SYSTEMATIC ANALYSIS OF THE SIGNAL RESPONSIVE GENE REGULATORY NETWORK GOVERNING MYXOCOCCUS XANTHUS DEVELOPMENT By Shreya Saha A DISSERTATION Submitted to Michigan State University in partial fulfil l ment of the requirements for the degree of Cell and Molecular Biology Doctor of Philosop hy 2020 ABSTRAC T SYSTEMATIC ANALYSIS OF THE SIGNAL RESPONSIVE GENE REGULATORY NETWORK GOVERNING MYXOCOCCUS XANTHUS DEVELOPMENT By Shreya Saha Studies of signal - induced gene expression in bacteria have contribute d to under standing of how bacteria cope with environmental stress. As an extensively studied model, Myxococcus xanthus provides fascinating insights into how changes at the level of gene expression enable which bacteria t o survive environmental insults such as nutrient limitation. Upon starvati on M. xanthus cells glide into aggregates and form mounds that mature into fruiting bodies as some cells form spores. Previously, our group defined 24 - 30 h poststarvation as the critical period for commitment to spore formation, when cells commit t o form s pores despite perturbation of the starvation signal by nutrient addition . The process of multicellular development that culminates in sporulation is governed by a network of signal - responsive transcription factors that integrate signals for starvation and cellular alignment . In this dissertation I present the first systematic approach to elucidate the network dynamics during the commitment period . In the network , MrpC is a starvation - responsiv e transcription factor, whereas FruA is a transcription factor th at responds to cellular alignment conveyed by C - signal ing . Transcription of fruA is dependent on MrpC binding , and FruA activity is proposed to be posttranslati onally regulated by C - signaling, although the mechanism is unknown . FruA and MrpC cooperatively regulate transcription of the dev operon. My systematic analysis of the network dynamics support ed a model in which posttranslational activation of FruA by C - signaling is critical for dev transcript ion and for commitment to sp ore formation. S imilar to dev , MrpC and C - signal - activated FruA combinatorially controlled transcription of the late - acting fadIJ operon involved in spore metabolism . Regulation of late - acting operons implicated in spore coat biogenesis ( exoA - I , n fsA - H , MXAN_3259 - MXAN_3263 ) was discovered to be under complex control by MrpC and FruA. My e v idence suggest s that transcription of these operons depend s at least in part on a C - signal - dependent switch from negative regulation by unactivated FruA to positi ve regulation by activated FruA during the period leading up to and including commitment to sporulation. MrpC negatively regulate d exo and MXAN_3259 during mound formation, but positively regulated nfs . During commitment to sporulation, MrpC continued to positively regulate nfs , switched to positive regulation of MXAN_3259 , and continued to negatively regulate exo . A third transcription factor, Nla6, appeared to be a positive regulator of all the late genes. We propose that in combination with regulation by Nla6, differential regulation by FruA in response to C - signaling and by MrpC controls late gene expression to ensure that spore resistance and surface characteristics meet environmental demands. Copy r i ght by SHREYA SAHA 2020 v This dissertation is dedicated to my parents Prabir Kumar Saha and Sipra Saha, who always inspired me by their life - choices, encouraged me to follow my dreams and work sincerely to accomplish those. vi ACKNOWLEDGEMENTS Firstly, I would like to thank my mentor Dr Lee R Kroos for providing constant support throughout this journey. During this entire journey he was patient, supportive, encouraging, and most of all a devoted mentor which helped me strive an d constantly grow as a scientist during my PhD. All these qualities not only make him one of the best mentors of Michigan State University, but also one of the best mentors I have come across in my research career so far. I will always look up to him as my role model for doing good science, dedication, sincerity, for taking care of my students and for supporting and respecting my colleagues. I want to extend my gratitude to all the scientists who have worked immensely hard to nurture the field of Myxococc us xanthus research which has been an excellent foundation for my PhD research. I would like to acknowledge my committee members for their valuable advice, my supportive peers at Michigan State University for making the place an excellent and fun working e nvironment. Finally, I wish to thank my parents, my younger si s ter who always motivated me to pursue my dreams and to dedicate myself towards doing good work and shouldering all the challenges they faced in India regarding the health and well - being of my family. I want to thank my husband Dr Pavel Roy and his family for being supporting throughout all the thick and thins of my life. My gratitude towards this meaningful journey of PhD at Michigan State University which added significantly towards the kind of person I am today. vii TABLE OF CONTENTS LIST OF TABLES i CHAPTER 1: Lessons from the study of signal induced gene expression in bacteria .. 1 Introduction . 1 R EFERENCES . 2 6 CHAPTER 2: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network suppo rts posttranslational regulation of FruA by C - signaling . . 3 6 Abstract 3 7 Introduction . 3 8 . 4 3 . 6 5 Materials and m . 7 4 APPENDIX 8 5 R EFERENCES .. 10 6 CHAPTER 3: Differential regulation of late - acting operons by FruA and MrpC during Myxococcus xanthus development . 11 4 . 1 1 4 11 5 Results 1 18 Discussion 1 37 Materials and m ethods . 1 49 APPENDIX . 15 6 R EFERENCES . . 1 7 1 CHAPTER 4: Conclusions and future directions 1 78 184 R EFERENCES . . 1 8 6 viii LIST OF TABLES Table S2.1 Cell and spore number s counted in chapter 2 .. 8 6 Table S2.2 Strains, plasmids and primers used in chapter 2 8 7 Table S3.1 Cell and spore numbers counted 1 57 Table S3.2 Changes in transcript levels during development 1 58 Table S3.3 Plasmids, strains and primers used in chapter 3 1 59 ix LIST OF FIGURES Figure 1 .1 Life cycle of . 14 Figure 1. 15 Figure 1. 3 The GRN governing sporulation with the output gene dev 2 1 Figure 1. 4 Model of differential regulation of late genes at two different morphological stages during development . 2 5 Figure 2.1 Simplified model of the gene regulatory network governing formation of fruiting 40 Figure 2. 2 Development of M. xanthus 4 4 Figure 2.3 Levels of MrpC and FruA during M. xanthus development 4 9 Figure 2.4 Transcript levels during M. xanthus development 50 Figure 2.5 Mathematical modeling of different hypotheses to explain the low dev transcript level in a csgA mutant 5 6 Figure 2.6 dev .. 60 Figure 2.7 dev transc ript levels 60 Figure 2.8 FruA protein and dev 6 2 Figure 2.9 Updated model of the gene regulatory network governing formation of fruiting bodies 6 Fig ure 2.10 Abbreviated summary 8 4 Figure S2.1 Cellular changes during M. xanthus 9 Figure S2.2 90 Figure S2.3 Validation of 16S rRNA as an internal standard for RT - qPCR analysis during M. xanthus developmen t . 9 1 Figure S2.4 fruA transcript stability in wild type and csgA 9 2 x Figure S2.5 Levels of dev and fmg transcripts in a ladA mutant . 9 3 Figure S2.6 Development al phenotype and quantification of sonication resistant spores of the ladA mutant . 4 Figure S2.7 Levels of MrpC and FruA in wild type, distal site mutant and mrpC mutant during M. xanthus development 5 Figure S2.8 mrpC, fruA and dev transcript levels in wild type, distal site mutant and mrpC mutant during M. xanthus development 9 6 Figure S2.9 mrpC transcript stability in wild type and mrpC mutant at 18 h poststarvation . 9 7 Figure S2.10 Developmental phenotype of wild type, distal site mutant and mrpC mutant 7 Figure S 2. 11 Cellular changes in wild type, distal site mutant and mrpC mutant . 9 8 Figure S2.12 Mathematical modeling prediction of the required reduction in FruA activity in the csgA mutant in comparison to wild type, to explain the experimental data . 98 Figure S2.13 mrpC and fruA transcript levels in wild type, csgA mutant and csgA (P van - fruA ) mutant .......................................................................................................... .. . . 99 Figure S 2. 14 MrpC protein level in wild type, csgA mutant and csgA (P van - fruA ) . . 10 0 Figure S 2. 15 Developmental phenotype of wild type, csgA mutant and fruA (P van - fruA ) mutant ..10 0 Figure S 2. 16 Cellular changes of wild type, csgA mutant and fruA (P van - fruA ) mutant . 10 1 Figure S 2.1 7 Developmental phenotype of fruA (P van - fruA ) mutant and fruA (P van - fruAD59E ) mutant . . 10 1 Figure S 2. 18 Levels of fmg transcripts in csgA mutant and csgA (P van - fruA ) mutant, csgA mutant and csgA (P van - fruAD59E ) mutant ..10 2 Figure S 2. 19. Models for regulation of fmgD and fmgE . 10 4 Figure S 2. 20. Representative MrpC and FruA immunoblots for wild type and . 10 5 Figure 3.1 Developmental phenotype of M. xanthus . . 1 2 1 xi Figure 3.2 Cellular changes during M. xanthus development 1 2 2 Figure 3.3 Transcript levels in wild type, csgA mutant and fruA mutant . 2 3 Figure 3.4 Transcript stability in wild type and csgA mutant . 12 5 Figure 3.5 Transcript levels in wild type, mrpC mutant and fruA mutant during M. xanthus 29 Figure 3.6 Transcript levels in wild type, mrpC (P van - fruA ) mutant, mrpC mutant and fruA mutant during early time points 3 3 Figure 3.7 Model of diffe rential late gene regulation . 1 47 Figure S3.1 Cellular changes during M. xanthus development 1 6 2 Figure S3.2 Transcript levels in wild type, csgA mutant, csgA (P van - fruA ) mutant, csgA (P van - fruAD59E ) during M. xanthus 1 6 3 Figure S3.3 Transcript stability in wild type and mrpC mutant 1 6 4 Figure S3.4 Protein levels in wild type, nla6 kmR mutant, nla6 tetR mutant and mrpC (P van - fruA ) mutants during early time points .. 16 6 Figure S3.5 Development of M. xanthus strains at early times .1 67 Figure S3.6 Developmental phenotype of wild type, , nla6 kmR mutant, nla6 tetR mutant, mrpC (P van - fruA ) mutant 67 Figure S3.7 Transcript levels in wild type, nla6 kmR mutant and nla6 tetR mutant during early time po . 1 68 Figure S3.8 Transcript levels in wild type, devI and dev S 1 69 Figure S3.9 Binding of FruA and MrpC to the dev , exo , and nfs upstream 7 0 1 C HAPTER 1: Lessons from the study of signal induced gene expression in bacteria Introduction How bacteria sense environmental c ues and initiate changes at the cellular and molecular level is an ever - fascinating question. In addition to environmental signals, bacteria send signals to each and respond appropriately . B acteria integrate multiple signals from each other and from the ir environment to chang e individual and community behavior . M olecular understanding of how bacterial communities coordinate their behavior in response to signals will enable us to manipulate these communities for several applications, such as biofuel production [4] and therapeutics [5] . An example of signal - induced gene expression that has been extensively studied is the stationary - phase response in E scherichia coli , during which sigma factor RpoS mediates global changes in gene expression [6] . Under growth conditions, RpoS is degraded by the protease ClpXP [7] because the response regulator protein RssB specifically targets RpoS to ClpXP for degradation [8] . Upon entering the stationary phase , in response to phosphate starvation, regulation of RssB by an anti - adaptor protein IraP ensures stability of RpoS [9] , promoting expression of RpoS - regulated stress response genes. R egulation of RpoS stability thus provides an example of signal - induced regulation of gene expression to promote a bacterial response. U pon starvation for carbon, nitrogen or phosphorous, Bacillus subtilis undergoes endospore formation , which provides a more complicated model system to study temporal regulation of gene expression induced by starvation signal s . During B . subtilis endosporulation, s ignals between the mother cell and forespore trigger differential gene expression and ensures coordination of gene regulation between the two compartments [10] . A critical step during 2 endo spore formation is the release of active K into the mother cel l by cleaving the precursor pro - K at the outer forespore membrane, in response to a signal from within the forespore [11] . This step promot es transcription of K RNA polymerase (RNAP) dependent genes in the mother cel l , products of which en sure formation of the cortex and coat layer s of the spore . Cleavage of pro - K serves as a critical checkpoint for forespore formation and is triggered by the interaction of pro - K with an intramembrane metalloprotease , SpoIVFB [12] . SpoIVFB is inhibited by comp lex formation with BofA and SpoIVFA until the signal comes from the forespore [13] . A fascinating model to study how a bac terial cell integrates multiple signals and initiates coordinated expression changes of multiple genes to alter community behavior is the multicellular developmental process of th e gram - negative bacterium Myxococcus xanthus [2] . Upon scarcity of nutrients a developmental process gets initiated with aggregation of r o d - shaped cells leading to formation of mounds followed by differentiation of cells within mounds to dormant spores , resistant to environment al insults . The majority of cells lyse during the formation of mounds [14] , perhaps providing nutrients to cells which are destined to form spores. About 15% of the developmental population remain outside of fruiting bodies as peripheral rods [3, 14, 15] , and only a few percent of the developing rods convert to round spores [3, 14] . The process of starvation - induced multicellular development of M. xanthus is governed by a signal - responsive gene regulatory network (GRN) [2, 11] . C ascades of signal - responsive transcription factors of the GRN coordinate ly and s equentially regulate target gene s temporally and spatially [2] . Among these transcription factors , some are involved in combinatorial regulation of the target genes , providing integration of multiple signals [16 - 18] . An intriguing aspect is that t ypical prokaryotic - like signaling and mechanisms of gene regulation 3 a ppear to be in suf f icient to control the multicellular developmental program of M. xanthus. I ntricate, eukaryotic - like components and mechanisms play crucial role s in reprogramming gene expression to regulate development of this bacterium [19] . For example, euk aryotic - like serine / threonine protein kinase (STPK) in bacteria was first identified in M. xanthus and it is required for normal development [20] . A breakthrough in the field was achieved when DNA microarray studies identified a group of 54 RNAP - dependent enhancer - binding proteins (EBPs), which are critical for fruiting body development [21] . Some of t hese EBPs have an N - terminal sensory domain predict ed to be phosphorylated by one of the abundant STPK s in M. xanthus , transduc ing the starvation signal to activate target gene transcription at the beginning of the developmental process [22] . Despite significant advance ments in the areas of signal transduction and gene regulation during M. xanthus fruiting body development, a systematic and quantitative understanding of gene expression changes ha d not yet been achieved when I began my research . Additionally, a mechanisti c understanding of the regulation of genes which are transcribed late in development and lead to completion of spore formation remain ed to be elucidated. In particular , how these late - acting genes are controlled by the upstream transcription factors of the GRN was unknown . My work has aimed to address t hese questions . Understanding s ignaling and its impact on gene expression is important to enable manipulat ion of bacterial lifestyle and community behavior Intercellular c ommunication or social interaction between bacteria allow s all members of a bacterial community to function in synchrony , thus act ing like a multicellular organism. In contrast to the harmonious behavior of a bacterial community , some social interactions between bacteria stimulate individual ity within a group of cells, thus promot ing diversity in a 4 bacterial community [23] . Exchange of information in the form of s ignal ing between bacteria is primarily responsible for the social interaction s between individual cells within a bacterial community. Hence, the production, release and exchange of signaling molecules is critical for formation, maintenance and function of bacteria l communities. Detection and e xchange of extracellular, diffusible signal molecules is a means to measure the density of a cell population (aka quorum sensing) to r egulate biofilm formation in pathogenic organisms [24 - 26] . Early in M. xanthus development, A - signal provides quorum sensing to measure accumulation of a certain cell density which serves as a checkpoint for the decision to begin aggregation [27] . The other type of signaling mechanism is short - range signaling , which requires close proximity or contact between bacterial cells [23] . This i ntimate conversation between bacteria is seen in reciprocal C - signaling between closely - packed M. xanthus cells within mounds , crisscross signaling between the mother cell and the forespore within a B. subtilis sporangium, and contact - dependent inh ibition of growth mediated by cell surface proteins CdiA and CsdiB in E. coli [23] . Upon association of extracellular s i gnaling molecules with cell surface receptors , a membrane - associated protein kinas e take s part in transduction of the signal via phosphorylation cascade s , which ev entually impact DNA - binding proteins (transcription factors) leading to regulation of the target genes as a response to the external stimuli [23] . Bacteria also respond to internal cues. The s tringent respo nse to nutrient limitation is extensively studied as a classic example of a broadly conserved reorientation of gene expression [28, 29] , during which the ribosome - associated protein RelA senses amino acid starvation and stimulates production of a secondary messen ger molecule (p)ppGpp in cell . Intracellular accumulation of (p)ppGpp modulates RNA polymerase activity , resulting in global 5 alteration of gene expression [28, 29] . The stringent response is crucial for regulation of processes like bacterial virulence [30] , resistance to anti microbial agents [31] , survival of pathogens upon host invasion [32] , and biofilm formation favoring environmental surviv al and host colonization of pathogens [33] . For example, e xpression of VpsT, t he transcriptional activator of biofilm genes in Vibrio cholera e , is dependent on the stationary - phase sigma factor RpoS , which is induced by the stringent response [33] . Bacterial adaptation and lifestyle alterations with changes in the environment are also mediated by the stringent response [34] , such as nutrient limitation - induced withdrawal from the biofilm lifestyle in order to switch to the free - swimming , planktonic lifestyle . This switch is mediated by a phosphodi esterase , BifA, expression of which is dependent on (p)ppGpp synthesis in P seudomonas putida [35] . Progression of the starvation - induced multicellular lifestyle of M. xanthus is promoted by a contact - dependent , intercellular signal (C - signal) encoded by CsgA, but production of the C - signal is linked to the stringent res ponse. T ranscription of csgA is positivel y regulated by (p)ppGpp upon starvation , whereas transcription of socE , which is high during vegetative growth, is negatively regulated by (p)ppGpp [36] . C - signal promo tes multicellularity by ensuring close proximity between developing cells in mounds before inducing gene expression that promotes further progression of the developmental process [37 - 39] . Hence, in this case, the starvation signal via the internal stringent response and (p)ppGpp signaling is coupled to intercellular C - signaling. U nderstanding how transduction of environmental and intercellular signals re - orient s gene expression is revealing how bacteria switch between lifestyles, and maintain both individual and communal life style s . Such understanding is important to be able to manipulate bacterial 6 communit ies for advancement of basic and applied research. For examp le , understanding bacterial communication within multispecies communit ies of gut microbiota aids probiotic intervention as a therapeutic strategy for gastrointestinal diseases [40] . Understanding signaling and gene regulation in multispecies biofil ms facilitates their manipulation for biofuel production by simple, cost - effective technologies like solid - state fermentation [4] . Successful application of biofilm - mediated bioremediation to clean up toxic effluents from industrial plants or treat public waste water requires understanding of interaction s within biofilm communities [41] . For example, a mixed biofilm formed by different species of Rhodococcus and Pseudomonas has been successfully use d to clean up toxic chlorophenols like 2,4,6 - tricholophenol, 2,3,4,6 - tetrachlorophenol, pentachlorophenol [42] . Features of signal - responsive gene regulatory networks and the s ignificance of studying gene regulatory networks Bacterial two - component systems are one of the major mechanism s of bacterial signal transduction and are t ypically comprised of a sensor histidine kinase for receiving an input signal and a response regulator which transm its the signal to the level of gene expression [43] . Proper interaction b etween the sensor kinase and its partner response regulator ensures fidelity in transmission of the signal leading to change s in expression of target genes. For example, t ransmission of a quorum signal by the ComA/Com P two - component system promotes transcription of srf genes leading to development of competence under inadequate nutritional condition s in B . subtilis [44] . The m ajority of the response regulators have a DNA - binding domain , and trigger transcription of genes by direct binding to promoter regions as transcription factors [45] . Some response regulators are involv ed in phosphorelay s involving phosphotransfer between proteins before being phosphorylated and activa ting target gene s 7 [43] . Mainly t wo sensor kinases, KinA and KinB , initiate a multicomponent cascade of sequential phosphorylation of downstream proteins, leading to phosphorylation of the transcription factor SpoA , which initiates sporulation in B. subtilis by activating transcription of sporulation genes [46, 47] . In comparison to the relatively simple phosphorelay network that control s initiat ion of sporu lation in B . subtilis , expression of mycobacterial E - dependent stress response genes during infection is under control of a complex GRN involving transcriptional, translational and posttranslational mechanisms [48] . An added layer of complexity in gene regulation is conferred when expression of a target gene of a downstream regulatory module is dependent on the protein product from an upstream regulatory module, leading to formation of enormous GRNs compris ed of multiple smaller regulatory modules. Some fascinating example s of such GRNs are those that control multicellular development of Myx ococcus xanthus and endospore formation of B . subtilis [2, 11] . For example, in the M . xanthus GRN the phosphorylated version of a product from the EBP module, Nla28~P , activates transcription of the mrpAB operon, leading to synthesis of MrpC, which is the product of the Mrp module [2] . An intriguing feature of such multilayered G RNs is regulatory loops, which can be positive or negative, feed - forward or feed - bac k, or autoregulatory [2, 11] . During B. subtilis sporulation, synthesis of the mother cell sigma factor K and the forespore sigma factor G are under the control of independent positive autoregulatory loops to elevate production of the respective regulatory proteins [11] . Autoregulatory loops typically involve a single transcription factor, whereas, feed - forward loops involve at least two, where expression of one transcription factor regulates expression of a second transcription factor, and the two regu lat e transcription of a target gene [11] . In B. 8 subtilis transcription of gerR is activated by E RNAP, and GerR negatively regulates 14 genes of the E regulon, result ing in 14 genes being expressed in a pulse [49] . In a different kind of feed - forward loop, expression of the target gene is under positive control by both transcription factors. For example , during B. subtilis sporulation, E RNAP activates transcription of SpoIIID , which together with E RNAP activates transcription of the gene encoding p ro - K , thus delaying expression of the K regulon until sufficient SpoIIID accumulation , a strategy referred to as [49] . Another example of a feed - forward loop will be discussed in detail in C hapter 2, which involves MrpC and activated FruA combinatorially activating transcription of the developmental dev operon in the M. xanthus GRN [3] . Also i n the M. xanthus GRN, a quorum sensing signal, A - signal, is produced under control of transcription factors in the EBP module, and A - signaling feeds back into the EBP module , thus forming a positive feed - back loop [2] . The regulatory loops of GRN s provide checkpoints, signal amplification, and combinatorial control of gene express ion in order to ensure spatial and temporal regulation of gene expression during stress response s , including development. At the early stage of M. xanthus multicellular development , sequential phosphorylation of EBPs ensures stage - to - stage transition into aggregation, providing checkpoints for furthering the decision to build mounds [50] . I n addition to starvation, the EBP module regulates produc tion of (p)ppGpp and A - signal, which positively feed back into the EBP module, resulting in amplification of the signal s , and providing a strategy to determ ine whether starvation is prolonged enough to initiate development [27, 51] . At the later stages of B. subtilis endospore formation, release of active K into the mother cell requires a signal from the forespore, thus ensur ing that coat prote in assembly around the forespore is suspended until the proper time , an example of strict 9 temporal regulation of gene expression [13, 52] . The regulatory loops significantly contribute to differential gene expression, a strategy to accomplish formation of two separate compartments within a sporangium (a larger mother cell and a smaller fore spore), during endospore formation in B. subtilis [11] . Expression of distinct sigma factor cascades beginning with H and A in the forespore and mother cell , respectively, ensures progression of differential gene expression between the two compartments in B. subtilis [11] . Some other fascinating features of GRN s which play critical role s in spatial and temporal regulation of gene expressi on are ultrasensitivity (nonlinear response s ), irreversibility, and bistability. A large change in the expression of a target gene (output) in response to a small change in the expression of a transcription factor (input) is considered an ultrasensitive or nonlinear response [53, 54] . In B. subtilis the decision to form a spore exhibits an ultrasensitive response t o a threshold concentration of the KinA histidine protein kinase [54] . This is an example of expression of cell fate - determining genes exhibit ing nonlinear outputs in response to a threshold level of the input sensor kinase of the phos phorelay that controls the sporulation decision [43, 47] . In bistable switche s the regulatory system switches between two ultrasensitive thresholds (Off to On and On to Off), rather than resting at an intermediate state [53, 55] . The mechanism of the O ff to O n switch often involves positive autoregulation of the regulator leading to amplification of the input [55] . In other cases of O ff to O n switch es , expression of a target gene is under the control of two mutually repressing repressors (R 1 and R 2 ), when addition of an inducer antagonizes production of R 2, derepressing the target gene, R 1 is produced ensuring repression of R 2 , leaving the target gene on [55] . The stress response mechanism of Mycobacterium tuberculosis exhibits bistability, where the active form of the 10 sigma factor E reaches - sigma factor (RseA), leading to a switch O ff mode of m prAB transcription . Release of active E from RseA, in combination with the positive feedback loop from MprAB to the E promoter , ensures that E reach es an irreversible switch On mode of MprAB production until a significant change in the input signal occurs [53] . Importantly , phenomena whe re expression of a target gene is controlled by multiple regulatory proteins in multiple layers i s not rare in bacterial system s . These systems often rely on alternative sigma factors and EBPs. The earlier discussed stringent response - induced accumulation of RpoS [6] in E. coli and the sporulation - specific sigma factors of B. subtilis are examples of alternative sigma factors, which are related to the major, housekeeping sigma factor (e.g. 70 in E. coli ), but possess distinct promoter recognition propert ies. U nlike 70 and related alternative sigma factors, 54 recognizes conserved sequences centered at - 12 and - 24 relative to the transcriptional start site, and works in coordination with EBPs to transcribe 54 RNAP - dependent genes. A well - studied example of 54 RNAP - dependent genes are activated by the EBP NtrC in response to limited nitrogen availability of cells [56] . Why is gene regulation often multilayered and why must gene expression be so tightly regulated? Expression of large regulons is energetically costly and in some cases becomes irreversible . For example, the E regulon of B. subtilis includes about 270 genes and irreversibly commits the cell to endospore formation [54] . Sporulation appear s to be the last resort for the cell in order to survive starvation . I n order to prevent wasteful usage of the limited resources available to a bacterial cell, conditional expression of stress - responsive genes is tightly regulated and often controlled at multiple levels. 11 The functions of GRN s with all the features mentioned above are not limited to regulating bacterial stress response s [7] , pathog enesis [33] , sporulation and multicellular development [2, 11] , but also community behavior and cell fate determination in bacteria. Multiple i nt erconnected regulon s of quorum - sensing genes regulate population wide gene expression , th ereby enabl ing community - wide coordination of collective behavior in P seudomonas aeruginosa [57] . This is just one ex ample of such coordination in bacterial community behavior , which leads to biofilm formation in many species [58] . Differential gene expression regulated by GRNs govern cell fate deter mination in bacteria. During endospore formation of B. subtilis , different level s of SpoA0~P within the developing population determine cell fate by down - regulating expression of the repressor SinR and inducing expression of the sporulation genes in cells destined to form endospores [59] . There are other bacterial systems where the mechanism of cell fate determination still remains to be understood [2] . Lineage commitment and determination of cell fate are critical proc esses during development of multicellular organism s [60, 61] . For example, p atterned expression of transcription factors under the control of HIPPO signaling determines cell fate decisions during embryonic stem cell development [60] . U nderstanding how GRNs govern decisi on during development of multicellular organisms not only advance s fundamental knowledge, it also provides a foundation to tackle diseases related to erroneous development. S ome of the significant challenges associated with studying relatively complicated GRNs of eukaryotes are the time - intensive genetic manipulation and comparatively higher cost , making it attractive to study simple model organism s to advance our understanding o f GRNs. Additionally , understanding 12 GRNs regulating bacterial stress response s , development, virulence, and antibiotic resistance can directly facilitat e invention of novel therapeutic strategies for infectious and other diseases . Myxococcus xanthus as a model for studying bacterial signaling and multicellularity Although many b acteria can l ead a unicellular life style , many bacteria spend a t least part of their lives in multicellular communities and some choose a multicellular lifestyle almost exclusively . Multicellular ity may arise by aggregation of single cells, chaining and clustering of cells by incomplete cell fission or filamentation from a single cell by cell division arrest [62] . Some of the examples of multicellular behavior in microbes are aerial mycelium formation during sporulation of S treptomyces , f ormation of heterocysts in chains in filamentous cyanobacteria Anabaena , and collective swarming of rod - shaped M. xanthus cells. S ince formation of multicellular structures require shared and unique molecular mechanisms within the population , the process is energetically expensive. Despite th e expense , bacteria can receive several benefits by forming multicellular structures , such as improved resource acquisition and resistance to predation or stress , thus offsetting the costs [62] . M . tuberculosis transitions from a unicellular mode of living to filamentous structures during proliferation inside macrophages as an adaptive response during phagocytosis [62] . In addition to the benefits a bacterium gains from multicellularity, the prevalence of intercellular signaling in bacterial communities makes them fascinating system s to study signal - induced gene reg ulation [23] . Among bacteria which predominantly lead a multicellular life style , the Myxobacteria are extensively studied. Myxobacteria are a group of gram - negative, soil dweller s and are involved in social predation . T he best - characteri z ed species is Myxococcus xanthus. During vegetative 13 growth, M. xanthus builds a multicellular community by organizing high cell density swarms [63] . Upon av ailability of nutrients , M. xanthus exhibits collective predation by secreting extracellular digestive enzymes to lyse the prey leading to cooperative feeding [63] . When p lenty of nutrients are available M . x anthus cells collectively spread over a solid surface to take advantage of the nutrients, a behavior referred as swarming [63] . The type IV p ili - dependent motility, referred as S or social motility , and motility dependent on focal adhesion complexes [64] , referred as A or a dv enturous motility , both are required for swarming of M. xa n thus in the presence of nutrients [63] . Leaving behind extracellular matrix slime trails at the lagging end of a cell for other c ell s to follow is associated with A - motility [20] . S - motility is characterized by coordinated swarming movement of large group s of cells and occurs by extension and retraction of the ty pe IV pili . Activation of highly coordinated S - motility requires close proximity between cells and is facilitated by ex o polysaccharide [20] . S - motility is essential for coordinated predation when plenty of nutrient is around and also for multicellular development when access to nutrient is limited [20] . 14 When the supply of nutrient s runs low , M. xanthus i nitiates a complex developmental program by coordinating cell movement s into mounds (Fig. 1). The program c ulminates in the formation of multicellular fruiting bodies filled with stress - resistant spores [1] . Under starvation conditions, developing cells adhere to the solid surface by forming biofilms, within which cells participate in organized wave - like movements referred as rippling un der certain conditions [1] . When two waves moving in opposite directions collide with each other, they may reflect, but imperfect refl ection may cause a traffic jam , forming high cell density stationary aggregates of cells that lead to mound building. Mounds also form upon fusion of adjacent aggregates [1] . As Figure 1.1 Life cycle of Myxococcus xanthus . Multicellular development is induced upon starvation. During development M. xanthus undergoes several morphological changes including aggregation, mound formation and maturation of fruiting bodies that involves differentiation of stress - resistant myxospores. Upon appearance of nutrients myxospores germinate and undergo vegetative growth. (Adapted from [1] ). 