Regulation of cell size and cell division in the filamentous cyanobacterium Anabaena: a paradigm of bacterial multicellularity

Regulation of cell size and cell division in the filamentous cyanobacterium Anabaena: a paradigm of bacterial multicellularity Cristina Velázquez Suárez Sevilla, 2023 Instituto de Bioquímica Vegetal y Fotosíntesis Departamento de Bioquímica Vegetal y Biología Molecular Consejo Superior de Investigaciones Científicas – Universidad de Sevilla TESIS DOCTORAL Regulation of cell size and cell division in the filamentous cyanobacterium Anabaena: a paradigm of bacterial multicellularity Trabajo presentado para optar al grado de Doctora en Biología por la graduada Cristina Velázquez Suárez Directora Director Dra. Antonia Herrero Moreno Dr. Ignacio Luque Romero Profesora de Investigación del CSIC Científico Titular del CSIC Tutora Dra. Mercedes García González Profesora Titular de Bioquímica y Biología Molecular – Universidad de Sevilla FUNDING This Ph.D. Thesis has been founded by FPI predoctoral contract BES-2017-082044 financed by MCIN/AEI /10.13039/501100011033 and by FSE invierte en tu futuro, associated to grants BFU2016-77097-P, financed by MCIN/ AEI /10.13039/501100011033/ and by FEDER Una manera de hacer Europa, and grant PID2020-118595GB-I00, financed by MCIN/ AEI /10.13039/501100011033. The work was also founded by grant P20_00032 funded by Junta de Andalucía and FEDER. This work has been conducted at the Instituto de Bioquímica Vegetal y Fotosíntesis (CSIC-US). A three-months stay at the Free University of Berlin (Germany), under the supervision of Dr. Dennis Nürnberg, founded by contract BES-2017-082044, was also made. A los tres pilares de mi vida: mi padre, mi madre y mi hermana AGRADECIMIENTOS Ahora que esta tesis llega a su fin quiero agradecer a todas las personas que han permanecido a mi lado, a los que han llegado y a los que se han ido antes de tiempo. A todos: GRACIAS. En primer lugar, quiero dar las gracias a mis directores de tesis, los Dres. Antonia Herrero e Ignacio Luque, por darme la oportunidad de formarme y de crecer como científica y como persona en su grupo. Por la paciencia, la dedicación y la motivación durante todos estos años, gracias de corazón. Al grupo de Biología de Cianobacterias Multicelulares: a Enrique por su visión crítica; a Ana y Gracia por cuidarme siempre; a Rocío, por ser como una hermana mayor; a Mercedes, Mireia, Sergio, Alicia, Miguel Ángel y Cristina-Zgz por aguantar mis dramas. A todo el personal de mi segunda casa, el IBVF, con una mención especial a Alicia Orea, a José Enrique, a Ana Jurado y a Toñi. I would like to thank Dr. Dennis Nürnberg for hosting me in his laboratory and thanks to Laura, Vera, Xiaoran, Shujie, Claudia, Kalman and Janis for the help during those months. A special thanks to Viktor and Mini, for making my stay in Berlin an unforgettable experience. Me gustaría agradecer a la primera persona que confió en mí en la ciencia, Eva Valdivia, que diez años después sigue siendo mi amiga y una fuente de inspiración. Por supuesto, a todos los que forman el BIO160 del departamento de Microbiología de la UGR y que prestan su apoyo desde la distancia: Rubén, Manolo, Platero, Sonia y Pirri. Sin mi formación con vosotros, no habría sido lo mismo. Mil gracias a mis amigas, por demostrar que los kilómetros que nos separan a todas, si se quiere, pueden ser solamente una cifra. Por las risas y por los llantos, por los consejos y por escucharme aunque no entendáis de qué hablo. María Isabel, Virginia, Cristina, Marién y Jesús. A bebé Blanca que me esperó hasta la vuelta de Berlín, algún día te lo contaremos. Gracias a Susana, a quien la microbiología trajo a mi vida y nos hemos visto crecer hasta compartir año de defensa de tesis. Gracias a mis primos, a mis tíos, a Mangel. Y, en especial, a mis abuelos por el amor incondicional y ser un ejemplo de superación y valentía; sé que estaréis orgullosos de mí. A Javi y a su familia, por haber sido parte de la mía durante tanto tiempo. Las últimas líneas de estos agradecimientos son para las tres personas más importantes de mi vida: mi padre, mi madre y mi hermana. Por el apoyo, por escuchar mis monólogos, por la confianza siempre, por creer en mí, por secar mis lágrimas y reír con mis tonterias. Gracias por hacer del pequeño renacuajo que fui la persona que hoy soy. Introduction 2 Introduction 3 1. BACTERIAL CELL MORPHOLOGY Billion years of evolution have made bacteria largely diverse, showing wide variations in lifestyle, metabolism and ultrastructure. The information for these particular characteristics is genome encoded and defines the identity of each phylogenetic group or species. Likewise, the morphology of bacterial cells is genetically encoded and it is an intrinsic feature of each bacterial species inherited from the mother cell. Thus, since all individuals of a particular species show a similar morphology, for more than a century microbiologists have used this trait as a taxonomic criterion. Cell size and cell shape contribute importantly to cell morphology, although other features (i.e. the existence of surface extensions or appendages) can also influence it. Despite the great progress made in the last decades, more research is still needed to elucidate in full the mechanisms of determination of bacterial cell morphology. 1.1. Cell size In the domain Bacteria, the variation in cell size expands more than four orders of magnitude, ranging from the 0.2 µm of Mycoplasma pneumoniae to the 1000 µm in length of the recently discovered giant bacteria. The fact that these giant bacteria are visible by the naked eye has challenged the concept that bacteria are microbes (Figure 1). The largest known giant bacteria are the free-living Thiomargarita namibiensis and T. magnifica, and Epulopiscium fishelsoni, a symbiont of the surgeonfish (Levin and Angert, 2015; Volland et al., 2022). Size variation in bacteria is limited by some constraints. The lower limit is dictated by the minimum space needed to house biochemical components such as proteins, lipids and nucleic acids, essential for survival and reproduction (Koch, 1996). The upper limit is thought to be set by the ratio between the cell surface and the cell volume (the S/V ratio). For spherical cells, the larger the cell, the lower its S/V ratio. Because the cell surface mediates the molecular exchange with the surrounding medium, large cells have, in relation to its volume, less surface for exchange with the environment than small cells, which would be detrimental beyond certain limits. However, rather than the surface size itself, the real limit appears to be imposed by the number of transporters to transfer nutrients and waste into and out of the cytoplasm than can be accommodated in the available envelope space (Young, 2006; Westfall and Levin, 2017). Cell size is also restricted by the limitations of molecular diffusion within the cytoplasm, which largely impinges on biochemical reactions, growth and reproduction (Beveridge, 1988; Schulz and Jørgensen, 2001). Nonetheless, bacteria have evolved different ways to overcome such limitations, such as adopting a shape that maximizes the surface area, i.e. the spiral shape of Spirochaetes or Helicobacter pylori, or the extensive vesiculation of the inner membrane of the giant bacterium E. fishelsoni (Young, 2006; Levin and Angert, 2015). The also giant T. namibiensis Introduction 4 shows a large vacuole at the center of the cytoplasm and confines the metabolic activity to peripheral vesicular compartments (Volland et al., 2022). Notwithstanding that cells of the same strain have the same average size, this parameter can change to some extent depending on certain conditions. The growth law proposed in 1958 by Schaechter was based on empirical observations in Salmonella thyphimurium and postulates that the cell size increases linearly with the growth rate and inversely correlates to the duplication time (reviewed in Morcinek-Orlowska et al., 2019). Thus, environmental conditions such as the availability of carbon, phosphate or nitrogen, which determine the growth rate, influence cell size in a variety of species (Levin and Angert, 2015). Some evidence indicates that in some rod-shaped bacteria such as Caulobacter crescentus, Escherichia coli or Listeria monocytogenes, the alarmone guanosine tetraphosphate (ppGpp) could play an important role on cell size determination (Büke et al., 2022). Also, a threshold of unincorporated cell wall precursors appear to signal the time of cell division initiation, which impacts on cell size (Chen and Good, 2016; Harris and Theriot, 2016). However, in general terms, the molecular mechanisms by which variations in bacterial cell size are determined remain largely unknown. Anabaena species Thiomargarita magnifica Mycoplasma pneumoniae Thiomargarita namibiensis 0.1 µm 1 µm 10 µm 0.1 mm 1 mm 1 cm Escherichia coli Lyngbya majuscula Epulopiscium fishelsoni Figure 1. Variation of bacterial size. The scheme is aimed to illustrate the different sizes of bacteria, from the smallest species of the genus Mycoplasma to the giant bacteria of the genus Thiomargarita. Figure generated with Biorender (https://biorender.com/). 1.2. Cell shape Species of the Bacteria domain exhibit an astonising vast variety of shapes (Figure 2). The bacterial cell shape is genetically determined and maintained through generations, although some modifications can take place throughout the cell cycle (van Teeseling et al., 2017). Although the most common shapes are cocci and rods, there are many other cell forms, including those resembling spirals or comma, lemon or teardrops, thin circular disks, hexagons or stars (Figure 2). There are also irregular morphologies derived from asymmetric cell division or the presence of appendages, which are used for attachment to surfaces or to increase the cell surface (Young, 2006; Willis and Huang, 2017). Introduction 5 The cell shape is important for different features of bacteria. For instance, as mentioned above, some shapes maximize the S/V ratio providing an adaptative advantage by increasing the accessibility to nutrients. Thin flat shapes provide maximal S/V ratio, but such morphology is not the most abundant, indicating that other parameters also influence on the evolution of cell morphology (Young, 2006; Ojkic and Banerjee, 2021). Cell shape is also important for the colonization of surfaces, as in C. crescentus; for biofilm formation, which requires a certain surface for contact with the substratum and the adjacent cells; for motility in social groups or as single cells, or for pathogenicity, as certain cell shapes may contribute to motility in viscous media and evasion from host defenses (van Teeseling et al., 2017). Also, the cell shape can impact antibiotic resistance by limiting the flow of antibiotics into the cell (Ojkic et al., 2022). Figure 2. Diversity of bacterial shapes. Although the most common bacteria are cocci, ovococci or bacilli, there are lemon-shaped, teardrop-shaped, curved, helical-shaped, star-shaped or irregularly-shaped cells, among others. Figure from Young, (2006). All bacterial cells show variations of their shape, including changes in cell length and width, troughout the cell cycle (Young, 2006). Changes associated with growth and division are in general moderate and the overall shape of the cell is maintained. However, some species undergo large alterations. For instance, some species exhibit morphological changes that involve the generation of new regions or appendages (van Teeseling et al., 2017). Besides, the so-called polymorphic, or dimorphic, bacteria show distinct shapes along their life cycle. A well-studied case is C. crescentus, which is dimorphic and whose cell cycle is dependent on morphological changes produced by an asymmetric division that gives rise to a stalked sessile cell and a flagellated swarmer cell, whose function is to search new places to colonize (Young, 2006; van Teeseling et al., 2017). The cell shape can also be altered by nutrient availability, leading to the formation of swarming motile cells that show a rounded shape (i.e. rod-shaped cells can turn into coccoids). In other cases, nutrient limitation induces a state of dormancy defined as viable but not culturable (VBNC) and the differentiation of spores, which are metabolically inert and show smaller size and rounder shape than normal cells (van Teeseling et al., 2017). Furthermore, shorter or longer cells can be generated when cell division and cytokinesis are altered, as Introduction 6 observed e.g. in E. coli mutants that form minicells or filaments by overexpression or downexpression of FtsZ, the protein initiating cell division (see below), respectively (Ward and Lutkenhaus, 1985; Dai and Lutkenhaus, 1991). Regarding the mechanisms of determination of cell shape, it is well known that the peptidoglycan (PG) sacculus that surrounds the cytoplasmic membrane is responsible for the specific shape of each bacterial species. Indeed, upon isolation, the sacculus of different bacteria maintains the cell shape. Furthermore, PG-less cells (spheroplasts) generated by treatment with hydrolases can synthesize their cell wall de novo and restore the original cell shape after a few generations (Ranjit and Young, 2013). This important observation indicates that the formation of the sacculus and in turn the determination of the cell shape does not require a preexisting template, but depends on biosynthetic complexes and their dynamics and topology (van Teeseling et al., 2017). 