. The Myxococcus xanthus developmental program can be delayed by inhibition of DNA replication. J Bacteriol 189 (24) 8793-8800. 10.1128/JB.01361-07 18. Harms A, Treuner-Lange A, Schumacher D, Søgaard-Andersen L. 2013. Tracking of Chromosome and Replisome Dynamics in Myxococcus xanthus Reveals a Novel Chromosome Arrangement. PLoS Genet. 9(9): e1003802. 10.1371/journal.pgen.1003802 19. Sager B, Kaiser D. 1993. Two cell-density domains within the Myxococcus xanthus fruiting body. Proc Natl Acad Sci U S A 90(8): 3690–3694. 10.1073/pnas.90.8.3690 20. Sager B, Kaiser D. 1993. Spatial restriction of cellular differentiation. Genes Dev 7:1645- 53. 10.1101/gad.7.9.1645 21. Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of hundreds of developmentally- regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. 10.1186/1471-2164-15-1123. Iniesta AA. 2014. ParABS system in chromosome partitioning in the bacterium Myxococcus xanthus. PLoS One 9(1):e86897. 10.1371/journal.pone.0086897 Santi PA. 2011. Light sheet fluorescence microscopy: A review. J Histochem Cytochem 9(2):129-38. 10.1369/0022155410394857. Tan IS, Ramamurthi KS. 2014. Spore formation in Bacillus subtilis. Environ Microbiol Rep 6(3): 212–225. 10.1111/1758-2229.12130 25. Wireman JW. 1979. Developmental induction of Myxococcus xanthus myxopores. J Bacteriol 140(1):147-53 26. Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736–2747. 10.1128/JB.01692-14 22. 23. 24. ! 192! 27. Saha S, Patra P, Igoshin O, Kroos L. 2019. Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C-signaling. Mol Microbiol 111: 1732-1752. 10.1111/mmi.14249 28. Boulos L, Prévost M, Barbeau B, Coallier J, Desjardins R. 1999. LIVE/DEAD(®) BacLight(TM): Application of a new rapid staining method for direct enumeration of viable and total bacteria in drinking water. J Microbiol Methods 37(1):77-86. 10.1016/s0167- 7012(99)00048-2 29. Berney M, Hammes F, Bosshard F, Weilenmann HU, Egli T. 2007. Assessment and interpretation of bacterial viability by using the LIVE/DEAD BacLight kit in combination with flow cytometry. Appl Environ Microbiol 73(10):3283-3290. 10.1128/AEM.02750-06 30. Nariya H, Inouye M. 2008. MazF, an mRNA Interferase, Mediates Programmed Cell Death during Multicellular Myxococcus Development. Cell 132:55–66. 10.1016/j.cell.2007.11.044 31. Lux R, Li Y, Lu A, Shi W. 2004. Detailed three-dimensional analysis of structural features of Myxococcus xanthus fruiting bodies using confocal laser scanning microscopy . Biofilms 1(4):293-303. 10.1017/S1479050505001559 32. Lau IF, Filipe SR, Søballe B, Økstad OA, Barre FX, Sherratt DJ. 2003. Spatial and temporal organization of replicating Escherichia coli chromosomes. Mol Microbiol 49(3):731-43. 10.1046/j.1365-2958.2003.03640.x 33. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, Smith HO. 2009. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods. 2009:343– 345. 10.1038/nmeth.1318 34. Kashefi K, Hartzell PL. 1995. Genetic suppression and phenotypic masking of a Myxococcus xanthus frzF- defect. Mol Microbiol 15:483–494. 10.1111/j.1365- 2958.1995.tb02262.x 35. Ueki T, Inouye S, Inouye M. 1996. Positive-negative KG cassettes for construction of multi-gene deletions using a single drug marker. Gene 183(1-2):153-57. 10.1016/S0378- 1119(96)00546-X 36. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9(7):671-5. 10.1038/nmeth.2089 37. Hanahan D. 1983. Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166(4):557-80. 10.1016/s0022-2836(83)80284-8 38. Kaiser D. 1979. Social gliding is correlated with the presence of pili in Myxococcus xanthus. Proc Natl Acad Sci U S A 6(11): 5952–5956. 10.1073/pnas.76.11.5952 39. ! Iniesta AA, Garcia-Heras F, Abellon-Ruiz J, Gallego-Garcia A, Elias-Arnanz M. 2012. Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl-!β-D-thiogalactopyranoside or vanillate. J Bacteriol 194:5875–5885. 10.1128/JB.01110-12 193! 40. Julien B, Kaiser D, Garza A. 2000. Spatial control of cell differentiation in Myxococcus xanthus. Proc Natl Acad Sci U S A 97:9098–9103. 10.1073/pnas.97.16.9098 ! 