15 more cells enter mounds , they increase in size and eventually contain around 10 5 cells [1] . Finall y, signal ing between closely - packed cells within mounds lead s to differentiation of rods into round spores resistant to environmental insults [1] . The spore - filled mound is called a Also during the developmental process , a majority of cells undergo lysis by a mechanism which is not completely understood [14, 65] and some cells persist as peripheral rods outside of fruiting bodies [15] . Th us , under starvation conditions , integration of intra cellular and extra cellular signals promotes a coordinated sequence of three highly organized multicellular structures (aggregates, mounds , fruiting bodies) and determination of three distinct cell fates (lysing cells, perip heral rods, stress - resistant spores) , in some ways comparable to eukaryotic development [19] . S tarvation - induced multicellular development and Figure 1.2 The GRN governing changes before and during aggregation . The four modules are shown in different colors [enhancer - binding protein cascade (blue), Nla24 module (red), MrpC module (orange), FruA module. Transcription factors are boxed. Arrows and lines with a barred end indicate positive and negative regulation, respectively (Adapter from [2] ). 16 cell fate determination (Fig . 1) of M . xanthus is governed by a highly signal - responsive GRN (Fig . 2) which is striki ngly unique compared to GRNs regulating sporulation of B . subtilis and three species of Streptomyces ( coelicolor , griseus , and venezuelae ) , the other commonly studied model s of bacterial development. The primary distinguishing feature is that u nlike the GRNs regulating sporulation in the other bacteria, in the M. xanthus GRN EBPs play unique roles in regulating transcription of developmental genes in coordination with 54 RNAP [2] . Compared to the sporulation GRN i n Stre ptomyces , the M. xanthus GRN exhibits more instances of combinatorial regulation of target genes [2] . Taken together, the other GRNs appear to be comparatively less signal intensive , perhaps because th ey did not evolve to govern developmental process es as complex as exhibited by M. xanthus [2] . M. xanthus provides an attractive model system to decipher the complications of multicellular development and cell fate decisions. With the added advantage of genetic manipulation of M. xanthus being less time and cost intensive , this bacterium is considered a premium model to study signal - responsive gene regulation in bacteria. The signal - responsive gene regulatory network governing multicellular development of M . xanthus The GRN governing changes before and during aggregation: Th is portion of the GRN can be summarized in four regulatory modules the EBP cascade module, the Nla24 module, the Mrp module and the FruA module [2] (Fig . 2) . U pon starvation, RelA activity leads to accumulation of the secondary messenger molecules penta and tetraphosphate [(p)ppGpp], when ribosome s stall due to amino acid limitation , which is called stringent response [2] . During the stringent response in M. xanthus ( p)ppGpp induces production of two extracellular signals, A - signal and C - signal [66] . The A - signal is a signal for quorum sensing and provides a measure of cell density 17 [51, 67] . W hen a particular quorum is reached the starving cells start expressing early developmental genes in resp onse to A - signal ing . T he identity of C - signal and its mode of action is still under investigation [38, 68] . According to one model , C - signal or C - factor (p17) is a proteolytic product generated by PopC mediated N - terminal cleavage of full - length CsgA ( p25 ) [38, 69] . An alternative model suggests that phospholipase activity of CsgA releases diacylglycerol s from the inner membrane which serves a C - signal and causes cell shortening [68] . In support of this model, purified p17 failed to rescue the developmental defect of a csgA mutant, but addition of purified M. xanthus diacylglyc erols induced formation of dark fruiting bodies by a csgA mutant [68] . During development two distinct threshold level s of C - signal ing are required to achieve aggregation and sporulation [70] , ensuring synchronized development [71] . The EBP cascade module: Within the GRN t he cascade of EBPs is the first module to respond to the starvation signal. EBPs typically bind 100 bp upstream of the promoter and activate transcription of 54 RNAP - dependent genes [56] . EBPs in the cascade module are likely activated by phosphorylation i n a sequential manner, w ith Nla18~P and Nla4~P activat ing transcription of the gene encoding Nla6 , and Nla6~P in turn activati ng the gene coding for Nla28 [50, 72] . Nla6 ~P and Nla28 ~P both are involved in positive autoregulation , providing signal amplification and serv ing to evaluate whether starvation is persistent enough to initiate aggregation [2] . Nla6~P and Nla28~P also regulate each other , providing positive feedback within the EBP cascade and regulat ing production of A - signal and ActB [2] . Products of the act operon regulate the rise of C - signal , which eventually feeds into the FruA module and lead s to aggregation and eventually fruiting bod y formation [73] . Nla28~P feeds into the Mrp module by 18 activating transcription of the mrpAB operon. Therefore, at the preagg regation stage expression of EBP - dependent genes ensures progression to the aggregation stage . The Nla24 module: This module is a relatively recent discovery and is activated by another secondary messenger molecule , cyclic diguanylate (c - di - GMP ), in response to starvation early in development. An i ncrease in the level of c - di - GMP is essential for fruiting body formation , as c - di - GMP binds with its receptor, the EBP Nla24 , leading to stimulation of exopolysaccharide (EPS) synthesis at the preaggregation stage [74] . The Mrp module: T his module is comprise d of three Mrp proteins and at least two starvation - responsive signal transduction pathway s (Pkn and Esp) [2] . MrpA acts as a phosphatase of MrpB~P, which encodes a n EBP, regulating transcription of mrpC [75] . MrpC encodes a transcription factor from the CRP (cAMP receptor protein ) family and is the key output of th e Mrp module [75] . The signal for starvation not only feeds into the EBP cascade but also impacts the Mrp module by posttranslationally affecting MrpC in two ways . T he Pkn STPK cascade can phosphoryla te MrpC, weaking its bind ing to DNA [76] , but recently reported results suggest this has a minor effect on development [77] . Starvation also triggers proteolysis of MrpC via the Esp signaling pathway [78] [79] . During the preaggregation stage, the Esp signaling pathway determine s the pace of development by regulating the concentration of MrpC in the starving cells. A ddition of nutrient medium results in rapid proteolysis of MrpC and blocks commitment to sporulation, su ggesting MrpC is a mediator of the starvation signal and serves as a checkpoint conveying persistent starvation [80] . MrpC accumulate s to a higher level in aggregating cells than in non - aggregating cells [14] . MrpC serves as a key transcriptional regulator of genes, including genes critical for aggregatio n , given that an mrpC mutant fails to 19 aggregate [3, 77] . MrpC negatively autoregulates its transcription at the level of synthesis [3, 77] and upregulates or downregulates transcri ption of nearly 300 developmental genes by direct binding [81] . Negative autoregulation of MrpC which is a major transcriptional regulator of M xanthus development perhaps shortens the response time of genetic regulatory circuits and decreases variability in gene expression between cells undergoing development [77] . MrpC activates transcription of the gene encoding FruA [3, 82] , a nother key transcription factor for developmental gene expression [3, 82] . The FruA module: FruA relies on C - signaling to activate it for transcription of downstream genes in the GRN [83, 84] . As described above, the mechanism of C - signaling is controversial, but it is clear that cells must be aligned to engage in C - signaling, which in turn promotes aggregation , f urther alignment of cells, and more C - signaling [37, 85, 86] . C - signal appear s to activate FruA posttranslationally [83, 84] , represented as formation of FruA * in Figure 2. FruA * is proposed to mediat e cellular response s to C - signal ing both individually [87] and combinatorially with MrpC [3, 18] by regulat ing transcription of genes whose products ensu re aggregation , mound formation, and eventually sporulation as the level s of C - signal ing and FruA * rise [70, 71] . H ow C - signal ing activates FruA remains to be elucidated . Earlier work s howed that FruA is similar to response regulator s of two - component system s and suggested that C - signal ing leads to phosphorylation of FruA [83, 84] . However, typically a response regulator is phos phorylated by a protein kinase, which ha s not been identified for FruA. The atypical response regulator domain of FruA lacks some aspartate residues normally required for phosphorylation. Additionally, treatment with small molecule phosphodonor s fail ed to increase DNA - binding ability of FruA [16] . Altogether , the evidence suggest ed that phosphorylation wa s un likely to be 20 the mechanism by which FruA is activated. In C hapter 2 we show additional evidence suggesting FruA is not activated by phosphorylation [3] . The GRN governing sporulation : Transcription factors from the modules described above ensure expression of developmental genes essential for the pre - aggregation and aggregation stages of development. In particular, MrpC and C - signal - dependent FruA * regulate transcription of genes whose products ensure progression of development from the sta ge of aggregation to the eventual completion of spore formation. MrpC and FruA integrate the two major signa ling inputs , starvation and C - signal ( serving as a spatial coordinator), respectively, and regulate expression of downstream genes of the GRN (Fig. 3) . Combinatorial regulation by MrpC and FruA * i ntegrates the signal for starvation and C - signal ing, respectively , thus ensur ing that only the starving cells capable of accumulating MrpC , and cells also in close proximity within mounds an d therefore capable of C - signaling and accumulating FruA * , commit to spore formation [16, 80] ) . Among the C - signal - dependent genes which are important for sporulation and are under combinatorial control of MrpC and FruA * a r e genes o f the dev operon [18] (Fig. 3) . The dev operon includes a CRISPR - Cas system and is proposed to protect M. xanthus cells from phage infection during multicellular development [3, 88] . Three genes of dev operon , devTRS , negatively autoregulat e transcription tenfold [3, 88] . The p roduct of the first gene , devI , is a small 40 - residue protein that inhibits sporulation [89, 90] . Hence, the mutant lacking devI forms sonication - resistant spore s 6 h earlier in comparison to the wild - type strain [3, 89, 90] . In contrast , mounds for med by the devTRS mutant s d o not darken and these mutants are impaired in spore formation [3, 88, 91] . These findings indicate that in devTRS mutants the lack 21 of the negative autoregulation leads to overproduction of DevI and result s in a sporulation defect . In agreem ent, sporulation of devTRS mutants can be restored by a null mutation in devI [89] . In C hapter 2 we investigate the effect s of C - signal ing, FruA, and MrpC on the dev transcript level [ 3] . Other C - signal - dependent genes of the GRN act late during the sporulation stage of the developmental process. In Fig ure 4 , exo and nfs represent nine ( exoA - I ) and eight (nfsA - H) gene operon s whose products help build the spore coat, and hence are critical for completion of spore formation [92, 93] . The protein products of three genes of exo operon ( exoA , exoB and exoC ) appear to form a terminal transport complex spanning from the cytoplasmic membrane Figure 1.3 The GRN governing sporulation with the output gene dev . Figure shows GRN governing sporulation in M. xanthus with one of the outputs dev . The key transcription factors MrpC and FruA integrate starvation signal and the C - signal ( the signal for spatial coordination) into the network. In response to starvation, MrpC negatively autoregulates at the level of transcription. Transcription of fruA is activated by MrpC. In response to C - signal, activated FruA (FruA * ) cooperatively with MrpC activate transcription of dev operon. One of the genes of dev operon, DevI, if overproduced delays spore formation, whereas, DevTRS proteins negatively autoregulate transcription. (Adapted from [3] ) . 22 to the outer membrane [93] . exoC encodes a PCP - 2a - like polysaccharide copolymerase family protein likely responsible for the export of spore coat material [92 - 94] . Developmental transcript ion of exo appears to be activated by direct binding of FruA * in response to C - signal ing [87] . Strikingly, the EBP Nla6 from the EBP module which regulates transcription of genes during the pre - aggregation stage , has been shown to bind to the exo promoter region [72] . Nla6 appears to activate transcription of exo at the stage when aggregation begins , but negatively regulate exo trans cription closer to the tim e of sonication - resistant spore formation [72] . In C hapter 3 we report novel aspects of regulatio n of exo transcription by MrpC and FruA which has not been activated by C - signaling (Fig . 4) . In C hapter 3 we also provide evidence that products of the exo operon are critical for sonication - resistant spore formation . Products of the nfs operon are critical for assembling the spore coat material once it has been exported to the cell surface [93] . In C hapter 3 we elucidate the role s of MrpC and unactivated FruA in regulating nfs transcription . In Chapter 3 w e also report studies of two other late - acting operons of the GRN (Fig . 4) . MXAN_3259 is predicted to encod e a polysaccharide deacetylase [72] and MXAN_5372 is predicted to encod e a FadI homolog involved in fatty acid - oxidation d uring spore formation [95] . Similar t o exo , transcription of MXAN_3259 is activated by Nla6 at the beginning of aggregation , but negatively regulated by Nla6 closer to the time of s p ore formation , and m utation s i n MXAN_3259 [72] or exoC [92, 96] caused similar sporulation defects . In C hapter 3 we elucidate the role s of MrpC , unactivated FruA, and activated FruA * in regulating developmental transcription of MXAN_3259 and MXAN _ 5372 ( fadI in Fig. 4) . Our findings reported in C hapter 3 indicate novel role s of unactivated FruA and MrpC in negative regulation, 23 and together with positive regulation by FruA * produce differential control of the four late - acting operon s (Fig . 4) . We propose that upon reaching a distinct threshold for sporulation, C - signal ing posttranslationally activate s Fru A, and FruA * increases transcription of genes essential for completion of spore formation. We hypothesize that by integrating the starvation signal and the signal for spatial coordina tion (C - signal) the GRN (Fig . 3 and 4) govern s the decision to form a spore during the commitment period. The decision to form a spore include s the molecular changes lead ing to spore formation. These changes are suggested to occur between 24 30 h after t he onset of starvation. The t ime between 24 - 30 h poststarvation is defined as the commitment period, during which an increasing number of M. xanthus cells actively convert to spores despite perturbing the starvation signal by adding nutrients [80] . Therefore, in order to understand changes leading to spore formation, work described in this dissertation attempt ed to elucidate the gene expression dynamics of the GRN (Fig . 3 and 4) during the commitment period. Previous attempts to study the GRN governing M. xanthus multicellular development involved usage of multiple strains, different conditions of development, and phenotypic rather than molecular approaches, without fine time resolution [75, 8 2, 87] . These factors made i t difficult to decipher the molecular complexity of M. xanthus development. To over come these challenges, systematic and quantitative experimental approaches need to be combined with computational methods to build mathematical models of GRNs that can predict novel outcomes. These outcomes are often testable by wet lab experiments , resulting in refinement of the existing mathematical models and formulation of novel testable hypothes e s. Thus, systematic and quantitative experimental approaches in combination with computational 24 model ing are advantageous in contributing to the understa nding of GRNs in a time efficient manner. C hapter 2 describes systematic and quantitative experimental approaches coupled with a computational model designed to help elucidate GRN function during the commitment period of M. xanthus development. Systematic analysis was performed with fine time resolution to build a computational model where the dev transcript level wa s the output and MrpC and FruA were inputs (Fig . 3) . The model was used to make predictions related to hypothes e s formulated to explain an observed large change in the dev transcript level in a csgA mutant despite a much small er change in the level of FruA [3] . Our systematic experimental analysis in combination with mathematical model ing supports the hypothesis that C - signal ing activa tes FruA at least n inefold posttranslationally in order to increase transcription of the dev operon and commit cells to spore formation [3] . The project took a striking turn when o ur systematic analysis rev eal ed unexpected changes in the expression levels of the late genes of the GRN ( exo, nfs, MXAN_3259 and fadI ) in the absence of MrpC and FruA , indicating novel roles of the se transcription factors in regulating the late genes during commitment (Fig . 4) . Th ese initial finding s were explored the work described in C hapter 3 to elucidate potential molecular mechanism s of the late gene regulation by Nla6, FruA and MrpC. Chapter 3 provides novel insights into the function and differential regulation of the late gene s . 25 Figure 1.4 Model of differential regulation of late genes at two different morphological stages during development . Starvation increases the MrpC level which in turn increases the FruA level. C - signal activates FruA to FruA*. Around the time of mound formation unactivated FruA predominates and around the time of spore formation activated FruA (FruA*) predominates. Posit ive regulation (yellow arrows) and negative regulation (blue line with blunt) of late genes (gray boxes) is indicated. During mound formation between 6 and 18 h poststarvation, Nla6 positively regulates transcription of all four late genes (dashed box), b ut unactivated FruA and MrpC negatively regulate certain late genes as indicated. During spore formation between 24 and 36 h, activated FruA* induces transcription of the dev operon gene. DevS (and DevT and DevR, which are not shown) negatively autoregula tes transcription of devI . DevI negatively regulates all four late genes (dashed box), but FruA* positively regulates their transcription, while MrpC positively or negatively regulates certain late genes as indicated. 26 REFERENCES 27 REFERENCES 1. Kaiser D: Coupling cell movement to multicellular development in myxobacteria. Nature reviews Microbiology 2003, 1(1):45 - 54. 2. Kroos L: Highly Signal - Responsive Gene Regulatory Network Governing Myxococcus Development. Trends in genetics : TIG 2017, 33(1):3 - 15. 3. Saha S, Patra, P, Igoshin, O, Kroos, L.: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C - signaling. Molecular microbiology 2019, 111(6):1732 - 1752. 4. Wang ZW, Chen, S.: Potential of biofilm - based biofuel production. Applied microbiology and biotechnology 2009, 83(1):1 - 18. 5. Hall - Sto odley L, Stoodley, P.: Evolving concepts in biofilm infections. Cellular microbiology 2009, 11(7):1034 - 1043. 6. Battesti A, Majdalani, N, Gottesman, S.: The RpoS - mediated general stress response in Escherichia coli. Annual review of microbiology 2011, 65:189 - 213. 7. Schweder T, Lee, K. H, Lomovskaya, O, Matin, A.: Regulation of Escherichia coli starvation sigma fac tor (sigma s) by ClpXP protease. Journal of bacteriology 1996, 178(2):470 - 476. 8. Zhou Y, Gottesman, S, Hoskins, J. R, Maurizi, M. R, Wickner, S.: The RssB response regulator directly targets sigma(S) for degradation by ClpXP. Genes & development 2001, 15 (5):627 - 637. 9. Bougdour A, Gottesman, S.: ppGpp regulation of RpoS degradation via anti - adaptor protein IraP. Proceedings of the National Academy of Sciences of the United States of America 2007, 104(31):12896 - 12901. 10. Stragier P, Losick, R.: Molecula r genetics of sporulation in Bacillus subtilis. Annual review of genetics 1996, 30:297 - 241. 11. Kroos L: The Bacillus and Myxococcus developmental networks and their transcriptional regulators. Annual review of genetics 2007, 41:13 - 39. 12. Halder S, Parr ell, D, Whitten, D, Feig, M, Kroos, L.: Interaction of intramembrane metalloprotease SpoIVFB with substrate Pro - sigma(K). Proceedings of the National Academy of Sciences of the United States of America 2017, 114(50):E10677 - E10686. 28 13. Cutting S, Driks, A, Schmidt, R, Kunkel, B, Losick, R.: Forespore - specific transcription of a gene in the signal transduction pathway that governs Pro - sigma K processing in Bacillus subtilis. Genes & development 1991, 5(3):456 - 466. 14. Lee B, Holkenbrink, C, Treuner - Lange, A , Higgs, P. I.: Myxococcus xanthus developmental cell fate production: heterogeneous accumulation of developmental regulatory proteins and reexamination of the role of MazF in developmental lysis. Journal of bacteriology 2012, 194(12):3058 - 3068. 15. O'Con nor KA, Zusman, D. R.: Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. Journal of bacteriology 1991, 173(11):3318 - 3333. 16. Mittal S, Kroos, L.: A combination of unusual transcription factors bin ds cooperatively to control Myxococcus xanthus developmental gene expression. Proceedings of the National Academy of Sciences of the United States of America 2009, 106(6):1965 - 1970. 17. Mittal S, Kroos, L.: Combinatorial regulation by a novel arrangement of FruA and MrpC2 transcription factors during Myxococcus xanthus development. Journal of bacteriology 2009, 191(8):2753 - 2763. 18. Campbell A, Viswanathan, P, Barrett, T, Son, B, Saha, S, Kroos, L.: Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. Journal of bacteriology 2015, 197(2):240 - 251. 19. Kroos L: Eukaryotic - like signaling and gene regulation in a prokaryote that undergoes multicellular development. Proceedings of the National Academy of Sciences of the United States of America 2005, 102(8):2681 - 2682. 20. Munoz - Dorado J, Marcos - Torres, F. J, Garcia - Bravo, E, Moraleda - Munoz, A, Perez, J.: Myxobacteria: Moving, Killing, Feeding, and Surviving Together. Frontiers in microbiology 2016, 7:781. 21. Jakobsen JS, Jelsbak, L, Jelsbak, L, Welch, R. D, Cummings, C, G oldman, B, Stark, E, Slater, S, Kaiser, D.: Sigma54 enhancer binding proteins and Myxococcus xanthus fruiting body development. Journal of bacteriology 2004, 186(13):4361 - 4368. 22. Jelsbak L, Givskov, M, Kaiser, D.: Enhancer - binding proteins with a forkhe ad - associated domain and the sigma54 regulon in Myxococcus xanthus fruiting body development. Proceedings of the National Academy of Sciences of the United States of America 2005, 102(8):3010 - 3015. 23. Bassler BL, Losick, R.: Bacterially speaking. Cell 20 06, 125(2):237 - 246. 29 24. Waters CM, Lu, W, Rabinowitz, J. D, Bassler, B. L.: Quorum sensing controls biofilm formation in Vibrio cholerae through modulation of cyclic di - GMP levels and repression of vpsT. Journal of bacteriology 2008, 190(7):2527 - 2536. 25 . de Kievit TR: Quorum sensing in Pseudomonas aeruginosa biofilms. Environmental microbiology 2009, 11(2):279 - 288. 26. Deziel E, Gopalan, S, Tampakaki, A. P, Lepine, F, Padfield, K. E, Saucier, M, Xiao, G, Rahme, L. G.: The contribution of MvfR to Pseudomonas aeruginosa pathogenesis and quorum sensing circuitry regulation: multiple quorum sensing - regulated genes are modula ted without affecting lasRI, rhlRI or the production of N - acyl - L - homoserine lactones. Molecular microbiology 2005, 55(4):998 - 1014. 27. Kuspa A, Plamann, L, Kaiser, D.: A - signalling and the cell density requirement for Myxococcus xanthus development. Journ al of bacteriology 1992, 174(22):7360 - 7369. 28. Hauryliuk V, Atkinson, G. C, Murakami, K. S, Tenson, T, Gerdes, K.: Recent functional insights into the role of (p)ppGpp in bacterial physiology. Nature reviews Microbiology 2015, 13(5):298 - 309. 29. Roberts JW: Promoter - specific control of E. coli RNA polymerase by ppGpp and a general transcription factor. Genes & development 2009, 23(2):143 - 146. 30. Dalebroux ZD, Svensson, S. L, Gaynor, E. C, Swanson, M. S.: ppGpp conjures bacterial virulence. Microbiology and molecular biology reviews : MMBR 2010, 74(2):171 - 199. 31. Poole K: Bacterial stress responses as determinants of antimicrobial resistance. The Journal of antimicrobial chemotherapy 2012, 67(9):2069 - 2089. 32. Geiger T, Francois, P, Liebeke, M, Fraunh olz, M, Goerke, C, Krismer, B, Schrenzel, J, Lalk, M, Wolz, C.: The stringent response of Staphylococcus aureus and its impact on survival after phagocytosis through the induction of intracellular PSMs expression. PLoS pathogens 2012, 8(11):e1003016. 33. He H, Cooper, J. N, Mishra, A, Raskin, D. M.: Stringent response regulation of biofilm formation in Vibrio cholerae. Journal of bacteriology 2012, 194(11):2962 - 2972. 34. Boutte CC, Crosson, S.: Bacterial lifestyle shapes stringent response activation. Trends in microbiology 2013, 21(4):174 - 180. 35. Diaz - Salazar C, Calero, P, Espinosa - Portero, R, Jimenez - Fernandez, A,, Wirebrand L, Velasco - Dominguez, M. G, Lopez - Sanchez, A , Shingler, V, Govantes, F.: The stringent response promotes biofilm dispersal in Pseudomonas putida. Scientific reports 2017, 7(1):18055. 30 36. Crawford EW, Jr, Shimkets, L. J.: The Myxococcus xanthus socE and csgA genes are regulated by the stringent respo nse. Molecular microbiology 2000, 37(4):788 - 799. 37. Kim SK, Kaiser, D.: C - factor: a cell - cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 1990, 61(1):19 - 26. 38. Lobedanz S, Sogaard - Andersen, L.: Identification of the C - signal, a contact - dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes & development 2003, 17(17):2151 - 2161. 39. Kroos L, Kaiser, D.: Expression of many developmentally regulated genes in Myxococcus depends on a sequence of cell interactions. Genes & development 1987, 1(8):840 - 854. 40. Rauch M, Lynch, S. V.: The potential for probiotic manipulation of the gastrointestinal microbiome. Current opinion in biotechnology 2012, 23(2):192 - 201. 41. Singh R, Paul, D, Jai n, R. K.: Biofilms: implications in bioremediation. Trends in microbiology 2006, 14(9):389 - 397. 42. Puhakka JA, Herwig, R. P, Koro, P. M, Wolfe, G. V, Ferguson, J. F.: Biodegradation of chlorophenols by mixed and pure cultures from a fluidized - bed reactor . Applied microbiology and biotechnology 1995, 42(6):951 - 957. 43. Mitrophanov AY, Groisman, E. A.: Signal integration in bacterial two - component regulatory systems. Genes & development 2008, 22(19):2601 - 2611. 44. Mueller JP, Bukusoglu, G, Sonenshein, A. L.: Transcriptional regulation of Bacillus subtilis glucose starvation - inducible genes: control of gsiA by the ComP - ComA signal transduction system. Journal of bacteriology 1992, 174(13):4361 - 4373. 45. Gao R, Stock, A. M.: Biological insights from structures of two - component proteins. Annual review of microbiology 2009, 63:133 - 154. 46. Sonenshein AL: Control of sporulation initiation in Bacillus subtilis. Current opinion in microbiology 2000, 3(6):561 - 5 66. 47. Tojo S, Hirooka, K, Fujita, Y.: Expression of kinA and kinB of Bacillus subtilis, necessary for sporulation initiation, is under positive stringent transcription control. Journal of bacteriology 2013, 195(8):1656 - 1665. 48. Manganelli R, Provvedi, R.: An i ntegrated regulatory network including two positive feedback loops to modulate the activity of sigma(E) in mycobacteria. Molecular microbiology 2010, 75(3):538 - 542. 31 49. Eichenberger P, Fujita, M, Jensen, S. T, Conlon, E. M, Rudner, D. Z, Wang, S. T, Fergu son, C, Haga, K, Sato, T, Liu, J. S, Losick, R.: The program of gene transcription for a single differentiating cell type during sporulation in Bacillus subtilis. PLoS biology 2004, 2(10):e328. 50. Giglio KM, Caberoy, N, Suen, G, Kaiser, D, Garza, A. G.: A cascade of coregulating enhancer binding proteins initiates and propagates a multicellular developmental program. Proceedings of the National Academy of Sciences of the United States of America 2011, 108(32):E431 - 439. 51. Kuspa A, Plamann, L, Kaiser, D. : Identification of heat - stable A - factor from Myxococcus xanthus. Journal of bacteriology 1992, 174(10):3319 - 3326. 52. Dong TC, Cutting, S. M.: The PDZ domain of the SpoIVB transmembrane signaling protein enables cis - trans interactions involving multiple partners leading to the activation of the pro - sigmaK processing complex in Bacillus subtilis. The Journal of biological chemistry 2004, 279(42):43468 - 43478. 53. Ray JC, Tabor, J. J, Igoshin, O. A.: Non - transcriptional regulatory processes shape transcript ional network dynamics. Nature reviews Microbiology 2011, 9(11):817 - 828. 54. Narula J, Devi, S. N, Fujita, M, Igoshin, O. A.: Ultrasensitivity of the Bacillus subtilis sporulation decision. Proceedings of the National Academy of Sciences of the United Sta tes of America 2012, 109(50):E3513 - 3522. 55. Dubnau D, Losick, R.: Bistability in bacteria. Molecular microbiology 2006, 61(3):564 - 572. 56. Bush M, Dixon, R.: The role of bacterial enhancer binding proteins as specialized activators of sigma54 - dependent transcription. Microbiology and molecular biology reviews : MMBR 2012, 76(3):497 - 529. 57. Schuster M, Greenberg, E. P.: A network of networks: quorum - sensing gene regulation in Pseudomonas aeruginosa. International journal of medical microbiology : IJMM 2 006, 296(2 - 3):73 - 81. 58. Svenningsen SL: Small RNA - Based Regulation of Bacterial Quorum Sensing and Biofilm Formation. Microbiology spectrum 2018, 6(4). 59. Tan IS, Ramamurthi, K. S.: Spore formation in Bacillus subtilis. Environmental microbiology repor ts 2014, 6(3):212 - 225. 60. Frum T, Ralston, A.: Cell signaling and transcription factors regulating cell fate during formation of the mouse blastocyst. Trends in genetics : TIG 2015, 31(7):402 - 410. 32 61. Levine M, Davidson, E. H.: Gene regulatory networks for development. Proceedings of the National Academy of Sciences of the United States of America 2005, 102(14):4936 - 4942. 62. Claessen D, Rozen, D. E, Kuipers, O. P, Sogaard - Andersen, L, van Wezel, G. P.: Bacterial solutions to multicellularity: a tale of biofilms, filaments and fruiting bodies. Nature reviews Microbiology 2014, 12(2):115 - 124. 63. Kaiser D, Robinson, M, Kroos, L .: Myxobacteria, polarity, and multicellular morphogenesis. Cold Spring Harbor perspectives in biology 2010, 2(8):a000380. 64. Mignot T: The elusive engine in Myxococcus xanthus gliding motility. Cellular and molecular life sciences : CMLS 2007, 64(21):27 33 - 2745. 65. Nariya H, Inouye, M.: MazF, an mRNA interferase, mediates programmed cell death during multicellular Myxococcus development. Cell 2008, 132(1):55 - 66. 66. Konovalova A, Lobach S, Sogaard - Andersen L: A RelA - dependent two - tiered regulated prote olysis cascade controls synthesis of a contact - dependent intercellular signal in Myxococcus xanthus. Molecular microbiology 2012, 84(2):260 - 275. 67. Kuspa A, Plamann L, Kaiser D: A - signalling and the cell density requirement for Myxococcus xanthus develop ment. Journal of bacteriology 1992, 174(22):7360 - 7369. 68. Boynton TO, Shimkets, L. J.: Myxococcus CsgA, Drosophila Sniffer, and human HSD10 are cardiolipin phospholipases. Genes & development 2015, 29(18):1903 - 1914. 69. Rolbetzki A, Ammon, M, Jakovljevi c, V, Konovalova, A, Sogaard - Andersen, L.: Regulated secretion of a protease activates intercellular signaling during fruiting body formation in M. xanthus. Developmental cell 2008, 15(4):627 - 634. 70. Kim SK, Kaiser, D.: C - factor has distinct aggregation and sporulation thresholds during Myxococcus development. Journal of bacteriology 1991, 173(5):1722 - 1728. 71. Kruse T, Lobedanz, S, Berthelsen, N. M, Sogaard - Andersen, L.: C - signal: a cell surface - associated morphogen that induces and co - ordinates multice llular fruiting body morphogenesis and sporulation in Myxococcus xanthus. Molecular microbiology 2001, 40(1):156 - 168. 72. Giglio KM, Zhu, C, Klunder, C, Kummer, S, Garza, A. G.: The enhancer binding protein Nla6 regulates developmental genes that are impo rtant for Myxococcus xanthus sporulation. Journal of bacteriology 2015, 197(7):1276 - 1287. 33 73. Gronewold TM, Kaiser, D.: The act operon controls the level and time of C - signal production for Myxococcus xanthus development. Molecular microbiology 2001, 40(3 ):744 - 756. 74. Skotnicka D, Smaldone, G. T, Petters, T, Trampari, E, Liang, J, Kaever, V., Malone JG, Singer, M, Sogaard - Andersen, L.: A Minimal Threshold of c - di - GMP Is Essential for Fruiting Body Formation and Sporulation in Myxococcus xanthus. PLoS gen etics 2016, 12(5):e1006080. 75. Sun H, Shi, W.: Genetic studies of mrp, a locus essential for cellular aggregation and sporulation of Myxococcus xanthus. Journal of bacteriology 2001, 183(16):4786 - 4795. 76. Nariya H, Inouye, S.: 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. Molecular microbiology 2006, 60(5):1205 - 1217. 77. McLaughlin PT, Bhardwaj, V, Feeley, B. E, Higgs, P. I.: MrpC, a CRP/Fnr homolog, functions as a negative autoregulator during the Myxococcus xanthus multicellular developmental program. Molecular microbiology 2018. 78. Higgs PI, Jagadeesan, S, Mann, P, Zusman, D. R.: EspA, an orphan hybrid histidine protein kinase, regulates the timing of expressi on of key developmental proteins of Myxococcus xanthus. Journal of bacteriology 2008, 190(13):4416 - 4426. 79. Schramm A, Lee, B, Higgs, P. I.: Intra - and interprotein phosphorylation between two - hybrid histidine kinases controls Myxococcus xanthus developm ental progression. The Journal of biological chemistry 2012, 287(30):25060 - 25072. 80. Rajagopalan R K, L.: Nutrient - regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. Journal of bacteriology 2014, 196(15):2736 - 2747. 81. Robinson M, Son, B, Kroos, D, Kroos, L.: Transcription factor MrpC binds to promoter regions of hundreds of developmentally - regulated genes in Myxococcus xanthus. BMC genomics 2014, 15:1123. 82. Ueki T, Inouye, S.: Identification of an activator protein required for the induction of fruA, a gene essential for fruiting body development in Myxococcus xanthus. Proceedings of the National Academy of Sciences of the United States of America 2003, 100(15):8782 - 8787. 83. Ogawa M, Fujitani, S, Mao, X, Inouye, S, Komano, T.: FruA, a putative transcri ption factor essential for the development of Myxococcus xanthus. Molecular microbiology 1996, 22(4):757 - 767. 34 84. Ellehauge E, Norregaard - Madsen, M, Sogaard - Andersen, L.: The FruA signal transduction protein provides a checkpoint for the temporal co - ordina tion of intercellular signals in Myxococcus xanthus development. Molecular microbiology 1998, 30(4):807 - 817. 85. Kim SK, Kaiser, D.: Cell motility is required for the transmission of C - factor, an intercellular signal that coordinates fruiting body morphog enesis of Myxococcus xanthus. Genes & development 1990, 4(6):896 - 904. 86. Kim SK, Kaiser, D.: Cell alignment required in differentiation of Myxococcus xanthus. Science 1990, 249(4971):926 - 928. 87. Ueki T, Inouye, S.: Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development. The Journal of biological chemistry 2005, 280(37):32279 - 32284. 88. Viswanathan P, Murphy, K, Julien, B, Garza, A. G, Kroos, L.: Regulation of dev, an operon that includes genes essential for Myxococcus xanthus development and CRISPR - associated genes and repeats. Journal of bacteriology 2007, 189(10):3738 - 3750. 89. Rajagopalan R, Wielgoss, S, Lippert, G, Velicer, G. J, Kroos, L.: devI is an evolutionarily young negative regulator of Myxococcus xanthus development. Journal of bacteriology 2015, 197(7):1249 - 1262. 90. Rajagopalan R, Kroos L: The dev Operon Regulates th e Timing of Sporulation during Myxococcus xanthus Development. Journal of bacteriology 2017, 199(10). 