2. THE BACTERIAL CELL ENVELOPE AND ITS EVOLUTION The bacteria cell envelope is constituted by the external layers of the cell and plays a fundamental role in the determination of the cell shape. Besides, the cell envelope is responsible for other functions, such as exchanges with the environment, including nutrient uptake and waste disposal, energy production, protection from mechanical stress or aggressions and adhesion to surfaces (Strahl and Errington, 2017; Egan et al., 2020; Sun et al., 2022). Bacteria have long been classified into two groups according to the structure of their envelope. Such classification was based on the differential staining of cells with the technique developed by Gram in 1884, but it was not until 1964 that the correlation between the differential staining pattern and the structure of the cell envelope was established (Megrian et al., 2020). Grampositive bacteria have an inner plasma membrane and a thick PG wall composed of multiple layers, decorated with teichoic and lipoteichoic acids facing the external medium. Gramnegative bacteria contain a thin layer of PG in between the inner membrane and an asymmetric outer membrane (OM), in the so-called periplasmic space or periplasm (Figure 3). The Gram test is based on staining with crystal violet followed by washing with ethanol and staining with safranin. Gram positive bacteria retain the crystal violet in its thick PG wall, thus appearing violet. In contrast, after the treatment with the alcohol, which degrades the OM and also makes their PG wall more porous, Gram-negative bacteria cannot retain the crystal violet in its thin PG; however, they can retain the safranin dye, thus appearing pink (Kumar et al., 2022). This nomenclature is still extensively used in the scientific literature, but since a major difference between these two groups of bacteria is the existence of one or two membranes in the cell envelope, a current tendency is to refer them as ‘monoderm’ or ‘diderm’ bacteria. Introduction 7 O-specific polysaccharide A Core polysaccharide Proteins Braun lipoprotein Porin Lipopolysaccharide Lipid A Outer membrane Phospholipid Periplasm Cytosol Pep4doglycan Plasma membrane B Teichoic acid Lipoteichoic acid Wall-associated protein Pep4doglycan Plasma membrane Proteins Phospholipid Figure 3. Types of bacterial cell envelopes. (A) Schematic of the cell wall of E. coli, a representative of Gram-negative or diderm bacteria, with a thin layer of peptidoglycan and an outer membrane, with lipopolysaccharides in its outer face, bound to the peptidoglycan by the Braun lipoprotein. A variety of proteins are embedded in the plasma membrane and the outer membrane. (B) Representation of the envelope of Gram-positive or monoderm bacteria containing a thick multilayered peptidoglycan sacculus in which teichoic and lipoteichoic acids are embedded. Both, the plasma membrane and the peptidoglycan present associated proteins. Figure generated with Biorender (https://biorender.com/). The basis for the existence of these two types of bacteria is a major question in Biology. It has been long assumed that monoderm and diderm bacteria are distinct phylogenetic groups, and an intense debate on their origin and evolution has led to propose that monoderm bacteria derived from diderm ones or vice-versa (Cavalier-Smith, 2002; Lake, 2009; Gupta, 2011; Tocheva et al., 2016). Recent investigations have provided a solid support for the notion that diderms were the first bacteria to evolve (Coleman et al., 2021; Witwinowski et al., 2022). Coleman and collaborators have accurately located the root of the bacterial evolutionary tree between two large phylogenetic groups known as ‘Gracilicutes’ and ‘Terrabacteria’, respectively. Gracilicutes include the vast majority of diderm bacteria, whereas Terrabacteria include phyla of diderm and monoderm bacteria (Figure 4), suggesting that the common ancestor of both groups was a diderm organism. Based on the analysis of the evolutionary path of bacterial genes, this study proposed that the Last Bacterial Common Ancestor (LBCA) was a rod-shaped, diderm bacteria (Coleman et al., 2021). Cytosol Introduction 8 In addition, some monoderm or diderm bacteria have an outermost envelope layer, the socalled S-layer. The S-layer is a paracristalline monolayer of proteins or glycoproteins that may have different roles such as anchoring to surfaces or structural strengthening, and even can serve as a sort of cell wall in some bacteria lacking it (Missiakas and Schneewind, 2017). Proteobacteria Desulfuromonadota Desulfobacterota Bdellovibrionota Myxococcota Nitrospirota Acidobacteriota ACD Dependen>ae FCB PVC Elusimicrobiota Spirochaetota Fusobacteriota DST Cyanobacteria Margullisbacteria Ac>nobacteriota Firmicutes Eremiobacterota Arma>monadota Chloroflexota Dormibacterota CPR Gracilicutes Terrabacteria Figure 4. Bacterial cell envelope evolution. Phylogenetic tree of bacteria showing division into two major groups, the so-called Gracilicutes and the Terrabacteria. For each phylum, the structure of the cell envelope is shown at the right side. Figure adapted from Coleman et al., (2021). Diderm Monoderm Introduction 9 2.1. The plasma membrane The innermost envelope layer that surrounds the cytoplasm and separates it from the environment is the cytoplasmic or inner or plasma membrane (PM). The bacterial PM is a lipid bilayer, the major constitutents of which are phosphatidylethanolamine, phosphatidylglycerol and cardiolipin (Epand and Epand, 2009; Sohlenkamp and Geiger, 2015). Proteins partially or completely embedded, or located at its periphery are also important constituents (Figure 3). The major function of the PM is to mediate the exchange of the cell with the environment, acting as a selective permeability barrier for the passage of molecules through it. The ability to traverse the PM is mainly dictated by the nature of molecules, in such a way that lipophylic and small neutral molecules can diffuse in a gradient-dependent manner. Other molecules pass through the PM by means of protein systems that allow their transport to the inside or the outside of the cell. The transport of nutrient and waste substances through the PM is essential and bacterial cells dedicate a plethora of systems to this role. Transport systems may mediate facilitated diffusion of molecules, which always happens in a gradient-dependent manner, or the transfer of substrates against a concentration gradient (active transport). Transport systems can be constituted by a single transmembrane protein or may be complex structures involving several proteins, such as those of the large category of ABC (ATP-binding cassette) transporters, which normally include a periplasmic substrate-binding protein. Energy for the transport of substances against a concentration gradient can be provided by the disipation of an ion gradient, the proton motive force or, as in the case of ABC transporters, ATP (Konovalova et al., 2017). Molecules exported from the cytoplasm are either secreted to the periplasm or to the extracellular medium and include waste molecules, constituents of the outermost layers of the cell envelope and molecules with signalling functions, mediating intercellular interactions or devoted to the scavenging of nutrients (i.e. siderophores). Some proteins of the PM may also participate in sensing environmental stimuli and signal transduction. 2.2. The outer membrane In diderm bacteria, the outermost layer of the cell envelope is the OM, which is a lipid bilayer but, at difference with the PM, it is asymmetric, being constituted by phospholipids, in its inner leaflet, and glycolipids, known as lipopolysaccharide (LPS), in its outer leaflet. In addition, the outer membrane possesses lipoproteins, with a lipid portion embedded in the membrane, and transmembrane proteins called outer membrane proteins (Konovalova et al., 2017; Choi and Lee, 2019; Macnair and Brown, 2020; Sun et al., 2022). The LPS is constituted by three components: a lipidic part, known as lipid A, a conserved core polysaccharide and the O-specific polysaccharide (‘O-antigen’), which is variable and possesses repeating units of different structures that can be linear or branched (Raetz and Whitfield, 2002) (Figure 3A). LPS molecules Introduction 10 are cross-linked by divalent cations, which contribute to reinforce the OM structure and somehow compensates for the weakness of the thin PG layer of diderm bacteria (Konovalova et al., 2017; Sun et al., 2022). The anchoring of the OM to the PG has traditionally been thought to be carried out by the Braun’s lipoprotein (Lpp). However, recent research has revealed that Lpp is restricted to the Proteobacteria clade, whereas the vast majority of bacteria use alternative anchoring proteins; e.g., OmpA (see below) and OmpM in Gracilicutes and Terrabacteria, respectively (Witwinowski et al., 2022). The OM also acts as a permeability barrier and delimits the periplasm, where numerous reactions take place. The OM is impermeble to large molecules, such as some antibiotics, but allows the pass of small hydrophilic molecules such as sugars and amino acids, which traverse the OM through porins (Choi and Lee, 2019; Macnair and Brown, 2020). Porins are transmembrane proteins with a  -barrel structure, which forms a channel through which the substrate can traverse. Porins are classified in monomeric, dimeric o trimeric, according to their functional structure; and in non-specific or specific according to their range of substrates. OmpA is the most studied porin; it is anchored to PG by a non-covalent bond and has a role in maintenaince of the integrity of the cell envelope (Choi and Lee, 2019). Other substrates are transported through the OM in an energy-dependent manner by dedicated TonB-dependent transporters (Silale and van den Berg, 2023). 2.3. The cell wall Bacterial cell morphology is dictated and maintained by the PG layer that composes the cell wall. This structure, also called the sacculus, surrounds the plasma membrane (it is located in the periplasmic space in diderm bacteria) and provides mechanical strength that precludes cells from bursting by turgor pressure. As explained above, the cell wall can be multilayered, as in monoderm bacteria, or monolayered as in canonical diderm bacteria (see section 2 and Figure 3). The specific arrangement of the cell wall components provides this structure with strength as well as flexibility (Egan et al., 2020; Rohs and Bernhardt, 2021; Kumar et al., 2022). 2.3.1. Structure of the cell wall The PG layer is a giant mesh-like molecule whose structure is based on long chains of alternating N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) units linked by  -1,4 bridges. In rod-shaped bacteria, the glycan chains are oriented perpendicular to the longitudinal axis of the cell. Each MurNAc amino sugar is covalently linked to a pentapeptide that is crosslinked to the pentapeptide of an adjacent glycan chain, which actually confers to PG its meshlike structure (Figure 5). In diderm bacteria, the pentapeptide composition is usually L-Ala-  -D- Glu-mDAP-D-Ala-D-Ala, where mDAP is meso-diamino pimelic acid. In monoderm bacteria, the Introduction 11 amino acid at the second and third positions are usually D-Gln and L-Lys, respectively. Crosslinking commonly occurs between the fourth and the third amino acids of different peptides (4-3 cross-link) or between amino acids at the third positions (3-3 cross-link); the proportion of each type varying between species and as function of the cellular state (Rohs and Bernhardt, 2021; Aliashkevich and Cava, 2022; Kumar et al., 2022). For instance, in E. coli the most common (90-98%) cross-link is 4-3, whereas in Agrobacterium tumefaciens 30% of crosslinks are 3-3 (Morè et al., 2019; Williams et al., 2021). In E. coli, 3-3 cross-links increase when the outer membrane is compromised or during PG maturation in the stationary phase of growth (Glauner et al., 1988; Typas et al., 2012; Morè et al., 2019). UPD– MurF UPD– MurE UPD– MurD UPD– Fructose 6-P GlmS GlmM GlmU UPD– MurA MurB UPD– MurC Figure 5. Peptidoglycan synthesis pathway and its cellular compartmentalization. Cell wall precursors and components are represented according to the code shown at the bottom. Arrows represent enzymatic reactions, where either the specific enzyme or the enzyme family is indicated. GlcNAc, N-acetylglucosamine; GT, glycosyltransferase enzymes; mDAP, meso-diaminopimelic acid; MurNAc, N-acetylmuramic acid; TP, transpeptidase enzymes. See the text for further details. D-Ala mDAP P P Undecaprenol D-Glu MurNAc L-Ala GlcNAc TP GT Periplasm Plasma membrane P P MraY MurG Cytoplasm P P P P MurJ P UPD– Lipid I Lipid II Introduction 18 role in cell morphogenesis (Divakaruni et al., 2007; Egan et al., 2020). Its main function is to regulate PG synthesis directed by the elongasome complex. In many bacteria including E. coli, MreC is a bitopic protein with a transmembrane  -helix in the N-terminal part and an extensive periplasmic domain, although in C. crescentus it is a soluble periplasmic protein (Divakaruni et al., 2005; 2007; Kruse et al., 2005). MreC is organized into patches of short filaments that colocalize with MreB and interact with periplasmic synthetic enzymes including PBP2 (Divakaruni et al., 2005; 2007; Contreras-Martel et al., 2017; Martins et al., 2021). In Pseudomonas aeruginosa, it has been shown that MreC self-associates in a head-to-tail fashion through interaction interfaces that are conserved in proteobacteria, alternating between self-association and interaction with the PG-synthesis machinery (Martins et al., 2021; see Rohs and Bernhardt, 2021; Xu et al., 2022). It has been suggested that the transition between monomeric and polymeric states could represent a regulatory mechanism for MreC function (see below) (Rohs et al., 2021). 