194! CHAPTER 6: Conclusions and Future Directions - An exciting opportunity to investigate bacterial interactions and cell fate determination inside a nascent fruiting body in situ Conclusions This dissertation has established a method to visualize the shape, nucleoid dynamics, and gene expression in individual developing cells in 3D by using confocal laser scanning microscopy and a custom image analysis pipeline. Figure 6.1 summarizes the main findings of this work. M. xanthus cells respond ultrasensitively to the addition of nutrient medium before the commitment period to form spores (Chapter 2). A two-fold change in nutrient medium concentration (12.5% vs. 25%) leads to an approximately 30-fold difference in the number of sonication-resistant spores. The level of the transcription factor MrpC correlates with multicellular mound formation and differentiation into spores. We propose that cells sense nutrients via the MrpC protein level. A threshold level of MrpC must be achieved for starving cells to stay in mounds and progress to sporulation. Positional information among cells is communicated via C-signaling. In Chapter 5, threshold effects of C-signaling were demonstrated by mixing wild-type cells with csgA mutant cells in different ratios. Wild-type cells rescued different emergent behaviors of the csgA mutant over a narrow range of ratios. No mound formation was observed at an overall ratio of 1 wild-type cell to 4 csgA cells (1:4). Mound formation but very little sporulation was seen at a ratio of 1:2. At a ratio of 1:1, mounds and spores were observed. We conclude that different threshold levels of C-signaling are required for new behaviors to emerge: first mound formation and then sporulation. We propose that a threshold level of MrpC may be necessary to meet a threshold requirement for C-signaling as MrpC positively regulates C-signaling. An mrpC null mutant is unable to rescue the development of a csgA mutant upon mixing (1). In addition, the level of CsgA protein is lower in an mrpC mutant than in the wild-type strain. Since MrpC starts to accumulated at 12 h PS (2) and streaming was observed at about 12 to 15 h PS, we propose ! 195! that MrpC increases the level of C-signaling so that streaming happens. Upon nutrient addition at 18 h PS, the MrpC level must recover to a certain threshold in order to meet a threshold requirement for C-signaling that is necessary for cells to stay in mounds. C-signaling activates FruA (denoted as FruA*), a transcription factor that is thought to bind to hundreds of developmentally-regulated genes (3). During mound formation, cells move closer to each other, enhancing short-range C-signaling and elevating the level of FruA*. Starva.on) Nutrients) MrpC) MrpC) C"signaling) FruA) FruA*) mound) rod) fmgE* irregularly"shaped) transi'oning*cell* spore) frui.ng)body) Figure 6.1 The multicellular starvation-induce development of M. xanthus. Starvation triggers the production of MrpC, which serves a mechanism to sense nutrients. Nutrient addition before the commitment period causes the MrpC level to decrease significantly. A threshold of MrpC level is required for starving cells to progress during development. Similarly, different thresholds of C-signaling are required for new behaviors to emerge: streaming, mound formation, and sporulation. MrpC positively regulates C-signaling and it also activates transcription of the fruA gene. C-signaling then activates FruA (denoted as FruA*), which in turn induces genes important for mound formation and sporulation. PfmgE is regulated by FruA* and MrpC cooperatively. PfmgE shows strong activity in transitioning cells and spores (in red) but not rods, suggesting cells committed to shape change show a higher level of C-signaling activity. The result is a fruiting body filled with transitioning cells and spores, which are more abundant in the inner domain. Cellular shape change inside mounds is correlated with the segregation of sister chromosomes. ! 