91. Thony - Meyer L, Kaiser, D.: devRS, an autoregulated and essential genetic locus for fruiting body development in Myxococcus xanthus. Journal of bacter iology 1993, 175(22):7450 - 7462. 92. Muller FD, Schink, C. W, Hoiczyk, E, Cserti, E, Higgs, P. I.: Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Molecular microbiology 2012, 83(3):486 - 505. 93. Holkenbrink C, Hoiczyk, E, Kahnt, J, Higgs, P. I.: Synthesis and assembly of a novel glycan layer in Myxococcus xanthus spores. The Journal of biological chemistry 2014, 289(46):32364 - 32378. 94. Cuthbertson L, Mainprize, I. L, Naismith, J. H, Whitfi eld, C.: Pivotal roles of the outer membrane polysaccharide export and polysaccharide copolymerase protein families in export of extracellular polysaccharides in gram - negative bacteria. Microbiology and molecular biology reviews : MMBR 2009, 73(1):155 - 177. 35 95. Bhat S, Boynton, T. O, Pham, D, Shimkets, L. J.: Fatty acids from membrane lipids become incorporated into lipid bodies during Myxococcus xanthus differentiation. PloS one 2014, 9(6):e99622. 96. Licking E, Gorski, L, Kaiser, D.: A common step for cha nging cell shape in fruiting body and starvation - independent sporulation of Myxococcus xanthus. Journal of bacteriology 2000, 182(12):3553 - 3558 . 36 C HAPTER 2: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C - signaling This chapter was published in the journal of Molecular Microbiology and can be found at the following citation. 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 (6): 1732 - 1752. I contributed significantly to the writing, editig and preparing all the figures for the manuscript. Data from all the wet - lab experiements were acquired and analy zed by me. Dr Pintu Patra from the research group of Dr Oleg Igoshin (Rice university, Housto n) prfommed all the work related to the mathematical mo del. Molecular micro provides authors the right to use and reprint published work without written permission as part of a dissertaion, give the work is properly cited (see above). The publication reprint and the supplementary material is provided on the following pages. 37 Abstr act Upon starvation Myxococcus xanthus undergoes multicellular development. Rod - shaped cells move into mounds in which some cells differentiate into spores. Cells begin committing to sporulation at 24 - 30 h poststarvation, but the mechanisms governing co mmitment are unknown. FruA and MrpC are transcription factors that are necessary for commitment. They bind cooperatively to promoter regions and activate developmental gene transcription, including that of the dev operon. Leading up to and during the commitment period, dev mRNA increased in wild type, but not in a mutant defective in C - signaling, a short - range signaling interaction between cells that is also necessary for commitment. T he C - signaling mutant exhibited ~20 - fo ld less dev mRNA than wild type at 30 h poststarvation, despite a similar level of MrpC and only twofold less FruA. Boosting the FruA level twofold in the C - signaling mutant had little effect on the dev mRNA level, and dev mRNA was not less stable in the C - signaling mutant. Neither did high cooperativity of MrpC and FruA binding upstream of the dev promoter explain the data. Rather, our systematic experimental and computational analyses support a model in which C - signaling activates FruA at least ninefol d posttranslationally in order to commit a cell to spore formation. 38 Introduction Differentiated cell types are a hallmark of multicellular organisms. Understanding how pluripotent cells become restricted to particular cell fates is a fascinating qu estion and a fundamental challenge in biology. In general, the answer involves a complex interplay between signals and gene regulation. This is true both during development of multicellular eukaryotes [1 - 3] and during transitions in microbial communities that lead to different cell types [4 - 7] . Bacterial cells in microbial communities adopt different fates as gene regulatory networks (GRNs) respond to a variety of signals, including some generated by other cells. Moreover, we now understand that microbial communities or microbiomes profoundly impact eukaryotic organisms, and vice versa [8, 9] . Yet the daunting complexity of microbiomes and multicellular eukaryotes impedes efforts to fully understand th eir interactions in molecular detail. By studying simpler model systems, paradigms can be discovered that can guide investigations of more complex interactions. A relatively simple model system is provided by the bacterium Myxococcus xanthus , which underg oes starvation - induced multicellular development [10] . In response to starvation, cells generate intr acellular and extracellular signals that regulate gene expression [7, 11] . The rod - shaped cells alter their movements so that thousands form a mound. Within a mound, cells differentiate into ovoid spores that resist stress and remain dormant until nutrients reappear. The spore - filled mound is called a fruiting body. Other cells adopt a diffe rent fate and remain outside the fruiting body as peripheral rods [12] . A large proportion of the cells lyse during the develop mental process [13] . What determines whether a given cell in the population forms a spore, remains as a peripheral rod, or undergoes lysis? M. xanthus provides an attractive 39 model system to discover how signaling between cells affects a GRN and determines cell fate. Here, we focus on a circuit that regulates commitment to sporulation. In a recent study, cells committed to spore formation prima rily between 24 and 30 h poststarvation (PS), because addition of nutrients to the starving population prior to 24 h PS blocked subsequent sporulation, addition at 24 h PS allowed a few spores to form subsequently, and addition at 30 h PS allowed about ten fold more spores to form [14] . At the molecular level, addition of nutrients before or during the commitment period caused rapid proteolysis of MrpC [14] , a transcription factor required for fruiting body formation [15, 16] . MrpC appears to directly regulate more than one hundred genes involved in development [17] , and one we ll - characterized MrpC target gene, fruA [18] , codes for another transcription factor required for fruiting body formation [19] . FruA and MrpC bind cooperatively to the promoter regions of many genes, and appear to activate transcription [17, 20 - 24] . In particular, transcription of the dev operon appears to be activated by cooperative binding of the two transcription factors at two sites located upstream of the promoter [20] . Because mutations in three genes of the dev operon ( devTRS ) strongly impair sporulation [25 - 27] , the fee d - forward loop involving MrpC and FruA regulation of the dev operon is an attractive molecular mechanism to control spore formation (Fig. 1). Recent work revealed that products of the dev operon act as a timer for sporulation [28] . DevTRS negatively autoregulate expression of DevI, which inhibits sporulation if overproduced, and delays sporulation by about 6 h when produced normally [28, 29] (Fig. 1). Expression of the dev operon and many other developmental genes depends on C - signaling [30] , which has been proposed to activate FruA [31] and/or MrpC [22] (Fig. 1), although the 40 mechanism of C - signal transduction remains a mystery. Null mut ations in the csgA gene block C - signaling and sporulation, but the mutants can be rescued by co - development with csgA + cells which supply the C - signal [32] . C - signaling appears to be a short - range signaling interaction that req uires cells to move into alignment [33 - 35] , as they do during mound formation [36] . Two theories about the identity of the C - signal have emerged. One theory states that the C - signal is a 17 - kDa fragment of CsgA produced by the specific proteolytic activity of PopC at the cell surface [37 - 39] . Figure 2. 1 Simplified model of the gene regulatory network governing formation of fruiting bodies . Starvation increases the level of MrpC early in the process [40 - 42] . MrpC ca uses an increase in C - signal [41] , the produc t of csgA [43, 44] . MrpC activates transcription of the gene for FruA [45] and C - signal somehow enhances FruA [46] and/or MrpC activity [47] . MrpC and FruA bind cooperatively to the promoter region of the dev operon and activate transcription [48] . The resulting DevTRS proteins negatively autoregulate [49 - 52] . DevI delays spore formation within nascent fruiting bodies [52] , but if overproduced, DevI inhibits sporulation [51] , which is promoted by MrpC [40] and FruA [53] activity. 41 The other theory is that diacylglycerols released from the inner membrane by cardiolipin phospholipase activity of intact CsgA are the C - signal [54] . However, in neither case has the signal receptor been identified, so our understanding of C - signaling is incomplete. Likewise, how C - signaling impacts recipient cells is unknown. One way that C - signaling has been proposed to affect recipient cells is to stimulate autophosphorylation of a histidine protein kinase, which would then transfer the phosphate to FruA [31] . This model was attractive because FruA is similar to response regulators of two - component signal transducti on systems [19, 31] . Typically, a response regulator is phosphorylated by a histidine protein kinase in response to a signal, thus activating the response regulator to perform a function [55] . The effects of substitutions at the predicted site of phosphorylation in FruA supported the model that FruA is activated by phosphorylation on D59 [31] . However, a histidine protein kinase capable of phosphorylating FruA has not been identified. Also, several observations suggest that FruA may not be phosphorylated. Most notably, D59 of FruA is present in an atypical receiver domain tha t lacks a conserved metal - binding residue normally required for phosphorylation to occur, and treatment of FruA with small - molecule phosphodonors did not increase its DNA - binding activity [22] . The receiver domain of FruA was shown to be necessary for cooperative binding with MrpC to DNA, so it was proposed that C - signaling may affect activity of MrpC and/or FruA [22] (Fig. 1). The regulation of MrpC has been reported to be complex, involving autoregulation, phosphorylation, proteolytic processing, binding to a toxin protein, and stability [14, 15, 56 - 60] . Also, since MrpC is similar to CRP family transcription factors that bind cyclic nucleotides [15] , 42 MrpC activity could be modulated by nucleotide binding, so there are many ways in which C - signaling could affect MrpC activity [22] . Here, using synergistic experimental and computational approaches, we investigate the impact of C - signaling on a circuit that regulates commit ment to sporulation by focusing on the feed - forward loop involving MrpC and FruA control of dev operon transcription (Fig. 1). We describe methods to systematically and quantitatively study the developmental process. Using these methods we measure the le vels of GRN components in wild type and in mutants (e.g., a csgA mutant unable to produce C - signal) during the period leading up to and including commitment to spore formation. We then formulate a mathematical model for the steady - state concentration of d ev mRNA and use the model to computationally predict the magnitude of potential regulatory effects of C - signaling that would be required to explain our data. By testing the predictions, some potential regulatory mechanisms are ruled out and at least ninef old activation of FruA by C - signaling is supported. 43 Results M. xanthus development can be studied systematically We first established quantitative assays to analyze cellular and molecular changes during M. xanthus development. To facilitate collection of sufficient cell numbers for counting, as well as for RNA and protein measurements, development was induced by starvation under submerged culture conditions. Cells adhere to the bottom of a plastic well or dish, an d develop under a layer of buffer. Prior to cell harvest, photos were taken to document phenotypic differences between strains. As expected, wild - type strain DK1622 formed mounds by 18 h poststarvation (PS) and the mounds matured into compact, darkened f ruiting bodies at 36 to 48 h PS (Fig. 2). In contrast, csgA and fruA null mutants failed to progress beyond forming loose aggregates. A devI null mutant was similar to wild type (WT), whereas a devS null mutant formed mounds slowly and they failed to dar ken. Developing populations were harvested at the times indicated in Figure 2 to measure cellular and molecular changes in the same populations. To quantify changes at the cellular level, we counted the total number of cells (after fixation and dispersal , so that rod - shaped cells, spores, and cells in transition between the two were counted) and the number of sonication - resistant spores in the developing populations. 44 We also counted the number of rod - shaped cells at the time when development was initiated by starvation (T 0 ). By subtracting the number of sonication - resistant spores from the total cell number, we determined the number of sonication - sensitive cells. About 30% of the wild - type cells present at T 0 remained as sonication - sensitive cells at 18 h PS (Fig. S1A), consistent with the suggestion that the majority of cells lyse early during development under submerged culture condit ions, which was based on the decrease in the total protein concentration of developing cultures [14] . The number of sonication - sensitive cells continued to decline after 18 Figure 2. 2 Development of M. xanthus strains. Wild - type DK 1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and images were obtained at the indicated number of hours poststarvation (PS). DK1622 formed mounds by 18 h PS (an arrow points to one). The csgA and fruA mutants failed to form mounds, the devI mutant was similar to DK1622 and the devS mutant formed mounds later, were observed in at least three biological replicate s. 45 h PS, reaching ~4% of the T 0 number by 48 h PS (Fig. S1A). Spores were first observed at 27 h PS an d the number rose to ~1% of the T 0 number by 48 h PS (Fig. S1B). The devI mutant was similar to WT, except spores were first observed 6 h earlier at 21 h PS, as reported recently [28] . The csgA , fruA , and devS mutants failed to make a detectable number of spores (at a d etection limit of 0.01% of the T 0 number) and appeared to be slightly delayed relative to WT and the devI mutant in terms of the declining number of sonication - sensitive cells (Fig. S1). We conclude that at the cellular level during the time between 18 and 30 h PS (when we measured RNA and protein levels as described below), the developing populations decline from ~30 - 40% to ~10 - 20% of the initial rod number and only ~0.5% (WT, devI ) or <0.01% ( csgA , fruA , devS ) of the cells form sonication - resistant spores (from which the RNAs and proteins we measured would not be recovered based on control experiments). We stopped collecting samples at 30 PS because thereafter the number of sonication - sensitive cells continues to decline and the spore number continues to rise, making RNA and protein more difficult to recover quantitatively, yet many cells are committed at 30 h PS to make spores by 36 h PS even if nutrients are added [14] . Hence, we focused on changes at the molecular level between 18 and 30 h PS, the period leading up to a nd including the time that many cells commit to spore formation. To measure RNA levels of a large number of samples, we adapted methods described previously [14] to a higher - throughput robotic platform for RT - qPCR analysis. Reproducibility of the analysis was tested among b iological replicates and two types of technical replicates as illustrated in Figure S2A, for each RNA to be measured, at 24 h PS, the midpoint of our focal period. No normalization was done in this experiment. Each transcript number was derived from a sta ndard curve of genomic DNA subjected to qPCR. For each RNA, we found that the 46 average transcript number and the standard deviation for three cDNA technical replicates from a single RNA sample, three RNA technical replicates from a single biological replic ate, and three - S2E). These results suggest that biological variation in RNA levels at 24 h PS is comparable to technical variation in preparing RNA and cDNA. In subsequent experiments, we measured RNA for at least three biological replicates and we did not perform RNA or cDNA technical replicates. We also note the high abundance of the mrpC transcript (~10%) relative to 16S rRNA, and the lower relative abundance of the fruA (~1%) and dev (~0.1%) transcripts. We have typically used 16S rRNA as an internal standard for RT - qPCR analysis during M. xanthus development [14] . The high abundance of mrpC transcript relative to 16S rRNA at 24 h PS (Fig. S2B and S2E) raised the possibility that rRNA decreases relative to total RNA at 18 to 30 h PS. To test this possibility, we measured the 16S rRNA level per 1 µ g of total RNA from 18 to 30 h PS. Figure S3A shows that 0.05), validating 16S rRNA as an internal standard for subsequent experiments. We also found that the total RNA yield per cell does not change significantly from 18 to 30 h PS (single fact or significantly, since the majority of total RNA is rRNA. To measure protein levels, a portion of each well - mixed developing population was immediately ad ded to sample buffer, boiled, and frozen for subsequent semi - quantitative immunoblot analysis [28] . The rest of the population was used for cell counting and RNA analysis as described above and in the Experimental Procedures. 47 Levels of MrpC and FruA fail to account for the low level of dev mRNA in a csgA mutant By systematically quantifying protein and mRNA levels during the period leading up to and including the time that cells commit to spore formation, we investigated whether the GRN shown in Figure 1 could account for observed changes over time in WT and in m utants. In particular, we were interested in whether changes in the levels of MrpC and/or FruA proteins could account for the observed changes in the level of dev mRNA, since MrpC and FruA bind cooperatively to the dev promoter region and activate transcr iption [20] . In WT, we found that the MrpC level decreased about 1.5 - fold on average from 18 to 30 h PS (Fig . 3A) and the FruA level rose about 1.5 - fold on average (Fig. 3B), whereas the dev mRNA level rose about threefold on average (Fig. 4A). In each case, the fold - change was small and the variation between biological replicates was large, so the result of a course did not support a significant difference. We reasoned that cooperative binding of MrpC and FruA could easily account for the threefold rise on average in dev mRNA. We also measured the levels of mrpC an d fruA mRNA. The mrpC mRNA level changed very little on average (Fig. 4B), but the fruA mRNA level decreased about twofold on average after 18 h PS (Fig. 4C), in contrast to the 1.5 - fold rise on average in the FruA protein level (Fig. 3B), suggesting weak positive posttranscriptional regulation of the FruA level during the period of commitment to spore formation. To investigate how C - signaling affects the GRN shown in Figure 1, we measured protein and mRNA levels in the csgA null mutant. In agreement wit h earlier studies suggesting that C - signaling activates FruA [31] and/or MrpC [22] , we found very little dev mRNA in the csgA mutant (Fig. 4A). Notably, the large decrease in the level of dev mRNA in the csgA mutant 48 compared with WT could not be accounted for by a large decrease in the level of MrpC or FruA. The MrpC level was elevated about 1.5 - fold on average in the csgA mutant relative to WT at most time points (Fig. 3A), but the differences were not statistically significant ( p > 0.05 in Stud - tailed t - tests comparing mutant to WT at each time point). The FruA level was diminished in the csgA mutant relative to WT, but only about twofold on average (Fig. 3B). The differences in the FruA level were statistically significant ( p < 0.05 - tailed t - tests) at each time point except 21 h PS ( p = 0.12), but alone the twofold lower levels of FruA in the csgA mutant fail to account for the very low levels of dev mRNA. We also investigated if CsgA regulates dev transcription via LadA instead of FruA. LadA is a LysR type of transcriptional activator, which was earlier shown to activate dev transcription by direct binding. In a ladA mutant, developmental expression of dev was shown to be impaired from a lacZ transcriptional fusion, dev transcription is dependent on LadA during development. In contradiction to the earlier findings, in our laboratory conditions, in the absence of LadA, dev expression was found to be unchange d in comparison to the wild type during commitment. Hence, in our laboratory condition developmental expression of dev does not appear to be dependent on LadA. Though earlier in commitment the ladA mutant was delayed in mound formation compared to the wild type, by 48 h PS the mutant formed dark fruiting bodies and 50% of the wild type sonication resistant spores. We compared the levels of fmg genes between wild type and ladA mutant. During the commitment period, the level of fmgE transcript was consistentl y low in the ladA mutant compared to the wild type. Both fmgA and fmgD showed significant decrease at 18 h PS, whereas, fmgB was significantly low at 18 h and 24 h PS in ladA mutant compared to wild type. 49 The mrpC and fruA mRNA levels were diminished abo ut twofold and 1.5 - fold on average, respectively, in the csgA mutant relative to WT (Fig. 4B and 4C), but at nearly all time points the differences were not statistically significant ( p - tailed t - tests, except p = 0.02 at 27 h for mr pC mRNA). Figure 2.3 Levels of MrpC and FruA during M. xanthus development. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of MrpC (A) and FruA (B) by immunoblot. Graphs show the data points and average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. two - tailed t - tests) from wild type at the corresponding time PS. 50 Figure 2.4 Transcript levels during M. xanthus development. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of dev (A), mrpC (B) and fruA (C) transcript levels by RT - qPCR. Graphs show the data points and average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS and error bars show one standard deviation. Asterisks indicate a significant difference ( p two - tailed t - tests) from wild type at the corresponding time PS. 51 The small differences in the level of fruA mRNA in the csgA mutant relative to WT are especially noteworthy, since they imply that C - signaling has little or no effect on MrpC activity. The results of our fruA mRNA measurements agree with published reports using fruA - lacZ fusions [31, 61] . Furthermore, we found that fruA mRNA stability is similar in the csgA mutant and in WT at 30 h PS (Fig. S4), indicating that the similar steady - state fruA mRNA level we observed (Fig. 4C) reflects a similar rate of synthesis, rather than altered synthesis compensated by altered stability. We conclude that C - signaling does not affect MrpC activity. Therefore, the low level of dev mRNA in a csgA mutant (Fig. 4A) could be due to failure to activate F ruA or to dev - specific regulatory mechanisms. To begin to characterize potential dev - specific regulatory mechanisms during the period leading up to and including commitment to sporulation, we measured protein and mRNA levels in the devS and devI null mutan ts. The MrpC and FruA levels were similar to WT (Fig. 3). The dev mRNA level ranged from 20 - fold higher in the devS mutant than in WT at 18 h PS, to 10 - fold higher at 30 h PS (Fig. 4A), consistent with negative autoregulation by DevS (and DevT and DevR) reported previously [28, 29] . Unexpectedly, the dev mRNA level in the devI mutant was about threefold lower than in WT at 30 h PS (Fig. 4A), suggesting that DevI feeds back positively on accumulation of dev mRNA, although the difference was not quite statistically significant at th e 95% confidence level ( p - tailed t - test). Other differences were that the fruA mRNA levels in the devI and devS mutants were about twofold lower than in WT at 27 and 30 h PS (Fig. 4C), and these were statistically significant ( p < - tailed t - tests comparing mutant to WT at each time point). Since the FruA levels in these mutants were 52 similar to those in WT (Fig. 3B), positive posttranscriptional regulation of FruA appeared to occur in the mutants, as well as in WT. To complete our characterization of the GRN shown in Figure 1, we also measured protein and mRNA levels in the fruA and mrpC null mutants. We did not collect samples of the mrpC mutant at as many time points since we expected little or no expression of GRN components. As expected, neither MrpC nor FruA were detected in the mrpC mutant (Fig. S5). In the fruA mutant, the MrpC level was similar to WT and, as expected, FruA was not detected (Fig. 3). Also as expected, in the fruA mutant the fruA mRNA w as not detected, the dev mRNA level was very low, and the mrpC mRNA level was similar to WT (Fig. 4). Since the mrpC mutant had an in - frame deletion of codons 74 to 229 [15] , we were able to design primers for RT - qPCR analysis that should detect the shorter mrpC transcript. Surprisingly, the mrpC mutant exhibited an elevated level of mrpC transcript compared with WT at 18 and 24 h PS (Fig. S6A). The result was surprising since expression of an mrpC - lacZ fusion had been reported to be abolished in the mrpC mutant, which had led to the conclusion that MrpC positively autoregulates [15] . We considered the possibility that the shorter transcript in the mrpC mutant is more stable than the WT transcript, but the transcript half - lives after addition of rifampicin did not differ significantly (Fig. S7). We conclude that MrpC negatively regulates the mrpC transcript level. While this work was in progress, McLau ghlin et al . reached the same conclusion [60] . In all other respects, the mrpC mutant yielded expected results. The fruA and dev transcripts were very low (Fig. S6B and S6C), consistent with the expectations that MrpC is required to activate fruA transcription [18] and that MrpC and FruA are required to activate dev transcription [20, 31, 62] . Also, the mrpC mutant failed to progress beyond forming loose aggregates (Fig. S8), 53 a ppeared to be slightly delayed relative to WT in terms of the declining number of sonication - sensitive cells (Fig. S9A), and failed to make a detectable number of spores (at a detection limit of 0.01% of the T 0 number) (Fig. S9B). Taken together, our systematic, quantitative measurements of components of the GRN shown in Figure 1 imply that failure to activate FruA and/or dev - specific regulatory mechanisms may account for the low level of dev mRNA in a csgA mutant. Given the complex feedback architecture of dev regulation (i.e., strong negative feedback by DevTRS and weak positive feedb ack by DevI at 30 h PS), delineating the effects of C - signaling on the dev transcript level requires a mathematical modeling approach. Mathematical modeling suggests several mechanisms that could explain the low level of dev mRNA in the csgA mutant The o bserved small differences in the levels of MrpC and FruA in the csgA mutant relative to WT do not account for the very low level of dev mRNA in the csgA mutant. To evaluate plausible mechanisms that may explain these experimental findings, we quantitative ly analyzed transcriptional regulation of dev by formulating a mathematical model that expresses the dev mRNA concentration as a function of the regulators MrpC, FruA, DevI, and DevS. MrpC and FruA bind cooperatively to the dev promoter region and activate transcription [20] . Our results sug gest that DevI is a weak positive regulator and DevS is a strong negative regulator of dev transcription by 30 h PS (Fig. 4A). Incorporating these effects into a transcriptional regulation model, we express the concentration of dev mRNA as a product of th ree regulation functions ( ) divided by the transcript degradation rate (see Experimental Procedures for detailed explanation): 54 Here, we use a quasi - steady state approximation for the mRNA levels by taking advantage of the fact that mRNA decay (with half - lives typically in minutes) is much faster than our experimental measurement times (in hours). This allows us to assume a rapid equilibrium between the rate of dev transcription and the decay of its mRNA, which leads to the above equation, in which and are parameters characterizing promoter regulation . We assume that these biochemical parameters are n ot a function of the genetic background and, therefore, in the strains in which dev mRNA was measured (e.g., the csgA mutant) , the concentration of dev mRNA is determined by the concentrations of proteins (indicated by square brackets in the equation), mor e specifically the concentrations of their transcriptionally active forms (in case there is a posttranslational regulation). To estimate how the different regulation parameters (such as transcription rate, degradation rate, cooperativity constant, etc.) af fect the dev mRNA level, we first constrain the model parameters by the experimental result shown in Figure 3B, and search for parameters that can result in the observed 22 - fold difference in in WT relative to the csgA m utant at 30 h PS (Fig. 4A). To estimate the contribution of autoregulation by Dev proteins to their own transcription (i.e., the terms ) in WT and the csgA mutant, we employ the data from the devI and devS mutants (Fig. 4A). Specifically, we take the ratio of the dev mRNA level in WT to that in devI and devS mutants to estimate the feedback regulation from DevI and DevS, respectively (see 55 Experimental Procedures for details). We find the contribution from DevI and DevS feedback regulation in WT to be and , respectively. Using these values, we find the contribution from FruA and MrpC regulation to be . In the csgA mutant, since the dev mRNA level is very low, we assume the DevI and DevS protein levels to be low. This gives the contribution of different regulation functions as , , and . In summary, this analysis reveals that the twofold reduction of FruA protein observed in the csgA mutant (Fig. 3B) leads to a change of - fold in the FruA - and MrpC - dependent transcript regulation term. We reasoned that the observed 22 - fold reduction in dev transcript in the csgA mutant relative to WT at 30 h PS (Fig. 4A) could result from a reduction in the FruA - and MrpC - dependent activa tion rate and/or an increase in the transcript degradation rate . In what follows we use the mathematical model to predict the magnitude of these effects that would be necessary to explain the observed 22 - fold difference in . Hyp othesis 1: Increase in dev transcript degradation rate in the csgA mutant First, we estimate the difference in dev transcript degradation rate necessary to explain the observed difference in transcript level between WT and the csgA mutant. For this, we make two assumptions. First, we assume that MrpC and FruA bind to the dev promoter region with a Hill cooperativity coefficient (i.e., the maximum for a single cooperative binding site). Second, we assume that the observed twofold difference in FruA protein level results in a twofold difference in transcriptionally active FruA. Under these assumptions, we vary t he remaining unknown parameters to compute the required fold difference in transcript degradation rate for different values of promoter saturation. Our results plotted in Figure 5A 56 show that at least a 20 - fold difference in transcript degradation rate is required to explain the transcript data. This experimentally testable prediction will be assessed in a subsequent section. Figure 2.5 Mathematical modeling of different hypotheses to explain the low dev transcript level in a csgA mutant. Plots showing the requi red fold change in dev transcript degradation rate in the csgA mutant in comparison to wild type (A), cooperativity coefficient for MrpC and FruA binding to the dev promoter region (B) and reduction in FruA activity in the csgA mutant in comparison to wi ld type (C), to explain the experimental data for different values of promoter saturation. 57 If the results are inconsistent with this prediction, we must conclude that at least one of the two assumptions above is invalid, resulting in the following two alternative hypotheses: the Hill coefficient of MrpC and FruA binding to the dev promoter regi on is much higher than and/or the amount of transcriptionally active FruA does not scale with the measured FruA protein level (e.g., if csgA - dependent C - signaling is also involved in posttranslational activation of FruA). Hypothesis 2: High cooperativity of MrpC and FruA binding to the dev promoter region Next, we test if a higher binding cooperativity can explain the difference in dev transcript level between WT and the csgA mutant. We compute the required cooperativity coefficient by assuming the degradation rate does not change between the two strains. Our results plotted in Figure 5B show that the minimum cooperativity coefficient required to explain the experimental results is six for low promoter saturation. In bi ologically realistic conditions, where promoter saturation is higher; the required cooperativity is even higher. Such a large cooperativity can only be explained if there is more than one site in the promoter region where MrpC and FruA bind with high coop erativity. We know that the dev promoter region has at least two MrpC and FruA cooperative binding sites; one is proximal upstream, whereas the other is distal upstream [20] . Interaction between the proximal and distal upstream binding sites by DNA looping may contribute to high cooperativity coefficient predicted by the model. The distal upstream binding site appeared to boost dev promoter activity after 24 h PS, based on - galactosidase activity from a lacZ reporter . Hence, in a subsequent section, we study the impact of a distal site deletion on different transcripts ( mrpC, fruA, dev ) and proteins (MrpC, FruA) to test if presence of the distal site contributes to higher cooperativity. If the results are 58 not consistent with the model predictions, we must conclude that the fold difference in active FruA exceeds that observed for the to tal concentration of each protein (i.e., csgA - dependent C - signaling is involved in posttranslational activation of FruA). Hypothesis 3 : Posttranslational regulation of FruA activity To assess the difference in active FruA level required to explain the o bserved difference in dev transcript level, in the absence of other effects, we fix the cooperativity coefficient at and assume the transcript degradation rate to be unchanged between WT and the csgA mutant. We then use our model to compute the fold difference in active FruA required to achieve a 22 - fold reduction in dev transcript in the csgA mutant relative to WT. Our results plotted in Figure 5C show that at least a ninefold reduction in active FruA is needed in the csgA mutant. The reduction in active FruA in the csgA mutant would presumably be due to the absence of C - signal - dependent posttranslational activation of FruA, not due to the twofold lower level of FruA protein we observed in the csgA mutant relative to WT (Fig. 3B). The reduction in active FruA may be considerably greater than ninefold if the dev promoter region approaches saturation (e.g., 20 - fold at 80% saturation in Fig. 5C). Also, mathematical modeling of our data at each time point from 18 to 30 h PS yields a similar result (Fig . S10), suggesting that in WT, FruA has already been activated by C - signaling at least ninefold by 18 h PS, and perhaps as much as 30 - fold if the dev promoter region approaches saturation (righthand panel in Fig. S10). Stability of the dev transcript is unchanged in a csgA mutant To measure the dev transcript degradation rate in WT and the csgA mutant, we compared the dev transcript levels after addition of rifampicin to block transcription at 30 h PS. The average half - life of the dev tran script in three biological replicates was 11 ± 6 min in WT and 7 ± 1 min in 59 the csgA mutant (Fig. 6), which is not a statistically significant difference ( p two - tailed t - test). We conclude that elevated turnover does not account for the low level of dev transcript in the csgA mutant. These results allow us to rule out Hypothesis 1. The distal upstream binding site for MrpC and FruA has little impact on the dev transcript level In a previous study, weak cooperative binding of MrpC an d FruA to a site located between positions - 254 and - 229 upstream of the dev promoter appeared to boost - galactosidase activity from a lacZ transcriptional fusion about twofold between 24 and 30 h PS, but deletion of the distal upstream site did not impair spore formation [20] . These findings suggested that the distal site has a modest impact on dev transcription that is inconsequential for sporulation. However, - galactosidase activity from lacZ fused to dev promoter segments with different amoun ts of upstream DNA and integrated ectopically may not accurately reflect the contribution of the distal site to the dev transcript level. Therefore, we measured the dev transcript level in a mutant lacking the distal site (i.e., DNA between positions - 254 and - 228 was deleted from the M. xanthus chromosome). The level of dev transcript in the distal site mutant was similar to WT measured in the same experiment, in this case increasing about twofold from 18 to 30 h PS (Fig. 7). Likewise, there were no sig nificant differences between the distal site mutant and WT in the levels of mrpC or fruA transcripts (Fig. S6) or the corresponding proteins (Fig. S5) ( p - tailed t - tests comparing mutant to WT at each time point). The distal site mu tant formed mounds by 18 h PS, which matured into compact, darkened fruiting bodies at later times, similar to WT (Fig. S8), and the percentages of sonication - sensitive cells and sonication - resistant spores observed for the distal site mutant were similar to WT (Fig. S9). We conclude that the distal site has little or no impact on the 60 developmental process. In particular, the distal site does not contribute to high cooperativity of MrpC and FruA binding to the dev promoter region that could explain the hi gher level of dev Figure 2.7 dev transcript levels in wild type and distal site mutant. Figure 2.6 dev transcript stability. Wild - type DK1622 and the csgA mutant were subjected to starvation under submerged culture conditions for 30 h. The overlay was replaced with fresh t 0 ) and at the times indicated ( t x ) for measurem ent of the dev transcript level by RT - qPCR. Transcript levels at t x were normalized to that at t 0 for each of three biological replicates and used to determine the transcript half - life for each replicate. The average half - life and one standard deviation ar e reported in the text. The graph shows the average ln( t x / t 0 ) and one standard deviation for the three biological replicates of wild type (black dashed line) and the csgA mutant (gray solid line). 