3.1.3. MreD The mreD gene is usually the last gene of the mreBCD operon. This gene is not always conserved; hence, some rod-shaped bacteria contain an mreBC operon. When present, MreD is involved in rod shape maintenance (Kruse et al., 2005; Leaver and Errington, 2005; den Blaauwen et al., 2008; Typas et al., 2012; Egan et al., 2020; Liu et al., 2020). It is a PM integral protein with five or six predicted transmembrane α-helices flanked by N- and C-terminal extensions located in the cytoplasm (Kruse et al., 2005; Leaver and Errington, 2005). In vitro, MreD interacts with MreC and PBP2, but not with MreB or itself (Kruse et al., 2005; Liu et al., 2020). 3.1.4. RodZ The rodZ gene, which was previously named yfgA, encodes a bitopic membrane protein that is widely conserved in rod-shaped bacteria, but is non-essential in E. coli (Newitt et al., 1999; Alyahya et al., 2009; Bendezú et al., 2009; Ago and Shiomi, 2019; Egan et al., 2020). Deletion of rodZ produces spherical cells with altered curvature and shape, a phenotype that can be supressed by high doses of the cell division protein FtsZ (Bendezú et al., 2009; van den Ent et al., 2010; Bratton et al., 2018; Liu et al., 2020). RodZ is thought to act as a scaffolding protein for the elongasome, being able to interact with itself and with MreB, MreC, RodA and PBP2 (Bratton et al., 2018; Egan et al., 2020; Liu et al., 2020; Özbaykal et al., 2020). The RodZ N-terminal domain is cytoplasmic and contains a helix-turn-helix motif for interaction with MreB, whereas the C- terminal domain is periplasmic and interacts with the PG synthesis machinery (van den Ent et al., 2010; Bratton et al., 2018; Rohs and Bernhardt, 2021). RodZ binds to monomeric and polymeric MreB and regulates MreB polymerization. In E. coli, the interaction with RodZ Introduction 19 contributes to anchor MreB to the PM and modulates its physical properties, influencing the curvature of MreB protofilaments (van den Ent et al., 2010; Typas et al., 2012; Morgenstein et al., 2015; Bratton et al., 2018; Shi et al., 2020). In vivo, the interaction of RodZ with MreB does not require the presence of other proteins, suggesting that RodZ may represent a link between MreB and other elongasome components (van den Ent et al., 2010; Egan et al., 2020; Özbaykal et al., 2020; Rohs and Bernhardt, 2021). Finally, in E. coli, in addition to being part of the elongasome, RodZ has been proposed to have a role in cell division, facilitating the polymerization of the division ring, which is dependent on the interaction of its cytoplasmic domain with FtsZ (Ago and Shiomi, 2019; Yoshii et al., 2019). 3.1.5. RodA and PBP2 In E. coli, RodA and PBP2 form the catalytic subcomplex of the elongasome (Liu et al., 2020; Mueller and Levin, 2020). The rodA gene forms an operon with the upstream pbpA gene encoding PBP2 (Matsuzawa et al., 1989). RodA was identified because its mutations produced round cells (de Pedro et al., 2001; Figge et al., 2004). It is an integral membrane protein, with multiple transmembrane domains, essential and widely conserved in rod-shaped bacteria (Kruse et al., 2005; Leaver and Errington, 2005; van den Ent et al., 2010; Mohammadi et al., 2011; Egan et al., 2020). Besides with PBP2, RodA also interacts with MreB and with itself, whereas PBP2 interacts with itself and with PBP1A, MreC and MreD (Egan et al., 2020; Liu et al., 2020; Chaudhary et al., 2021). In addition to the RodA-PBP2 pair, aPBPs may play a role in lateral PG synthesis (see paragraph 2.3.3.2. above). In both, monoderm and diderm bacteria, aPBPs interact with proteins of the elongasome, as MreB in B. subtilis or PBP2 in E. coli, although in both models aPBPs do not rotate with the elongasome and their inactivation affects neither PG synthesis by the elongasome nor the elongasome dynamics (Cho et al., 2016; Meeske et al., 2016). 3.2. Regulatory role of MreC and MreD in PBP2-RodA activity Recent data indicate that in T. thermophilus and E. coli, PBP2 oscillates between an inactive and an active state in which the so-called pedestal domain is in its open conformation. In the RodA- PBP2 subcomplex, PBP2 in its open conformation interacts with RodA, promoting activation of its GT activity (Sjodt et al., 2020). Interestingly, in the co-crystal of MreC and PBP2 from H. pylori, MreC interacts with the pedestal domain of PBP2 leading to an opening of the two pedestalinteracting subdomains (Contreras-Martel et al., 2017; Rohs et al., 2018). All these results have led to a current model in which the interaction of MreC with PBP2 changes the conformation of the latter, which in turn is communicated to RodA to stimulate PG synthesis by the elongasome (Contreras-Martel et al., 2017; Rohs et al., 2018; Rohs and Bernhardt, 2021) (Figure 6C). For its Introduction 20 part, MreD has been described as a negative regulator of the activity of the elongasome by interaction with MreC and PBP2, preventing the conformational change of the latter and, therefore, the activation of RodA and PG synthesis (Liu et al., 2020). Hence, MreC and MreD would play opposite roles in modulation of the interaction between the components of the bPBP-SEDS pair of the elongasome (Liu et al., 2020; Rohs et al., 2021). 3.3. Relationships of endopeptidases with the elongasome Endopeptidase enzymes that cleave the crosslinks in the PG and could function in association with the elongasome have been identified. However, the coordination between these enzymes and the elongasome is not yet well defined. In diderm bacteria, little is known about endopeptidases, with some data available in E. coli and Vibrio cholerae (Rohs and Bernhardt, 2021). By contrast, in monoderm bacteria like B. subtilis, the LytE and CwlO endopeptidases have been studied in depth in regard to their regulation and connection with the elongation machinery (Rohs and Bernhardt, 2021). LytE appears to act in concert with the MreB isoform, whereas CwlO have an apparent relationship with Mbl (Carballido-López et al., 2006; Domínguez-Cuevas et al., 2013; Meisner et al., 2013). However, the simultaneous inactivation of CwlO and LytE does not affect the circumferential movement of the elongasome, illustrating that many aspects concerning the role of PG hydrolytic enzymes in cell wall elongation still need clarification (Meisner et al., 2013). 3.4. Elongasome dynamics In vivo, MreB form short filaments, with disperse localization in the PM, that rotate uncoordinately, and often in opposite orientations, perpendicular to the long axis of the cell (Dominguez-Escobar et al., 2011; Garner et al., 2011; van Teeffelen et al., 2011). The other components of the elongasome also rotate in the same direction with a similar speed, indicating that it is the whole complex that rotates. The observation that inhibition by antibiotics, or by depletion or mutation of PG synthesis enzymes provoked cessation of elongasome movement led to the notion that PG synthesis fuels the motion of the entire elongasome complex (Dominguez-Escobar et al., 2011; Garner et al., 2011; van Teeffelen et al., 2011). This raises the question on the role of MreB in the elongasome. A critical point being that synthesis of new PG must occur in a direction perpendicular to the long cell axis, the emerging picture is that MreB governs the orientation of the elongasome, ensuring that its rotation, and thus PG synthesis, occurs in the correct direction. Crucial for this is the intrinsic curvature of MreB filaments, which makes energetically favorable their alignment along the greatest principal membrane curvature, which in rod-shaped bacteria occurs around the rod circumference (Hussain et al., 2018; Wong et al., 2019; Rohs and Bernhardt, 2021) (Figure 6A). Introduction 21 D C F B t= 90 s t= 180 s t= 0 Single molecule Filament E 4. THE DIVISOME Cell division is a fundamental process governed by a multiprotein complex known as the divisome that involves numerous components, including scaffolding proteins, regulatory factors and PG synthesis enzymes. This complex coordinates chromosome segregation with the formation of the new poles of the resulting daughter cells, envelope constriction and daughter cell separation. In diderm bacteria, the divisome expands the three layers of the cell envelope, so that the dynamics of the OM is coordinated to the re-modelling of the PG and the constriction of the PM (Du and Lutkenhaus, 2017; Barrows and Goley, 2021; Levin and Janakiraman, 2021). The structure and function of the divisome has been studied mainly in E. coli and B. subtilis, although in recent years a good deal is being learned on other model bacteria. Figure 7. FtsZ, the initial protein of the divisome. (A) Linear representation of FtsZ with its different domains depicted as lines or cylinders. C-terminal linker (CTL), C-terminal peptide (CTP), C-terminal variable region (CTV). Note that the CTV is included in the CTP domain. (B) Crystal structure of FtsZ from S. aureus in its open and closed conformations. (C) Model of a single-stranded FtsZ filament from E. coli. Note that subunits are in their open conformation. (D) Scheme of the cooperativity between subunits during FtsZ polymerization. Association of monomers in a closed state promote a conformational change to an open state that fosters further recruitment of subunits with increased affinity. (E) Scheme of the movement of the divisome driven by the treadmilling dynamics of FtsZ. (F) Representation of the apparent rotation of FtsZ-GFP fusion proteins. Note that FtsZ subunits are static during treadmilling. (B) and (C) adapted from Wagstaff et al., 2017; (E) adapted from Lamanna & Maurelli, (2022); (F) adapted from Bisson-Filho et al., (2017). Globular core CTL CTP CTV C A N Introduction 22 4.1. FtsZ structure and dynamics One of the main proteins of the divisome is FtsZ, which is conserved in bacteria, chloroplasts and some archaea and mitochontria (Levin and Janakiraman, 2021). FtsZ is homologous to the eukaryotic cytoskeletal protein tubulin, showing a similar three-dimensional structure in the common core domain (Löwe and Amos, 1998). Much of the divisome functioning depends on the physico-chemical properties and the dynamics of FtsZ. FtsZ has been mainly studied in E. coli and B. subtilis and it has been shown to contain five domains (Levin and Janakiraman, 2021) (Figure 7A). The N-terminal part shows a variable length depending on the species, ranging from 10, in E. coli, to 68 amino acids in some cyanobacteria (see below) (Corrales-Guerrero et al., 2018; Levin and Janakiraman, 2021). The next domain is a tubulin-like core with GTP binding and hydrolytic activity (Schumacher et al., 2020). The C- terminal linker (CTL) is the domain that binds the proteins that anchor FtsZ to the membrane (Buske and Levin, 2012; Levin and Janakiraman, 2021). Finally, the C-terminal peptide (CTP) is widely conserved and interacts with some proteins that participate in cell division. CTP comprises the domain known as the C-terminal variable region (CTV), which varies in length according to species and seems to mediate self-interactions between FtsZ protofilaments (Buske and Levin, 2012; Ortiz et al., 2015). FtsZ is a soluble protein that in vitro polymerizes in different forms (single-stranded filaments, sheets, bundless or rings) depending on buffer conditions. However, it does not polymerize with the typical microtubular structure of tubulin. In vivo, FtsZ is a cytoplasmic protein that at the initiation of cell division gets anchored to the PM by interaction with a variety of PM-located proteins, such