196! FruA* and MrpC bind cooperatively to activate the transcription of developmentally- regulated genes, such as the dev operon (4) and the fmgE gene (5). In Chapter 3, we found that Pdev activity was similar in all cell types inside mounds, while PfmgE activity was strongly induced in transitioning cells and spores, but not in rods. Both PfmgE and Pdev employ a cooperative binding site, located just upstream of the promoter, which activates transcription. However, the architecture of the PfmgE region is more complex as it has a distal upstream site with higher affinity for cooperative binding of MrpC and FruA, which acts negatively by competing for binding with the lower-affinity site just upstream of the promoter (5). These results suggest that Pdev has a relatively low threshold for FruA*, whereas PfgmE has a higher threshold. In other words, Pdev is more sensitive to C-signaling than PfmgE. However, none of these genes are essential for the initial cellular shape change during sporulation. Indeed, such genes are currently unknown. We hypothesize that the genes responsible for initiating shape change have a relatively high threshold for FruA* and therefore depend strongly on C-signaling. Methods to identify these genes will be discussed later in this chapter. We propose that when the thresholds for both MrpC and C-signaling are met, unidentified genes responsible for initiating cellular shape change are expressed. Such genes may be expressed in the inner domain of a nascent fruiting body, since we found that transitioning cells and spores are more abundant in the inner domain (Chapter 3). However, the circular movements of rod-shaped cells in the outer domain have been proposed to move spore precursor cells to the inner domain (6), and we have observed movement of transitioning cells by rods in our time-lapse microscopy efforts (data not shown), so it is possible that expression of genes responsible for initiating shape change may not be localized to the inner domain. Pattern formation is an important phenomenon in developmental biology that is observed in prokaryotes as well as in eukaryotes. Some examples of pattern formation in bacteria are swarm ring and aggregate formation by E. coli (7), mucus veils on top of sulfidic marine sediment by Thiovulum majus (8), and the knotted-branching pattern of the Bacillus circulans ! 197! colony (9). Compared with other bacteria, the three-dimensional fruiting body structure of M. xanthus is strikingly uniform in size and shape. Cellular shape change inside mounds is correlated with chromosome segregation (Chapter 4). At 24 h PS, the base of the mound consists of mostly rod-shaped cells. The nucleoids occupy primarily the mid region of starving rods. By 30 h PS, many cells are transitioning to spores. The nucleoids look either like a crescent along one side of the cell, or are segregated into two loci. At 42 h, most of the cells are spores. Segregated nucleoids are still visible at opposite sides of some spores, but the nucleoids appear to be decondensed and distributed evenly within other spores. Future directions Identify genes responsible for initiating cellular shape change. Although the GRN governing development has been intensively studied in M. xanthus, genes important for initiating cellular shape change have not yet been identified. Mutations in the known genes either block mound formation or affect spore maturation (e.g. ability to resist heat and sonication). The method to quantify cellular shapes and gene expression inside nascent fruiting bodies (Chapter 3) paves the way for the identification and study of genes responsible for initiating shape change during sporulation. The information we learned about cellular shapes and gene expression can be incorporated into future studies to guide RNA-seq, which would be designed to identify genes that are up- or down-regulated at the onset of the rod-to-spore transition, and Tn-seq, which would employ a selection for spore formation. Comparison of RNA-seq data from the wild-type strain and mutants that are delayed or advanced for sporulation may provide useful information to identify candidate genes responsible for initiating shape change during sporulation. A rapid and simple method to guide the timing of sample collection for RNA-seq is to use the FM 4-64 membrane stain, which helps to visualize cells in vivo. We have classified mutants into four groups based on their timing of cellular shape change compared with the wild-type strain (data not shown). A devS mutant was the only strain ! 