61 Figure 2.7 ( c ) transcript in WT than in the csgA mutant. These results allow us to rule out Hypoth esis 2. Boosting the FruA level in the csgA mutant has no effect on the dev transcript level Having ruled out the first two hypotheses, our modeling predicts that the only viable option to explain the effect of the csgA null mutation on the dev transcript level is Hypothesis 3: at least a ninefold reduction in active FruA is needed in the csgA mu tant as compared with WT. Specifically, our model showed that the low dev transcript level in the csgA mutant is not due to its twofold lower FruA level (Fig. 3B), but rather due to a failure to activate FruA in the absence of C - signaling (Fig. 5C and S10 ). As a result, the model predicts that in the csgA mutant most of the FruA remains inactive. To test this prediction, we integrated fruA transcriptionally fused to a vanillate - inducible promoter ectopically in the csgA mutant. Upon induction the csgA P van - fruA strain accumulated a similar level of FruA as WT (Fig. 8A), but the dev transcript level Wild - type DK1622 and its indicated mutant derivative were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of dev transcript levels by RT - qPCR. Graphs show the data points and average of three biological replicates, relative to wild - type DK1622 at 18 h PS and error bars show one standard deviation. 62 Figure 2.8 FruA protein and dev transcript levels. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of FruA levels by immunoblot (A) and dev transcript levels by RT - qPCR (B). Graphs show the data points and average of three biological replicates, relative to wild - type DK1622 at 18 h PS and error bars show one standard deviation. Asterisks indicate a significant difference ( p two - tailed t - tests) from wild type at the corresponding time PS. 63 remained as low as in the csgA mutant (Fig. 8B). Hence, boosting the FruA level in the csgA mutant had no effect on the dev transcript level . Additionally, we tested the previously proposed idea of phosphorylation to be a potential mechanism of CsgA dependent activation of FruA. FruA was earlier suggested to be phosphorylated on the residue D59. We created a phosphomimetic form of FruA by substi tuting E for D at residue 59. We ectopically expressed the phosphomimetic form of FruA ( fruA( D59E) by transcriptionally fusing it to a vanillate - inducible promoter in the csgA mutant background. Upon induction the csgA P van - fruA (D59E) strain accumulated a similar level of FruA as WT and csgA P van - fruA . (Fig. 8A), but the dev transcript level remained as low as observed in csgA mutant and csgA P van - fruA strain (Fig. 8B), consistent with our prediction and supporting the hypothesis that C - signaling activate s FruA at least ninefold. Additionally, boosting the level of the phosphomimetic form of FruA (FruA - D59E), in the csgA mutant fails to recover the level of dev transcript, suggesting phosphorylation not to be a mechanism for CsgA dependent activation of FruA. The boost in FruA level correlated with a boost in fruA transcript level in the csgA P van - fruA and csgA P van - fruA (D59E) strain(Fig. S11A). As expected, the mrpC transcript (Fig. S11B) and MrpC protein (Fig. S12) levels were similar in the csgA P van - fruA and csgA P van - fruA (D59E) strain as in the csgA mutant at 18 and 24 h Post starvation. Induction of the csgA P van - fruA strain did not rescue its development sinc e it failed to progress beyond forming loose aggregates (Fig. S13), failed to make a detectable number of spores by 48 h PS (at a detection limit of 0.01% of the T 0 number; data not shown), and appeared to be slightly delayed relative to WT in terms of the declining number of sonication - sensitive cells, like the csgA mutant (Fig. S14). 64 As a control, P van - fruA and P van - fruA (D59E) were integrated ectopically in the fruA mutant. Upon induction the fruA P van - fruA strain formed mounds by 18 h PS and the mounds matured into compact, darkened fruiting bodies at later times , similar to WT without or with vanillate added (Fig. S15). Also, the induced fruA P van - fruA strain exhibited a similar number of sonication - resistant spores as WT at 36 h PS (data not shown). fruA Pvan - fruA (D59E) was delayed in mound formation by 6 h PS, but eventually formed darkened fruiting bodies by 30 h PS followed by similar number of sonication resistant spores as wild type by 36 h PS. These results show that ectopic induction of the f ruA and fruA (D59E) from P van - fruA rescued development of a fruA mutant, presumably because C - signaling activated FruA produced from P van - fruA . Expressions of FruA regulated genes are dependent on CsgA induced activation of FruA In order to investigate this, we asked if the effect of CsgA dependent activation of FruA is specific for dev transcription? Four other genes ( fmgA, fmgB, fmgD and fmgE ), which were earlier suggested to be combinatorially regulated by MrpC and FruA. We com pared expression levels of all four genes between csgA mutant, csgA P van - fruA and csgA P van - fruA D59E. The csgA mutant showed significantly low level of all four fmg genes compared to the wild type. Similar to dev, despite boosting the level of FruA, expre ssions of the fmg genes were not recovered to the wild type level in csgA P van - fruA and csgA P van - fruA D59E. Hence, CsgA dependent activation of FruA is critical for induction of genes combinatorially regulated by MrpC and FruA. 65 Discussion Our systematic, quantitative analysis of a key circuit in the GRN governing M. xanthus fruiting body formation implicates posttranslational regulation of FruA by C - signaling as primarily responsible for dev transcript accumulation during the period leading up to and including commitment to spore formation. Mathematical modeling of the dev transcript level allowed us to predict the magnitude of potential regulatory mechanisms. Experiments ruled out C - signal - dependent stabilization of dev mRNA or highly coo perative binding of FruA and MrpC to two sites in the dev promoter region as the explanation for the much higher dev transcript level in WT than in the csgA mutant. Although the FruA level was twofold lower in the csgA mutant than in WT (Fig. 3B and 8A), boosting the FruA level in the csgA mutant had no effect on the dev transcript level (Fig. 8B). Taken together, our experimental and computational analyses provide evidence that C - signaling activates FruA at least ninefold posttranslationally during M. xa nthus development (Fig. 9). The activation of FruA may be considerably greater than ninefold if the dev promoter region approaches saturation (Fig. 5C and S10). Since efficient C - signaling requires cells to move into close proximity [33 - 35] , we propose that activation of FruA by C - signaling acts as a checkpoint for mound formation during the developmental process (Fig. 9). Regulation of FruA by C - signaling If activation of FruA by C - signaling acts as a checkpoint for mound formation, then active FruA should be present at 18 h PS since mound formation is well underway (Fig. 2). In agreement, mathematical modeling of our data using the assumptions of hypothesis 3 at each time point from 18 to 30 h PS yields a similar result (Fig. S10). Thi s analysis implies that FruA has already 66 been activated by C - signaling at least ninefold by 18 h PS, if the assumptions of hypothesis 3 apply. Figure 2.9 Updated model of the gene regulatory network governing formation of fruiting bodies . Relative to the simplified model shown in Fig. 1 (see legend), this model also includes phosphorylated MrpB (MrpB - P) which appears to activate transcription of mrpC , and negative autoregulation by MrpC which appears to involve competition with MrpB - P for b inding to overlapping sites in the mrpC promoter region; proteolysis of MrpC, which is regulated by the Esp signal transduction system that normally slows the developmental process and is regulated by nutrient addition that can halt development; posttransl ational activation of FruA to FruA* by C - signaling and promotion of mound formation by FruA*, thus enhancing short - range C - signaling by bringing cells into proximity; the possibility that DevI inhibits negative autoregulation by DevTRS; and speculation t hat the feed - forward loop involving MrpC and FruA* not only controls transcription of the dev operon, but that of genes involved in cellular shape change as well, committing cells to spore formation and resulting in spore - filled fruiting bodies. This mode l deletes activation of MrpC by C - signaling, which was included as a possibility in Fig. 1, but was not supported by our data. See the text for details and references. 67 The assumption that the distal site does not contribute to high cooperativity of MrpC and FruA binding to the dev promoter region applies since the dev transcript level did not differ significantly in the distal site mutant as compared with WT at 18 or 24 h PS (Fig. 7). We did not measure dev transcript stability at 18 to 27 h PS, but at 30 h PS there was no significant difference between WT and the csgA mutant (Fig. 6). Therefore, C - signaling may have already activated FruA at least ninefold by 18 h PS, and perhaps as much as 30 - fold if the dev promoter region a pproaches saturation (90% saturation in the righthand panel of Fig. S10). We note that during the period from 18 to 30 h PS , the dev transcript level rises, but the rise is due to positive autoregulation by DevI (Fig. 4A). Hence, active FruA may not be t he limiting factor for dev transcription during this period (i.e., the dev promoter region may indeed approach saturation binding of active FruA and MrpC). The proximal upstream site in the dev promoter region, which is crucial for transcriptional activat ion, exhibits a higher affinity for cooperative binding of FruA and MrpC than the distal upstream site [20] o r several other sites [17, 24] , perhaps conferring on dev transcription a re latively low threshold for active FruA. The mechanism of FruA activation by C - signaling is unknown. Since FruA is similar to response regulators of two - component signal transduction systems, phosphorylation by a histidine protein kinase was initially proposed to control FruA activity [19, 31] . While this potential mechanism of posttranslational control cannot be ruled out, a kinase capable of phosphorylating FruA has not been identified despite considerable effort. Moreover, the atypical receiver domain of FruA and the inability of small - molecule phosphodonors to increase its DNA - binding activity suggest that FruA may not be phosphorylated [ 22] . 68 Several atypical response regulators have been shown to be active without phosphorylation and a few are regulated by ligand binding [63, 64] . For example, the atypical receiver domain of Streptomyces ve nezuelae JadR1 is bound by jadomycin B, causing JadR1 to dissociate from DNA, and the acylated antibiotic undecylprodigiosin of Streptomyces coelicolor may use a similar mechanism to modulate DNA - binding activity of the atypical response regulator RedZ [65] . Conceivably, FruA activity could likewise be regulated by binding of M. xanthus diacylglycerols, which have been implicate d in C - signaling [54] . Alternatively, FruA could be regulated by a posttranslational modification other than phosphorylation or by binding to another protein (i.e., sequestration). In addition to regulating FruA activity posttranslationally, C - signaling appears to regulate th e FruA level posttranscriptionally. The FruA level was reproducibly twofold lower in the csgA mutant than in WT (Fig. 3B and 8A), but the fruA transcript level was not significantly different (Fig. 4C and S11A). These results suggest that positive posttr anscriptional regulation of the FruA level requires C - signaling. C - signaling may increase synthesis (i.e., increase fruA mRNA accumulation slightly and also increase translation of fruA mRNA) and/or decrease turnover of FruA. We did not investigate this further because the FruA deficit in the csgA mutant could be overcome with P van - fruA , yet there was very little effect on the dev transcript level (Fig. 8). This demonstrates that the activity of FruA, rather than its level, primarily controls the level of dev transcript. Additionally, boosting the level of FruA with P van - fruA failed to recover level of fmg transcripts in the csgA mutant, suggesting activity of FruA but the level is critical for regulating expressions of FruA dependent genes. Overexpressing the phosphomimetic form of FruA (FruA D59E), also failed to recover the level of dev and fmg transcripts. This finding contradicts the 69 previously proposed idea of phosphorylation to be the potential mechanism for CsgA induced activation of FruA. Regulation by Dev proteins DevI inhibits sporulation if overexpressed, as in the devS mutant [29] (Fig. 2 and S1). Deleti on of devI or the entire dev operon allows spores to begin forming about 6 h earlier than normal, but does not increase the final number of spores [28] (Fig. S1). The level of MrpC was about twofold higher on average in the de vI mutant than in WT at 15 h PS, perhaps accounting for the observed earlier sporulation, although the difference diminished at 18 - 24 h PS [28] , as reported here (Fig. 3A). It was concluded that Dev I may transiently and weakly inhibit translation of mrpC transcripts during the period leading up to commitment, delaying sporulation [28] . As noted above, DevI positively autoregulates, causing a small rise in the dev transcript level by 30 h PS (Fig. 4A, 7, and 8B). Although the mechanism of this feedback loop is unknown, one possibility is that DevI inhibits negative autoregulation by DevTRS (Fig. 9). In pr evious studies, mutations in devT , devR , or devS relieved negative autoregulation, resulting in ~10 - fold higher dev transcript accumulation at 24 h PS [28, 29] . In this study, a devS mutant likewise accumulated ~10 - fold more dev transcript than WT at 24 - 30 h PS, and the difference was ~20 - fold at 18 and 21 h PS (Fig. 4A), suggesting that negative autoregulation mediated by DevS has a stronger effect leading up to the commitment period than during commitment. Strong negative a utoregulation may promote commitment to sporulation by lowering the level of DevI, which would raise the MrpC level by relieving inhibition of translation of mrpC transcripts [28] . Our data suggest that negative autoregulation by DevTRS weakens during the commitment period, perhaps accounting for the observed small rise in the 70 dev transcript level (Fig. 4A, 7, and 8B). If the elevated dev transcript level is accompanied by a small increase in the level of DevI, then DevI may inhibit translation of mrpC transcripts, causing the MrpC level to decrease slightly by 30 h PS in WT (Fig. 3A). DevI is predicted to be a 40 - residue polypeptide [29] and currently no method has been devised to measure the DevI level. This is a worthwhile goal of future research, as is und erstanding how cells overcome DevI - mediated inhibition of sporulation (depicted in Fig. 9 as inhibition of cellular shape change). In addition to regulating the timing of commitment to spore formation, Dev proteins appear to play a role in maturation of spores. Mutations in dev genes strongly impair expression of the exo operon [28, 66] , which encodes proteins that help form the polysaccharide spore coat necessary to maintain cellular shape change and form mature spores [67, 68] . The role of MrpC Our results add to a growing list of observations that indicate MrpC functions differently during M. xanthus development than originally proposed. We found that MrpC negatively autoregulates accumulation of mrpC mRNA about twofold at 18 and 24 h PS (Fig. S6A), and it does so at 18 h PS without significantly altering transcript stability (Fig. S7). This contradicts an earlier study that concluded MrpC positively autoregulates, based on finding that expression of an mrpC - lacZ fusion was abolished in an mrpC mutant [15] . Recently, and in agreement with our result, it was reported that MrpC is a negative autoregulat or that competes with MrpB for binding to the mrpC promoter region [60] . MrpB, likely when phosphorylated, binds to two sites upstream of the mrpC promoter and activates transcription. MrpC binds to multiple sites upstream of the mrpC promoter [57, 60] , including two that overlap the MrpB binding sites [60] . Purified MrpC competes with the MrpB DNA - binding domain for binding to the 71 overlapping sites, supporting a model in which MrpC negatively autoregulates by d irectly competing with phosphorylated MrpB for binding to overlapping sites [60] (Fig. 9). The role of MrpC in cellular lysis during development appears to be less prominent than originally proposed. MrpC was reported to function as an antitoxin by binding to and inhibiting activity of the MazF toxin protein, an mRNA interferase shown to be imp ortant for developmental programmed cell death [58] . However, the effect of a null mutation in mazF on developmental lysis depends on the presence of a pilQ1 mutation [13, 69] . In pilQ + backgrounds such as our WT strain DK1622, MazF is dispensable for lysis. Here, we found on ly a slight delay of the mrpC mutant relative to WT in terms of the declining number of sonication - sensitive cells at 18 - 48 h PS (Fig. S9A), comparable to other mutants ( csgA , fruA , devS , csgA P van - fruA ) that were unable to form spores (Fig. S1 and S13; data not shown). Under our conditions, MrpC appears to play no special role in modulating the cell number during development. Both the synthesis and the degradation of MrpC are regulated. Synthesis is regulated by phosphorylated MrpB and MrpC acting positively and negatively, respectively, at the level of transcription initiation as described above [60] (Fig. 9). Degradation is regulated by the complex Esp signal transduction system [59, 70, 71] , which presumably senses a signal and controls the activity of an unidentified protease involved in MrpC turnover, thus ensuring that development proceeds at the appropriat e pace (Fig. 9). Interestingly, preliminary results suggest that the Esp system does not govern the proteolysis of MrpC observed when nutrients are added at 18 h PS [14] (Y. Hoang, R. Rajagopalan, and L. Kroos; unpublished data). This implies that another system senses nu trients and degrades MrpC to halt development (Fig. 9). 72 Combinatorial control by MrpC and FruA Nutrient - regulated proteolysis of MrpC provides a checkpoint for starvation during the period leading up to and including commitment to sporulation [14] (Fig. 9). If activation of FruA by C - signaling acts as a checkpoint for mound formation as we propose (Fig. 9), then combinatorial control by MrpC and activate d FruA could ensure that only starving cells in mounds express genes that commit them to spore formation. MrpC and FruA bind cooperatively to the promoter regions of five C - signal - dependent genes [20 - 24] . In each case, cooperative b inding to a site located just upstream of the promoter appears to activate transcription. Hence, MrpC and FruA form a type 1 coherent feed - forward loop with AND logic [72] . This type of loop is abundant in GRNs and can serve as a sign - sensitive delay element [72, 73] . The sign sensitivity refers to a difference in the network means for the feed - forward loop created by MrpC, FruA, and their target genes is that target gene expression is delayed as MrpC accumulates, awaiting FruA activated by C - signaling (i.e., - range C - signaling, activated FruA would bind cooperatively with MrpC, stimulating transcription of target genes that eventually commit cells to spore formation (depicted in Fig. 9 as cellular shape change). However, if nutrients reappear prior to commitment, MrpC is degraded and nce MrpC binds to the promoter regions of hundreds of developmental genes based on ChIP - seq analysis, and in 13 of 15 cases cooperative binding of MrpC and FruA was observed [17] . 73 In addition to the feed - forward loop involving cooperative binding of MrpC and FruA to a site located just upstream of the promoter, the promoter regions of some genes h ave more complex architectures that confer greater dependence on C - signaling for transcriptional activation. For example, in the fmgD promoter region, binding of MrpC to an additional site that overlaps the promoter and the FruA binding site appears to re press transcription, and it has been proposed that a high level of active FruA produced by C - signaling is necessary to outcompete MrpC for binding and result in transcriptional activation [21] (Fig. S16A). In the fmgE promoter region, a distal upstream site with higher affinity for cooperative binding of MrpC and FruA appears to act negatively by competing for binding with the lower affinity site just upstr eam of the promoter [24] (Fig. S16B). In addition to fmgD and fmgE , other genes depend more strongly on C - signaling and are expressed later during development than dev [30] . We infer that such genes require a higher level of active FruA than dev in order to be transcribed. In contrast to the dev promoter region, which may have a relatively low threshold for active FruA and therefore approach satura tion binding of active FruA and MrpC at 18 h PS (Fig. S10), we predict that the promoter regions of genes essential for commitment to sporulation have more complex architectures and a higher threshold for active FruA. According to this model, C - signal - dep endent activation of FruA continues after 18 h PS and the rising level of active FruA triggers commitment beginning at 24 h PS. We speculate that genes governing cellular shape change are under combinatorial control of MrpC and FruA (Fig. 9), and have a h igh threshold for active FruA. 74 Materials and methods Bacterial strains, plasmids and primers The strains, plasmids, and primers used in this study are listed in Table S1. Escherichia coli strain DH5 was used for cloning. To construct p SS10 , primers FruA - F - NdeI - Gibson and FruA - R - EcoRI - Gibson were used to generate PCR products using chromosomal D NA from M. xanthus strain DK1622 as template. The products were combined with NdeI - EcoRI - digested pMR3691 in a Gibson assembly reaction to enzymatically join the overlapping DNA fragments [74] . The cloned PCR product was verified by DNA sequencing. M. xanthus strains with P van - fruA integrated ectopically were constructed by electroporati on [75] of p SS10 , selection of transformants on CTT agar containing 15 µg/ml of tetracycline [76] , and verification by colony PCR using primers pMR3691 MCS G - F and pMR3691 MCS G - R. To express fruA (D59E) from the vanillate inducible promoter P van - fruA was subjected to site directed mutagenesis using primers D59E F and D59E R , followed verification by PCR and sequencing using pMR3691 MCS G - F and pMR3691 MCS G - R. M. xanthus strains with P van - fruA(D59E) integrated ectopically were constructed by electroporation [75] , followed by selection of transformants on CTT agar cont aining 15 µg/ml of tetracycline [76] , and further verification by colony PCR using primers pMR3691 MCS G - F and pMR3691 MCS G - R. Growth and development of M. xanthus Strains of M. xanthus were grown at 32°C in CTTYE liquid medium (1% Casitone, 0.2% yeast extract, 10 mMTris - HCl [pH 8.0], 1 mM KH 2 PO 4 - K 2 HPO 4 , 8 mM MgSO 4 [final pH 7.6]) with shaking at 350 rpm. CTT agar (CTTYE lacking yeast extract and solidified with 1.5 % agar) was used for growth on solid medium and was supplemented with 40 µg/ml of kanamycin sulfate or 75 15 µg/ml of tetracycline as required. Fruiting body development under submerged culture conditions was performed using MC7 (10 mM morpholinepropanesulfo nic acid [MOPS; pH 7.0], 1 mM CaCl 2 ) as the starvation buffer as described previously [14] . Briefly, log - phase CTTYE cultures were centrifuged and cells were resuspended in MC7 at a density of approximately 1,000 Klett units. A 100 l sample (designated T 0 ) was removed, g lutaraldehyde (2% final concentration) was added to fix cells, and the sample was stored at 4°C at least 24 h before total cells were quantified as described below. For each developmental sample, 1.5 ml of the cell suspension plus 10.5 ml of MC7 was added to an 8.5 - cm - diameter plastic petri plate. Upon incubation at 32°C, cells adhere to the bottom of the plate and undergo development. At the indicated times developing populations were photographed through a microscope and collected as described below. Microscopy Images of fruiting bodies were obtained using a Leica Wild M8 microscope equipped with an Olympus E - 620 digital camera. In order to quantify cells in samples collected and dispersed as described below, high resolution images were obtained with an Olympus BX51 microscope using a differential interference contrast filter and a 40× objective lens, and equipped with an Olympus DP30BW digital camera. Sample collection At the indicated times the submerged culture supernatant was replac ed with 5 ml of fresh MC7 starvation buffer with or without inhibitors as required. Developing cells were scraped from the plate bottom using a sterile cell scraper and the entire contents were collected in a 15 - ml centrifuge tube. Samples were mixed tho roughly by repeatedly (three times total) vortexing for 76 15 s followed by pipetting up and down 15 times. For quantification of total cells, 100 l of the mixture was removed, glutaraldehyde (2% final concentration) was added to fix cells, and the sample w as stored at 4°C for at least 24 h before counting as described below. For measurement of sonication - resistant spores, 400 µ l of the mixture was removed and stored at - 20°C. For immunoblot analysis, 100 µ l of the mixture was added to an equal volume of 2 × 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. Immediately after collecting the three samples just described, the remaining 4. 4 ml of the developing population was mixed with 0.5 ml of RNase stop solution (5% phenol [pH < 7] in ethanol), followed by rapid cooling in liquid nitrogen until almost frozen, centrifugation at 8,700 × g for 10 min at 4°C, removal of the supernatant, fre ezing of the cell pellet in liquid nitrogen, and storage at - 80°C until RNA extraction. Control experiments with a sample collected at 30 h PS indicated that the majority of spores remain intact after boiling in 2 × sample buffer or RNA extraction as described below (data not shown), so the proteins and RNAs analyzed are from developing cells that have not yet formed spores. Quantification of total cells and sonication - resistant spores During development a small percentage of the rod - shaped cells transition to ovoid spores that become sonication - resistant. The number of sonication - resistant spores in developmental samples was quantified as described previously [14] . Briefly, each 400 - µ l sample collected as described above was th awed and sonicated using a model 450 sonifier (Branson) at output setting 2 for 10 - s intervals three times with cooling on ice in between. A 60 µ l sample was removed and ovoid spores were counted microscopically using a Neubauer counting chamber. 77 A remai ning portion of the sample was used to determine total protein concentration as described below. The total cell number, including rod - shaped cells, ovoid spores, and cells in transition between the two, was determined using the glutaraldehyde - fixed sample s collected as described above. Each sample was thawed and mixed by vortexing and pipetting, then 10 or 20 µ l was diluted with MC7 to 400 µ l , sonicated once for 10 s, and all cells were counted microscopically. The total cell number minus the number of s onication - resistant cells was designated the number of sonication - sensitive cells (consisting primarily of rod - shaped cells) and was expressed as a percentage of the total cell number in the corresponding T 0 sample (consisting only of rod - shaped cells). RNA extraction and analysis RNA was extracted using the hot - phenol method and the RNA was digested with DNase I (Roche) as described previously [71] . One g of total RNA was subjected to cDNA synthesis using Superscript III reverse transcriptase (InVitrogen) and random primers (Promega), according to the instructions provided by t he manufacturers. Control reactions were not subjected to cDNA synthesis. One l of cDNA at the appropriate dilution (as determined empirically) and 20 pmol of each primer were subjected to qPCR in a 25 l reaction using 2× reaction buffer ( 20 mM Tris - HC l [pH 8.3], 13 mM MgCl 2 , 100 mM KCl, 400 M dNTPs, 4% DMSO, 2 × SYBR Green I [Molecular Probes], 0.01% Tween 20, 0.01% NP40, and 0.01 g/ l of Taq polymerase) as described previously [77] . qPCR was done in quadruplicate for each cDNA using a LightCycler ® 480 System (Roche). A standard curve was generated for each set of qPCRs using M. xanthus wild - type strain DK1622 genomic DNA and gene expression was quantified using the relative standard curve method (user bulletin 2; Applied Biosystems). 16S rRNA wa s used 78 as the internal standard for each sample. Relative transcript levels for mutants are the average of three biological replicates after each replicate was normalized to the transcript level observed for one replicate of WT at 18 h PS in the same expe riment. Transcript levels for WT at other times PS were likewise normalized to that observed for WT at 18 h PS in the same experiment. For WT at 18 h PS, the transcript levels of at least three biological replicates from different experiments were normal ized to their average, which was set as 1. Immunoblot analysis A semi - quantitative method of immunoblot analysis was devised to measure the relative levels of MrpC and FruA in many samples collected in different experiments. Equal volumes (10 µ l for mea surement of MrpC and 15 µ l for measurement of FruA) of samples prepared for immunoblot analysis as described above were subjected to SDS - PAGE and immunoblotting as described previously [14, 78] . 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 sho rt enough to ensure signals were not saturated, and signal intensities were quantified using Image Lab 5.1 (Bio - Rad) software. After normalization to the internal control, each signal intensity was divided by the total protein concentration of a correspon ding sample that had been sonicated for 10 - s intervals three times as described above. After removal of a sample for spore quantification, the remaining portion was centrifuged at 10,000 × g for 1 min and the total protein concentration of the supernatant was determined using a Bradford [79] assay kit (Bio - Rad). The resulting values of normalized signal intensity/total protein concentration were 79 further normalized to the average value for all biological replicates of WT at 18 h PS, which was set as 1. Mathematical modeling Activation of dev transcription by FruA and MrpC FruA and MrpC bind cooperatively to the dev promoter region and activate transcription [20] . In agreement, no dev mRNA was detected in either the fruA mutant (Fig. 4A) or the mrpC mutant (Fig. 7). We represent the activation of dev transcript by FruA and MrpC using a where denotes the maximal d ev transcription rate, is the half - saturation constant, and denotes the cooperativity of binding. Note that this expression will give when or (i.e., we have neglected any basal transcription rate as we did not detect dev mR NA in the fruA or mrpC mutant. The expression in brackets can be thought as the promoter occupancy probability ( in the equation below), a dimensional parameter telling what fraction of the promoters will be occupied by the transcription factors for a given value of . Note that the sensitivity of this expression to changes in the c oncentrations of FruA and MrpC are maximal when and minimal near saturation when . In Figure 5 we assess how 80 different hypotheses about the role of C - signaling in dev regulation play out at different levels of . To facilitate the biological i nterpretation of the findings, we plot these as a function of dev promoter saturation. Feedback regulation by Dev proteins The dev mRNA level is further regulated by Dev proteins DevI and DevS. Our finding that the dev transcript level is lower in the dev I mutant than in WT (Fig. 4A) indicates that DevI is a positive regulator of dev mRNA accumulation. In contrast, the dev transcript level in the devS mutant is significantly higher than in WT (Fig. 4A), indicating that DevS is a negative regulator of dev mRNA accumulation. Since the exact mechanisms of regulation by DevI and DevS are unclear, we assume for simplicity that these proteins regulate the dev transcript level through independent mechanisms. We model these regulation functions as follows: Here, I is a dimensionless parameter characterizing the feedback strength (i.e., the fold - increase in transcription of the dev operon due to DevI), K I is the half - saturation constant, and b denotes cooperativity of DevI binding. Likewise, K S is the half - saturation constant and c denotes the cooperativity of DevS binding. Note that these functions are normalized so that for the devI mutant and for the devS mutant (i.e., when or . We ass ume that regulation by the Dev proteins is independent of that by FruA and MrpC, and the effects will be multiplicative: 81 where, K FM , K I , and K S are the saturation constants for regulation by [FruA][MrpC], [DevI], and [DevS], respectively. Numerical procedure to estimate unknown regulation parameters To explain the difference in the dev mRNA level in the csgA mutant as compared with WT, in terms of perturb ation of potential regulatory mechanisms, we use a mathematical approach where we constrain the FruA ratio and find the regulation parameters that can result in the observed - fold difference in [ . Specifically, we use the expression of dev transcript ratio between WT and the csgA mutant below: where, and . First, we estimat e the contribution from Dev protein regulation terms in determining the dev transcript level in WT and the csgA mutant. Since we did not measure the Dev proteins 82 explicitly in our experiments, we estimate their contribution in regulating dev transcription in WT by comparing the changes in transcript level in their absence (i.e., in the devI and devS mutants). Based on our transcript data for WT, and the devI and d evS mutants (Fig. 4A), we have the following relations between the regulation functions; and . Using these relations, we obtain . For the csgA mutant, assuming regulation by Dev proteins is absent due to the low dev transcript level, we have and . With these estimates, the above expression for dev transcript ratio has three unknown parameters . Next, we determine the required fold change in degradation rate for different promoter saturation probability values that explains the observed 22 - fold difference in dev transcript. To estimate this, we set the cooperativity constant to 2 and take the fold change in FruA from the experiments, while assuming MrpC is unchanged between WT and the csgA mutant. The result is plotted in Fig. 5A. Then, we determine the r equired cooperativity for different values with the FruA fold change from the experiments and assuming no change in the degradation rate ( ). The result is plotted in Fig. 5B. Finally , we compute the fold change in FruA with and for different values. The result is shown in Fig. 5C. RNA stability At the indicated time the submerged culture supernatant was replaced with fresh MC7 starvation buffer supplemented with 50 g/ml of rifampicin to inhibit RNA synthesis. Samples were collected immediately (designated t 0 ) and 8 and 16 min later for RNA extraction and analysis as described above, except for each biological replicate the transcript levels after 8 and 83 16 min were normalized to the transcript level at t 0 , which was set as 1, and the natural log of the resulting values was plotted versus minutes after rifampicin treatment and the slope of a linear fit of the data was used to compute the mRNA half - life. Induction of P van - fruA To induce expression of fruA and fruA (D59E) fused to a vanillate - inducible promoter in M. xanthus , the CTTYE growth medium was supplemented with 0.5 mM vanillate when the culture reached 50 Klett units. Growth was continued until the culture r eached 100 Klett units, then the culture was centrifuged and cells were resuspended at a density of approximately 1,000 Klett units in MC7 supplemented with 0.5 mM vanillate, followed by submerged culture development as described previously [14] . Acknowledgements We thank Monique Floer for advice about high - throughput qPCR and for use of the LightCycler® 480 System. We thank Montserrat Elias - Arnanz and Penelope Higgs for sharing strains. This work was supported by the National Science Foundation (award MCB - 1411272) and by salary support for L.K. from Michigan State University AgBioResearch. Author contributions Conception or design of the study: LK, OI, SS, PP Acquisition of the data: SS, PP Analysis or interpretation of the data: SS, PP, LK , OI Writing of the manuscript: LK, SS, PP, OI 84 Figure 2.10 Abbreviated summary. Starvation promotes MrpC accumulation, whereas nutrients favor proteolysis. MrpC activates transcription of fruA , but FruA protein appears to be activated by short - range C - signaling in a cycle leading to mound formation and lysis of some cells. Activated FruA* and MrpC are proposed to cooperatively stimulate transcription of the dev operon and genes that commit starving rod - shaped cells to form spores, while Dev proteins slow commitment, resulting in a spore - filled fruiting body surrounded by peripheral r ods. 85 APPENDIX 86 Table S 2. 