as FtsA and ZipA (see below), which also promote FtsZ polymerization into filaments near the inner layer of the PM. Filaments eventually aggregate into a structure known as the Z-ring that encicles the cell around the circumference of the division plane and acts as a scaffold for the recruitment of further divisome proteins. FtsZ polymerization is spatiotemporally regulated and occurs at specific times in the cell cycle prior to cell division (Bramhill and Thompson, 1994; Du and Lutkenhaus, 2017; Levin and Janakiraman, 2021). Z-ring assembly requires a threshold concentration of FtsZ protein above its average cytoplasmic concentration, which is otherwise very stable. Local variations of the FtsZ concentration are observed due to crowding of FtsZ molecules, particularly at the future septal site. The increase of FtsZ concentration and the subcellular localization of Z-ring assembly are key events in cell division, and these events are extensively regulated by both negative (e.g., the Min and nucleoid occlusion systems) and positive (e.g., Ter-MatP-linkage) regulatory circuits (Si et al., 2019; Barrows and Goley, 2021; Levin and Janakiraman, 2021). FtsZ polymerization relies on its dynamic oscillation between two conformational states, a closed state, commonly adopted in individual monomers, and an open state that favours interaction between monomers and is commonly found in the subunits of filaments (Figures 7B and 7C). Introduction 23 As it is the case with tubulin, FtsZ polymerizes by head-to-tail interactions between subunits and, importantly, polymerization is associated to a phenomenon of cooperativity such that, whereas individual FtsZ monomers in the closed conformation show little affinity for each other, once polymerization has started binding of new monomers occurs with higher affinity (Figure 7D) (Michie and Löwe, 2006; Wagstaff et al., 2017). Also similar to tubulin, FtsZ filaments exhibit ‘treadmilling’ dynamics, which consists in the polymerization of FtsZ GTP-bound monomers at one end of the filament (plus end) and simultaneous dissociation of subunits at the other end (minus end) (Mukherjee and Lutkenhaus, 1998; Loose and Mitchison, 2014; Bisson-Filho et al., 2017; Yang et al., 2017; Ramirez-Diaz et al., 2018). Treadmilling of FtsZ depends on its GTPase activity, which active site is assembled from two consecutive subunits in the filament (Scheffers et al., 2002; Oliva et al., 2004). This way, recently incorporated subunits maintain for some time its GTP-bound state at the plus end of filamens. Hydrolysis of GTP favor the conformational change of the subunits to a closed state, which is prone to dissociation. Treadmilling give rise to the apparent rotational movement of FtsZ-GFP fusions observed in vivo; however, it has been shown that individual subunits in the filaments remain static (Figure 7E). This dynamic property of some cytoskeletal proteins is useful to direct the movement of accompanying molecules without the action of motor proteins. Indeed, in E. coli and B. subtilis, FtsZ has been shown to direct the movement of the divisome complex to ensure the correct building of the septum. It has been propossed that the FtsZ GTPase activity and treadmilling affect the spatiotemporal distribution of PG synthesis to initiate constriction, rather than the PG synthesis rate (Yang et al., 2017; Whitley et al., 2021). In B. subtilis, FtsZ treadmilling also appears essential for the initial condensation of the Z-ring (Whitley et al., 2021). It should be pointed out that, notwithstanding the conservation of the basic composition of the divisome in most bacterial species, whether FtsZ treadmilling is always required for the localization and dynamics of the septal PG synthesis machinery is a matter of current debate. Thus, in contrast to its essential role in B. subtilis, in S. pneumoniae the constriction mechanism is independent of FtsZ, being instead directed by the TG and TP activities of the PG synthesis machinery. There are also organisms with a biphasic division process, whereby a first part is dependent on treadmilling and a second part independent of it. S. aureus and E. coli have been described to show this biphasic behavior, although in the latter this issue is currently under debate (Lamanna and Maurelli, 2022). 4.2. Divisome assembly The E. coli divisome is composed of 36 proteins, among which a dozen are essential and widely conserved in bacteria. Divisome components are found in the cell in almost constant amounts, but concentrate at the site of the future septum for divisome assembly (Du and Lutkenhaus, 2017; Levin and Janakiraman, 2021). The assembly of the divisome is considered to take place Introduction 24 FtsZ, FtsA, ZipA Proto-ring FtsEX FtsK FtsQ FtsLB Late proteins FtsWI FtsN B C D PM PM stepwise, with an early phase, where the Z-ring is formed by FtsZ polymerization and tethering to the PM, and a later phase where the rest of the proteins join to organize the divisome. A Figure 8. Assembly and activation of the divisome and envelope constriction. (A) The proposed linear hierarchical pathway for divisome assembly is depicted, with arrows separating sequential steps. (B) Scheme of the regulatory events that promote divisome activation. Arrows connect the components involved, and numbers indicate the temporal order of events. (C) Scheme of peptidoglycan remodeling at the septum. The action of PG hydrolases, including PG amidases cleaving septal peptidoglycan (sPG) and producing denuded PG is depicted in the inset. Green and dashed arrows indicate activation by denuded glycans. (D) Representation of the proteins required for constriction of the cell envelope and separation of the daughter cells, including PG amidases and the Tol-Pal complex. (A) adapted from Attaibi & der Blaauwen, (2022); (B) from Tsang & Bernhardt, (2015); (C) adapted from Navarro et al., (2022); (D) adapted from Levin & Janakiraman, (2021). Many data about the order in which proteins are assembled in the divisome derive from genetic approaches involving the inactivation of one of the components, which permits the identification of proteins acting upstream and downstream of the inactivated one. These studies led to the proposal that the second phase of divisome assembly follows a linear hierarchical pathway (Figure 8A) (Attaibi and Den Blaauwen, 2022). However, the operation of a sequential assembly is currently under disscusion. Some groups have shown that forced premature incorporation of some late proteins promotes the recruitment of earlier components, a Introduction 25 phenomenon known as back recruiting, which speaks of a more complex pathway (Goehring and Beckwith, 2005; Goehring et al., 2005; Du and Lutkenhaus, 2017). The emerging picture is that the divisome assembles following aproximately the proposed linear hierarchical pathway, but with the occurrence of dynamic interactions between some components, which changes during the assembly process. In addition, some proteins can fulfill different functions at different stages of assembly (see e.g. FtsA or FtsEX below) (Du and Lutkenhaus, 2017; Barrows and Goley, 2021). A brief description of the components of the E. coli divisome is shown in Table 2. FtsA is an actinlike protein with ATPase activity, linked to the PM by a C-terminal amphipatic helix (Pichoff and Lutkenhaus, 2005; Pichoff et al., 2012; Szwedziak et al., 2012), and ZipA is a bitopic protein with a large cytoplasmic domain. Both, FtsA and ZipA, bind to the CTP domain of FtsZ to anchor the FtsZ protofilaments to the PM. Both proteins are necessary for the recruitment of downstream proteins, but FtsA has a broader role, being also responsible for regulation of the divisome activity (Liu et al., 2015; Tsang and Bernhardt, 2015; Du and Lutkenhaus, 2017). FtsA can be found in two conformational states, but only in its monomeric state it is active for downstream protein recruitment (Du and Lutkenhaus, 2017). FtsE and FtsX form a FtsEX subcomplex homologous to ABC transporter components, where FtsE is an ATPase-like protein and FtsX a transmembrane protein. This subcomplex fulfils at least two functions: regulation of FtsA polymerization and activation of PG amidases (see below) to separate daugher cells (Arends et al., 2009; Du et al., 2016; Attaibi and Den Blaauwen, 2022). FtsK is a DNA translocase that pumps DNA from the septal zone (Chen and Beckwith, 2001; Männik et al., 2017). FtsQ, FtsL and FtsB are three bitopic proteins that form a subcomplex whose main function is to recruit the PG synthases FtsW and FtsI (PBP3) (Du and Lutkenhaus, 2017). As mentioned above, FtsW is a glycosyl transferase of the SEDS family and FtsI is its partner bPBP transpeptidase. The last protein to join the divisome is FtsN (SPOR domain), which triggers PG synthesis and septum constriction (Levin and Janakiraman, 2021). Introduction 26 Table 2. Essential components of the E. coli divisome. Protein (complex) Incorporation timing Localization Comments Role(s) References FtsZ Early 1. Cell division initiation C Tubulin-like GTPase 2. Guiding septal PG synthesis Bisson-Filho et al., (2017) Levin and Janakiraman, (2021) FtsA Early 1. Membrane anchor 2. Recruitment of downstream C Actin-like proteins 3. Regulation of PG synthesis activity Pichoff et al., (2012) Du and Lutkenhaus, (2017) ZipA Early PM / C 1. Membrane anchor Levin and Janakiraman, (2021) FtsEX Mid time 1. FtsA oligomerization regulation ABC transporter-like 2. Septal cell wall hydrolysis PM / C ATPase and control. membrane subunits 3. EnvC regulation Arends et al., (2009) Du et al., (2016) FtsK Late 1. Exclusion of DNA from the PM / C DNA translocase division site 2. Linkage of downstream proteins to the Z-ring Chen and Beckwith, (2001) Männik et al., (2017) FtsQLB Late PM / P 1. FtsW and FtsI regulation Du and Lutkenhaus, (2017) FtsW Late PM Glycosyl transferase 1. PG synthase Levin and Janakiraman, (2021) FtsI Late PM / P Transpeptidase 1. PG synthase Levin and Janakiraman, (2021) FtsN Late PM / P C- terminal SPOR 1. PG synthesis regulation domain Du and Lutkenhaus, (2017) PM, plasma membrane; C, cytoplasmic; P, periplasmic. 4.3. Divisome activation and function In diderm bacteria, the constriction of the septum initiates once the last essential divisome protein, FtsN, is recruited. PG remodelling involves coupled synthesis of new PG close to the PM and degradation of the outermost layer by PG hydrolases to allow OM invagination. At this step, PG degradation and new synthesis are tightly coordinated, which is necessary to prevent deleterious events. Among the PG hydrolases, the amidases AmiA and AmiB, which break the stem peptides of the glycan chains of the PG, have main roles in septal PG splitting (Figures 8B and 8C). These amidases are strictly regulated by the divisome factor EnvC. EnvC is anchored to the periplasmic side of the PM and its C-terminal domain, which is involved in amidase activation, contains the so-called restraining arm. This part is subjected to an autoinhibitory mechanism that is released upon interaction of the EnvC N-terminal coiled-coil domain with the periplasmic domain of FtsX (Cook et al., 2020). Such interaction is fostered by the ATPase activity of FtsE in the cytoplasm, which promotes a conformational change in the periplasmic side of FtsX (Yang et al., 2011; Tsang et al., 2017; Cook et al., 2020; Shaku et al., 2020), with the FtsE ATPase activity being in turn regulated by FtsX (Heidrich et al., 2001; Bernhardt and de Boer, 2004; Uehara et al., 2009; Yang et al., 2011). The amidase activity produces denuded glycans (PG strands stripped of the peptide Introduction 27 stems), which are recognized by the periplasmic SPOR domain of FtsN, which then promotes the FtsQLB subcomplex to be in an ON state able to activate the PG synthases FtsW and FtsI. In this way, it is ensured that new PG can be added at the points of hydrolysis. 4.4. Constriction mechanism During constriction, simultaneous and coordinated events occur in the three layers of the cell envelope, as the PM undergoes constriction while PG is remodelled in the periplasm and the OM invaginates at the septal site (Du and Lutkenhaus, 2017; Levin and Janakiraman, 2021; Navarro et al., 2022). Coordination is achieved in several ways. For instance, while the AmiA and AmiB amidases are regulated by the divisome from the PM, the activity of a third amidase, AmiC, is regulated by the OM protein NlpD (Tsang et al., 2017). Besides, the Tol-Pal system, a large complex that traverses the cell wall connecting the OM to the PM, activates PG remodelling enzymes (Figure 8D) and provides a physical linkage that couples invagination of the OM to the PM constriction (Petiti et al., 2019; Szczepaniak et al., 2020; Yakhnina and Bernhardt, 2020). Traditionally it has been thought that the constriction force was provided by FtsZ dynamics, but recent data have shown that in B. subtilis and S. aureus FtsZ disassembles before the separation of the daughter cells, so at least the last stages of constriction appear not to be dependent on FtsZ (Söderström et al., 2014; Whitley et al., 2021). However, early dissasembly of FtsZ does not seem to occur in all species (Lamanna and Maurelli, 2022). In some cases, the driving force for constriction is provided by PG synthesis in the septal zone, which pushes the PM towards the cell center (Coltharp et al., 2016; Monteiro et al., 2018). Once cytokinesis is completed, the divisome disassembles following approximately the reverse pathway of that followed for assembly, except that FtsN remains in place after cytokinesis has completed (Söderström et al., 2016). 