198! that showed a delay in cellular shape change among the mutants examined. devS mutant cells were still rod-shaped at 30 h PS, and we observed transitioning cells at 42 h PS. On the other hand, mutations in MXAN_3259 and MXAN_5372 advanced the timing of cellular shape change, which happened before 24 h PS. Genes differentially expressed in the wild-type strain as compared with devS, MXAN_3259, and MXAN_5372 mutants are candidates for involvement in initiating shape change. Another method to guide RNA-seq is to monitor gene expression in situ during development. As there is a large increase in PfmgE activity in transitioning cells compared with rod-shaped cells (Chapter 3), PfmgE activity can serve as a guide to determine the timing of RNA- seq analysis. Samples for RNA-seq could be collected before and during the high activity of PfmgE, which can be visualized using the transcriptional fusion of PfmgE to tdTomato. In addition, comparison of RNA-seq analysis from the wild-type strain and a csgA mutant can lead to identification of new C-signal-dependent genes. Genes regulated in transitioning cells and strongly dependent on C-signaling are likely to be responsible for the initiating shape change. Tn-seq analysis can be used to determine which candidate genes from RNA-seq analysis are important for sporulation. Tn-seq relies on a library of transposon (Tn) insertion mutants, a selection scheme, and sequencing of Tn flanking regions en masse (10). Samples for Tn-seq analysis could be collected at 18 h PS, when all cells are rod-shaped, and at 42 h PS, when most cells have converted into spores. The latter sample needs to be sonicated twice. The first sonication treatment disrupts rods. DNase I would be used to degrade DNA released from rods, then EDTA would be added to inhibit DNase I. The second sonication, with glass beads, disrupts spores. Genes underrepresented in 42-h spores as compared with the 18-h sample, and differentially regulated during shape change in mutants defective in C-signaling, DevS, MXAN_3259, and MXAN_5372, would be candidate genes for involvement in initiating cellular shape change. ! 199! Knockout mutants of candidate genes could be examined using various methods. Confocal fluorescence microscopy with the FM 4-64 membrane stain is a simple and rapid method to detect defects in initiating cellular shape change during development. For mutants that exhibit defects, further studies to measure mRNA levels from candidate genes in the wild- type strain, a csgA mutant, and a devS mutant using high-throughput RT-qPCR could be done to verify RNA-seq results. Dependence of candidate genes on a high level of C-signaling could be tested by fusing the promoter region to a gene coding for a fluorescent protein and measuring fluorescence intensity of cells in nascent fruiting bodies (e.g. in comparison with Pdev and PfgmE fusions, which appear to have low and high thresholds, respectively, for FruA* resulting from C-signaling. As peptidoglycan likely decreases during the rod-to-spore transition (11) (12), we think that genes encoding peptidoglycan remodeling enzymes may be good candidates for involvement in initiating shape change. Investigate how C-signal transmission leads to emergent behaviors of streaming, mound formation, and sporulation. We have shown that different threshold levels of C-signaling are required for different behaviors to emerge by co-developing wild-type cells and csgA mutant cells at different ratios (Chapter 4). Pdev and PfmgE activity were measured in single csgA mutant cells before and during mound formation and sporulation. However, we were not able to make time-lapse movies to track single cells from when they move into a mound to when they sporulate, which hinders our ability to investigate what motility behaviors are important for development. As cells in mounds can move up and down, they will rapidly move out of the focal plane. To solve that problem, z-stacks need to be collected for a time-lapse movie. However, our preliminary data show that collecting a z-stack too often with confocal laser scanning microscopy leads to photobleaching and developmental defects. Light-sheet microscopy offers a method to obtain 3D, live images of large volumes much faster than the point-by-point scanning of a typical confocal microscope, therefore expose individual cells to less light. Light-sheet microscopy illuminates an entire plane of the sample, ! 