1 Cell and spore numbers counted in chapter 2 . Strain Sonication - sensitive cells at T 0 (10 7 /mL) Sonication - resistant spores at T 48 (10 7 /mL) wild type 140 ± 16 1.5 ± 0.3 csgA 150 ± 30 < 0.05 fruA 150 ± 25 < 0.05 devI 150 ± 27 2 ± 0.4 devS 140 ± 17 < 0.05 ladA 150 ± 5 0.7 ± 0.3 distal site 150 ± 6 1.4 ± 0.2 mrpC 130 ± 16 < 0.05 csgA P van - fruA 150 ± 7 < 0.05 csgA P van - fruA D59E 140 ± 8 < 0.05 Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions. The number of rod - shaped sonication - sensitive cells at T 0 and the number of sonication - resistant spores at 48 h poststarvation were counted microscopically using a Neubauer chamber. Values indicate the average of at least 3 biological replicates and one standard deviation. 87 Table S 2 .2 Strains, plasmids and primers used in chapter 2 . Bacterial strain Description Source E. coli DH5a l - f80d lacZD M15 D ( lacZYA - argF ) U169 recA1 endA1 hsdR17 (r K - m K - ) supE44 thi - 1 gyrA relA1 [80] M. xanthus DK1622 Laboratory strain [81] SW2808 mrpC [15] DK5285 fruA ::Tn 5 lac Km r ) [30] DK11209 devS [27] MRR7 devI [29] DK5208 csgA ::Tn 5 - 132 W205 (Tc r ) [82] MRR33 csgA ::pRR028 (Km r ) [28] MSS1 A deletion of chromosomal DNA between positions - 254 and - 228 relative to the dev transcriptional start site [20] MSS3 csgA ::pRR028 (Km r ) MXAN_0018 - MXAN_0019::p SS10 (Tc r ) This study MSS5 csgA ::pRR028 (Km r ) MXAN_0018 - MXAN_0019::pSS9 (Tc r ) This study M SS6 fruA ::Tn 5 lac r ) MXAN_0018 - MXAN_0019::p SS10 (Tc r ) This study MSS7 fruA ::Tn 5 lac r ) MXAN_0018 - MXAN_0019::pSS9 (Tc r ) This study MRR027 DK1622:: ladA This study Plasmids Description Source p SS10 Tc r ; pMR3691 with fruA inserted at MCS_G This study pSS9 Tc r ; pMR3691 with fruA (D59E) inserted at MCS_G This study pMR3691 Tc r ; M. xanthus MXAN_0018 - MXAN_0019 - P R3 - 4 :: vanR - P van - MCS_G [76] Primers Description Source FruA - F - NdeI - Gibson GATGCGAGGAAACGCATATGGCAACCAATCAAGCAG CGATTCGTG This study FruA - R - EcoRI - Gibson GTACGCGTAACGTTCGAATTCCTAGAGGTCCGGCGGC GGCCGGA This study pMR3691 MCS G - F CACGATGCGAGGAAACGCA This study pMR3691 MCS G - R CACCGGTACGCGTAACGTTC This study 88 Table S2.2 ( c Primers Description Source 16S rRNA fwd CAAGGGAACTGAGAGACAGG [83] 16S rRNA rev CTCTAGAGATCCACTACTTGCG [83] fruA oPH252 CGTCACGGAAGGCATCAATC [28] fruA oPH253 CGAGATGATTTCCGGTGTGC [28] mrpC qPCR F GGAGGCCATCGACTTCAAGG [14] mrpC qPCR R GGCCGGACTTCAGCAGGTAG [14] cas6 - F TGGGG AAATCTAATGGTGTTTG This study cas6 - R GAGAACAGCAGATAGGCATGGT This study D59E F CCGCAGGTCGCGGTGATGGAGGTGGAGGGCGACAGCGAG This study D59E R CTCGCTGTCGCCCTCCACCTCCATCACCGCGACCTGCGG This study FmgA - F9 AAGACGCGCATCAAGGACG This study FmgA - R9 CCAGACTTCGAAGCCATCCGAG This study FmgB - F3 TGCGCTGCTGTACGACTCC This study FmgB - R3 GATGGCCTGGACGGGGCA This study FmgD - F3N TTACGGTGGCACCGCATTC This study FmgD - R3N CTGGGCTTCCGTCATCTTG This study FmgE - F3N CTCATCTGTCGCGGCCAA This study FmgE - R3N ACAGCGGTCAGTTCTGAATG This study LadA - F2 TTCACCTCGCCCTGCGCC This study LadA - R2 GATGGACAACGTGGAGAC This study 89 Figure S2.1 Cellular changes during M. xanthus development. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for quantification of (A) sonication - sensitive cells and (B) s onication - resistant spores. Values are expressed as a percentage of the number of rod - shaped cells present at the time starvation initiated development (T 0 ). Bars show the average of three biological replicates and error bars show one standard deviation. 90 Figure S2.2 Reproducibility of RNA measurements. 91 Figure S2.2 (A) Experimental scheme. Three biological replicates of wild - type DK1622 were subjected to starvation under submerged culture conditions and samples were collected at 24 h poststarvation. One biological replicate sample was used to prepare RNA in triplicate and one of these RNA s amples was used to prepare cDNA in triplicate. (B - E) Variation in transcript numbers among cDNA technical replicates, RNA technical replicates (the average of cDNA technical replicates was used as one of the values), and biological replicates. Transcript n umbers are per mg total RNA. Bars show the average and error bars show one standard deviation. Figure S2.3 Validation of 16S rRNA as an internal standard for RT - qPCR analysis during M. xanthus development . Four biological replicates of wild - type DK1 622 were subjected to starvation under submerged culture conditions and RNA was prepared from samples collected at the indicated times poststarvation. (A) Transcript numbers per mg total RNA. (B) Total RNA yield per cell. The RNA yield in femtograms (fg) was divided by the number of rod - shaped cells in the sample prior to RNA preparation. Bars show the average and error bars show one standard deviation. 92 . Figure S2.4 fruA transcript stability in wild type and csgA mutant . Wild - type DK1622 and the csgA mutant were subjected to starvation under submerged culture conditions for 30 h. The overlay was replaced with fresh starvation buffer containing rifampicin (50 mg/ml) and samples were collected immediately (t 0 ) and at the times indicated (t x ) for measurement of the fruA transcript level by RT - qPCR. Transcript levels at t x were normalized to that at t 0 for each of three biological replicates and used to determine the transcript half - life for each replicate. The average hal f - life (Average t 1/2 ) and one standard deviation are shown, and the difference is not statistically significant ( p - tailed t - test) . The graph shows the average ln(t x / t 0 ) and one standard deviation for the three biological replicates of wild type (black dashed line) and the csgA mutant (gray solid line). 93 Figure S2.5 Levels of dev and fmg transcripts in a ladA mutant. Wild type DK1622 and its ladA mutant derivative were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of (A) dev , (B) fmgA , (C) fmgBC , (D) fmgD , and (E) fmgE transcript levels by RT - qPCR. Graphs show the data points and average of three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p < 0.05 in Stu - tailed t - tests) from wild type at the corresponding time PS. 94 Figure S2.6 Developmental phenotype and quantification of sonication resistant spores of the ladA mutant . Wild - type DK1622 and its ladA mutant derivative were subjected to starvation under submerged culture conditions. (A) Microscopy. Images were obtained at the indicated number of hours poststarvation (PS). DK1622 formed mounds by 18 h PS (an arrow points to one) and the mounds darkened at 36 to 48 h. The ladA mutant formed mounds at 30 h, and the mounds did not darken until 48 h. Bar, 100 mm. Similar results were observed in at least three biological replicates. (B) Quantification of sonication - resistant spores. Values are expressed as a percentage of the number of rod - shaped cells present at the time starvation initiated development (T 0 ) (Table S1). Bars show the average of three biological replicates and error bars show one standard deviation. 95 Figure S2.7 Levels of MrpC and FruA in wild type, distal site mutant and mrpC mutant during M. xanthus development. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of (A) MrpC and (B) FruA levels by immuno blot. Graphs show the data points and average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p - tailed t - test s) from wild type at the corresponding time PS. 96 Figure S2.8 mrpC, fruA and dev transcript levels in wild type, distal site mutant and mrpC mutant during M. xanthus development . Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of (A) mrpC , (B) fruA , and (C) dev transcript levels by RT - qPCR. Graphs show the data points and average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p < 0.05 in Stu - tailed t - tests) from wild type at the corresponding time PS. 97 Figure S2.9 mrpC transcript stability in wild type and mrpC mutant at 18 h poststarvation . Wild - type DK1622 and the mrpC mutant were subjected to starvation under submerged culture conditions for 18 h. The overlay was replaced with fresh starvation buffer containing rifampicin (50 mg/ml) and samples were collected immediately (t 0 ) and at the times indicated (t x ) for measur ement of the mrpC transcript level by RT - qPCR. Transcript levels at t x were normalized to that at t 0 for each of three biological replicates and used to determine the transcript half - life for each replicate. The average half - life (Average t 1/2 ) and one s tandard deviation are shown, and the difference is not statistically significant ( p two - tailed t - test) . The graph shows the average ln(t x / t 0 ) and one standard deviation for the three biological replicates of wild type (black dashed line) and the mrpC mutant (gray solid line). Figure S2.10 Developmental phenotype of wild type, distal site mutant and mrpC mutant . Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions an d images were obtained at the indicated number of hours poststarvation (PS). The wild type and the distal site mutant formed mounds by 18 h PS (an arrow points to one) and the mounds darkened at 36 to 48 h. The mrpC mutant failed to form mounds. Bar, 100 µm. Similar results were observed in at least three biological replicates. 98 Figure S2.11 Cellular changes in wild type, distal site mutant and mrpC mutant . Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hour s poststarvation for quantification of (A) sonication - sensitive cells and (B) sonication - resistant spores. Values are expressed as a percentage of the number of rod - shaped cells present at the time starvation initiated development (T 0 ) (Table S1). Bars sho w the average of three biological replicates and error bars show one standard deviation Figure S2.12 Mathematical modeling prediction of the required reduction in FruA activity in the csgA mutant in comparison to wild type, to explain the experimental da ta. Bars show the average of 108 datasets representing all possible combinations of four biological replicates of wild type and three biological replicates of each mutant ( csgA, devI, devS ), and error bars show one standard deviation. 99 Figure S2.13 mrpC and fruA transcript levels in wild type, csgA mutant and csgA (P van - fruA ) mutant. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of (A) fruA and (B) mrpC transcript level s by RT - qPCR. Bars show the average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p - tailed t - tests) from wi ld type at the corresponding time PS. 100 Figure S2.14 MrpC protein level in wild t ype, csgA mutant and csgA (P van - fruA ) mutant . Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of MrpC levels by immunoblot. Bars show the average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p two - tailed t - tests) from wild type at the corresponding time PS. Figure S2.15 Developmental phenotype of wild type, csgA mutant and fruA (P van - fruA ) mutant. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and images were obtained at the indicated number of hours poststarvation (PS). The wild type formed mounds by 18 h PS (an arrow points to one) and the mounds darkened at 36 to 48 h. The csgA P van - fruA , csgA P van - fruA D59E , and csgA mutants failed to form mounds. Bar, 100 mm. Similar results were observed in at least three biological replicates. 101 Figure S2.16 Cellular changes of wild type, csgA mutant and fruA (P van - fruA ) mutant . Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for quantification of sonication - sensitive cells. Values are ex pressed as a percentage of the number of rod - shaped cells present at the time starvation initiated development (T 0 ) (Table S1). Bars show the average of three biological replicates and error bars show one standard deviation. Figure S2.17 Developmental phenotype of fruA (P van - fruA ) mutant and fruA (P van - fruAD59E ) mutant . Wild - type DK1622 without or with vanillate induction and its indicated mutant derivatives with vanillate induction were subjected to starvation under submerged culture conditions and images were obtained at the indicated number of hours poststarvation (P S). The wild type without or with vanillate, and the fruA P van - fruA and fruA P van - fruA D59E strains, formed mounds by 18 h PS (arrows point to mounds) and the mounds darkened at 36 to 48 h. The fruA mutant failed to form mounds. Bar, 100 mm. Similar resul ts were observed in at least three biological replicates. 102 Figure S 2.18 Levels of fmg transcripts in csgA mutant and csgA (P van - fruA ) mutant, csgA mutant and csgA (P van - fruAD59E ) mutant. 103 Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of (A) fmgA , (B) fmgBC , (C) fmgD , and (D) fmgE transcript levels by RT - qPCR. Bars show the average of at least three biological replicates, relative to wild - type DK1622 at 18 h PS, and error bars show one standard deviation. Asterisks indicate a significant difference ( p - tailed t - tests) from wild type at the corresponding time PS. 104 Figure S2.19 Models for regulation of fmgD and fmgE . C - signaling causes the level of activated FruA* to rise as development proceeds (triangles). (A) Coopera tive binding of two MrpC initially represses fmgD transcription, but eventually FruA* outcompetes the downstream MrpC for binding to the upstream MrpC, activating transcription. (B) MrpC and activated FruA* bind cooperatively first to a higher affinity cen tered at - 100 bp relative to the fmgE transcriptional start site. As FruA* rises, the lower affinity site just upstream of the promoter is also cooperatively bound by FruA* and MrpC, activating transcription. In both panels, boxes indicate the promoter - 3 5 and - 10 regions. 105 Figure S2.20 Representative MrpC and FruA immunoblots for wild type and mutants. Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation (PS) for measurement of MrpC and FruA levels by immunoblot. Equ al volumes (10 µ l for measurement of MrpC and 15 µ l for measurement of FruA) of whole - cell extract samples were subjected to semi - quantitative immunoblot analysis as described in the Experimental Procedures. 106 REFERENCES 107 REFERENCES 1. Davidson EH, Levine MS: Properties of developmental gene regulatory networks. Proc Natl Acad Sci USA 2008, 105(51):20063 - 20066. 2. Frum T, Ralston A: Cell signaling and transcription factors regulating cell fate during formation of the mouse blastocyst. Trends Genet 2015, 31(7):402 - 410. 3. Drapek C, Sparks EE, Benfey PN: Uncovering gene regulatory networks controlling plant cell diff erentiation. Trends Genet 2017, 33(8):529 - 539. 4. van Gestel J, Vlamakis H, Kolter R: Division of labor in biofilms: the ecology of cell differentiation. Microbiol Spectr 2015, 3(2):MB - 0002 - 2014. 5. Norman TM, Lord ND, Paulsson J, Losick R: Stochastic sw itching of cell fate in microbes. Annu Rev Microbiol 2015, 69:381 - 403. 6. Bush MJ, Tschowri N, Schlimpert S, Flardh K, Buttner MJ: c - di - GMP signalling and the regulation of developmental transitions in streptomycetes. Nat Rev Microbiol 2015, 13(12):749 - 76 0. 7. Kroos L: Highly signal - responsive gene regulatory network governing Myxococcus development. Trends Genet 2017, 33:3 - 15. 8. Barratt MJ, Lebrilla C, Shapiro HY, Gordon JI: The gut microbiota, food science, and human nutrition: a timely marriage. Cell Host Microbe 2017, 22(2):134 - 141. 9. Jansson JK, Hofmockel KS: The soil microbiome - from metagenomics to metaphenomics. Curr Opin Microbiol 2018, 43:162 - 168. 10. Yang Z, Higgs P (eds.): Myxobacteria: genomics, cellular and molecular biology. Norfolk, U K: Caister Academic Press; 2014. 11. Bretl DJ, Kirby JR: Molecular mechanisms of signaling in Myxococcus xanthus development. J Mol Biol 2016, 428:3805 - 3830. 12. O'Connor KA, Zusman DR: Development in Myxococcus xanthus involves differentiation into two cell types, peripheral rods and spores. J Bacteriol 1991, 173(11):3318 - 3333. 13. Lee B, Holkenbrink C, Treuner - Lange A, Higgs PI: Myxococcus xanthus developmental cell fate production: heterogeneous accumulation of developmental regulatory proteins and reexamination of the role of MazF in developmental lysis. J Bacteriol 2012, 194(12):3058 - 3068. 14. Rajagopalan R, Kroos L: Nutrient - regulated proteoly sis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 2014, 196:2736 - 2747. 108 15. Sun H, Shi W: Genetic studies of mrp , a locus essential for cellular aggregation and sporulation of M yxococcus xanthus . J Bacteriol 2001, 183(16):4786 - 4795. 16. Sun H, Shi W: Analyses of mrp genes during Myxococcus xanthus development. J Bacteriol 2001, 183(23):6733 - 6739. 17. Robinson M, Son B, Kroos L: Transcription factor MrpC binds to promoter region s of many developmentally - regulated genes in Myxococcus xanthus . BMC Genomics 2014, 15:1123. 18. Ueki T, Inouye S: 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 USA 2003, 100(15):8782 - 8787. 19. Ogawa M, Fujitani S, Mao X, Inouye S, Ko mano T: FruA, a putative transcription factor essential for the development of Myxococcus xanthus . Mol Microbiol 1996, 22(4):757 - 767. 20. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L: Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 2015, 197:240 - 251. 21. Lee J, Son B, Viswanathan P, Luethy P, Kroos L: Combinatorial regulation of fmgD by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 2011, 193:1681 - 1689. 22. Mittal S, Kroos L: A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proc Natl Acad Sci USA 2009, 106:1965 - 1970. 23. Mittal S, Kroos L: Combinatorial regulation by a novel arra ngement of FruA and MrpC2 transcription factors during Myxococcus xanthus development. J Bacteriol 2009, 191:2753 - 2763. 24. Son B, Liu Y, Kroos L: Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcu s xanthus development. J Bacteriol 2011, 193:2756 - 2766. 25. Boysen A, Ellehauge E, Julien B, Sogaard - Andersen L: The DevT protein stimulates synthesis of FruA, a signal transduction protein required for fruiting body morphogenesis in Myxococcus xanthus . J Bacteriol 2002, 184(6):1540 - 1546. 26. Thony - Meyer L, Kaiser D: devRS , an autoregulated and essential genetic locus for fruiting body development in Myxococcus xanthus . J Bacteriol 1993, 175:7450 - 7462. 27. Viswanathan P, Murphy K, Julien B, Garza AG, Kroos L: Regulation of dev , an operon that includes genes essential for Myxococcus xanthus development and CRISPR - associated genes and repeats. J Bacteriol 2007, 189:3738 - 3750. 28. Rajagopalan R, Kroos L: The d ev operon regulates the timing of sporulation during Myxococcus xanthus development. J Bacteriol 2017, 199(10):e00788 - 00716. 109 29. Rajagopalan R, Wielgoss S, Lippert G, Velicer GJ, Kroos L: devI is an evolutionarily young negative regulator of Myxococcus xanthus development. J Bacteriol 2015, 197(7):1249 - 1262. 30. Kroos L, Kaiser D: Expression of many developmentally regulated genes in Myxococcus depends on a sequence of cell interactions. Genes Dev 1987, 1:840 - 854. 31. Ellehauge E, Norregaard - Madsen M, Sogaard - Andersen L: The FruA signal transduction protein provides a checkpoint for the temporal co - ordination of intercellular signals in Myxococcus xanthus development. Mol Microbiol 1998, 30:807 - 817. 32. Shimkets LJ, Gill RE, Kaiser D: Develo pmental cell interactions in Myxococcus xanthus and the spoC locus. Proc Natl Acad Sci USA 1983, 80:1406 - 1410. 33. Kim SK, Kaiser D: Cell motility is required for the transmission of C - factor, an intercellular signal that coordinates fruiting body morphog enesis of Myxococcus xanthus . Genes Dev 1990, 4:896 - 905. 34. Kim SK, Kaiser D: Cell alignment required in differentiation of Myxococcus xanthus . Science 1990, 249:926 - 928. 35. Kroos L, Hartzell P, Stephens K, Kaiser D: A link between cell movement and ge ne expression argues that motility is required for cell - cell signaling during fruiting body development. Genes Dev 1988, 2:1677 - 1685. 36. Sager B, Kaiser D: Two cell - density domains within the Myxococcus xanthus fruiting body. Proc Natl Acad Sci USA 1993, 90:3690 - 3694. 37. Kim SK, Kaiser D: C - factor: a cell - cell signaling protein required for fruiting body morphogenesis of M. xanthus . Cell 1990, 61:19 - 26. 38. Lobedanz S, Sogaard - Andersen L: Identification of the C - signal, a contact - dependent morphogen co ordinating multiple developmental responses in Myxococcus xanthus . Genes Dev 2003, 17(17):2151 - 2161. 39. Rolbetzki A, Ammon M, Jakovljevic V, Konovalova A, Sogaard - Andersen L: Regulated secretion of a protease activates intercellular signaling during frui ting body formation in M. xanthus . Dev Cell 2008, 15(4):627 - 634. 40. Sun H, Shi, W.: Genetic studies of mrp, a locus essential for cellular aggregation and sporulation of Myxococcus xanthus. Journal of bacteriology 2001, 183(16):4786 - 4795. 41. Sun H, Shi, W.: Analyses of mrp genes during Myxococcus xanthus development. Journal of bacteriology 2001, 183(23):6733 - 6739. 42. Nariya H, Inouye, S.: Identification of a protein Ser/Thr kinase cascade that regulates essential transcriptional activators in Myxococcus xanthus development. Molecular microbiology 2005, 58(2):367 - 379. 110 43. Hagen TJ, Shimkets, L. J.: Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. Journal of bacteriology 1990, 172(1):15 - 23. 44. Kim SK, Kaiser, D.: C - factor: a cell - cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 1990, 61(1):19 - 26. 45. Ueki T, Inouye, S.: Identification of an activator protein required for the induction of fruA, a gene essential for fruiting body development in Myxococcus xanthus. Proceedings of the National Academy of Sciences of the United States of America 2003, 100(1 5):8782 - 8787. 46. Ellehauge E, Norregaard - Madsen, M, Sogaard - Andersen, L.: The FruA signal transduction protein provides a checkpoint for the temporal co - ordination of intercellular signals in Myxococcus xanthus development. Molecular microbiology 1998, 3 0(4):807 - 817. 47. Mittal S, Kroos, L.: A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proceedings of the National Academy of Sciences of the United States of America 2009, 10 6(6):1965 - 1970. 48. Campbell A, Viswanathan, P, Barrett, T, Son, B, Saha, S, Kroos, L.: Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. Journal of bacteriology 2015, 197(2):240 - 251. 49. Thony - Meyer L, K aiser, D.: devRS, an autoregulated and essential genetic locus for fruiting body development in Myxococcus xanthus. Journal of bacteriology 1993, 175(22):7450 - 7462. 50. Viswanathan P, Murphy, K, Julien, B, Garza, A. G, Kroos, L.: Regulation of dev, an ope ron that includes genes essential for Myxococcus xanthus development and CRISPR - associated genes and repeats. Journal of bacteriology 2007, 189(10):3738 - 3750. 51. Rajagopalan R, Wielgoss, S, Lippert, G, Velicer, G. J, Kroos, L.: devI is an evolutionarily young negative regulator of Myxococcus xanthus development. Journal of bacteriology 2015, 197(7):1249 - 1262. 52. Rajagopalan R, Kroos L: The dev Operon Regulates the Timing of Sporulation during Myxococcus xanthus Development. Journal of bacteriology 2017, 199(10). 53. Ogawa M, Fujitani, S, Mao, X, Inouye, S, Komano, T.: FruA, a putative transcription factor essential for the development of Myxococcus xanthus. Molecular microbiology 1996, 22(4):757 - 767. 54. Boynton TO, Shimkets LJ: Myxococcus CsgA, Drosop hila Sniffer and human HSD17B10 are cardiolipin phospholipases. Genes Dev 2015, 29:1903 - 1914. 55. Stock AM, Robinson VL, Goudreau PN: Two - component signal transduction. Annu Rev Biochem 2000, 69:183 - 215. 111 56. Nariya H, Inouye S: Identification of a protein Ser/Thr kinase cascade that regulates essential transcriptional activators in Myxococcus xanthus development. Mol Microbiol 2005, 58(2):367 - 379. 57. Nariya H, Inouye S: A protein Ser/Thr kinase cascade negat ively regulates the DNA - binding activity of MrpC, a smaller form of which may be necessary for the Myxococcus xanthus development. Mol Microbiol 2006, 60(5):1205 - 1217. 58. Nariya H, Inouye M: MazF, an mRNA interferase, mediates programmed cell death durin g multicellular Myxococcus development. Cell 2008, 132(1):55 - 66. 59. Schramm A, Lee B, Higgs PI: Intra - and inter - protein phosphorylation between two hybrid histidine kinases controls Myxococcus xanthus developmental progression. J Biol Chem 2012, 287:250 60 25072. 60. McLaughlin PT, Bhardwaj V, Feeley BE, Higgs PI: MrpC, a CRP/Fnr homolog, functions as a negative autoregulator during the Myxococcus xanthus multicellular developmental program. Mol Microbiol 2018, Epub ahead of print. 61. Srinivasan D, Kro os L: Mutational analysis of the fruA promoter region demonstrates that C - box and 5 - base - pair elements are important for expression of an essential developmental gene of Myxococcus xanthus . J Bacteriol 2004, 186(17):5961 - 5967. 62. Viswanathan P, Ueki T, Inouye S, Kroos L: Combinatorial regulation of genes essential for Myxococcus xanthus development involves a response regulator and a LysR - type regulator. Proc Natl Acad Sci USA 2007, 104:7969 - 7974. 63. Bourret RB: Receiver domain s tructure and function in response regulator proteins. Curr Opin Microbiol 2010, 13(2):142 - 149. 64. Desai SK, Winardhi RS, Periasamy S, Dykas MM, Jie Y, Kenney LJ: The horizontally - acquired response regulator SsrB drives a Salmonella lifestyle switch by re lieving biofilm silencing. eLife 2016, 5:e10747. 65. Wang L, Tian X, Wang J, Yang H, Fan K, Xu G, Yang K, Tan H: Autoregulation of antibiotic biosynthesis by binding of the end product to an atypical response regulator. Proc Natl Acad Sci U S A 2009, 106( 21):8617 - 8622. 66. Licking E, Gorski L, Kaiser D: A common step for changing cell shape in fruiting body and starvation - independent sporulation of Myxococcus xanthus . J Bacteriol 2000, 182(12):3553 - 3558. 67. Muller FD, Schink CW, Hoiczyk E, Cserti E, Hig gs PI: Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Mol Microbiol 2012, 83(3):486 - 505. 112 68. Ueki T, Inouye S: Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development. J Biol Chem 2005, 280(37):32279 - 32284. 69. Boynton TO, McMurry JL, Shimkets LJ: Characterization of Myxococcus xanthus MazF and implications for a new point of regulation. Mol Microbiol 2013, 87(6):1267 - 1276. 70. Cho K, Zusman DR: Sporulation timing in Myxococcus xanthus is contolled by the espAB locus. Mol Microbiol 1999, 34:714 - 725. 71. Higgs PI, Jagadeesan S, Mann P, Zusman DR: EspA, an orphan hybrid histidine protein kinase, regulates the timing of expression of key developmental proteins of Myxococcus xanthus . J Bacteriol 2008, 190:4416 - 4426. 72. Mangan S, Alon U: Structure and function of the feed - forward loop net work motif. Proc Natl Acad Sci USA 2003, 100(21):11980 - 11985. 73. Mangan S, Zaslaver A, Alon U: The coherent feedforward loop serves as a sign - sensitive delay element in transcription networks. J Mol Biol 2003, 334(2):197 - 204. 74. Gibson DG, Young L, Chu ang RY, Venter JC, Hutchison CA, 3rd, Smith HO: Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 2009, 6(5):343 - 345. 75. Kashefi K, Hartzell PL: Genetic suppression and phenotypic masking of a Myxococcus xanthus frzF - defec t. Molecular microbiology 1995, 15(3):483 - 494. 76. Iniesta AA, Garcia - Heras F, Abellon - Ruiz J, Gallego - Garcia A, Elias - Arnanz M: Two systems for conditional gene expression in Myxococcus xanthus inducible by isopropyl - - D - thiogalactopyranoside or vanillat e. J Bacteriol 2012, 194(21):5875 - 5885. 77. Bryant GO, Prabhu V, Floer M, Wang X, Spagna D, Schreiber D, Ptashne M: Activator control of nucleosome occupancy in activation and repression of transcription. PLoS Biol 2008, 6(12):2928 - 2939. 78. Yoder - Himes D, Kroos L: Regulation of the Myxococcus xanthus C - signal - dependent 4400 promoter by the essential developmental protein FruA. J Bacteriol 2006, 188:5167 - 5176. 79. Bradford M: A rapid and sensitive method for the quantitation of microgram qu antities of protein utilizing the principle of protein - dye binding. Anal Biochem 1976, 72:248 - 254. 80. Hanahan D: Studies on transformation of Escherichia coli with plasmids. J Mol Biol 1983, 166:557 - 580. 81. Kaiser D: Social gliding is correlated with t he presence of pili in Myxococcus xanthus . Proc Natl Acad Sci USA 1979, 76:5952 - 5956. 82. Shimkets LJ, Asher SJ: Use of recombination techniques to examine the structure of the csg locus of Myxococcus xanthus . Mol Gen Genet 1988, 211:63 - 71. 113 83. Ossa F, Diodati ME, Caberoy NB, Giglio KM, Edmonds M, Singer M, Garza AG: The Myxococcus xanthus Nla4 protein is important for expression of stringent response - associated genes, ppGpp accumulation, and fruiting body development. J Bacteriol 2007, 189(23):8474 - 8483. 114 C HAPTER 3: Differential regulation of late - acting operons by FruA and MrpC during Myxococcus xanthus development Abstract Upon nutrient depletion Myxococcus xanthus undergoes multicellular development . R od - shaped cells coordinate their movement s to build mounds . W ithin mounds , rods d ifferenti ate in to round stress - resistant spores . Short - range C - signal ing is proposed to activate FruA, which binds DNA cooperatively with MrpC to increase transcription of many genes. This mechanism likely regulates t ranscription of the late - acting fad IJ operon involved in spore metabolism, based on comparisons of transcript levels and degradation rates in wild - type cells and mutants. Regulation of late - acting operons implicated in spore coat biogenesis ( exo A - I , nfs A - H , MXAN_3259 - MXAN_3263 ) was found to be more complex. These operons are negatively regulated by unactivated FruA during mound formation, then positively regulated by C - signal - activated FruA during sporulation. MrpC also negatively regulated exo and MXAN_32 59 during mound formation, but positively regulated nfs . During sporulation, MrpC continued to positively regulate nfs , switched to positive regulation of MXAN_3259 , and continued to negatively regulate exo . DNA - binding studies suggest that FruA exerts its effects by binding to promoter regions, whereas the effects of MrpC may be indirect. A third transcription factor, Nla6 , was shown previously to bind to the exo and MXAN_3259 promoter regions. Here, transcript measurements indicated that Nla6 is a positiv e regulator of all four late - acting operons during mound formation, whereas the small protein DevI is a negative regulator during sporulation. We conclude that multiple regulators control expression of late - acting operons and we propose that differential regulation by FruA in response to C - signaling and by MrpC ensures that spore resistance and surface characteristics meet environmental demands. 