5. MULTICELLULARITY IN BACTERIA Classically, bacteria have been considered and studied as unicellular organisms, but multicellularity has emerged during evolution in all three Domains of life (Figure 9A), even with losses and secondary gains of this character (Shapiro, 1988; Bonner, 1998; Carroll, 2001; Grosberg and Strathmann, 2007). Multicellularity is a form of social organization whereby cells cooperate to favor survival of the ensemble instead of the individual cells. This relies on a coordinated behavior of cells within the multicellular entity, which is based on cell-cell adhesion, intercellular communication and, in some cases, cell differentiation that allows a division of labor between different cell types (Flores and Herrero, 2010; Lyons and Kolter, 2015). Multicellularity can be achieved in two ways: by aggregation of individual cells or by incomplete cell division that give rise to cell chains or to the formation of other types of aggregates (Bonner, Introduction 34 the uptake of sugars such as glucose by cyanobacteria could not only have a nutritional role but be also important for the establishment of symbiosis with plants, as described for Nostoc punctiforme (Ekman et al., 2013; Picossi et al., 2013). 6.3.3. Nitrogen assimilation Nitrogen constitutes 5-10% of the dry weight of the cyanobacterial cells, which preferably use inorganic nitrogen, as ammonium or nitrate, and can also use nitrite, urea, cyanate and some amino acids as nitrogen sources (Stanier and Cohen-Bazire, 1977; Fay, 1992; Flores and Herrero, 1994; Herrero et al., 2001). In addition, some cyanobacteria of different taxonomic groups can perform N2 fixation (Rippka and Waterbury, 1977; Bergman et al., 1997). Nitrogen fixation is a costly process in terms of energy and resources and it is easily inactivated by low concentrations of oxygen (Fay, 1992). Therefore, different cyanobacterial strains have developed different strategies to segregate nitrogen fixation from the oxygen produced by photosynthesis, as will be detailed below (see 6.4.2.1). The diversity of nitrogen sources that cyanobacteria can assimilate are metabolized to generate intracellular ammonium, which is incorporated into carbon skeletons by the glutamine synthethase-glutamate synthase (GS-GOGAT) cycle (Figure 11) (Stanier and Cohen-Bazire, 1977; Fay, 1992). 2-OG is the molecule that connects the pathways for nitrogen and carbon assimilation. Its intracellular levels depend on the available nitrogen and carbon sources, being highest in the absence of ammonium and the presence of abundant carbon. Indeed, the 2-OG levels signal the C-to-N balance of the cell (Muro-Pastor et al., 2001; Vázquez-Bermúdez et al., 2003). In cyanobacteria, the assimilation of nitrogen is tightly regulated. For most strains, ammonium is the preferred nitrogen source and, when available, it provokes a negative regulation of the pathways for the assimilation of other nitrogen sources. NtcA is a global transcriptional regulator, of the Crp/CAP (catabolite gene activator protein) family, of universal distribution in cyanobacteria. It regulates the expression of multiple genes involved in nitrogen assimilation and other cellular functions, acting as an activator or as a repressor depending on the position of its DNA binding site relative to the regulated genes (Herrero et al., 2001; 2004; Picossi et al., 2014; Giner-Lamia et al., 2017). NtcA binds 2-OG, which is indeed required for NtcA-dependent transcription activation (Valladares et al., 2008; Zhao et al., 2010). The PII protein also has an important role in nitrogen regulation in cyanobacteria (Flores et al., 2005). It is a widespread protein that senses the C-to-N balance, also binds 2-OG, and the energetic status of the cell (see Herrero et al., 2019). In Synechococcus sp. PCC 7942, PII sequesters PipX, another protein involved in nitrogen regulation, which interacts with NtcA stimulating its activity (Burillo et al., 2004; Forchhammer and Selim, 2019). However, PII interacts with PipX only in its 2-OG-free form. Thus, PII and PipX modulate the NtcA activity in a 2-OG-dependent manner (Espinosa et al., 2006; Llácer et al., 2010). In Anabaena, PipX plays a specific role in the activation Introduction 35 RuBisCO of NtcA-dependent promoters activated during the differentiation of the heterocysts (Valladares et al., 2011). Figure 11. Schematic of the pathways for nitrogen and carbon assimilation in Anabaena sp. PCC 7120. Uptake of nitrogenous compounds by Nrt, Urt and Amt transporters, and of inorganic carbon by the NDH-1 systems, in the thylakoid membrane, and BicA, SbtA and Cmp is represented. Intracellular nitrate reduction depends on reduced ferredoxin and is catalyzed by nitrate reductase (NarB) and nitrite reductase (NirA). Urea reduction is mediated by the urease (UreABC) complex. Ammonium is incorporated to the GS/GOGAT pathway, which also incorporates 2-oxoglutarate (2-OG). In the carboxysome, bicarbonate is converted to CO2 by carbonic anhydrase (CA) and incorporated into organic compounds by RuBisCO, producing 3-phospho-glycerate (3-PG) that is catabolized with the production of 2-OG. The orange line represents the plasma membrane. Figure adapted from Herrero and Flores, (2019). 6.4. Multicellularity in cyanobacteria 6.4.1. Intercellular communication As mentioned above, multicellularity relies on cell aggregation and intercellular communication. In filamentous cyanobacteria, the cells are held together by a continuous outer membrane, and PG sacculus, and by septal proteins that bind contiguous cells in the filament (Mariscal et al., 2007; Mullineaux et al., 2008). The continuity of the OM conforms a continuous filament periplasm, which is also functionally continuous and allows the movement through it of molecules of between 25 and 50 kDa in size (Mariscal et al., 2007; Haselkorn, 2008; Zhang et al., 2008; Flores and Herrero, 2010). Intercellular exchange though this compartment requires the export and import of molecules involving PM transporters (Picossi et al., 2005; López-Igual et al., 2010; Pernil et al., 2010; Nieves-Morión et al., 2017). In order for the periplasm to be a conduit for efficient intercellular exchange, the OM should represent an efficient permeability barrier, to avoid leakage of molecules to the extracellular milieu. In Anabaena, this property has been demonstrated for the molecules glutamate and sucrose (Nicolaisen et al., 2009b). Introduction 36 A PM PM A second path for intercellular communication is represented by structures known as septal junctions (SJs), which are proteinaceous complexes that traverse the septum between neighboring cells in the filament connecting their cytoplasms (Figure 12). Although these connections were already observed decades ago, their structure has been resolved recently (Kieninger and Maldener, 2021). Molecular exchange between cells have been evidenced mainly through the monitorization of low MW fluorescent tracers such as calcein, 5-carboxyfluorescein or esculin. A number of proteins including SepJ, FraC, FraD and FraE have been identified to localize at the intercellular septa and to be involved in cell-cell adhesion and intercellular molecular exchange (Mullineaux et al., 2008; Flores and Herrero, 2010; Merino-Puerto et al., 2010; Merino-Puerto et al., 2011; Nürnberg et al., 2015). However, only FraD and SepN have been demonstrated to be structural constituents of SJs (Weiss et al., 2019; Kieninger et al., 2022). Communication through SJs occurs by diffusion and is regulated by gating (Weiss et al., 2019; Arévalo et al., 2021). B Figure 12. Structure of septal junctions (SJs). (A) Cryotomography of the septum between two Anabaena cells. Arrows indicate SJs crossing the septum. Plasma membrane (PM), outer membrane (OM), thylakoid membrane (TM), phycobilisome (PB), peptidoglycan (PG). (B) Schematic of SJs. i) Modules of the complex: tube, plug and cap; ii) 3D representation of open (left) and closed (right) conformations. (A) adapted from Weiss et al., (2019); (B) adapted from Kieninger and Maldener, (2021). 6.4.2. Cell differentiation: the heterocyst When confronted with conditions of nitrogen scarcity, cyanobacteria of sections IV and V of Rippka’s classification can differentiate heterocysts, which are cells specialized for the fixation of atmospheric nitrogen in oxic environments (Rippka et al., 1979; Murry and Wolk, 1989; no stress stress ii) i) Introduction 37 Corrales-Guerrero et al., 2013). Some of those cyanobacteria can also differentiate other specialized cells like resistance spores, known as akinetes, or hormogonia, short motile filaments involved in the dispersal in natural environments and in host colonization for the establishment of symbiotic associations (Flores and Herrero, 2010). 6.4.2.1. Heterocyst structure and function Heterocysts allow the spatial separation of the incompatible processes of oxygenic photosynthesis and N2 fixation, which is catalyzed by the enzyme nitrogenase, a large metalloenzymatic complex highly sensitive to oxygen. The structure and metabolism of the heterocyst present remarkable differences in comparison to the original cell from which it differentiates, aimed to maintain intracellular microoxic conditions and generate the higher input of ATP and reductants required by the nitrogenase activity (see e.g., Kumar et al., 2010; Zeng and Zhang, 2022). Heterocysts are easily recognizable by their larger size and elongated shape and the presence of refringent granules at the cell poles (Wildon and Mercer, 1963) (Figure 13A). Active PG synthesis allows cell elongation and enlargement during differentiation (Zhang et al., 2018). The poles of the heterocysts are differentiated structures, known as the heterocyst “neck”, which reduces the contact surface with the adjacent vegetative cell, and hence intercellular exchanges. Indeed N2 enters the heterocyst mainly through the neck (Walsby, 2007). The polar granules are made of cyanophycin, a polymer of arginine and aspartic acid resulting from the nitrogen fixation activity (Lang et al., 1972; Flores and Herrero, 2010). Other structural characteristics of heterocysts are the absence of carboxysomes and the reorganization of the thylakoid membranes, which convolute and accumulate around the cell poles forming the so-called “honeycomb” membranes (Valladares et al., 2007; Flores and Herrero, 2010). During heterocyst differentiation, additional layers of the cell envelope are deposited, which restrict the entrance of gasses, O2 but also N2, into the cell cytoplasm (Murry and Wolk, 1989; Fay, 1992). These include an inner laminated layer of glycolipids (HGL), which functions as a gas permeability barrier, and a thicker and more homogeneous outer layer of polysaccharides (HEP), which protects the HGL from physical damage (Cardemil and Wolk, 1979; Kumar et al., 2010; Flores and Herrero, 2010). Both HGL and HEP are engrossed around the cell poles (Nicolaisen et al., 2009a; Herrero et al., 2016). Regarding metabolic differentiation, the endogenous production of oxygen is abrogated by the loss of PSII, photosynthesis being restricted to cyclic electron flow through PSI (Magnuson and Cardona, 2016). Honeycomb membranes show a high respiratory activity, due to the presence of specific cytochrome oxidases, that readily consume free oxygen that may enter the heterocyst from neighboring vegetative cells (Valladares et al., 2007; Walsby, 2007). Heterocysts have lost the ability to fix CO2 due to the loss of the RuBisCO enzyme and carboxysomes. This adaptation Introduction 38 PM PM Honeycomb Neck creates the necessity of receiving fixed carbon from adjacent vegetative cells, which transfer to the heterocyst glutamate, alanine and sucrose (Winkenbach and Wolk, 1973; Valladares et al., 2007; Flores and Herrero, 2010; Herrero et al., 2016). The GOGAT enzyme is also absent in heterocysts. Hence, glutamine produced by GS is exported to vegetative cells (Wolk et al., 1976; Thomas et al., 1977; Martín-Figueroa et al., 2000). Heterocysts also transfer to the adjacent cells  -aspartyl-arginine dipeptide, which derives from the hydrolysis of cyanophycin (Burnat et al., 2014). Given the metabolic compartmentalization, molecular exchange between heterocysts and vegetative cells is an absolute requirement for the operation and survival of the diazotrophic