200! captures a wide-field image, then moves the plane (13). This approach may enable us to track individual cells in mounds, and investigate how C-signaling leads to emergent behaviors. It has been shown that C-signaling decreases the reversal frequency of cell motility and increases the net-distance traveled by a cell per minute (14) (15). Live-cell imaging and data-driven modeling will continue to advance our knowledge about what motility behaviors are important for development. C-signal transmission occurs by a short-range, possibly contact-dependent, mechanism (16) (17). It should be possible to measure instances of close cell-cell proximity (end-to-end, end-to-side, side-to-side) of individual csgA mutant cells when mixed with the wild-type strain by labeling a portion of each type of cells with different fluorescent proteins. If individual cells can be tracked during development (e.g. using light-sheet microscopy), the instances of close cell- cell proximity of csgA mutant cells that succeed or fail to sporulate could be compared in order to determine the types of short-range interactions that are important for C-signal transmission. If it proves impossible to track individual cells, proximity measured in many z-stack “snapshots” could be used in modeling efforts to understand the interaction dependence of C-signaling. The interaction of cgsA mutant cells and wild-type cells at different ratios that permit different behaviors (e.g. mound formation but not sporulation) could also be examined. Furthermore, a gene responsible for initiating cellular shape change could be used as a marker of commitment to sporulation in these studies. We could track gene expression in csgA mutant cells and determine how motility and cell-cell interactions affect C-signal-dependent gene expression. The mechanism of chromosome segregation during cellular shape change. The nucleoids segregate to form two loci when cells change shape during M. xanthus development (Chapter 5). However, the mechanism of chromosome segregation in transitioning cells is unknown. M. xanthus uses a parABS system for chromosome segregation in vegetative cells (18). ParB proteins bind to parS DNA sequences. ParA proteins are P-loop ATPases that interact with the ParB/parS complex. We hypothesize that chromosome segregation during ! 201! developmental shape change is also regulated by the parABS system. To test this hypothesis, we would need to create parA, parB, and parAB conditional mutants because ParA and ParB are essential for cell viability in M. xanthus (18) (19). To create the conditional mutants, the first step would be to ectopically integrate Pvan-parA and/or Pvan-parB in the genome (20). Then, allelic exchange using a positive-negative selection method would be used to delete the native copy of parA and/or parB (21). We expect to see no chromosome segregation in transitioning cells in parA and parB single and double mutants, which may result in sporulation and/or germination defects. In vegetative cells, ParA clusters at the poles and, in some cells, at the DNA-free cell division plane between the two nucleoids (18) (19). This ParA localization pattern depends on ParB. ParB inhibits the nonspecific interaction of ParA with DNA, and ParA co-localizes with chromosomal DNA only when ParB is depleted. Future work investigating subcellular localization of ParA and ParB inside developing cells in situ would advance our knowledge of parABS system function during starvation-induced development. By constructing a strain producing both ParA-mNeonGreen and ParB-tdTomato fusion proteins under the control of vanillate-inducible promoter (optimally, both fusion proteins would be functional and the native genes would be deleted), we could observe the localization of the fusion proteins in developing cells. We would expect to see changes in ParA and ParB co-localization when cells undergo shape change. Staining cells with DAPI would help visualize any correlation of ParA and ParB localization with chromosome condensation and segregation. It has been shown that during DNA replication in vegetative cells, one ori region remains in the original subpolar region, while the second copy segregates unidirectionally to the opposite subpolar region, followed by the rest of the chromosome (18). At the same time, the ter region of the mother chromosome relocates to midcell. After completion of replication and segregation, the two chromosomes show an ori-ter-ter-ori arrangement with mirror symmetry about a transverse axis at midcell. Although the chromosome is replicated in both vegetative and ! 