115 Introduction The gram - negative soil bacterium Myxococcus xanthus provides an attractive model system to study signal - induced gene regulation and bacterial community behavior. Under starvation conditions, cells move on solid surfaces and form mounds , within which some of the rod - shaped cells differentiate in to round spores [3] . During this multicellular de velopmental process of fruiting body formation, a majority of the population undergoes lys is, while some cells remain outside of fruiting bodies as peripheral rods [4] . T he period between 24 and 30 h poststarvation (PS) is critical for commitment to sporulation , since during this period cells commit to forming spores despite pertur bation of the starvation signal by addition of nutrient medium [5] . The developmental process of M . xanthus is governed by a signal - responsive gene regulatory network [6] . S tarvation triggers produ ction of the intracellular secondary messenger molecule (p)ppGpp , which lead s to production of the extracellular A - and C - signal s [7, 8] . The short - range C - signal is suggested to be a proteolytic fragment of the CsgA protein [9] , or di acylglycerol s produced by enzymatic activity of full - length CsgA [10] . C - signal appears to posttranslationally activate a transcription factor , F ruA [1, 11] , by a n unknown mechanism . FruA is similar to response regul ator s of two - component signal - transduction system s [12] . Response regulators are typically activated by phosphorylation by a histidine kinase [11] . However, recent studies suggest that phosphorylation is unlikely to be the mechanism by which FruA is activated in response to C - signal ing [1, 13] . T ranscription of fruA is regulated by a cascade of starvation - responsive transcription factors . Among these transcription factors , MrpC bind s upstream of the fruA promoter and 116 activate s transcription [14, 15] . MrpC undergoes proteolysis if nutrient medium is added to develop ing cells [5] . This response is ultrasensitive to th e concentration of nutrient medium added [16] . However, by 24 PS, cells begin to commit to sporulation, resisting the e ffects of adding nutrient medium [5] . MrpC and FruA combinatorially regulate transcription of the dev operon [17] , which includes eight genes comprising a CRISPR - C as system that may protect dev eloping M . xanthus against phage infection [18] . The first gene in the dev operon , dev I , codes for a 40 - residue protein that inhibits sporulation when over produc ed [19] , delays sporulation of wild - type strain DK1622 by about 6 h [1, 20] , and exerts we a k positive autoregulation on dev transcript accumulation [1] . In contrast, in - frame deletion s in three genes of the de v operon ( devTRS ) increase accumulation of the dev transcript tenfold during development [1, 20] , indicat ing that Dev TRS proteins exert strong negative autoregulation . Systematic experimental analysis in combination with mathematical modeling of the dev transcript level suggested that C - signal - dependent posttranslational activ ation of FruA is critical for expression of the dev operon and for commitment to spore formation [1] ( see Fig. 3 in C hapter 1). Wh ile transcription of the dev ope ron is critical for the timing of spore formation [20] , the products of other operon s ( exo A - I , nfs A - H , MXAN_3259 - 3263 , fadIJ ) act late during the developmental process, but regulation of the se operons is not we ll - understood [21 - 25] [26] . Here , we report systematic investigat ion of the regulatio n of these late - acting operons, as well as the phenotypes of mutants. Previous work on the regulation and function of the late - acting operons has provided some insights. T ranscription of the exo A - I operon depend s on FruA binding to the promoter region 117 [27] . Additionally, the enhancer - binding protein [28] , Nla6, bind s to the exo promoter region [25] . The p rotein products of the exo operon are involved in the export of polysaccharide chains that f orm the spore coat , which is necessary to generate compact and rigid stress - bearing spores [29] . A transposon insertion mutation in exoC causes a defect in spore morphogenesis upon chemical induction of sporulation [30] . Additionally, a plasmid insertion exoC mutant failed to complete the rod to spore transition during chemical ly - induced sporulation [22] . In this study we investigated the impact of an exoC insertion mutation on sporulation induced by starvation . W e also systematically analyzed the effects of mutations in regulatory genes on the exo transcript le vel during development . The p reviously identified nfsA - H locus [31] was found to be an operon critical for spore morphogenesis upon chemical induction of sporulation [22] . The Nfs proteins appear to arrang e the polysaccharide o f the spore coat after secret ion by Exo proteins [21] . Studies with a reporter fusion to nfsA suggested that C - signaling positively regulates nfs transcription and, unusually, that FruA negatively regulates nfs transcription [31] . We examined these effects, as well as others, by measuring the nfs tran script level in mutants. An i nsertion mutation in the predicted polysaccharide deacetylase encoding gene MXAN_3259 was shown to phenocopy an exo mutant by forming mounds but not mature spores [25] . Nla6 binds to the MXAN_3259 upstream region and posi tive ly regulates transcription early in development , then negatively regulates later , based on comparison of MXAN_3259 transcript levels in an nla6 mutant with a WT strain [25] . We further explored the regulation and function of the putative MXAN_3259 - MXAN_3263 operon. 118 The putative fadIJ operon ( MXAN_5372 - MXAN_5371 ) is induced twofold during development [32] , specifically in sporulating cells [24] , and appears to code for a fatty acid - oxidation pathway that impacts spore structure and resistance properties [24] . A reporter fusion to fadI failed to be induced in the absence of C - signaling, FruA, or MrpC during development [24] . We report systematic investigation of the fadI insertion mutant phenotype and of fadI transcriptional regulation. Results In previous work, w e established methods to systematically analyze M. xanthus development under submerged culture conditions [1] . We used those methods to investigate the effects of mutations in so - [16] , which lie in operons induced later during development than mrpC and fruA . Specifically, we examined t he effects of an insertion in exoC of the exoA - I operon [22] , a deletion of the entire nfsA - H operon [23] , an insertion in MXAN_3259 of the putative MXAN_3259 - MXAN_3263 o peron [25] , and an insertion in fadI of the fadIJ operon ( MXAN5372 - MXAN5371 ) [24] . Figure 1 shows ima ges of wild - type [33] strain DK1622 and each mutant from 18 - 48 h poststarvation (P S). As expected, the WT strain formed distinct mounds by 18 h PS and the mounds began to darken by 30 h [1] . Darkening typically correlates with spore formation. The fadI mutant was indistinguishable fro m the WT strain. The exoC and MXAN _3259 mutants formed normal - looking mound s by 18 h, but the mounds did not darken as much at the WT strain, suggesting a sporulation defect. The nfsA - H mutant was delayed by about 3 h in mound formation and darkening of the mounds appear ed to be delayed and reduced. 119 To quantify changes at the cellular level , samples harvested from submerged culture were treated with glutaraldehyde to fix cells or were left untreated [1] . The untreated samples were used to quantify sonication - resistant spores and mature spor es that are heat - and sonication - - sonication - resistant spores observe d in the corresponding untreated sample). The majority of sonication - sensitive cells are rod - shaped cells, but cells in transition from rods to sonication - resistant spores may also be observed. In agreement with our published data [34] , sonication - resis tant spores were first observed for the WT strain at 27 h PS , and as a percentage of the rod - shaped cells present at the time starvation initiated development ( T 0 ), rose from 0.2 % to nearly 2% by 48 h (Fig . 2) . The fadI mutant formed about two - fold more spores than the WT strain. Consistent with the mound darkening defects we observed (Fig. 1), the nfsA - H mutant formed about two - fold less spores than the WT strain, and the exoC and MXAN _3259 mutants failed to make a det ectable number of spores (at a detection limit of 0.04% of the T 0 number) (Fig. 2). The exoC and MXAN _3259 mutants also failed to make a detectable number of mature spores by 72 h PS , while the nfsA - H mutant made about ten - fold less mature spores than th e WT strain, and the fadI mutant made a similar number of mature spores as the WT strain (Table S1). The WT strain exhibited a similar decline of sonication - sensitive cells during development as reported previously [1] , with only 31% of the cells present at T 0 remaining by 18 h PS and only 6% remaining by 48 h (Fig. S1A). The decrease in cell number correlates with a decrease in the 120 total protein concentration of developing cultures, which was suggested to reflect lysis of the majority of cells early during dev elopment under submerged culture conditions [35] . The mutants showed simil ar decreases in cell number as the WT strain (Fig. S1A). Interestingly, we observed a small percentage of cells in the fixed samples of the exoC and MXAN _3259 mutants that were not rod - shaped. These cells could be premature spores that do not achieve son ication resistance. Previously, exo and nfs mutants have been reported to fail to complete morphogenesis from rods to spores, instead transiently exhibiting deformed cell morphology before reverting into rods [22] . Those observations were made upon chemical induction of sporulation, rather than starvation - induced development, as used in our experiments. We counted the number of cells that were not rod - shaped in the fixed samples. For the WT strain and the fadI and nfsA - H mutants, the number of sonication - resistant spores observed in the corresponding untreated sample (Fig. 2) was subtracted. Using this approach, about 3% of the WT cells present at T 0 were neither rods nor sonication - resistant spores at 24 h (Fig. S1B). T he percentage decreased at later times, presumably as cells in transition from rods to spores became sonication - resistant (Fig. 2). The fadI mutant showed about half as many cells in transition as the WT strain at 24 h, and the number declined to a greater extent in the fadI mutant than in the WT strain by 48 h (Fig. S1B), consistent with the fadI mutant forming a larger number of sonication - resistant spores (Fig. 2). The other mutants exhibited less cells in transition at 24 h (0.4 - 0.7%), but in each case the percentage rose to at least 1% later (Fig. S1B). These results suggest that the exoC and MXAN_3259 mutants, which fail to form a detectable 121 number of sonication - resistant spores, as well as the nfsA - H mutant, which forms about half as many sonication - resistant spores as the WT strain (Fig. 2), begin to change shape during starvation - induced development, but are impaired in their ability to mak e spores, as reported previously for exo and nfs mutants upon chemical induction of sporulation [22] . In agreement, WT cells in transition from rods to spores have been visualized in mounds using a fluorescent membrane s tain and confocal laser scanning microscopy, and all the mutants examined in this Figure 3.1 Development of M. xanthus strains. Wild - type strain DK1622 and its indicated mutant derivates were subjected to starvation under submerged culture condition and images were obtained at the indicated number of hours poststarvation. The wild - type strain and all the mutants except nfsA - H forme d compact mounds by 18 h (an arrow points to one in each panel). The nfsA - H mutant formed compact mounds by 21 h. Mounds of the wild - type strain began to darken by 30 h. Mounds of the exoC , nfsA - H , and MXAN_3259 mutants darkened by 30 h but remained slight ly less dark than mounds of the wild - type strain and the fadI mutant by 48 h. Bar, 100 mm. Similar results were observed in at least three biological replicates. 122 study exhibit transitioning cells in mounds, which are distinguishable from cells undergoing lysis (Y Hoang and Lee Kroos, unpublished data). We conclude that genes in the exoA - I and nfsA - H operons, and in the putative MXAN3259 - MXAN3263 operon, are important for normal spore formation. Late gene transcript levels are very low in the absence of C - signaling In our previous work, we also established methods to systematically analyze transcript levels during M. xanthus development [34] . Those methods were used to investigate the effect of nutrient medium addition to developing cells, and it was found that transcript levels of late genes decrease within 1 h after nutrient medium addition at 18 h PS [16] . Here, we report using Figure 3.2 Cellular changes during M. xanthus development. Wild - type strain DK1622 and its mutant derivatives were subjected to star vation under submerged culture conditions. Samples were collected at the indicated hours post - starvation for quantification of sonication - resistant spores. Values are expressed as a percentage of the number of rod - shaped cells present at the time when star vation initiated development (T 0 ) (Table S1). Bars show the average of three biological replicates and error bars indicate one standard deviation. 123 the same methods to determine the effects of mutations on late gene transcript levels. We measured transcript levels of the WT strain and mutants at 18 - 30 h PS, the period leading up to and including the time that many cells commit to spore formation [35] and sonication - resistant spores begin to be observed (Fig. 2). In agreement with a prior study [16] , in the WT strain late gene transcript levels rose between 18 and 30 h PS (Fig. 3). As noted previously [20] , the fold - increase of the exo transcript level varied greatly between biological replicates (Fig. 3A), which we do not understand. The MXAN_3259 transcript level is likewise greatly variable (Fig. 3C). The nfs (Fig. 3B) and fadI (Fig. 3D) transcript levels varied less between biological replicates and on average both increased fourfold between 18 and 30 h, much less than the exo (24 - fold) and MXAN_3259 (70 - fold) transcript levels. Figure 3.3 Transcript levels in wild type, csgA mutant and fruA mutant. 124 Figure 3.3 ( c In a csgA null mutant unable to produce C - signal, the late gene transcript levels do not increase during development as in the WT strain (Fig. 3, asterisks indicate p - tailed t - tests comparing the mutant to the WT strain at each time point). Asterisks are absent above the exo and nfs transcript levels in the csgA mutant because the statistical test yielded p > 0.05, but th is is due to the large variation between biological replicates in the WT strain. In the csgA mutant, the exo and nfs transcript levels were low in all biological replicates at each time point. We conclude that late gene transcript levels are very low in t he absence of C - signaling. In agreement, reporter activity from fusions to exo [30] , nfs [23] , and fadI [24] was very low in csgA mutants relative to WT strains during development. The low transcript levels could be due to decreased synthesis and/or increased degradation in the csgA mutant compared with the WT strain. To measure the transcript degradation rate s , we added rifampicin to block transcription at 30 h PS and determined the transcript levels at intervals thereafter. The degradation rates of the late gene transcripts did not differ significantly between the csgA mutant and the WT strain (Fig. 4). These results suggest that decreased synthesis of late gene transcripts primarily accounts for the low transcript levels in the absence of C - signaling. Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs , (C) MXAN_3259 and ( D) fadI transcript levels by RT - qPCR. Graphs show the data points and average of at least three biological replicates, relative to wild - type strain at 18 h, and error bars indicate one standard deviation. Asterisks indicate a significant difference (p < 0.05 in - tailed t - tests) from the wild - type at the corresponding time poststarvation. 125 Figure 3.4 Transcript stability in wild type and csgA mutant . 126 Figure 3.4 ( c Considerable evidence supports a model in which C - signaling activates FruA posttranslationally in order to in crease transcription of genes during M. xanthus development [11] [1] . We showed previously that the FruA level is about two - fold lower in the csgA mutant than in the WT strain at 18 - 30 h PS [1] . Boosting the FruA level in the csgA mutant to the WT level us ing a vanillate - inducible promoter (P van ) fused to fruA did not increase the transcript levels of five genes or operons ( fmgA , fmgBC , fmgD , fmgE , dev ) [1] known to be under combinatorial control of FruA a nd MrpC [36] [37 - 39] . Neither did boosting the level FruA D59E (with a phosphomimetic substitution in its receiver domain) using a P van - fruA D59E fusion, increase the transcript levels [1] . To test whether boosting the level of native FruA or its D59E variant in the csgA mutant increases the late gene transcript leve ls at 18 - 30 h, we used the same approach. The late gene transcript levels remained low in all cases (Fig. S2) , suggesting that neither the two - fold lower level of FruA nor a lack of D59 phosphorylation causes late gene transcript levels to remain low in t he csgA mutant. Rather, we propose that C - signaling Wild - type DK1622 and a csgA mutant were subjected to starvation under submerged culture conditions for 30 h. The overlay was replaced with fresh starvation buffer containing rifampicin (50 mg/ml) and samples were collected immediately ( t 0 ) and at the times indicated ( t x ) for measure ment of the exo (A), nfs (B), MXAN_3259 (C) and fadI (D) transcript level by RT - qPCR. Transcript levels at t x were normalized to that at t 0 for each of three biological replicates and used to determine the transcript half - life for each replicate. The graph shows the average ln( t x / t 0 ) and one standard deviation for the three biological replicates of wild type (black dashed line) and the csgA mutant (gray dashed line). The average half - life and one standard deviation are reported in (E). Natural log of the t ranscript levels was plotted vs minutes of rifampicin treatment for each biological replicate and the slope of the linear fit of the graph were used to calculate the half - life of the transcripts. No significant difference in transcript stability between th e wild type and t - tests as indicated by the respective P values on the table ( P > 0.05). 127 regulates FruA by a mechanism other than phosphorylation of its receiver domain, allowing increased transcription of the late genes as well as the fmg and dev genes known to be under combinatorial control of FruA and MrpC. The exo, nfs, and MXAN_3259 transcript levels are elevated in a fruA mutant To determine whether C - signal - dependent activation of FruA might explain the failure of late gene transcript levels to rise in the csgA mutant between 18 and 30 h PS (Fig. 3), we measured transcript levels in a fruA null mutant. We expected the late gene transcript levels to remain low, as observed previously for the dev transcript level in the fruA mutant [1] . Instead, we observed that exo , nfs , and MXAN_3259 transcript levels were elevated in the fruA m utant relative to the WT strain at 18 h (Fig. 3). The exo and MXAN_3259 transcript levels were also elevated at 21 and 24 h. In contrast, the fadI transcript level, like the dev transcript level [1] , re mained low in the fruA mutant (Fig. 3D), consistent with a model in which C - signaling activates FruA and activated FruA increases fadI and dev transcription. Activated FruA may also increase nfs transcription at 27 and 30 h, since the nfs transcript level is lower in the fruA mutant than in the WT strain at those times (Fig. 3B). On average, the exo transcript level is lower in the fruA mutant than in the WT strain at 30 h, but this difference is due to one biological replicate of the WT strain with a much greater transcript level than the other three replicates (Fig. 3A), so the evidence that activated FruA increases exo transcription at 30 h is weak. The MXAN_3259 transcript level stays elevated in the fruA mutant through 30 h (Fig. 3C), providing no evidence that activated FruA is necessary to increase MXAN_3259 transcription. Comparison of the fruA and csgA mutants suggests that FruA which has not been activated by C - signaling negatively regulates late gene transcript levels. The avera ge levels of all four late 128 gene transcripts were elevated in the fruA mutant compared with the csgA mutant at 18 - 30 h (Fig. 3), with the exception of the fadI transcript level at 27 and 30 h (Fig. 3B). Since FruA is present but cannot be activated by C - si gnaling in the csgA mutant, unactivated FruA appears to be responsible for the observed negative regulation. As noted above, the FruA level is about two - fold lower in the csgA mutant than in the WT strain at 18 - 30 h [1] . In the WT strain, unactivated FruA may account for the lower exo (Fig. 3A) and MXAN_3259 (Fig. 3C) transcript levels at 18 - 24 h, and the lower nfs transcript level at 18 h (Fig. 3B), than in the fruA mutant. The fadI transcript level in the WT strain exceeded that in the fruA mu tant (Fig. 3D), suggesting that positive regulation by activated FruA overcomes negative regulation by unactivated FruA in this case. Taken together, the effects of mutations in fruA and csgA on late gene transcript levels suggest that regulation of these genes during the period leading up to and including commitment to spore formation depends at least in part on a C - signal - dependent swi tch from negative regulation by unactivated FruA to positive regulation by activated FruA. The exo, nfs, and MXAN_3259 transcript levels differ in mrpC and fruA mutants MrpC appears to directly activate transcription from the fruA promoter [40] . In agreement , Fru A was undetectable in an mrpC null mutant at 18 - 30 h PS [34] . Hence, the mrpC mutant lacks both MrpC and FruA. To compare the effects of losing both transcription factors with the effects of losing only FruA, we measured late gene transcript levels of the mrpC mutant in parallel with the fruA mutant and the WT strain at 18, 24, and 30 h. 129 Figure 3.5 Transcript levels in wild type, mrpC mutant and fruA mutant during M. xanthus development. Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs , (C) MXAN_3259 and ( D) fadI transcript levels by RT - qPCR. Graphs show the data points and the average of at least three biological replicates, relative to the wild - type strain at 18 h, and error bars indicate one standard deviation. Asterisks indicate a significant differenc e ( p - tailed t - tests) from the wild - type strain at the corresponding time poststarvation, or a significant difference between the mutants. 130 Strikingly, for each late gene, the pattern of effects on transcript levels differed (Fig . 5). The exo transcript level wa s on average elevated in the mrpC mutant compared with both the WT strain and the fruA mutant at all times (Fig. 5A, instances of p < 0.05 indicated by asterisks) . The nfs transcript level remained low in the mrpC mutant, unlike either the WT strain or the fruA mutant (Fig. 5B). The MXAN_3259 transcript level was elevated in the mrpC mutant relative to the WT strain at 18 and 24 h (Fig. 5C). Relative to the fruA mutant, th e MXAN_3259 transcript level was on average elevated in the mrpC mutant at 18 h, but lower in the mrpC mutant at 24 and 30 h. The fadI transcript level remained low in the mrpC mutant, unlike the WT strain, but similar to the fruA mutant (Fig. 5D). Among the late genes, only the fadI transcript level did not differ between the mrpC and fruA mutants. To determine whether the absence of MrpC and FruA affects the degradation rates of the late gene transcripts , we added rifampi cin to block transcriptio n at 18 h PS and determined the transcript levels at intervals thereafter. We chose 18 h for this analysis since the exo (Fig. 5A) and MXAN_3259 (Fig. 5C) transcript levels were elevated in the mrpC mutant relative to the WT strain at that time. The degra dation rates did not differ significantly between the mrpC mutant and the WT strain (Fig. S3). These results suggest that increased synthesis of the exo and MXAN_3259 transcripts primarily accounts for the elevated transcript levels in the absence of MrpC and FruA. Our results suggest that MrpC negatively regulates transcription of exo and MXAN_3259 independently of unactivated FruA at 18 h PS, because the average transcript levels were elevated in the mrpC mutant compared with the fruA mutant at that time (Fig. 5A and 5C). At later times, MrpC appears to continue to negatively regulate exo transcription independently of 131 FruA, b ut MrpC appears to positively regulate MXAN_3259 transcription. MrpC also appears to positively regulate transcription of nfs and fadI (Fig. 5B and 5D). The DNA - binding domain of FruA has been shown to bind in vitro to two sites in the exo upstream regi on, which appears to contain three promoters [41] . Deletion of a site spanning from - 89 to - 64 bp upstream of the apparent start site of transcription from P D1 reduced - galactosidase activity from a lacZ fusion abou t threefold at 20 h PS, suggesting that FruA binding to the site activates transcription. Binding of MrpC to the exo upstream region was not tested. Conversely, binding of FruA to the nfs upstream region has not been tested, but ChIP - seq analysis suggest ed that MrpC is bound at 18 h PS to a site located at - 137 bp relative to the predicted nfsA translation start codon [42] . Since our results suggest that both FruA and MrpC regulate transcription of both exo and nfs (Fig. 3 and 5), we tested the binding of purified pro teins to upstream DNA fragments using electrophoretic mobility shift assays (EMSAs). As a control, we performed EMSAs with a dev upstream DNA fragment (Fig. S9). As expected, the fragment was bound separately by His 6 - MrpC (lane 1) and FruA - His 6 (lane 3), a nd cooperatively by the two proteins, producing a complex that migrated more slowly and was more abundant (lane 6) [37] . The FruA DNA - binding domain (FruA - DBD - His 8 ) bound separately (lane 2), but there was little or no indication of cooperative binding with MrpC (lane 5), suggesting that the FruA N - terminal region is important for cooperative binding with MrpC to the dev promoter region , as observed previously for th e fmgA promoter region [36] . The exo upstream DNA fragment was bound by FruA - DBD - His 8 (lan e 8), as expected [41] , and by FruA - His 6 (lane 9), but there was no detectable binding by His 6 - MrpC separately (lane 7) or cooperatively (lanes 11 and 12). The nfs upstream DNA fragment was bound separately by His 6 - 132 M rpC very weakly (lane 13), in qualitative agreement with a suggestion based on ChIP - seq analysis [42] . The fragment was also bound separately by FruA - DBD - His 8 (lane 14) and FruA - His 6 (lane 15), but there was no indication of cooperative binding with MrpC (lanes 17 and 18). Altogether, the EMSA results suggest that FruA may directly regulate transcription of exo , but regulation of exo by MrpC is l ikely indirect, whereas both proteins may directly regulate nfs transcription, with MrpC acting positively and FruA acting negatively (see Discussion). FruA can positively regulate exo and MXAN_3259 transcript levels in the absence of MrpC To examine the effects of FruA on late gene transcript levels in the absence of MrpC, we used the P van - fruA fusion mentioned earlier to produce FruA in the mrpC mutant. The inducer (vanillate) was added during growth and at 0 h PS. By 6 h, the FruA level was about thre e - fold greater than in the WT strain, but FruA reached that level at 12 and 18 h in the WT strain, while the level did not change in the mrpC mutant containing the P van - fruA fusion (Fig. S4A). To our surprise, the exo transcript level was greatly elevated in the mrpC P van - fruA strain relative to the WT strain at 6 - 18 h (Fig. 6A). The result was unexpected since unactivated FruA appeared to negatively regulate the exo transcript level based on comparison of the fruA and csgA mutants at 18 - 30 h (Fig. 3A). However, MrpC appeared to negatively regulate exo transcription independently of unactivated FruA at 18 h (Fig. 5A and S3A), so relief of negative regulation by MrpC likely explains in part the elevated exo transcript level in the mrpC P van - fruA strain. In addition, positive regulation by FruA activated in response to C - signaling may also explain in part the elevated exo transcript level in the mrpC P van - fruA strain, since the level was elevated relative to the mrpC mutant at 6 h and on average at 12 and 18 h (Fig. 6A) (see Discussion). The MXAN_3259 transcript level was also elevated in the mrpC P van - fruA strain relative to the WT 133 strain at 6 - 18 h and relative to the mrpC mutant on average at 6 and 12 h (Fig. 6C), so similar explanations may apply (see Discussion). Figure 3.6 Transcript levels in wild type, mrpC (P van - fruA ) mutant, mrpC mutant and fruA mutant during early time points. 134 Figure 3.6 ( c Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs, (C) MXAN_3259 and ( D) fadI transcript levels by RT - qPCR. Graphs show the data points and the average of at least three biological replicates, relative to the wild - type strain at 6 h, and error bars indicate one standard. Asterisks indicate a significant difference ( p < 0.05 - tailed t - tests) from wild type at the corresponding time poststarvation, or a significant difference between the mutants. 135 The nfs (Fig. 6B) and fadI (Fig. 6D) transcript levels were not elevated in the mrpC P van - fruA strain compared to the WT strain at 18 h PS, suggesting that FruA was unable to overcome the apparent need for MrpC to positively regulate transcription of these genes (Fig. 5B, 5D, S3B, and S3D). The nfs transcript level was slightly elevated in the mrpC P van - fruA strain relative to the WT strain at 6 h and relative to the mrpC mutant at 6 - 18 h (Fig. 6B), perhaps due to positive regulation by FruA activated in response to C - signaling. Late gene transcript levels are low in the absence of Nla6 The Nla6 transcripti on factor appears to be a direct regulator of exo and MXAN_3259 . The Nla6 DNA - binding domain binds to the exo and MXAN_3259 promoter regions in vitro , and the transcript levels suggest positive regulation by Nla6 during the first 8 h PS, and negative regu lation by Nla6 at 24 h [25] . Since our results showed that FruA and MrpC impact late gene transcript levels at 18 h (Fig. 3 and 5) and in some cases earlier during development (Fig. 6), we examined the effects of null mutations in nla6 on t he FruA and MrpC protein levels and late gene transcript levels at 6, 12, and 18 h. We constructed a new nla6 mutant and compared it with one described previously [43] . The new mutant is tetracycline - resistant (Tc r ) and the one described previously is kanamycin - resistant (Km r ). Both mutants formed immature mounds by 12 h, but failed to progress to more mature mounds with distinct, round edges by 18 h (Fig. S5). Later during development, the Km r nla6 mutant mounds matured somewhat at 24 - 30 h, but failed to darken by 36 - 48 h (Fig. S6). The Tc r nla6 mutant mounds did not mature until 36 h and also failed to darken by 48 h. The two nla6 mutants were indistinguishable in terms of the molecular markers we tested. The FruA and MrpC protein levels were similar to the WT strain at 6 - 18 h PS (Fig. S4). The late 136 gene transcript levels remained low in both nla6 mutants (Fig. S7). These results suggest that Nla6 positively regulates the exo , nfs , and fadI transcript levels by 18 h in the WT strain (note that the MXAN_3259 transcript level h ad not increased by 18 h). Because the exo , nfs , and MXAN_3259 transcript levels were elevated in the mrpC and/or fruA mutants at 18 h (Fig. 3 and 5) and in some cases earlier during development (Fig. 6), we tried to construct mrpC nla6 and fruA nla6 doub le mutants, but our efforts were unsuccessful (see Discussion). Therefore, we were unable to determine whether positive regulation by Nla6 could account for the elevated exo and MXAN_3259 transcript levels observed in both the mrpC and fruA mutants, and th e elevated nfs transcript level in the fruA mutant (Fig. 3, 5, and 6). Late gene transcript levels are low in the absence of DevS Using reporter fusions, transcription of exo [30] and nfs [44] appeared to be very low in a devRS mutant compared with a WT strain during development. The exo transcript level was also very low in a devS null mutant compared with WT strain DK1622 at 30 h PS [20] . Here, we report late gene transcript levels at 18 - 30 h in devI and devS null mutants. We chose these mutants because sporulation occurs about 6 h earlier than normal in the devI mutant [20] and sporulation is severely impaired in the devS mutant [18] . DevI appears to delay sporulation of the WT strain, and overproduction of DevI in the absence of DevS (or DevR or DevT), due to loss of negative autoregulation of dev transcription, appears to strongly inhibit sporulation [19] [20] . We found that late gene transcript levels remain low in the devS mutant at 18 - 30 h PS (Fig. S8). These results suggest that in the absence of Dev S, overproduction of DevI inhibits expression of several late genes that are important for sporulation (Fig. S8 ). We acknowledge that comparing transcript levels in the devS mutant with the WT strain at each time point rarely 137 yielded p < 0.05 in a Student - tailed t - test (indicated by an asterisk in Fig. S8), but this is due to the large variation between biological replicates in the WT strain. We emphasize that transcript levels were low in all biological replicates of the devS mutant at all times. On average, late gene transcript levels were elevated in the devI mutant compared with the WT strain at most times from 18 - 30 h PS (Fig. S8), consistent with the notion that in the absence of DevI, late genes important for sporulation may be expressed earl ier than normal. However, our evidence is weak on this point owing to large variation between biological replicates of both the devI mutant and the WT strain, yielding p - tailed t - tests at most time points. Discussion Our systemati c cellular and molecular analysis of the function and regulation of late genes during M . xanthus development provides several new insights. First, we found that mutations in late genes do not prevent the initial cellular shape change associated with sporulation, but mutations in exoC and MXAN_3259 prevent formation of sonication - resistant spores and mature spores, while a mutation in nfsA - H reduces sonication - resistant spore formation about twofold and reduces mature spore formation about tenfold. Second, our analysis of late gene transcript levels in a csgA mutant and a derivative engineered to prod uce a phosphomimetic form of FruA is consistent with a model in which posttranslational regulation of FruA by C - signaling allows increased transcription of the late genes and may involve a mechanism other than phosphorylation of the FruA receiver domain. T hird, we discovered that FruA which has not yet been activated by C - signaling negatively regulates the transcript levels of three late genes ( exo, nfs , MXAN_3259 ) during the period leading up to spore formation . Fourth, our 138 results suggest that MrpC also n egatively regulates transcription of exo and MXAN_3259 prior to sporulation, independently of unactivated FruA, but the effects of MrpC differ for the two genes later during development, and MrpC appears to positively regulate nfs and fadI transcription bo th leading up to and including the period of spore formation. Fifth, purified FruA bound in vitro to exo and nfs upstream DNA fragments, consistent with direct regulation, whereas binding of MrpC was very weak or undetectable, suggesting indirect regulatio n. Sixth, although production of FruA normally requires MrpC, ectopic production of FruA in an mrpC mutant prematurely elevated the exo and MXAN_3259 transcript levels, but the levels of nfs and fadI transcripts remained low, further supporting differentia l regulation of late gene transcription by FruA and MrpC. Seventh, our results also implicate Nla6 as a positive regulator and DevI as a negative regulator of late gene transcript levels. We incorporate these new insights into a model of the regulatory n etwork governing mound and spore formation (Fig. 7). We propose that multiple regulators act in concert to differentially control late genes and thus ensure proper formation of mature spores. New insights into late gene function Our results show that mutations in exoC , nfsA - H , and MXAN_3259 do not prevent the initial cellular shape change associated with sporulation (Fig. S1B) but do impact the formation of sonication - resistant spores beginning at 27 h PS (Fig. 2). Previously, mut ants were examined for starvation - induced spore formation at 120 h [23, 25] . Our findings indicate a much earlier role of Exo, Nfs, and MXAN_3259 proteins during starvation - induced sporulation than established previously. Glutaraldehyde fixation of cells followed by brief sonication allowed us to visualize and enumerate cells that were not rod - shaped in samples of the exoC an d MXAN_3259 139 mutants. These cells appear to be changing shape from rods to spores (Fig. S1B) and likely resemble cells of exoC and nfsA - H mutants that fail to complete morphogenesis upon chemical induction of sporulation in liquid culture [22] . Under such conditions, neither glutaraldehyde nor sonication are necessary to observe individual cells. In contrast, starvation - induced submerged culture results in mounds of developing cells (Fig. 1) that are difficult to disperse. For the WT st rain and the nfsA - H and fadI mutants, we estimated the number of cells in transition from rods to spores (Fig. S1B) by subtracting the number of sonication - resistant spores in a sample taken at the same time, but not fixed with glutaraldehyde, and sonicate d longer (Fig. 2). The WT strain may exhibit a higher percentage of cells changing shape than the fadI mutant (Fig. S1B) because the mutant makes more sonication - resistant spores (Fig. 2). The fadI insertion mutant presumably has a reduced rate of fatty acid - oxidation, as appeared to be the case for a fadIJ ( MXAN5372 - MXAN5371 ) deletion mutant [24] , so perhaps altered metabolism enhanced formation of sonication - resistant spores by the fadI mutant at 27 - 48 h (Fig. 2), albeit not mature spores at 72 h (Table S1). The nfsA - H mutant made about two - fold less sonication - resistant spores than the WT strain, and the exoC and MXAN_3259 mutants made less than the detection limit (Fig. 2). For these three mutants, the lower percentag e of cells changing shape as compared with the WT