filament. Net transfer of carbon compounds from vegetative cells to heterocysts, and of nitrogen compounds in the opposite direction, ensures the correct nutrition of all cells in the filament. NanoSIMS (nanometer-scale secondary ion mass spectrometry) technology has recently helped to monitor this exchange, showing that vegetative cells are progressively more nitrogen limited as the distance from the nearest heterocysts increases (Popa et al., 2007; Kumar et al., 2010; Herrero et al., 2016; Nieves-Morión et al., 2021). Molecular transfer between heterocysts and vegetative cells take place through SJs (see 6.4.1), although the flow rate is slower than that between two vegetative cells, and the SJ array also differ (Nicolaisen et al., 2009a; Arévalo and Flores, 2021) (see section 7 below). The presence of the cyanophycin granules located at the poles of the heterocysts appears to represent a factor limiting transfer (Mullineaux et al., 2008). A B Figure 13. Heterocyst ultrastructure and regulatory circuit for heterocyst differentiation. (A) Transmission electron microscopy image and schematic of a heterocyst (left) and a vegetative cell (right). The main structures of the heterocyst are indicated: cyanophycin granule polypeptide (CGP), glycolipid (HGL) and polysaccharide (HEP) layers of the cell envelope. (B) Sequence of regulatory events during heterocyst differentiation indicating the the participation of the transcription factors NtcA and HetR, and their mutual activation. The inhibitory signaling peptides derived from PatS and HetN, produced in the differentiating cells and mature heterocysts, respectively, are transferred to inhibit differentiation of adjacent cells. PatA is a positive regulator of differentiation. Transfer of carbon compounds from vegetative cells to the heterocyst and of nitrogen compounds in the opposite direction is also indicated. Adapted from Flores and Herrero, (2010). Introduction 39 6.4.2.2. Heterocyst differentiation The process of heterocyst differentiation is triggered by the perception of nitrogen scarcity. In different strains the pattern of heterocysts distribution along the filament can differ, with some species showing heterocysts only at the filament ends and others showing both terminal and intercalary heterocysts. The regulatory mechanisms controlling heterocyst differentiation has been particularly well characterized in Anabaena, whereby heterocysts are separated by tracts of ca. 10-12 vegetative cells (Flores and Herrero, 2010) (Figure 13B). Differentiation involves the sequential activation and repression of multiple genes directed by two wide-range transcription factors: NtcA, the global nitrogen regulator, and the heterocyst-differentiation factor HetR. In addition, other regulators, including sRNAs, participate in specific steps (Buikema and Haselkorn, 1991; Flores and Herrero, 2010; Flaherty et al., 2014; Herrero et al., 2016; Brenes-Álvarez et al., 2019; Xu et al., 2020; Zeng and Zhang, 2022). Under conditions of low nitrogen availability, intracellular levels of 2-OG increase and activate NtcA, which in turn is required for the expression of the nrrA gene. The NrrA factor represents a link between NtcA and HetR and it is required for the full induction of the hetR gene (Xu et al., 2008). The expression of NtcA and HetR are mutually activated in a regulatory loop that seems essential for the establishment of differentiation (Muro-Pastor et al., 2002). Although all cells might sense nitrogen deficiency, not all cells in the filament differentiate. The decision on which cells undergo differentiation appears to rely on positive signals in the cell to differentiate and negative signals that would diffuse from it to the neighboring cells, inhibiting their differentiation (Wolk, 1967; Herrero et al., 2016, Muñoz-García and Ares, 2016). One positive signal is identified with the induction of hetR, whereas negative signals correspond to diffusible peptides of 5-7 amino acids excised from longer peptides, PatS, PatX, HetN, produced in the differentiating cells and processed during export from them (Yoon and Golden, 2001; Corrales-Guerrero et al., 2013; 2014; Videau et al., 2015; Zhang et al., 2017; Elhai and Khudyakov, 2018). Whereas patS is expressed together with patX at an early stage of differentiation and would participate in the establishment of the heterocyst distribution pattern, hetN is expressed at a later stage and would participate in pattern maintenance. In the vegetative cells, the inhibitory peptides would interact with HetR, inhibiting its DNA binding activity and thereby HetR-dependent transcription activation, blocking the differentiation of these cells (Yoon and Golden, 2001; Kim et al., 2011). 6.5. Cell Morphology in Cyanobacteria Cyanobacterial cells display a wide variety of sizes and morphologies. Size ranges between the 0.5 µm diameter of Prochorococcus strains to the 69 µm of Oscillatoria princeps (Schulz-Vogt et al., 2007). Cell morphologies include coccoidal, elipsoidal, discoidal, rod-shaped or spiral forms, and some species like Mastigocladus laminosum or Fischerella muscicola are pleomorphic Introduction 40 (Flores and Herrero, 2014) (Figure 10A) (see e.g. https://wwwcyanosite.bio.purdue.edu/images/images.html for some images of cyanobacterial species). In addition, some environmental cues can induce changes in cell morphology. For instance, in Fremyella diplosiphon the quality of light has been reported to induce the transition from rodshaped cells (predominant in green light) to more coccoid cells (predominant in red light) (Bordowitz and Montgomery, 2008). In Anabaena, the nutritional status as well as the light intensity provoke changes in cell shape and size (this work, Velázquez-Suárez, in preparation). The mechanisms underlying the determination of these cell shapes in different strains are still poorly understood, but they surely involve the PG synthesis complexes. Multicellularity adds an additional layer of complexity, as multicellular strains show unique features of the cell envelope (see below). In addition, morphological changes associated to cell differentiation require remodelling of the cell wall (Singh and Montgomery, 2011). Knowledge on the mechanisms governing cyanobacterial cell shape and size could also be of interest for biotechnological applications. Recent reports have shown that morphology can be manipulated resulting in advantages for some applications; e.g., making easier cell harvesting from scaled-up cultures (Jordan et al., 2017). 7. THE CELL ENVELOPE OF CYANOBACTERIA The cyanobacteria are diderm bacteria with a thick PG layer, ranging in thickness from 10 to 35 nm (although in some cases it can be up to 700 nm thick) in the periplasmic space between the PM and the OM (Hoiczyk and Hansel, 2000; Springstein et al., 2020c). Traditionally classified as Gram-negative bacteria, the high degree of crosslinkage and the anomalous thickness of their PG layer, in comparison to model Gram-negative bacteria, led to propose that the cyanobacterial envelope was intermediate between those of canonical Gram-positive and Gram-negative bacteria, showing features specific to each type (see Hahn and Schleiff, 2014 and references therein). The presence in cyanobacterial genomes of cell wall-processing and cell division genes homologous to those of Gram-negative and Gram-positive bacteria was thought to further support this proposal (Hoiczyk and Hansel, 2000; Springstein et al., 2020c). Recent advances in the knowledge on the early evolution of bacterial lineages have helped to understand their current features. Cyanobacteria are clasified in the Terrabacteria group that includes diderm and all known monoderm bacteria (Coleman et al., 2021) (see Figure 4). In this group, the monoderm phenotype arose as a result of multiple independent losses of OM genes. Thus, given the phylogenetic position of Cyanobacteria as a sister clade of monoderm bacteria, it is not surprising to observe genes and features in common to them (Witwinowski et al., 2022). Introduction 41 A PM B C D E FtsQ FtsX FtsK FtsZ ZipN ZipS FtsW SepJ SepI FtsI SepF MreB Zick/ Zack Outer membrane Pep>doglycan mesh Plasma membrane Figure 14. The cyanobacterial cell envelope and divisome components. (A) The cell envelope of Anabaena with focus on an intercellular septum: plasma membrane (PM), peptidoglycan layer (PG) and continuous outer membrane (OM). (B) Continuous peptidoglycan sacculus from Anabaena. (C) Distribution of mre genes in representative strains from different habitats showing different morphologies. (D) Interaction network of proteins of the Anabaena divisome (blue), elongasome (green) and septal complexes (orange). (E) The Anabaena divisome. (A) Adapted from Flores et al., (2006), (B) from Arévalo et al., (2021); (C) and (D) adapted from Springstein et al., (2020c). In addition to a thick PG layer, the cyanobacterial cell envelope shows other specific features. For instance, the OM contains carotenoids that are thought to protect the cell from excess radiation (Resch and Gibson, 1983). Besides, the OM proteome of cyanobacteria is mainly compossed of beta barrel proteins that show significant differences in the overall structure and pore size with respect to OM proteins of E. coli (Huang et al., 2004; Moslavac et al., 2005). Of note, some cyanobacterial strains have an additional outermost proteinaceous S-layer, and some others produce a muscillage capsule that confers protection against desiccation and UV- radiation (Hoiczyk and Hansel, 2000). As mentioned above, in filamentous cyanobacteria, whereas the PM is individual for each cell, the OM is continuous, enclosing a periplasmic space that is common to all the cells of the filament and constitutes a conduit for molecular exchanges between them (Flores et al., 2006; Mariscal et al., 2007; Wilk et al., 2011) (Figure 14A). The continuity of the OM relies on the lack Introduction 42 of OM invagination and fission during cell division. During cell division the newly synthetized septal PG does not split either, so that the resulting daughter cells are separated by a single PG layer, named septal disk, that is thicker than the lateral PG layer. This makes the PG saculus to be also continuous for the whole filament and, indeed, it can be isolated as an unit (Figure 14B). PG septal disks present an average of ca. 20 nm-wide perforations, termed nanopores, that likely serve to lodge the SJ protein complexes that comunicate the cytoplasms of adjacent cells. Nanopores in septal PG disks are drilled by amidases, whose inactivation provoke complete loss or alterations in the number and size of septal nanopores, and impairment in cell-cell communication (Lehner et al., 2011; Berendt et al., 2012; Bornikoel et al., 2017; Zheng et al., 2017). The number of nanopores and the nanopore diameter are larger in the septa between a vegetative cell and a heterocyst than in septa between two vegetative cells (Nicolaisen et al., 2009a; Arévalo and Flores, 2021). 7.1. PG synthesis in cyanobacteria The pattern of PG growth has been studied in a variety of cyanobacteria with diverse morphologies by labeling with Van-FL. Van-FL binds to the D-Ala-D-Ala in lipid II, either not incorporated or incorporated in nascent PG chains but not yet involved in cross-linking (Lehner et al., 2013; Springstein et al., 2020b). Also, a fluorescent analog of D-alanine, HADA (D-Ala, 7- Hydroxycoumarin-amino-D-alanine), which incorporates in the cell wall has been used (Zhang et al., 2018). Significant differences were observed associated with different morphotypes. Both septal and peripheral labelling were observed in unicellular strains such as the coccoid Synechocystis sp. PCC 6803 and the rod-shaped Synechococcus sp. PCC 7942. In addition, the coccoid Microcystis sp. PCC 7806 showed double staining rings at midcell typical of pre-septal PG growth. Filamentous strains, like Arthospira sp. FACHB 792, Oscillatoria animalis, Leptolyngbya sp. CB006, Spirulina subsalsa or Anabaena, all showed both peripheral and septal staining, although in the former two labeling was confined to cells at the tip of the filament, indicating that cell elongation and division are restricted to these cells (Zhang et al., 2018). The pleomorfic strain Fischerella muscicola, a filamentous branching cyanobacteria, showed Van-FL staining indicative of PG growth at peripheral and septal position throught the filament and at polar position in the apical cells of the branches. Most bacteria contain a long cluster of genes, named dcw (division and cell wall), involved in PG synthesis and cell division, which is not observed in cyanobacteria, whereby such genes are commonly found as individual genes or in short clusters (Megrian et al., 2022). Some genes involved in PG synthesis in cyanobacteria, including the intracellular synthesis of PG precursors (Videau et al., 2016; Liu et al., 2021) and the extracellular PG polymerization (Lázaro et al., 2001; Leganés et al., 2005; Marbouty et al., 2009) have been identified and their products characterized. Of note, it was observed that the number of genes encoding PBPs increases with the size of the genome, with a minimum of four in marine unicellular strains and up to 14 in Introduction 43 Nostoc punctiforme. Filamentous strains include a higher number of class A PBPs than unicellular ones, and it was proposed that these enzymes are involved in the differentiation of specific cells such as heterocysts (Leganés et al., 2005). 