202! developing rods, compartmentalization does not follow during development as it does during division of vegetative cells. It would be interesting to see whether replication and segregation of chromosomes in developing cells is similar as in vegetative cells. The fluorescent reporter operator system utilized in Chapter 5 could be used, and development of a method to track individual cells over time by light-sheet microscopy would offer a powerful approach to investigate the replication process as well as ParAB localization. ! 203! REFERENCES 204! ! 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Sun H, Shi W. 2001. Analyses of mrp genes during Myxococcus xanthus development. J Bacteriol 183(23):6733-39. 10.1128/JB.183.23.6733-6739.2001 Rajagopalan R, Kroos L. 2014. Nutrient-regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 196:2736–2747. 10.1128/JB.01692-14 Robinson M, Son B, Kroos D, Kroos L. 2014. Transcription factor MrpC binds to promoter regions of hundreds of developmentally- regulated genes in Myxococcus xanthus. BMC Genomics 15:1123. 10.1186/1471-2164-15-1123 Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L. 2015. Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 197:240–251. 10.1128/JB.02310-14 Son B, Liu Y, Kroos L. 2011. Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 193:2756–2766. 10.1128/JB.00205-11 Sager B, Kaiser D. 1993. Spatial restriction of cellular differentiation. Genes Dev 7:1645- 53. 10.1101/gad.7.9.1645 Brenner MP, Levitov LS, Budrene EO. 1998. Physical mechanisms for chemotactic pattern formation by bacteria. Biophys J 74(4):1677-93. 10.1016/S0006-3495(98)77880-4 Thar R, Kühl M. 2005. Complex pattern formation of marine gradient bacteria explained by a simple computer model. FEMS Microbiol Lett 246(1):75-9. 10.1016/j.femsle.2005.03.036 9. Wakano JY, Maenosono S, Komoto A, Eiha N, Yamaguchi Y. 2003. Self-Organized Pattern Formation of a Bacteria Colony Modeled by a Reaction Diffusion System and Nucleation Theory. Phys Rev Lett 90(25 Pt 1):258102. 10.1103/PhysRevLett.90.258102 10. van Opijnen T, Bodi KL, Camilli A. 2009. Tn-seq: high-throughput parallel sequencing for fitness and genetic interaction studies in microorganisms. Nat Methods 6:767–72. 10.1038/nmeth.1377 11. Muller FD, Schink CW, Hoiczyk E, Cserti E, Higgs PI. 2012. Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Mol Microbiol 83:486–505. 10.1111/j.1365-2958.2011.07944.x 12. Bui NK, Gray J, Schwarz H, Schumann P, Blanot D, Vollmer W. 2009. The peptidoglycan sacculus of Myxococcus xanthus has unusual structural features and is degraded during glycerol-induced myxospore development. J Bacteriol 191(2):494-505. 10.1128/JB.00608-08 13. Santi PA. 2011. Light sheet fluorescence microscopy: A review. J Histochem Cytochem ! 205! 59(2):129-38. 10.1369/0022155410394857 14. Jelsbak L, Søgaard-Andersen L. 2002. Pattern formation by a cell surface-associated morphogen in Myxococcus xanthus. Proc Natl Acad Sci U S A 99(4):2032-37. 10.1073/pnas.042535699 15. Jelsbak L, Søgaard-Andersen L. 1999. The cell surface-associated intercellular C-signal induces behavioral changes in individual Myxococcus xanthus cells during fruiting body morphogenesis. Proc Natl Acad Sci U S A. 16. Kim SK, Kaiser D. 1990. C-factor: A cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 96(9):5031-6. 10.1073/pnas.96.9.5031 17. Kim SK, Kaiser D. 1990. Cell aligmnent required in differentiation of Myxococcus xanthus. Science 249(4971):926-8. 10.1126/science.2118274 18. Harms A, Treuner-Lange A, Schumacher D, Søgaard-Andersen L. 2013. Tracking of Chromosome and Replisome Dynamics in Myxococcus xanthus Reveals a Novel Chromosome Arrangement. PLoS Genet. 10.1371/journal.pgen.1003802 Iniesta AA. 2014. ParABS system in chromosome partitioning in the bacterium Myxococcus xanthus. PLoS One. 10.1371/journal.pone.0086897 Iniesta AA, Garcia-Heras F, Abellon-Ruiz J, Gallego-Garcia A, Elias-Arnanz M. 2012. Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl- β-D-thiogalactopyranoside or vanillate. J Bacteriol 194:5875–5885. 10.1128/JB.01110-12 21. Ueki T, Inouye S, Inouye M. 1996. Positive-negative KG cassettes for construction of multi-gene deletions using a single drug marker. Gene 183(1-2):153-7. 10.1016/S0378- 1119(96)00546-X 19. 20. ! 206!