strain at 24 and 27 h (Fig. S1B) may reflect reduced ability to initiate and/or maintain the cellular shape change associated with sporulation. Upon chemical induction of sporulation, exo and nfs mutants ap peared to initiate the transition from rods to spores, but then revert into rods [22] . Given the evidence that Exo and Nfs proteins function in spore coat polysaccharide export [21, 140 22] and deposition [21, 22] , respectively, our results suggest that defective spore coat biogenesis of the exoC and nfsA - H mutants reduces their ability to maintain cellular shape change as early as 24 h (Fig. S1B) and blocks or reduces their ability to form sonication - resistant spo res by 27 h (Fig. 2) and mature spores by 72 h (Table S1). The MXAN_3259 mutant was indistinguishable from the exoC mutant in our cellular assays. As noted previously, MXAN_3259 (renamed MXAN_RS15785 in NCBI) is predicted to be a polysaccharide deacetyla se [ 25] . The downstream genes of the predicted MXAN_3259 - MXAN_3263 operon are predicted to code for an oligosaccharide flippase (MXAN_RS15790), a serine acetyltransferase (MXAN_RS15795), and glycosyltransferases (MXAN_RS15800 and MXAN_RS15805). Since all th ese proteins may function in polysaccharide export and modification, defective spore coat biogenesis of the MXAN_3259 mutant likely explains its behavior in our cellular assays. MrpC and unactivated FruA negatively regulate certain late genes Our syst ematic analysis of late gene transcript levels revealed differential regulation by MrpC and by FruA in response to activation by C - signaling (Fig. 7). The fadI transcript levels in csgA , fruA , and mrpC mutants (Fig. 3 and 5) were consistent with a model in which C - signaling activates FruA posttranslationally in order to increase transcription cooperatively with MrpC. This model is based on analysis of dev [1, 11, 37] and fmg [1, 36, 38, 39, 45] genes. An early study of - galactosidase activity from lacZ fusions indicated positive regulation of many genes after about 6 h PS by C - signaling [46] . Subsequent analysis of dev [11, 47] and fmg [38, 39, 48, 49] fusions in fruA mutants, and similarity of FruA to response regulators [50] , suggested that phosphorylation of FruA in response to C - signaling might increase transcription during development. However, several observations suggest that FruA may not be phosphorylated 141 [36] , but is activated by C - signaling via a different posttranslational mechanism [1] . Although the mechanism of FruA activation remains a mystery, our results suggest that fadIJ transcription may be increased by cooperative binding of MrpC and activated FruA, like transcription of dev and fmg genes appears to be. This model is also in agr eement with measurements of fluorescence intensity from a fadI - tdTomato fusion in a WT strain and csgA , fruA , and mrpC mutants during development [24] . On the other hand, ChIP - seq analysis did not detect MrpC binding upstream of fadI [42] , so positive regulation of fadIJ transcription by MrpC could be indirect, perhaps relying on activated FruA. Differential regulation by MrpC and by FruA in response to C - signaling, of the late genes implicated in spore coat biogenesis ( exoA - I , nfsA - H , MXAN_3259 - MXAN_3263 ) (Fig. 7), is more complex than described above for fadIJ , whose products appeared to have little impact on sporulation in our assays. Strikingly, comparison of transcript levels in the WT strain with those in the csgA , fruA , and mrpC mutants (Fig. 3, 5, and 6) and comparison of transcript degradation rates (Fig. 4 and S3) suggests that FruA which has not been activated by C - signaling negatively regulates exo, nfs , and MXAN_3259 transcription at 18 h PS and that MrpC negatively regulates exo and MXAN_3259 transcription independently of unactivated FruA at 18 h. Negative regulation of nfs by FruA was observed previously by comparing fluorescence intensity from a P nfsA - mCherry fusion in a WT strain versus a fruA mutant during development [23] . Negative regulation by unactivated FruA appears to reduce transcription of late genes important for spore coat biogenesis before their products are needed (Fig. 7). Under our cond itions of submerged culture development, the WT strain forms compact mounds by 18 h [35] ( Fig. 1), but cells are not yet changing shape (Fig. S1B). Transcript levels from exo, nfs , and 142 MXAN_3259 rise by 24 h (Fig. 3 and 5), coincident with the beginning of detectable cellular shape change (Fig. S1B). Relative to the transcript level at 6 h, the exo and nfs levels, but not the MXAN_3259 level, rose at 18 h (Fig. 6 and Table S2), so temporal regulation of these genes differs slightly. Using a fluorescent membrane stain and confocal laser scanning microscopy, some cells in mounds of the WT strain begin to change shape by 21 h, but not by 1 8 h (Y Hoang and Lee Kroos, unpublished data). We did not detect cells changing shape at 21 h (Fig. S1B). In any case, exo and nfs transcript levels rose by 18 h, and the MXAN_3259 level rose later, between 18 and 30 h (Fig. 3 and 5, and Table S2), close to the time cellular shape change begins. Negative regulation by MrpC independently of unactivated FruA appears to further differentiate transcription of late genes that play a role in spore coat biogenesis (Fig. 7). Negative regulation by MrpC was stron gest for exo . The exo transcript level wa s on average elevated in the mrpC mutant compared with the fruA mutant at all times and the differences yielded p < 0.05 in a test of statistical significance at 6, 18, and 30 h (Fig. 5A and 6A). This comparison s uggests that MrpC negatively regulates the exo transcript level during the entire period leading up to and including the time that many cells commit to spore formation [35] . Negative regulation by MrpC independently of unactivated FruA was weaker for MXAN_3259 than for exo . The MXAN_3259 transcript level was on average elevated in the mrpC mut ant relative to the fruA mutant only at 12 and 18 h (Fig. 5C and 6C). Although the differences did not yielded p < 0.05 in a test of statistical significance, the average was elevated about threefold in the mrpC mutant relative to the fruA mutant at 18 h in both experiments. Interestingly, MrpC appeared to regulate MXAN_3259 transcription positively at 24 and 30 h 143 (Fig. 5C), in clear contrast to the persistent negative regulation of exo transcription by MrpC (Fig. 5A and 7). MrpC did not appear to regula te nfs transcription negatively at any time tested, but rather MrpC appeared to regulate nfs transcription positively at 18 - 30 h (Fig. 5B, 6B, and 7). Potential mechanisms of differential regulation of spore coat biogenesis genes by MrpC , FruA The negative regulation of exo and MXAN_3259 transcript levels by MrpC, and the independent negative regulation of these transcript levels and the nfs transcript level by unactivated FruA (Fig. 7), raised the question whether increased transcript levels in the WT strain could be explained solely by relief from negative regulation. In the case of nfs , this scenario would not explain the low transcript level in the mrpC mutant (which lacks MrpC and FruA) compared with the WT strain at 24 and 30 h (Fig. 5B). Nor would the scenario explain the low nfs transcript level in the fruA mutant relative to the WT strain at 27 and 30 h (Fig. 3B and 5B). Rather, both MrpC and activated FruA appear to positively regulate nfs transcription (Fig. 7). The effect of MrpC coul d be indirect since binding in vitro of purified MrpC was very weak and there was no indication of cooperative binding with FruA (Fig. S9). On the other hand, ChIP - seq analysis suggested that MrpC binds to the nfs upstream region in vivo [42] , so a direct effect of MrpC is possible. A direct effect of FruA is supported by binding to an nfs upstream DNA fragment in vitro (Fig. S9). Positive regulation by activated FruA in the absence of MrpC in vivo is supported by the slightly elevated nfs transcript level in the mrpC P van - fruA strain relative to the WT strain at 6 h and relative to the mrpC mutant at 6 - 18 h (Fig. 6B). The mrpC P van - fruA strain produces FruA ectopically at a three - fold elevated level relative to the WT strain by 6 h (Fig. S4A). Relative to the mrpC mutant, the mrpC P van - fruA strain forms nascent mounds earlier (Fig. S5), suggestive of ongoing C - signaling, although the mo unds fail to become compact by 18 h or 144 later, as they do in the WT strain (Fig. S5 and S6), so short - range C - signaling is likely impaired. As a result, the level of activated FruA in the mrpC P van - fruA strain may only be sufficient to elevate nfs transcri ption slightly (Fig. 6B). Relief from negative regulation by MrpC and unactivated FruA could explain the increasing exo and MXAN_3259 transcript levels in the WT strain at 24 and 30 h PS, since the levels of these transcripts in the mrpC and fruA mutants a re comparable or elevated relative to the WT strain at 18 - 30 h (Fig. 3A, 3C, 5A, and 5C). The only exceptions were that on average the exo transcript level was lower in the fruA mutant than in the WT strain at 30 h, which we initially considered weak evide nce that activated FruA increases exo transcription (since one biological replicate of the WT strain had a much greater transcript level than the other three replicates) (Fig. 3A), but the evidence was strengthened by a second experiment (Fig. 5A). Moreove r, the exo transcript level was elevated in the mrpC P van - fruA strain relative to both the WT strain and the mrpC mutant at 6 - 18 h (Fig. 6A), suggesting that activated FruA can greatly increase exo transcription in the absence MrpC. We propose that in the WT strain negative regulation by MrpC is partly relieved and negative regulation by unactivated FruA is switched to positive regulation by activated FruA, increasing exo transcription at 18 - 30 h (Fig. 7). The effect of MrpC is likely indirect, since neithe r binding to an exo upstream DNA fragment in vitro (Fig. S9) nor binding to the exo upstream region in vivo [42] was detected. On the other hand, FruA likely exerts its effects directly, since binding was observed in vitro [41] (Fig. S9). Presumably, recombinant FruA purified from E. coli represents unactivated FruA in M. xanthus. How C - signaling activates FruA is unknown. The answer is key to und erstanding the switch from 145 negative regulation by unactivated FruA to positive regulation by activated FruA, which our results suggest occurs for both exo and nfs transcription (Fig. 7). Although relief from negative regulation by MrpC and unactivated FruA could explain the increasing MXAN_3259 transcript level in the WT strain at 24 and 30 h PS for the reasons mentioned above, our data suggest that both MrpC and FruA switch from negative to positive regulation of MXAN_3259 transcription (Fig. 7). The MXAN _3259 transcript level is elevated in the fruA mutant relative to the mrpC mutant at 30 h (Fig. 5C), suggesting positive regulation by MrpC. Positive regulation by activated FruA is supported by the elevated MXAN_3259 transcript level in the mrpC P van - fruA strain relative to both the WT strain at 6 - 18 h and the mrpC mutant on ave rage at 6 and 12 h (Fig. 6C). Therefore, we propose that in the WT strain negative regulation by MrpC and unactivated FruA during mound formation is relieved, and switches to positive regulation by MrpC and activated FruA during spore formation (Fig. 7). T he effects of MrpC may be indirect since binding to the MXAN_3259 upstream region was not observed in ChIP - seq analysis at 18 h [42] . Binding of FruA remains to be test ed and the transcriptional start site remains to be identified. Roles of Nla6 and Dev proteins in late gene regulation Our results implicate Nla6 as a positive regulator of late gene transcript levels during mound formation (Fig. 7). Late gene transcript levels remained very low in nla6 mutants (Fig. S7). Since the DNA - binding domain of Nla6 binds to the exo and MXAN_3259 promoter regions in vitro [25] , Nla6 may directly activate transcription of these genes. One of the two Nla6 binding sites in the exo promoter region partially overlaps with the FruA - binding site [41] , so negative regulation by unactivated FruA could involve competition for binding with Nla6. Later, during 146 spore formation, positive regulation by activated FruA could depend on a diminished Nla6 level. This potential mechanism would explain the elevated exo transcript level observed in an nla6 mutant compared w ith the WT strain at 24 h PS [25] , since Nla6 would be absent from the nla6 mutant but perhaps only diminished in the WT strain. The nfs and fadI upstream regions were not identified as potential targets of Nla6 binding using bioinformatics [25] , so perhaps positive regulation by Nla6 is indirect for nfsA - H and fadIJ. Since we discovered that MrpC and unactivated FruA negatively regulate certain late genes during mound formation (Fig. 3, 5, 6, and 7), we predicted that transcript levels of those late genes wo uld remain low in mrpC nla6 and fruA nla6 double mutants. However, we were unable to construct the double mutants in order to test our prediction. Unexpectedly, a null mutation in nla6 appeared to create a synthetic lethal phenotype in combination with a n ull mutation in mrpC or fruA , suggesting that Nla6 functions redundantly with MrpC and FruA to express genes required for growth. This outcome was unexpected since none of the three transcription factors have been reported to function during growth, althou gh MrpC has been shown to be present [15] . Our results implicate DevI as a negative regulator of late gene transcript levels during spore formation (Fig. 7). In the absence of DevI, sporulation occurs about 6 h earlier [20] and late gene transcript levels are slightly elevated relative to the WT strain at most times from 18 - 30 h PS (Fig. S8). In the absence of DevS (or DevR or DevT), DevI is overproduced and strongly inhibits sporulation [19, 20] and late gene transcription [20, 23, 30] (Fig. S8). Given the similarities mentioned above between fadI and fmg genes with respect to positive regulation by MrpC and activated FruA, and given that the fadI transcript level was slightly elev ated in the devI mutant and low in the devS mutant compared with the WT strain (Fig. S8), it would be 147 interesting to determine whether mutations in devI and devS have similar effects on fmg transcript levels as on fadI and the other late genes. Perhaps Dev I overproduction broadly inhibits transcription of genes positively regulated by MrpC and activated FruA, and among those genes are one or more required for cellular shape change, since DevI overproduction greatly delays and reduces the shape change associ ated with sporulation [19, 20] . Figure 3.7 Model of differential late gene regulation. Starvation increases the MrpC level which in turn increases the FruA level. C - signal activates FruA to FruA*. Positive regulation (yellow arrows) and negative regulation (blue line with blunt) of late genes (gray boxes) is indicated. During mound formati on between 6 and 18 h poststarvation, Nla6 positively regulates transcription of all four late genes (dashed box), but unactivated FruA and MrpC negatively regulate certain late genes as indicated. During spore formation between 24 and 36 h, activated FruA * induces transcription of the dev operon gene. DevS (and DevT and DevR, which are not shown) negatively autoregulates transcription of devI . DevI negatively regulates all four late genes (dashed box), but FruA* positively regulates their transcription, wh ile MrpC positively or negatively regulates certain late genes as indicated. 148 Differential regulation of late genes Nla6 and Dev proteins regulated all four late genes similarly, whereas MrpC and FruA mediated differential regulation (Fig. 7). C - signaling appears to switch FruA from negative regulation of certain late genes during mound formation to positive regulation of all the late genes during sporulation. FruA and MrpC did not bind cooperatively to exo or nfs upstream DNA fragments, as observed for dev (Fig. S9) [37] and fmg [36, 38, 39, 45] genes. Systematic experimental and computational analyses of dev transcript levels support a model in which C - signaling activates FruA at least ninefold by 18 h PS [1] , and different arrangements and affinities of cooperative binding sites for activated FruA and MrpC have been proposed to explain differential dependence on C - signaling and timing of transcription of dev and individual fmg genes [1, 36 - 39, 45] . Cooperative binding of activated FruA and MrpC may likewise explain the C - signal - dependence (Fig. 3) and timing of fadI transcription (Table S2). The late genes implicated in spore coat biogenesis appear to be regulated uniqu ely. Most of the evidence so far points to indirect regulation by MrpC (Fig. S9), which can be positive ( nfs ), negative ( exo ), or switching from negative to positive during mound formation and sporulation, respectively ( MXAN_3259 ) (Fig. 5, 6, and 7). Regul ation by FruA appears to be direct ( nfs , exo ) (Fig. S9), although binding remains to be tested in the MXAN_3259 upstream region. Unactivated FruA acts negatively during mound formation and activated FruA acts positively during sporulation (Fig. 3, 5, 6, an d 7). Different arrangements and affinities of binding sites for the two forms of FruA, and for Nla6 acting positively during mound formation (Fig. S7 and 7) and negatively during spore formation [25] (not shown in Fig. 7 since we only measured late 149 gene t ranscript levels at 6 - 18 h), may account for differential transcription of the late genes implicated in spore coat biogenesis (Table S2). In summary, multiple signals and transcription factors appear to act in concert to differentially control late genes . This strategy presumably prevents starving cells from wasting resources during mound formation and finely - tunes expression of genes involved in metabolism, spore coat biogenesis, and other functions during spore formation. Multiple transcription factors likewise positively and negatively fine - tune the expression of hundreds of genes during Bacillus subtilis endospore formation [51] [52] [5 3] , ensuring that the resulting spores are endowed with resistance and surface properties tailored for their environment [54 - 56] . MrpC and/or FruA likely regulate hundreds of genes during M. xanthus development, including genes involved in protein phosphorylation and fate, transcription, signal production, and motility, as well as other proteins important for spor e formation inch addition to those studied here [42] . Identifying the key genes for mound formation and the cellular shape change associated with sporulation, and eluc idating the molecular mechanisms of regulation by MrpC and FruA for those genes, including the mechanism by which C - signaling activates FruA, are important goals for the future. Materials and methods Bacterial strains, plasmids and primers The strains, plasmids and primers used in this study are listed in T able S 3. E . coli strain D H w as used for cloning . M . xanthus strains with ectopically integrated P van - fruA and P van - fruA D59E were constructed by electroporation [57] followed by selection of transformants on CTT agar with 15 g/m L tetracycline [58] . To construct pSS11, primer pair Nla6 Fwd and Nla6 Rev 150 w as used to generate a PCR produc t using chromosomal DNA from M . xanthus strain DK1622 as a template . The product w as combined with DNA amplified from p MR3487 using PMR3487G Fwd and PMR3487G Rev primers , and a Gibson assembly reaction was used to enzymatically join the overlapping DNA f ragments [59] . The cloned DNA sequence was verified using primers 3487 seq Fwd1, 3487 seq Fwd2, 3487 seq Fwd3, 3487 seq Fwd4, and 3487 seq Fwd5 . M . xanthus strain MSS10 was created by electroporating pSS11 into strain DK1622 . The transformants w ere selected on CTT agar with tetracycline (15 µg/m L ) followed by verification by colony PCR using PMR3487 Rev , Nla6 Fwd4 , and Nla6 Fwd5 primers . To create pSS13 and pSS14 , 315 bp and 373 bp DNA fragments were amplified from M . xanthus strain DK1622 genomic DNA using primer pairs Exo - 267G and Exo +108G , and Nfs - 290G and Nfs +83G , respectively . The products were combined with DNA amplified from p MR3487 using PMR3487G Fwd and PMR3487G Rev primers , and joined using a Gibson assembly reaction [59] . The cloned DNA sequences w ere verified using the same primers as for pSS11. Growth and development E . coli strains containing plasmids were grown at 37 ° C in Luria Burt a ni broth supplemented with 15 µg/m L of tetracycline or 50 µg/m L of kanamycin sulfate as needed . Strains of M . xanthus were grown at 32 ° C in CTTYE liquid medium (1% Casitone, 0.2% yeast extract, 10 mM Tris - HCl [pH 8.0], 1 mM KH 2 PO 4 - K 2 HPO 4 , 8 mM MgSO 4 [final pH 7.6]) with shaking at 350 rpm . CTT agar (CTTYE without yeast extract and solidified with 1.5% agar) was used for growth on solid medium and was supplemented with 40 µg/m L of kanamycin sulfate or 15 µg/m L of tetracycline as required . Fruiting body development under submerged culture conditions was performed using MC7 (10 mM morpholinepropanesulfonic acid [MOPS; pH 7.0], 1 mM CaCl 2 ) as 151 the starvation buffer as described previously [5] . Briefly, at mid - exponential growth, cells were collected by centrifug ation CTTYE medi um was removed. The c ell pellet w as resuspended in MC7 buffer at a density of approximately 1,000 Klett units and fruiting body development was initiated in submerged culture . Upon incubation at 32°C, cells adhere to the bottom of the plate and undergo development . A 9 6 - L sample (designated T 0 ) was removed and was stored at 4°C for at least 24 h with glutaraldehyde (2% final concentration) to fix cells, followed by quantification of cells by counting microscopically as described previously [1] . For each developmental sample, 1.5 m L of the cell sus pension plus 10.5 m L of MC7 buffer was added to an 8.5 - cm - diameter plastic petri plate . At the indicated times , developing populations were photographed using a Leica Wild M8 microscope equipped with an Olympus E - 620 digital camera. Sample collection At the indicated times PS, the MC7 buffer overlay was replaced with 5 m L of fresh MC7 buffer with or without inhibitors as required . Developing cells were scraped from the bottom of the plates , the entire contents were collected in a 15 - m L centrifuge tube , and s amples were mixed thoroughly as described previously [1] . For quantification of rods and cells changing shape , 96 µ L of the mixture was removed and 4 µ L of glutaraldehyde was added from a 50% stock solution to achieve a 2% final concentration in order to fix the developing cells . T he sample was stored at 4°C for at least 24 h before counting as described below . For measurement of sonication - resistant sp ores, 400 µ L of the mixture was removed and stored at - 20°C . Immediately after collecting the two samples just described, the remaining 4.4 m L of the developing population was mixed with 0.5 m L of RNase stop solution (5% phenol [pH < 7] in 152 ethanol), follo wed by rapid cooling in liquid nitrogen until almost frozen, centrifugation at 8,700 × g for 10 min at 4°C, removal of the supernatant, freezing of the cell pellet in liquid nitrogen, and storage at - 80°C until RNA extraction . Quantification of total cell s, sonication - resistant spores, and cells changing shape During starvation - induced development a small percentage of the rod - shaped cells convert to round spores that become sonication - resistant . The number of sonication - resistant spores in developmental samples was quantified as described previously [5] . E ach 400 - µ L sample was also used for determination of total protein concentration as described earlier [1] . The total number of cells, including rod - shaped cells and round spores, as well as cells that appeared to be in transition between the two, was determined using the glutaraldehyde - fixed samples collected as describ ed above . Each sample was thawed and mixed by vortexing and pipetting , dilut ed with MC7 buffer , sonicat ed once for 10 s, and then all cells were counted microscopically as described previously [1] , except taking note of the number of cells that were not rod - shaped (i.e. cells changing shape plus round spores). The total cell number minus the number of sonication - resistant spores was designated the number of sonication - sensitive cells (consisting primarily of rod - shaped cells) and was expressed as a percentage of the total cell number in the corresponding T 0 sample (consisting only of rod - shaped cells) . The number of cells that were not rod - shaped minus the number of sonication - resistant spores was designated the number of cells changing shape and was also expressed as a percentage of the total cell number in the corresponding T 0 sample . 153 RNA extraction and analysis Total RNA was extracted using th e hot - phenol method followed by digestion with DNase I (Roche) as described previously [60] . T otal RNA (1 µg) was subjected to cDNA synthesis using Superscript III reverse transcriptase (Invitrogen) and random primers (Promega), as instructed by the manufacturers . In parallel , total RNA (1 µg) was subj ected to cDNA synthesis reaction conditions without Superscript III reverse transcriptase , as a control . One µ l of cDNA at the appropriate dilution (as determined empirically) and 20 pmol of each primer were subjected to qPCR in a 25 µ l reaction using 2× reaction buffer as described previously [1] . qPCR was done in quadruplicate for each cDNA using a LightCycler® 480 System (Roche) . In parallel , a standard curve was generated for each pair of qPCR primers using the gen omic DNA of M . xanthus WT strain DK1622 and gene expression was quantified using the relative standard curve method (user bulletin 2; Applied Biosystems) . 16S rRNA was used as the internal standard for each sample . Transcript levels fo r the WT strain at each time except 18 h PS, and for mutants at each time, were normalized to the transcript level observed for one replicate of the WT strain at 18 h in the same experiment, as describe previously [1] . For the WT strain at 18 h, the transcript le vels of at least three biological replicates from different experiments were normalized to their average, which was set as 1 [1] . RNA stability after addition of rifampicin ( 50 µ g/m L) to inhibit transcription was also determined as described previously [1] . Preparation of FruA - DBD - His 8 , FruA - His 6 and His 6 - MrpC E. coli strain BL21(DE3) (Novagen) was freshly transformed individually with plasmids pET28a/H 6 - MrpC [4, 61] , pET11a/FruA - DBD - H 8 [49] and pET11a/FruA - H 6 [13] . For each transformation an isolated k anamy cin - resistant colony was used to inoculate 10 m L of Luria - 154 Bertani broth supplemented with kanamycin followed by overnight incubation at 37 ° C with shaking . The cultures ( 5 m L) w ere used to inoculate 500 m L of the same medium , followed by continued incubation at 37 o C with shaking until the cultures reached 60 - 80 K lett units . IPTG (1 mM final concentration) was added to induce synthesis of the recombinant proteins . After 2 h, cells were harvested as reported previously [49] and stored at - 8 0° C unt il further purification . Each c ell pellet was resuspended in 35 ml of lysis buffer (50 mM Na - phosphate buffer [pH 8.0], - mercaptoethanol) supplemented with protease inhibitor cocktail (Roche Mini EDTA - free table t s) and sonicated 4 time s for 1 min to disrupt the cells with intermittent cooling on ice . After centrifugation at 18,000 × g for 10 min at 4 ° C , the supernatant was mixed with lysis buffer supplemented with 10% w/v Triton X - 100 to make the volume upto 50 ml. Lysis buffer supplemented with 20 mM imidazole (pH 8 .0 ) was used to wash 3 times and finally resuspend Ni - NTA beads (Qiagen) followed by addition of 1/100 volume to the supernatant for binding on a rotator for 1 h at 4 ° C . The unbound fraction was collected by centrifu gation at 700 × g for 3 min at 4 ° C . The Ni - NTA beads were washed 4 times with 50 m L of wash buffer A (50 mM Na - phosphate buffer [pH 8.0], 500 mM NaCl, 5 - mercaptoethanol, 20 mM imidazole [pH 8.0], 20% v / v glycerol) . Proteins were eluted from the beads with 10 m L of elution buffer (50 mM Na - phosphate buffer [pH 8.0], 500 mM NaCl, 2 mM - mercaptoethanol, 250 mM imidazole [pH 8.0], 20% v / v glycerol) supplemented with protease inhibitor cocktail (Roche Mini EDTA - free table t s) on a rotator for 30 min at 4 ° C . Eluates were dialyzed overnight at 4 ° C against a buffer containing 10 mM Tris - HCL [pH 8.0], 100 mM NaCl, 1 - mercaptoethanol and 10% w/ v glycerol . The c oncentration of each protein preparation was determined using the Bradf ord method [62] . 155 EMSA 32 P - labelled DNA fragments were generated from the dev , exo and nfs promoter regions using primers labelled with [ - 32 P ] ATP using T4 polynucleotide kinase as mentioned previously [17] . A DNA fragment from the dev promoter region from bp - 19 to - 114 was generated by PCR using primers LK1298 and LK1331 and plasmid pPV391 as template [17] . Plasmids pSS13 and pSS14 were used as templates to generate DNA fragmen t s spanning the exo and nfs promoter region s, respectively, from bp +1 to - 120 and from bp +1 to - 20 1 by using primer pairs Exo +1 Rev and Exo - 120 Fwd , and Nfs +1 Rev and Nfs - 20 1 Fwd . The labeled DNA fragments were purified by electrophoresis on 15% poly acrylamide gel s followed by visualization using autoradiography , excision and overnight elution by soaking in TE buffer as described previously [49] . Binding reactions (10 L) were performed as reported previously [49] , except the reaction mixtures were incubated for 10 min at room temperature prior to loading on 8% polyacrylamide gels . Gels were dried and exposed to X - ray film for autoradiography as described earlier [49] . 156 APPENDIX 157 Table S 3. 1 Cell and spore numbers counted in chapter 3 . Strain Sonication sensitive cells at T 0 (10 7 / ml) Sonication resistant spores at 48 h PS (10 7 / ml) Mature spores at 72 h PS (10 6 / ml) Wild type 140 2.3 2 0.1 exoC 140 < 0.05 0 N fs A - H 142 1.42 0.3 0.1 MXAN_3259 160 < 0.05 0 MXAN_5372 142 4.27 2 3 mrpC ( P van - fruA ) 120 < 0.05 0 Km r nla6 120 < 0.05 0 Tc r nla6 110 < 0.05 0 Wild - type DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions. Rod - shaped sonication - sensitive cells at T 0 and sonication - resistant spores at 48 h PS were counted microscopically using a Neubauer chamber. Mature spores at 72 h PS were quantified by subjecting samples to heat - and sonication - treatments followed by plating on nutrient agar medium and counting o f colonies after 5 days. Values indicate average of at least 3 biological replicates and one standard deviation. 158 Table S 3. 2 Changes in transcript levels during development . a The ratio of average transcript levels at 18 and 6 h PS from Fig. 6. b The ratio of average transcript levels at 30 and 18 h PS from Fig. 3 .3 , 3. 5, and S 3. 2 is reported as the average and standard deviation of the three experiments. Transcript level Ratio 18/6 h PS a Ratio 30/18 h PS b exo 73 14 ± 9 nfs 10 3.4 ± 0.6 MXAN_3259 0.85 53 ± 21 fadI 2.7 3.0 ± 0.8 159 Table S3 .3 P lasmids, strains and primers used in chapter 3 . Plasmids Description Source pSS10 Tc r ; pMR3691 with fruA inserted at MCS_G [34] pSS9 Tc r ; pMR3691 with fruA (D59E) inserted at MCS_G [34] pMR3691 Tc r ; M. xanthus MXAN_0018 - MXAN_0019 - P R3 - 4 :: vanR - P van - MCS_G [58] pMR3487 Tc r ; M. xanthus 1.38 - kb - P IPTG - MCS_A - PR4:: lacI [58] pSS11 Tc r ; ColE1 amplified from pMR3487 using PMR3487G Fwd and PMR3487G Rev combined with 600 bp nla6 fragment starting from +81 till + 699 of Nla6 ORF. [34] pSS13 Tc r ; ColE1 amplified from pMR3487 using PMR3487G Fwd and PMR3487G Rev combined with 315 bp fragment starting from the upstream region of exo promoter (bp - 207 to bp +108). This study pSS14 Tc r ; ColE1 amplified from pMR3487 using PMR3487G Fwd and PMR3487G Rev combined with 373 bp fragment starting from the upstream region of nfs pr omoter (bp - 290bp to bp +83) This study pPV391 pCR 2.1 TOPO with dev DNA spanning bp - 321 to +71 generated by PCR [17] Strains Description Source LS3950 DK1622:: Mxan_5372 (Km 40 ) (note that MXAN_5372 is referred to as fadI herein, although M. xanthus has a fadI paralog that is not up - regulated during development. [64] AG1152 DK1622:: Mxan_3259 (Km 40 ) [65] DK10524 DK1622:: Tn 5 lac exoC (Km 40 ) [26] PH1200 DK1622:: ( nfsA - H ) [23] AG306 DK1622:: nla6 (Km 40 ) [66] MSS2 mrpC ::pRR028 (Km r ) MXAN_0018 - MXAN_0019::pSS10 (Tc r ) This study MSS3 csgA ::pRR028 (Km r ) MXAN_0018 - MXAN_0019:: pSS10 (Tc r ) [34] MSS5 csgA ::pRR028 (Km r ) MXAN_0018 - MXAN_0019::pSS9 (Tc r ) [34] MSS7 fruA ::Tn 5 lac r ) MXAN_0018 - MXAN_0019::pSS9 (Tc r ) [34] MSS10 DK1622::pSS11 This study DK1622 Laboratory strain [67] 160 Table 3.3 ( c Strains Description Source DK5208 csgA ::Tn 5 - 132 r ) [68] SW2808 mrpC [69] DK5285 fruA ::Tn 5 lac r ) [70] DK11209 devS [71] MRR7 devI [72] DK5285 fruA ::Tn 5 lac r ) [70] Primers Description Source Nla6 Fwd ATTGATTCCATTTTTACACTGATGAGGTACCGAATTCTGACACAAGG TCGAGAT CGCATT This study Nla6 Rev TCTCCTTACGCATCTGTGCGGTATTCTCGAGCCCGGGTCACATCTCG AACACG CCGGG This study PMR3487 G Fwd AATACCGCACAGATGCGTAA This study PMR3487 G Rev TCATCAGTGTAAAAATGGAATCAATAAA This study 3487 seq Fwd1 GTAAAAAGGCCGCGTTGCTGG This study 3487 seq Fwd2 CCTTTGATCTTTTCTACGGGG This study 3487 seq Fwd3 GTCCATTCCGACAGCATCGCC This study 3487 seq Fwd4 ACCAAACGTTTCGGCGAGAAG This study 3487 seq Fwd5 CTGGATACCGCGCGGCTCAAG This study LK1298 CGAGGACCAGCGCTCGTC This study LK1331 CCAAGCTTGCTCACGTTGCAGACGGGG This study exo - 120 fwd CCTGCTCAGAGCAATGCCTG This study exo + 1 rev CCTTGGATCGCAGTGGGTTAC This study Exo - 14 TGGGTTACGAAGTGCCCTTC This study Exo - 161 AAATGGGAAGCGGGAGGGGC This study nfs - 201 fwd CTGCCCCGCGTGACGACC This study nfs + 1 rev CTGCCCCGCGTGACGACC This study 161 Table 3.3 ( c Primers Description Source Exo - 267G ATTGATTCCATTTTTACACTGATGAGGTACCGAATTCCTTCCT CATCCGACCATCCCC This study Exo +108G TCTCCTTACGCATCTGTGCGGTATTCTCGAGCCCGGGCTCGT CTTGCCCATCGTCAGC This study Nfs - 290G ATTGATTCCATTTTTACACTGATGAGGTACCGAATTCCGCTTC CGGGCCCGATTCCTC This study Nfs +83G TCTCCTTACGCATCTGTGCGGTATTCTCGAGCCCGGGGACGG CCAACGAAGCAAAGACG This study D59E (F) CCGCAGGTCGCGGT GATGGAGGTGGAGGGCGACAGCGAG [34] D59E (R) CTCGCTGTCGCCCTCCACCTCCATCACCGCGACCTGCGG [34] ExoA - NF4 CAGCAAGGGCGGACAGAT This study ExoA - NR4 CGGAGCATGACCTCGTGT This study NfsA - NF TTCTTCATCCTGGACAAGCAC This study NfsA - NR TCCAGGTTGACGCGGTAG This study Mxan_5372 F1 CTGGAGTCTTCACGGACGAT This study Mxan_5372 R1 TCTGTTCGACAACGAGGTCA This study Mxan_3259 F3 TCCTCTCCGGGCAGAAGAC This study Mxan_3259 R3 GCATCGATGATCTCCGTCA This study 16S rRNA fwd CAAGGGAACTGAGAGACAGG [73] 16S rRNA rev CTCTAGAGATCCACTACTTGCG [73] pMR3691MCS G - F CACGATGCGAGGAAACGCA [34] pMR3691 MCS G - R CACCGGTACGCGTAACGTTC [34] 162 Figure S3.1 Cellular changes during M. xanthus development. Wild - type strain DK1622 and its mutant derivatives were subjected to starvation under submerged culture conditions. Samples were collected at the indicated hours post - starvation for quantification of (A) sonication - sensitive cells and (B) cells changing shape. Values are expressed as percentage of the number of rod - shaped cells present at the time when starvation - initiated development (T 0 ) (Table S1). Bars show the average of three biological replicates and error bars indicate one standard deviation. 163 Figure S3.2 Transcript levels in wild type , csgA mutant, csgA (P van - fruA ) mutant, csgA (P van - f ruAD59E ) during M. xanthus development . 164 Figure S3.2 ( c Figure S3.3 Transcript stability in wild type and mrpC mutant . Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs, (C) MXAN_3259 and (D) fadI transcript levels by RT - qPCR. Induction of P van with vanillate (0.5 mM) during growth and development was as described previously [1] . Graphs show the data points and average of three biological replicates, relative to the wild - type strain at 18 h, and error bars indicate one standard deviation. Asterisks indicate a significant difference ( p - tailed t - tests) from the wild - type strain at the corresponding time poststarvation . 