7.1.1. The elongasome As mentioned above, the protein MreB is generally present in rod-shaped and absent in coccoid bacteria. In cyanobacteria, some correlation has been observed between the presence of mre genes and the cell morphology (Springstein et al., 2020c), but the rules that determine this correlation are not fully understood. In cyanobacteria the mre genes generally group in a mreBCD cluster, although this cluster in not universally conserved in the phylum. In Anabaena this cluster has been shown to constitute an operon (Hu et al., 2007). Genomic analyses have revealed that out of 131 genomes surveyed, only four, corresponding to coccoid bacteria, lacked the three mreB, mreC and mreD genes (Figure 14C). Some spherical cyanobacteria such as Prochlorococcus marinus bear mreBC genes, whereas some others, such as the unicellular Cyanobium sp. and even the filamentous Trichodesmium erythraeum, maintain a rod-shape in the absence of mreD or, in the case of Arthrospira maxima, of mreC and mreD. Filamentous cyanobacteria that form heterocysts, including those that also form branched or multiseriated filaments, present a mreBCD gene cluster, although Chlorogloeopsis fritschii features coccoid cells (Springstein et al., 2020b). The obligate symbiont Atelocyanobacterium thalassa (also known as UCYN-A) present coccoid cells and lack not only the mre genes, but also rodA (Cassier- Chauvat and Chauvat, 2014; Springstein et al., 2020c). In the unicellular, rod-shaped strain Synechococcus sp. PCC 7942, mreB is essential, and partially segregated mutants showed a rounder shape, indicating that this gene is involved in the determination of cell morphology (Savage et al., 2010). In Anabaena sp. PCC 7120, MreB is not essential but it is also involved in cell morphology and cell wall stability (Hu et al., 2007). MreB from Anabaena was reported to interact with Zick and Zack, two coiled-coil-rich proteins (CCRPs) that form heteropolymeric filaments, and this interaction was shown to impact the MreB localization (Springstein et al., 2021). In F. diplosiphon, the expression of mreB is regulated by BolA, a protein that under red light binds to the mreB promoter region repressing transcription. This might explain that under red light F. diplosiphon cells are rounder than under green light (Singh and Montgomery, 2014). In species that differentiate hormogonia, such as N. punctiforme, the mreBCD operon is involved in the transition from coccoid vegetative cells to rod-shaped hormogonia (Gonzalez et al., 2019). Regarding the relationship between elongasome and divisome proteins, interactions have been reported to occur in several species (Springstein et al., 2020b), and in Synechocystis it has been suggested that MreB could be, directly or indirectly, negatively regulated by cell division factors Objectives 50 Objectives 51 1. To study the influence of the nitrogen and carbon availability on cell size and morphology in Anabaena sp. PCC 7120. 2. To investigate the spatiotemporal pattern of PG incorporation and the role of the mre genes in filaments of Anabaena sp. PCC 7120 growing with combined nitrogen or diazotrophically. 3. To unravel specific interactions between cell growth, PG synthesis and cell division in Anabaena sp. PCC 7120. Objectives 52 Summary of results 53 SUMMARY OF RESULTS Summary of results 54 Summary of results 55 Out of the great diversity of existing bacteria, PG synthesis has mainly been studied in a few model organisms that share an unicellular character. Therefore, expanding our knowledge to other bacteria with different morphologies and life styles appears necessary for a deeper understanding of this essential issue. Towards this goal, we have chosen a representative of the clade of multicellular cyanobacteria, which present a distinct organismal organization. We have addressed the pattern of PG synthesis, including the influence of environmental factors and the relationships with the process of cell division, in Anabaena sp. PCC 7120. The work and the results obtained in this thesis have been divided into three chapters, each corresponding to a published research article. Chapter I: The Inorganic Nutrient Regime and the mre Genes Regulate Cell and Filament Size and Morphology in the Phototrophic Multicellular Bacterium Anabaena. Velázquez-Suárez, C, Luque, I, Herrero, A. (2020). mSphere, 5(5):e00747-20. https://doi.org/10.1128/msphere.00747-20 Cell shape and size are inheritable traits that can be modified under certain environmental conditions. The information on PG synthesis in Anabaena was scarce and little was known on the influence of the carbon and nitrogen sources and on the role of MreB, MreC and MreD in the determination of cell size and shape, and the filament length, in this cyanobacterium. Results in this work showed that in Anabaena the carbon and nitrogen regime impact the cell size, the filament length and the aspect ratio (the ratio between the length of the cell axis parallel to the filament and the one perpendicular to the filament). When growing with combined nitrogen, these morphological parameters fluctuated throughout the cell cycle. In general, cells growing in the presence of nitrogen compounds, specially with abundant inorganic carbon supply, are larger and longer than vegetative cells from diazotrophic cultures, and filaments are longer in the former conditions. The mreB, mreC and mreD genes form an operon whose expression is highest in the exponential phase of growth, when the cells were increasing the aspect ratio, and is negatively regulated by NtcA, consistent with the higher aspect ratio observed in cells of an ntcA mutant in comparison to the wild type (WT). mreBCD expression increases during heterocyst differentiation upon N- stepdown and, transiently, in the differentiating cells. Mutants lacking mreB, mreC or mreD exhibit impaired growth, especially under diazotrophic conditions. Filaments of these mutants are longer than those of the WT, and their cells are bigger and rounder; indeed, in the mutants the cell axis parallel to the filament is shorter than that perpendicular to the filament. Also, heterocysts of the mutants are larger and rounder than the wild-type heterocysts. In addition, intercellular septa along the filament are wider and thicker than in the WT. These results indicate a role of MreB, MreC and MreD in the synthesis of peripheral PG for cell elongation during growth with combined nitrogen or diazotrophically, as well as during differentiation of the heterocysts, affecting also the intercellular PG along the filament. Summary of results 56 Chapter II: The Role of Mre Factors and Cell Division in Peptidoglycan Growth in the Multicellular Cyanobacterium Anabaena. Velázquez-Suárez, C, Valladares, A, Luque, I, Herrero, A. (2022). mBio, 13(4):e0116522. https://doi.org/10.1128/mbio.01165-22 PG synthesis is required for cell growth and subsequent division, giving rise to the specific morphology of each bacterial species. PG growth has been studied in depth in model unicellular rod-shaped bacteria. However, the envelope structure of Anabaena is peculiar since cytokinesis does not proceed to completion and OM invagination does not occur, which results in a continuous OM, enclosing all cells of the filament, and a continuous periplasm containing the PG sacculus, which is not split during cytokinesis and remains fused in the intercellular septa along the filament. This work focused on the growth of PG in Anabaena and in mreB, mreC and mreD mutants. It permited to conclude that Anabaena presents three modes of PG growth: lateral, divisional and intercellular septal. The intercellular septal synthesis is permanent and represent a distinctive mode that likely is shared by other filamentous multicellular cyanobacteria. In Anabaena, the MreB, MreC and MreD proteins are not only localized in the cylindrical part of the cell periphery, but also in the intercellular septa and, notably, in the divisome localization during all the stages of cell division. Continuous septal PG synthesis can contribute to the positioning and remodeling of septal protein complexes for intercellular communication. Besides alterations in the cell morpholgy, the absence of the mreB, mreC or mreD genes provoke alterations in the positioning of the Z-ring and the division septa, resulting in impared orientation of the intercellular septal PG along the filament. Molecular interactions have been detected between MreB or MreD and some divisome components. These results led to the proposal that the elogasome Mre proteins are required for the correct positioning of the divisome. Conversely, mutants affected in the expression of FtsZ or ZipN not only are deficient in cell division, but also present impaired uncheked peripheral PG growth, leading to giant cells prone to lysis. Thus, the divisome appear not only to direct the septal synthesis, but also to regulate the lateral PG growth. These results show unprecedented mutual connections between the elongasome and the divisome, as well as between the different modes of cellular PG growth. Summary of results 57 Chapter III: The Role of MreB, MreC and MreD in the Morphology of the Diazotrophic Filament of Anabaena sp. PCC 7120. Velázquez-Suárez, C., Luque, I., & Herrero, A. (2022). Life, 12(9):1437. https://doi.org/10.3390/life12091437 In Anabaena, the absence of nitrogenous compounds in the external medium triggers the differentiation of heterocysts, a process that has been widely studied. However, the role of the elongasome proteins MreB, MreC and MreD during heterocyst differentiation and in the diazotrophic filament was not known. In filamentous multicellular cyanobacteria, intercellular communication through the shared periplasm and the SJs, including the exchange of nutritional compounds and signalling molecules, is pivotal for the organism performance under diazotrophic conditions. This work evidenced that during heterocyst differentiation there is a transient incorporation of PG in the periphery of the differentiating cells and in the septal regions between heterocys and the adjacent vegetative cells, the latter involved in the remodeling of the cell wall to form the hetrocyst neck. Polar PG incorporation persists in mature heterocysts, likely involved in maintenance of the structure of the septa between both cell types. MreB, MreC and MreD localize to the periphery and the poles of the differentiating cells, but they go away from the periphery of old heterocysts. Inactivation of mreB, mreC or mreD leads to instances of abnormally high and aberrant PG incorporation in the periphery and the poles of the differentiating cells, and the presence of heterocysts with conspicuous morphological alterations, including deformed poles and instances of filament breackage at vegetative cellheterocyts connections. Moreover, septa devoid of SepJ can be detected in the mutants. These results involve the Mre proteins in the formation of septal communication structures in the heterocyst poles. Also, apparently mature heterocyst still presenting a Z-ring could be detected in the mre mutants, consistent with a role of Mre factors in the disassembly of the divisome in the differentiating cells. Summary of results 58 59 Chapter I CHAPTER I: The inorganic nutrient regime and the mre genes regulate cell and filament size and morphology in the phototrophic multicellular bacterium Anabaena Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 6 FIG 2 Distribution of filament length in Anabaena and strains CSE2 (ntcA mutant) and CSSC2 (hetR mutant) grown with different nitrogen and carbon supplies. At the indicated times, aliquots of cultures treated as described in the legend to Fig. 1 were taken with care to avoid filament breakage and photographed. Filaments from three independent cultures of each condition were counted. Filaments longer than 120 cells were counted as of 120. Violin-plot representations of the data are shown. Mann-Whitney tests were performed to assess significance of differences with data from filaments up to 119 cells long (214 to 387 filaments) (Data Set S1). different nutrient conditions in each of the considered growth periods (Fig. S2). Anabaena showed a positive correlation during the phases of more active growth (see FEG and first week in Fig. S2), although this correlation was nonlinear. Notably, a linear negative correlation was observed with data corresponding to the 2nd week of growth, which corresponded