165 Figure S3.3 ( c Wild - type strain DK1622 and the mrpC mutant were subjected to starvation under submerged culture conditions for 18 h. The overlay was replaced with fresh starvation buffer containing rifampicin (50 mg/mL) and samples were collected immediately ( t 0 ) and at the times indicated ( t x ) for measurement of the exo (A), nfs (B), MXAN_3259 (C) and fadI (D) transcript level by RT - qPCR. Transcript levels at t x were nor malized to that at t 0 for each of three biological replicates and used to determine the transcript half - life for each replicate. The graph shows the average ln( t x /t 0 ) and one standard deviation for the three biological replicates of the wild - type strain an d the mrpC mutant [2] . The average half - li fe and one standard deviation, as well as p values - tailed t - tests, are reported in (E). 166 Figure S3.4 Protein levels in wild type, nla6 kmR mutant, nla6 tetR mutant and mrpC (P van - fruA ) mutants during early time points. Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for immunoblot analy sis as described previously [1] to measure (A) MrpC and (B) FruA levels. Graphs show the data points and average of three biological replicates, relative to the wild - type strain at 6 h, and error bars indicate one standard deviation. The asterisk in panel A indicates a significant dif ference ( p - tailed t - test) from the wild - type strain at the corresponding time poststarvation. 167 Figure S3.6 Developmental phenotype of wild type, nla6 kmR mutant, nla6 tetR mutant, mrpC (P van - fruA ) mutant. Figure S3.5 Development of M. xanthus strains at early times. Wild - type strain DK1622 and its indicated mutant derivates were su bjected to starvation under submerged culture conditions and images were obtained at the indicated number of hours poststarvation. The wild - type strain, both nla6 mutants, and the mrpC P van - fruA strain formed nascent mounds by 12 h (black arrows); however, only the wild - type strain formed compact mounds by 18 h (blue arrow). The mrpC and fruA mutants failed to form mounds. 168 Figure S3.6 ( c Wild - type strain DK1622 and its indicated mutant derivates were subjected to starvation under submerged culture condition and images were obtained at the indicated number of hours poststarvation. The w ild - type strain formed compact mounds (blue arrows) by 18 h, which darkened by 36 h. The Km r nla6 mutant formed nascent mounds (black arrows) at 18 h and compact mounds at 24 and 30 h, but the mounds failed to darken and became less compact at 36 and 48 h. The Tc r nla6 mutant formed nascent mounds by 18 h, but the mounds did not become compact until 36 h and did not darken by 48 h. The mrpC P van - fruA strain formed nascent mounds by 18 h, but the mounds did not become compact and did not darken by 48 h. Figure S3.7 Transcript levels in wild type, nla6 kmR mutant and nla6 t etR mutant during early time points. Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs, (C) MXAN_3259 and (D) fadI transcript levels by RT - qPCR. Graphs show the data points and average of at least three biological replicates, relative to the wild - type strain at 6 h, and error bars indicate one standard deviation. Asterisks indicate a significant difference ( p < - tailed t - tests) from wild type at the corresponding time poststarvation. 169 Figure S3.8 Transcript levels in wild type, devI and devS mutants. 170 Figure S3.8 ( c Wild - type strain DK1622 and its indicated mutant derivatives were subjected to starvation under submerged culture conditions and samples were collected at the indicated number of hours poststarvation for measurement of (A) exo , (B) nfs , (C) MXAN _ 3259 and ( D) fadI transcript levels by RT - qPCR. Graphs show the data points and the average of at least three biological replicates, relative to the wild - type strain at 18 h, and error bars indicate one standard deviation. Asterisks indicate a significant differenc e ( p - tailed t - tests) from wild type at the corresponding time poststarvation. Figure S3.9 Binding of FruA and MrpC to the dev, exo and nfs upstream regions. EMSAs with 32 P - labeled DNA fragments (2 nM) of the dev ( - 119 to + 1), exo ( - 120 to + 1) and nfs ( - 201 to +1) promoter regions and H 6 - MrpC (1 mM), FruA - H 6 (3 mM) and FruA - DBD - His 8 (3 mM) as indicated. Black arrows indicate shifted complexes produced by the individual pr oteins. The gray arrow points to a faint complex indicative of non - cooperative binding by both proteins, whereas the white arrow points to and abundant complex indicative of cooperative binding of the two proteins. 171 REFERENCES 172 R EFERENCES 1. Saha S, Patra P, Igoshin O, Kroos L: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C - signaling. Mol Microbiol 2019, 111:1732 - 1752. 2. B ui NK, Gray J, Schwarz H, Schumann P, Blanot D, Vollmer W: The peptidoglycan sacculus of Myxococcus xanthus has unusual structural features and is degraded during glycerol - induced myxospore development. J Bacteriol 2009, 191(2):494 - 505. 3. O'Connor KA, Zu sman DR: Behavior of peripheral rods and their role in the life cycle of Myxococcus xanthus . J Bacteriol 1991, 173(11):3342 - 3355. 4. Lee B, Holkenbrink, C, Treuner - Lange, A, Higgs, P. I.: Myxococcus xanthus developmental cell fate production: heterogeneou s accumulation of developmental regulatory proteins and reexamination of the role of MazF in developmental lysis. Journal of bacteriology 2012, 194(12):3058 - 3068. 5. Rajagopalan R, Kroos, L.: Nutrient - regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 2014, 196(15):2736 - 2747. 6. Kroos L: Highly signal - responsive gene regulatory network governing Myxococcus development. Trends Genet 2017, 33:3 - 15. 7. Crawford EW, Shimkets LJ: The stringent response in Myxococcus xanthus is regulated by SocE and the CsgA C - signaling protein. Genes Dev 2000, 14(4):483 - 492. 8. Garza AG, Pollack JS, Harris BZ, Lee A, Keseler IM, Licking EF, Singer M: SdeK is required for early fruiting body development in Myxococcus xanthus . J Bacteriol 1998, 180(17):4628 - 4637. 9. Lobedanz S, Sogaard - Andersen L: Iden tification of the C - signal, a contact - dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus . Genes Dev 2003, 17(17):2151 - 2161. 10. Boynton TO, Shimkets LJ: Myxococcus CsgA, Drosophila Sniffer and human HSD17B10 are cardio lipin phospholipases. Genes Dev 2015, 29:1903 - 1914. 11. Ellehauge E, Norregaard - Madsen M, Sogaard - Andersen L: The FruA signal transduction protein provides a checkpoint for the temporal co - ordination of intercellular signals in Myxococcus xanthus developm ent. Mol Microbiol 1998, 30:807 - 817. 12. Ogawa M, Fujitani, S, Mao, X, Inouye, S, Komano, T.: FruA, a putative transcription factor essential for the development of Myxococcus xanthus. Molecular microbiology 1996, 22(4):757 - 767. 173 13. Mittal S, Kroos, L.: A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proceedings of the National Academy of Sciences of the United States of America 2009, 106(6):1965 - 1970. 14. Ueki T, Inouye S: I dentification of an activator protein required for the induction of fruA , a gene essential for fruiting body development in Myxococcus xanthus . Proc Natl Acad Sci USA 2003, 100(15):8782 - 8787. 15. Nariya H, Inouye S: A protein Ser/Thr kinase cascade negati vely regulates the DNA - binding activity of MrpC, a smaller form of which may be necessary for the Myxococcus xanthus development. Mol Microbiol 2006, 60(5):1205 - 1217. 16. Hoang Y, Kroos L: Ultrasensitive response of developing Myxococcus xanthus to the ad dition of nutrient medium correlates with the level of MrpC. J Bacteriol 2018, 200:e00456 - 00418. 17. Campbell A, Viswanathan, P, Barrett, T, Son, B, Saha, S, Kroos, L.: Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. Journal of bacteriology 2015, 197(2):240 - 251. 18. Viswanathan P, Murphy K, Julien B, Garza AG, Kroos L: Regulation of dev , an operon that includes genes essential for Myxococcus xanthus development and CRISPR - associated genes and repeats. J Bacteriol 2007, 189:3738 - 3750. 19. Rajagopalan R, Wielgoss S, Lippert G, Velicer GJ, Kroos L: devI is an evolutionarily young negative regulator of Myxococcus xanthus development. J Bacteriol 2015, 197(7):1249 - 1262. 20. Rajagopalan R, Kroos L: The dev operon regulates the timing of sporulation during Myxococcus xanthus development. J Bacteriol 2017, 199(10):e00788 - 00716. 21. Holkenbrink C, Hoiczyk E, Kahnt J, Higgs PI: Synthesis and assembly of a novel glycan layer in Myxococcus xanthus spores. J Biol Chem 2014, 289(46):32364 - 32378. 22. Muller FD, Schink CW, Hoiczyk E, Cserti E, Higgs PI: Spore formation in Myxococcus xanthus is tied to cytoskeleton functions and polysaccharide spore coat deposition. Mol Microbiol 2012, 83(3):486 - 505. 23. Muller FD, Treuner - Lange A, Heider J, Huntley SM, Higgs PI: Global transcriptome analysis of spore formation in Myxococcus xanthus reveals a locus necessary for cell differentiation. BMC Genomics 2010, 11:264. 24. Bullock HA, Shen H, Boynton TO, Shimkets LJ: Fatty acid oxidation is required for Myxococcus xanthus development. J Bacteriol 2018, 200(10):e00572 - 00517. 25. Giglio KM, Zhu C, Klu nder C, Kummer S, Garza AG: The enhancer binding protein Nla6 regulates developmental genes that are important for Myxococcus xanthus sporulation. J Bacteriol 2015, 197(7):1276 - 1287. 174 26. Licking E, Gorski, L, Kaiser, D.: A common step for changing cell sha pe in fruiting body and starvation - independent sporulation of Myxococcus xanthus. Journal of bacteriology 2000, 182(12):3553 - 3558. 27. Ueki T, Inouye, S.: Identification of a gene involved in polysaccharide export as a transcription target of FruA, an ess ential factor for Myxococcus xanthus development. The Journal of biological chemistry 2005, 280(37):32279 - 32284. 28. Bush M, Dixon R: The role of bacterial enhancer binding proteins as specialized activators of 54 - dependent transcription. Microbiol Mol B iol Rev 2012, 76(3):497 - 529. 29. Higgs P, Hartzell PL, Holkenbrink C, Hoiczyk E: Myxococcus xanthus vegetative and developmental cell heterogeneity. In: Myxobacteria: genomics, cellular and molecular biology. Edited by Yang Z, Higgs P. Norfolk, UK: Cais ter Academic Press; 2014: 51 - 77. 30. Licking E, Gorski L, Kaiser D: A common step for changing cell shape in fruiting body and starvation - independent sporulation of Myxococcus xanthus . J Bacteriol 2000, 182(12):3553 - 3558. 31. Blaylock B, Jiang X, Rubio A , Moran CP, Jr., Pogliano K: Zipper - like interaction between proteins in adjacent daughter cells mediates protein localization. Genes Dev 2004, 18(23):2916 - 2928. 32. Bhat S, Ahrendt T, Dauth C, Bode HB, Shimkets LJ: Two lipid signals guide fruiting body d evelopment of Myxococcus xanthus . MBio 2014, 5(1):e00939 - 00913. 33. Cavicchioli R, Ripple WJ, Timmis KN, Azam F, Bakken LR, Baylis M, Behrenfeld MJ, Boetius A, Boyd PW, Classen AT et al : Scientists' warning to humanity: microorganisms and climate change. Nat Rev Microbiol 2019, 17(9):569 - 586. 34. Saha S, Patra, P, Igoshin, O, Kroos, L.: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational regulation of FruA by C - signaling. Molecular microbiology 2019, 111(6):1732 - 1752. 35. Rajagopalan R, Kr oos L: Nutrient - regulated proteolysis of MrpC halts expression of genes important for commitment to sporulation during Myxococcus xanthus development. J Bacteriol 2014, 196:2736 - 2747. 36. Mittal S, Kroos L: A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proc Natl Acad Sci USA 2009, 106:1965 - 1970. 37. Campbell A, Viswanathan P, Barrett T, Son B, Saha S, Kroos L : Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 2015, 197:240 - 251. 38. Mittal S, Kroos L: Combinatorial regulation by a novel arrangement of FruA and MrpC2 transcription factors during Myxo coccus xanthus development. J Bacteriol 2009, 191:2753 - 2763. 175 39. Lee J, Son B, Viswanathan P, Luethy P, Kroos L: Combinatorial regulation of fmgD by MrpC2 and FruA during Myxococcus xanthus development. J Bacteriol 2011, 193:1681 - 1689. 40. Ueki T, Inouye, S.: Identification of an activator protein required for the induction of fruA, a gene essential for fruiting body development in Myxococcus xanthus. Proceedings of the National Academy of Sciences of the United States of America 2003, 100(15):8782 - 8787. 41. Ueki T, Inouye S: Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development. J Biol Chem 2005, 280(37):32279 - 32284. 42. Robinson M, Son B, Kroos L: Transcriptio n factor MrpC binds to promoter regions of many developmentally - regulated genes in Myxococcus xanthus . BMC Genomics 2014, 15:1123. 43. Caberoy NB, Welch RD, Jakobsen JS, Slater SC, Garza AG: Global mutational analysis of NtrC - like activators in Myxococcus xanthus : identifying activator mutants defective for motility and fruiting body development. J Bacteriol 2003, 185(20):6083 - 6094. 44. Huntley S, Hamann N, Wegener - Feldbrugge S, Treuner - Lange A, Kube M, Reinhardt R, Klages S, Muller R, Ronning CM, Nierman WC et al : Comparative genomic analysis of fruiting body formation in Myxococcales. Mol Biol Evol 2010, 28:1083 - 1097. 45. Son B, Liu Y, Kroos L: Combinatorial regulation by MrpC2 and FruA involves three sites in the fmgE promoter region during Myxococcus xanthus development. J Bacteriol 2011, 193:2756 - 2766. 46. Kroos L, Kaiser D: Expression of many developmentally regulated genes in Myxococcus depends on a sequence of cell interactions. Genes Dev 1987, 1:840 - 854. 47. Viswanathan P, Ueki T, Inouye S, Kroo s L: Combinatorial regulation of genes essential for Myxococcus xanthus development involves a response regulator and a LysR - type regulator. Proc Natl Acad Sci USA 2007, 104:7969 - 7974. 48. Viswanathan K, Viswanathan P, Kroos L: Mutational analysis of the Myxococcus xanthus 4406 promoter region reveals an upstream negative regulatory element that mediates C - signal dependence. J Bacteriol 2006, 188:515 - 524. 49. Yoder - Himes D, Kroos L: Regulation of the Myxococcus xanthus C - signal - dependent 4400 promoter by the essential developmental protein FruA. J Bacteriol 2006, 188:5167 - 5176. 50. Ogawa M, Fujitani S, Mao X, Inouye S, Komano T: FruA, a putative transcription factor essential for the development of Myxococcus xanthus . Mol Microbiol 1996, 22(4):757 - 767. 51. Eichenberger P, Fujita M, Jensen ST, Conlon EM, Rudner DZ, Wang ST, Ferguson C, Haga K, Sato T, Liu JS et al : The program of gene transcription for a single differentiating cell type during sporulation in Bacillus subtilis . PLoS Biol 2004, 2(10):1664 - 1683. 176 52. Kroos L: The Bacillus and Myxococcus developmental networks and their t ranscriptional regulators. Ann Rev Genet 2007, 41:13 - 39. 53. Arrieta - Ortiz ML, Hafemeister C, Bate AR, Chu T, Greenfield A, Shuster B, Barry SN, Gallitto M, Liu B, Kacmarczyk T et al : An experimentally supported model of the Bacillus subtilis global trans criptional regulatory network. Mol Syst Biol 2015, 11(11):839. 54. McKenney PT, Driks A, Eichenberger P: The Bacillus subtilis endospore: assembly and functions of the multilayered coat. Nat Rev Microbiol 2013, 11(1):33 - 44. 55. Shuster B, Khemmani M, Abe K, Huang X, Nakaya Y, Maryn N, Buttar S, Gonzalez AN, Driks A, Sato T et al : Contributions of crust proteins to spore surface properties in Bacillus subtilis . Mol Microbiol 2019, 111(3):825 - 843. 56. Shuster B, Khemmani M, Nakaya Y, Holland G, Iwamoto K, Abe K, Imamura D, Maryn N, Driks A, Sato T et al : Expansion of the spore surface polysaccharide layer in Bacillus subtilis by deletion of genes encoding glycosyltransferases and glucose modification enzymes. J Bacteriol 2019, 201(19):e00321 - 00319. 57. Kas hefi K, Hartzell P: Genetic supression and phenotypic masking of a Myxococcus xanthux frzF - defect. Molec Microbiol 1995, 15:483 - 494. 58. Iniesta AA, Garcia - Heras, F, Abellon - Ruiz, J, Gallego - Garcia, A, Elias - Arnanz, M.: Two systems for conditional gene e xpression in Myxococcus xanthus inducible by isopropyl - beta - D - thiogalactopyranoside or vanillate. Journal of bacteriology 2012, 194(21):5875 - 5885. 59. Gibson DG, Young, L, Chuang, R. Y, Venter, J. C, Hutchison, C. A., 3rd, Smith HO: Enzymatic assembly of DNA molecules up to several hundred kilobases. Nature methods 2009, 6(5):343 - 345. 60. Higgs PI, Jagadeesan, S, Mann, P, Zusman, D. R.: EspA, an orphan hybrid histidine protein kinase, regulates the timing of expression of key developmental proteins of Myx ococcus xanthus. Journal of bacteriology 2008, 190(13):4416 - 4426. 61. McLaughlin PT, Bhardwaj, V, Feeley, B. E, Higgs, P. I.: MrpC, a CRP/Fnr homolog, functions as a negative autoregulator during the Myxococcus xanthus multicellular developmental program. Molecular microbiology 2018. 62. Bradford M: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein - dye binding. Anal Biochem 1976, 72:248 - 254. 63. Britton RA, Eichenberger P, Gonzalez - Pastor JE, Fawcett P, Monson R, Losick R, Grossman AD: Genome - wide analysis of the stationary - phase sigma factor (sigma - H) regulon of Bacillus subtilis. J Bacteriol 2002, 184(17):4881 - 4890. 64. Bullock HA, Shen, H, Boynton, T. O, Shimkets, L. J.: Fatty Acid Oxidation Is Required for Myxococcus xanthus Development. Journal of bacteriology 2018, 200(10). 177 65. Giglio KM, Zhu, C, Klunder, C, Kummer, S, Garza, A. G.: The enhancer binding protein Nla6 r egulates developmental genes that are important for Myxococcus xanthus sporulation. Journal of bacteriology 2015, 197(7):1276 - 1287. 66. Caberoy NB, Welch, R. D, Jakobsen, J. S, Slater, S. C, Garza, A. G.: Global mutational analysis of NtrC - like activators in Myxococcus xanthus: identifying activator mutants defective for motility and fruiting body development. Journal of bacteriology 2003, 185(20):6083 - 6094. 67. Kaiser D, Manoil C: Myxobacteria: cell interactions, genetics and development. Ann Rev Microb iol 1979, 33:595 - 639. 68. Shimkets LJ, Asher, S. J.: Use of recombination techniques to examine the structure of the csg locus of Myxococcus xanthus. Molecular & general genetics : MGG 1988, 211(1):63 - 71. 69. Sun H, Shi, W.: Genetic studies of mrp, a loc us essential for cellular aggregation and sporulation of Myxococcus xanthus. Journal of bacteriology 2001, 183(16):4786 - 4795. 70. Kroos L, Kaiser, D.: Expression of many developmentally regulated genes in Myxococcus depends on a sequence of cell interacti ons. Genes & development 1987, 1(8):840 - 854. 71. Viswanathan P, Murphy, K, Julien, B, Garza, A. G, Kroos, L.: Regulation of dev, an operon that includes genes essential for Myxococcus xanthus development and CRISPR - associated genes and repeats. Journal of bacteriology 2007, 189(10):3738 - 3750. 72. Rajagopalan R, Wielgoss, S, Lippert, G, Velicer, G. J, Kroos, L.: devI is an evolutionarily young negative regulator of Myxococcus xanthus development. Journal of bacteriology 2015, 197(7):1249 - 1262. 73. Ossa F, Diodati, M. E, Caberoy, N. B, Giglio, K. M, Edmonds, M, Singer, M, Garza, A. G.: The Myxococcus xanthus Nla4 protein is important for expression of stringent response - associated genes, ppGpp accumulation, and fruiting body development. Journal of bacterio logy 2007, 189(23):8474 - 8483. 178 C HAPTER 4: Conclusion and f uture d irections Studies of bacterial gene regulatory networks (GRNs) have significantly progressed in the recent past. By understanding the mechanisms of gene regulation, this work is not only building a base of fundamental knowledge, but in the case of genes critical for pathogenesis, is elucidating potential targets for the advancement of therapeutic strategies. Additionally, understanding of GRNs governing bacterial sporulation aids elucidation of mechanisms underlying disease transmission [1] and resistance against host immunity [2] . With the advancement of interdisciplinary research, systems biology combining experimental and computational methods has emerged as a powe rful approach to study GRNs [3] . By using M yxococcus xanthus as a model system, my work has involved the first such systematic approach to elucidate the dynamics of the GRN governing a bacterial multicellular developmental pr ocess that culminates in commitment to sporulation. Our systematic analysis uncovered a novel role of the long - known transcriptional activator FruA in negatively regulating expression of network output genes before FruA is activated by C - signaling. From th e systematic analysis, the project progressed to mechanistic approaches (i.e. DNA - binding studies) aimed at further elucidating the differential role of upstream transcription factors (MrpC and FruA) in regulating the network output genes. In this chapter, the key findings of the work will be discussed in the context of the outstanding questions and potential future directions. The GRN governing multicellular development of M. xanthus can be studied systematically Significant progress was made toward establishing systematic and quantitative methods to study the GRN governing multicellular development of M. xanthus [4] . One major accomplishment was the establishment of a higher - throughput, robotic platform for qRT - PCR 179 analysis to measu re RNA levels of large numbers of sample [4] . Reproducibility of the qRT - PCR analysis was tested between biological and technical replicates. In the field of M . xanthus development, 16S rRNA has been commonly used as an internal standard for mRNA measure ments. By reporting that the yield of total RNA per cell (which is primarily rRNA) remains unchanged during the commitment period, we validated 16S rRNA as a reliable internal control to be used to measure mRNA transcript levels. Another significant accomp lishment was the establishment of systematic methods to quantify cellular changes during the commitment period. These methods involve quantification of cell numbers and types followed by computation of the percentage of the starving population committing t o the three developmental cell fates (lysis, peripheral rods and spores) [4] . Systematically collected data on transcript and protein levels was used to build a computational model of part of the GRN, which was used to make predictions about experimental ly testable hypotheses. A potential extension of the work would be refinement of the computational model by implementing the cell fate data reported in Chapter 2. The magnitude of the molecular changes predicted by the computational model can perhaps be re fined by incorporating the percentages of cells in the developing population adopting different cell fates. Systematic analysis of the M. xanthus GRN supports C - signal - dependent posttranslational activation of FruA resulting in commitment to form spores A significant finding from the systematic analysis was support for a model in which C - signaling posttranslationally activates FruA at least ninefold in order to increase dev transcription and commit cells to form spores [4] . C - signaling was earlier suggested to posttranslationally activate FruA by phosphorylation [5] . However, we found that boosting the level of native FruA or a phosphomimetic form of FruA by ectopic expression from a vanillate - inducible promoter in a 180 mu tant defective of C - signaling, did not increase the dev transcript level (Chapter 2) [4] . This finding perhaps rules out phosphorylation to be the mechanism by which FruA is activated in response to C - signal. A fascinating extension of this work would be to elucidate the mechanism by which C - signaling activates FruA. Investigations of C - signaling have led to two models discussed in Chapter 1 [6, 7] . One of those models suggests that phospholipase activity of CsgA releases diacylglycerols (DAGs) from the inner membrane, which serve as the C - signal and account for cell shortening during development [7] . DAGs released by CsgA are eventually converted t o triacylglycerols (TAGs) by acyltransferases, resulting in formation of cytosolic lipid bodies that may store energy for use later during development [7] . It is possible that - oxidation of fatty acids released from TAGs elevates cellular acetyl - CoA and FruA is activated by acetylation. In order to test whether Fr uA gets activated by acetylation, FruA expressed in and purified from E. coli could be subjected to in vitro acetylation [8] . The DNA - binding ability of acetylated FruA alone and or in combination with MrpC would be tested in EMSA with 32 P - labelled dev DNA, and compared with non - acetylated FruA. Greater affinity of acetylated FruA for dev DNA would suggest that acetylation of one or more lysine residues of FruA is the mechanism of activation by C - signaling. If it appears that C - signaling ac tivates FruA by acetylation, the investigation could be further extended by measuring the levels of CsgA and activated FruA (FruA * ) during development by immunoblot. Commercially available acetylation - specific antibody would be used to determine the level FruA * . Anti - FruA antibody would be used to quantify the level of total FruA (activated and unactivated). Using a method devised by Tye Boynton and Larry Shimkets (University of Georgia), I purified CsgA and it was used to generate polyclonal antibodies in rabbits. By 181 quantifying the levels of CsgA, total FruA and FruA * during development, the ratio of FruA * to total FruA as a function of the CsgA could be determined. It would be exciting to observe a steady increase in the FruA * /total FruA ratio, perhaps co rrelating with an increasing level of CsgA, during the period leading up to and including commitment to spore formation. If it appears that C - signaling does not activate FruA by acetylation, other approaches to investigate the mechanism of FruA activatio n could be tried. For example, native FruA has been substituted by a functional histidine - tagged version in M. xanthus , both in the wild - type strain and in a csgA mutant. Purification of the recombinant protein from both strains followed by mass spectromet ry approaches identify a modification and its precise location in the protein from the wild - type strain that is not present in the protein from the csgA mutant. Understanding of the mechanism of C - signal - dependent posttranslational modification of FruA would solve a long - standing mystery in the field of M. xanthus development and would open up avenues to test the effect of FruA activation on gene express ion. Our systematic analysis in combination with computational modeling suggests that C - signaling activates FruA at least ninefold for cells to increase dev transcription and commit to spore formation [4] . It would be fascinating to determine the minimu m level of activated FruA required to induce dev expression in individual cells committing to form spores. In order to accomplish this, methods to measure gene expression and visualize cellular shape change at the single - cell level are being developed. A functional mNeonGreen - FruA fusion expressed from the native promoter in M. xanthus has been created and studied by Y Hoang in our group. Y has also fused the dev promoter region and the fmgE promote region (which appears to require a higher level of activa ted FruA for expression) to tdTomato. Both the wild - type strain and a csgA 182 mutant bearing both the mNeonGreen - FruA fusion and a promoter - tdTomato fusion have been created. 3D confocal laser scanning microscopy (CLSM) is being used to measure the green and red fluorescence intensity of individual cells in mound during development. Because tdTomato is cytosolic, red fluorescence also indicates cell shape. Based on the data the computational model can be refined where dev serves as the reporter of FruA activit y in individual cells undergoing shape change. FruA is both a negative and a positive regulator of developmental genes in Myxococcus xanthus Ou r systematic analysis revealed a novel role of unactivated FruA in negatively regulating three of the output genes of the network ( exo , nfs , MXAN_3259 ) (Chapter 3). The EBP Nla6 appears to activate developmental genes at the preaggregation stage of M . xanthus development [9] and differentially regulate exo expression at different times during development [10] . By binding to the exo promoter region, Nla6 is proposed to positively and negatively regulate exo transcription early and late in development, respectively [10] One of the two Nla6 binding sites in the exo promoter region partially overlaps with the FruA - binding site [11] , so negative regulation by unactivated FruA could involve competition for binding with Nla6 (Chapter 3). Experiments are plan ned to test this model using purified MBP fused to the Nla6 DNA - binding domain (MBP - Nla6 DBD) and FruA in EMSAs. Since the patte r n s of nfs and MXAN_3259 transcript levels in a fruA mutant are similar to exo (Chapter 3 ) , it would be a worthwhile future direction to test if MBP - Nla6 DBD and FruA interfere with each other to regulate nfs and MXAN_3259 . An additional motivation for pursuing these competition EMSAs is the earlier evidence showing MBP - Nla6 DBD binding to the M XAN_3259 promoter region in vitro [10] . An 183 added challenge is that neither the nfs nor the MXAN _ 3259 transcriptional start site has been mapped yet, which will be important for interpretation of binding results. We propose that during mound formation unactivated FruA negatively regulates all three output genes by interfering with Nla6 binding, whereas later during development, in response to a higher threshold level of C - signaling [11] , FruA * ac tivates transcription of these operons, whose products are involved in spore coat biogenesis. To further extend this work it would be fascinating to elucidate the mechanism by which FruA * replaces inactive FruA at the promoter regions of these operons. It is likely that the promoter regions of exo, n f s and MXAN_3 259 would differ in their affinities for FruA and FruA * . Presumably, a promoter with a relatively high binding affinity for FruA and/or a relatively low binding affinity for FruA * would require a high level of C - signaling and FruA * in order for transcripti on to occur. It would be intriguing to perform in vitro competition EMSAs to determine whether FruA * competes with FruA for binding to the same site(s). For example, if acetylation is the mechanism by which C - signaling activates FruA, then acetylated FruA would be used in competition EMSAs with non - acetylated FruA. Developmental expression of fadI was di stinct compared to the other three outputs. Evidence provided in Chapter 3 suggests that transcription of fadI is not regulated by unactivated FruA, but is positively regulated by FruA * and MrpC. The close proximity between a putative FruA binding site cen tered at - 110 and a putative MrpC binding site centered at - 90 (the centers of both sites are relative to the translation start codon since the transcriptional start site has not been mapped) strongly suggests cooperative binding between the two transcript ion factors , as observed for dev [12] and fmg genes [13 - 15] . Alternatively, MrpC may 184 regulate fadI indirectly by activating fruA transcription, resulting in FruA * binding to the fad I p romoter region in response to C - signaling. Hence, it would be an interesting future direction to test whether MrpC and FruA, individually and/or cooperatively bind to the fadI promoter region using EM SAs. Upon detection of MrpC and/or FruA binding , smaller DNA fragments would be used to localize the binding sites, then the sites would be examined for sequences matching the consensus binding sites for MrpC and FruA ) [16] . Mutations designed to eliminate binding of each transcription factor would be introduced into M . xanthus by allelic exchange, followed by testing the effect in vivo on the fadI transcript level by RT - qPCR. Another interesting future direction of this work would be to design an allele of fruA that makes FruA unable to be activated by C - signaling. For example, if FruA appears to be activated by acetylation, substitution of one or more lysine residues w ith a residue that cannot be acetylated (accomplished by allelic exchange of fruA in M. xanthus ), would be tested by measuring transcript levels of the output genes by RT - qPCR during development. Our results presented in Chapter 3 suggest that activated Fr uA * positively regulates all four output genes, so in a strain making FruA that cannot be activated, output gene transcript levels are predicted to remain low. This approach has potential to provide additional evidence in support the mechanism of FruA acti vation and in support of our model for regulation of late gene transcription (Chapter 3). Closing remarks The work presented in this dissertation has relied on the huge amount of fantastic work that was already done on M. xanthus development. My work benefited greatly from established genetic and molecular approaches, and knowledge, in publications contributed by many 185 scientists. In particular, I thank the scientists who provided us with antibodies, strains and advice, which was important for my work. Our findings have contributed systematic methods and a better understanding of commitment to spore formation in M. xanthus . We both elucidated the dynamics of the GRN and discovered a novel role of the previously known transcriptional activator FruA in negatively regulating transcription of network output genes. In the future, there is tremendous potential to extend my work toward a deeper mechanistic understanding of the role of MrpC and FruA in differentially regulating the output genes. I look forward to the discovery of additional novel mechanisms that will advance the fields of signal transduction and gene regulati on. 186 REFERENCES 187 R EFERENCES 1. Shen A, Edwards, A. N, Sarker, M. R, Paredes - Sabja, D.: Sporulation and Germination in Clostridial Pathogens. Microbiology spectrum 2019, 7(6). 2. Manganelli R, Provvedi, R.: An integrated regulatory network including two positive feedback loops to modulate the activity of sigma(E) in mycobacteria. Molecular microbiology 2010, 75(3):538 - 542. 3. Durmus S, Cakir, T, Ozgur, A, Guthke, R.: A review on c omputational systems biology of pathogen - host interactions. Frontiers in microbiology 2015, 6:235. 4. Saha S, Patra, P, Igoshin, O, Kroos, L.: Systematic analysis of the Myxococcus xanthus developmental gene regulatory network supports posttranslational r egulation of FruA by C - signaling. Molecular microbiology 2019, 111(6):1732 - 1752. 5. Ellehauge E, Norregaard - Madsen, M, Sogaard - Andersen, L.: The FruA signal transduction protein provides a checkpoint for the temporal co - ordination of intercellular signals in Myxococcus xanthus development. Molecular microbiology 1998, 30(4):807 - 817. 6. Lobedanz S, Sogaard - Andersen, L.: Identification of the C - signal, a contact - dependent morphogen coordinating multiple developmental responses in Myxococcus xanthus. Genes & development 2003, 17(17):2151 - 2161. 7. Boynton TO, Shimkets, L. J.: Myxococcus CsgA, Drosophila Sniffer, and human HSD10 are cardiolipin phospholipases. Genes & development 2015, 29(18):1903 - 1914. 8. Christensen DG, Baumgartner, J. T, Xie, X, Jew, K. M, Basisty, N, Schilling, B, Kuhn, M. L, Wo lfe, A. J.: Mechanisms, Detection, and Relevance of Protein Acetylation in Prokaryotes. mBio 2019, 10(2). 9. Giglio KM, Caberoy, N, Suen, G, Kaiser, D, Garza, A. G.: A cascade of coregulating enhancer binding proteins initiates and propagates a multicellu lar developmental program. Proceedings of the National Academy of Sciences of the United States of America 2011, 108(32):E431 - 439. 10. Giglio KM, Zhu, C, Klunder, C, Kummer, S, Garza, A. G.: The enhancer binding protein Nla6 regulates developmental genes that are important for Myxococcus xanthus sporulation. Journal of bacteriology 2015, 197(7):1276 - 1287. 11. Ueki T, Inouye, S.: Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development. The Journal of biological chemistry 2005, 280(37):32279 - 32284. 12. Campbell A, Viswanathan, P, Barrett, T, Son, B, Saha, S, Kroos, L.: Combinatorial regulation of the dev operon by MrpC2 and FruA during Myxococcus xanthus development . Journal of bacteriology 2015, 197(2):240 - 251. 188 13. Lee JS, Son, B, Viswanathan, P, Luethy, P. M, Kroos, L.: Combinatorial regulation of fmgD by MrpC2 and FruA during Myxococcus xanthus development. Journal of bacteriology 2011, 193(7):1681 - 1689. 14. Mitt al S, Kroos, L.: A combination of unusual transcription factors binds cooperatively to control Myxococcus xanthus developmental gene expression. Proceedings of the National Academy of Sciences of the United States of America 2009, 106(6):1965 - 1970. 15. Mi ttal S, Kroos, L.: Combinatorial regulation by a novel arrangement of FruA and MrpC2 transcription factors during Myxococcus xanthus development. Journal of bacteriology 2009, 191(8):2753 - 2763. 16. Robinson M, Son, B, Kroos, D, Kroos, L.: Transcription fa ctor MrpC binds to promoter regions of hundreds of developmentally - regulated genes in Myxococcus xanthus. BMC genomics 2014, 15:1123.