to the transit to phases of less active growth and, in the case of nitrate-grown cells, of larger cell size. Regarding the ntcA and hetR mutants, cells of the former, with ammonium and LC, were generally similar to those of the wild type (Fig. 1B). Under these conditions, cells of the hetR mutant were smaller than those of the wild type or the ntcA mutant. Also, during fast growth in the presence of nitrate, with LC or HC, cells of the hetR mutant were similar or somewhat smaller than those of the wild type (Fig. 1C). Filament length responds to the C and N nutrition. The number of cells per filament was counted through the growth cycle under the different nutritional conditions described above, and the distribution of filament lengths was analyzed (Fig. 2; the significance of comparisons can be found in Data Set S1). In the wild type, filaments tended to be longer during the phases of more active growth. Perhaps senescence and Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 7 death of random intercalary cells, which would be more frequent during less active growth, could contribute to shorten the filaments. Filaments incubated with nitrate and LC were similar in size to those incubated with ammonium and tended to be shorter than those incubated with nitrate and HC. A larger difference was observed between filaments incubated with nitrate or ammonium and the shorter diazotrophic filaments. Perhaps the septa between vegetative cells and heterocysts are more fragile than the septa between vegetative cells, contributing to shortening of filaments under diazotrophic conditions. Moreover, because heterocysts are terminally differentiated cells, senescence and lysis of old heterocysts would increase filament breakage. This is consistent with the fact that filaments of 5 to15 cells, roughly coincident with the size of vegetative-cell intervals separating two heterocysts, were the most frequently observed under diazotrophic conditions. Finally, both in the ntcA mutant and in the hetR mutant, the filaments were much shorter than in the wild type. Dynamics of cell morphology during growth using different C and N supplies. The lengths of the longitudinal (parallel to the filament) and the transversal (perpendicular to the filament) axes of the cells were determined in Anabaena under the different growth conditions used and, as an indication of cell aspect, the ratio between the two cell axes (aspect ratio) was calculated (Fig. 3; the significance of comparisons can be found in Data Set S1). In nitrate- and ammonium-supplemented medium, with LC, the aspect ratio increased through the phases of more active growth and then decreased more pronouncedly in the presence of ammonium. This same trend was observed for cells incubated under diazotrophic conditions, although in this case the aspect ratio increase was smaller and delayed, consistent with a delayed FEG (see above), compared to growth in combined nitrogen-supplemented medium. The aspect ratio of cells incubated with nitrate or ammonium reached higher values than in diazotrophic cultures, and HC supplement generally had a positive effect. Because in rod-shaped cells an increase in length would increase the total surface area, it could have larger benefits when growth depends on nutrients taken up from the external medium (see reference 1), such as nitrate, ammonium or bicarbonate, and during phases of faster growth, and could facilitate light adsorption also. In contrast, under diazotrophic conditions, it is assumed that most N2 enters the heterocysts from the neighboring vegetative cells by intercellular connections (38). Moreover, because elongation is likely more costly than making short rods, the energetic balance would represent another factor favoring Anabaena cells being longer when combined nitrogen is available. Notably, in ammonium-supplemented medium, the aspect ratio of the ntcA and hetR mutants was considerably higher than that of the wild type (Fig. 3B). In the hetR mutant, the aspect ratio was higher in cultures with ammonium than in those with nitrate (see Fig. 3B and C). The mre gene cluster of Anabaena. In the Anabaena genomic sequence, the mreB gene (all0087) and a homologue to mreC (all0086) are clustered together, being separated by 95 nucleotides. Downstream of mreC, open reading frame (ORF) all0085 is found, which is separated from the putative mreC by 114 nucleotides (see Fig. 4A). Anabaena MreB has 347 residues with 52% identity to E. coli MreB. No transmembrane helix is present in Anabaena MreB (checked with TMHMM) although, similar to E. coli MreB (39), it may include an amphipathic in-plane membrane anchoring segment (Amphipaseek, NPS@ [Network Protein Sequence Analysis]) for interaction with the cytoplasmic membrane. MreC is a bitopic protein with a transmembrane helix and a large periplasmic part (40). The predicted product of all0086 has 273 amino acids (aa) and 19% identity to E. coli MreC (367 aa), bears an N-terminal signal peptide locating almost the whole protein in the periplasm (Phobius), and includes a coiled-coil motif encompassing residues ca. 60 to 90 (Network Protein Sequence Analysis [41]). Finally, all0085 would encode a 170-aa product with 19% identity to E. coli MreD (162 aa). However, in spite of the low overall sequence identity, All0085 would include 5 transmembrane helices with a periplasmic N terminus and cytoplasmic C terminus, Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 8 4 FIG 3 Aspect ratio of cells of Anabaena and strains CSE2 (ntcA mutant) and CSSC2 (hetR mutant) grown with different nitrogen and carbon supplies. In the same cells used in Fig. 1, the lengths of the longitudinal and transversal cell axes were measured as described in Materials and Methods. The aspect ratio is the result of dividing the length of the axis parallel to the filament by the length of the axis perpendicular to the filament. Notched boxplot representations of the data are shown. The mean values are represented by black dots. Tukey tests were performed to assess significance of differences (Data Set S1). (A) WT; (B) ntcA and hetR mutants in BG11 0 + NH + ; (C) hetR mutant in BG11. similar to the organization of E. coli MreD (Phobius, TMHMM). Thus, we will consider all0085 to be a putative mreD gene of Anabaena. Expression of the Anabaena mre gene cluster. Expression of the Anabaena mreBCD cluster was studied by means of Northern blot analysis. RNA was extracted from filaments of the wild type and the ntcA and the hetR mutants in the FEG phase (24 h) and in a period of slower growth (360 h) using ammonium or N2 as the nitrogen source. Consistent with previous results (42), different transcripts hybridized with the probe of each gene. We found transcripts of ca. 3 and 1.3 kb (mreB), of ca. 3 and 1.7 kb (mreC), and of ca. 3, 1.7 and 1.1 kb (mreD) (Fig. 4A). For the different strains and conditions, the levels of mreB transcripts were always higher than those of mreC, which were higher than those of mreD (Table 2). For the different nitrogen sources, total Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 9 4 FIG 4 Expression of the Anabaena mre gene cluster. (A) Northern-blot analysis of the expression of the mreB, mreC, and mreD genes (genomic cluster represented in the upper part). RNA was extracted from filaments of the indicated strains grown in BG110 or BG110 + NH + medium, refreshed in the same medium at a cell density of 0.2 µg chlorophyll/ml, and incubated under culture conditions for 24 h (lighter colors) or 360 h (darker colors). Hybridization was performed with probes of the indicated mre gene (upper panels) or of rnpB used for normalization (lower panels). Three different cultures were used for each strain and condition (one representative filter of each strain and condition is shown). The mean and standard deviation of total transcript abundance, normalized for the rnpB signal for each lane, are represented, and significance of differences was assessed with Student’s t test (**, P < 0.01; *, P < 0.05; ns, P > 0.05). Colored arrows at the right side of each panel point to the main transcripts detected (size in kb indicated). A size standard (RiboRuler High Range, Thermo) is shown at the left. (B) Schematic model of transcription in the mre gene cluster. In the upper part, the localization and size of expected transcripts is depicted (dashed segments indicate unprecise localization or transcript end). TSPs located at 171 nucleotides upstream from mreB and 37 nucleotides upstream from mreC are indicated, and a gray barrel represents a putative NtcA-binding sequence. Nucleotide sequences upstream from mreB and in the mreB-mreC intergenic region are depicted. Colored sequences denote putative - 10 and - 35 promoter elements; colored bold, transcription start points and the ATG start of mreC. Shadowed GTA and TAC sequences represent a putative NtcA-binding site (consensus sequence GTAN8TAC), and framed GTA and TAG sequences represent an imperfect NtcA-binding site. Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 10 4 4 TABLE 2 Comparisons of the expression of the Anabaena mreB, mreC, and mreD genes by Northern blot analysisa Strain/condition Time (h) Ratio of total transcript levels mreB/mreD mreC/mreD mreB/mreC PCC7120 (WT)/N2 24 4.21 ± 1.03 3.52 ± 1.43 1.28 ± 0.26 360 3.69 ± 0.09 4.21 ± 1.79 1.11 ± 0.56 PCC 7120 (WT)/NH + 4 24 4.08 ± 0.85 3.44 ± 0.83 1.29 ± 0.46 360 4.50 ± 1.49 3.77 ± 1.31 1.29 ± 0.40 CSE2 (ntcA)/NH + 4 24 4.91 ± 1.42 3.23 ± 0.96 1.61 ± 0.49 360 4.85 ± 2.21 3.07 ± 0.98 1.51 ± 0.35 CSSC2 (hetR)/NH + 4 24 5.20 ± 0.40 3.21 ± 1.01 1.76 ± 0.46 360 5.87 ± 1.32 3.64 ± 0.48 1.59 ± 0.15 a RNA was extracted from filaments of the indicated strains grown in BG11 0 (N 2 ) or BG11 0 + NH + (NH + ) medium for 24 or 360 h (see Fig. 4A for further details). The figures represent the mean and standard deviation of the ratio of total transcript levels of the corresponding gene pair in three independent cultures of each strain and condition. transcript levels of any of the three genes were higher in the FEG phase than during slow growth (Fig. 4A; Student’s t test P values between 0.019 and 0.001, except for mreB and mreC in the hetR mutant, which gave P values higher than 0.05). No significant differences were detected comparing ammonium-grown and diazotrophic filaments (Student’s t test P > 0.05). In the presence of ammonium, total transcript levels were higher in the ntcA mutant than in the wild type (Student’s t test P = 0.003 for comparison of mreB transcript levels in the exponential phase; P = 0.050 for mreD). This result, together with the fact that the longer (ca. 3 kb) transcript was better detected in the ntcA mutant than in the wild type (WT), indicates a negative effect of NtcA on the expression of the mre genes. No significant effect of the hetR mutation was detected. We also studied the spatial pattern of the activity of the mreB promoter along the Anabaena filament. In a previous report, a gfp gene under the control of the mreB promoter introduced in Anabaena in a shuttle vector established with several copies per chromosome copy was reported to produce higher fluorescence in heterocysts than in vegetative cells (42). Here, we generated strain CSCV3, an Anabaena derivative that includes a gfp-mut2 gene fused after the 6th codon of the mreB sequence, thus preserving the transcription and translation start signals of native mreB, in the chromosomal mreB locus. In addition, an intact copy of mreB, mreC, and mreD preceded by their native promoter region was preserved (Fig. 5A; see Materials and Methods for details). Green fluorescent protein (GFP) fluorescence was monitored through the growth cycle in filaments of strain CSCV3 incubated with nitrate or ammonium or in the absence of combined nitrogen. In the latter case, the filaments included vegetative cells and cells in different stages of differentiation into heterocysts. Importantly, growth of strain CSCV3 was not impaired in comparison to the wild type (not shown). In cultures containing nitrate or ammonium, green fluorescence above the background in the wild-type strain was homogenously observed along the filament during active growth (Fig. 5B). Maximum fluorescence was observed during the first 24 to 48 h of incubation. After that time, fluorescence decreased, so that after 168 h, only background fluorescence was detected (Fig. S3). Upon N-stepdown, fluorescence levels increased in differentiating cells, which exhibited higher levels than vegetative cells (see 24 h in Fig. 5B). However, this increase was transitory, so that in mature heterocysts, GFP fluorescence levels were similar to those of vegetative cells (see 24 h in Fig. 5B). Indeed, 48 h after the transfer, mature heterocysts exhibited very low or undetectable GFP fluorescence, lower than that in vegetative cells. (Of note, overexpression of the PmreB-gfp reporter utilized previously [42] could have led to maintenance of increased fluorescence for longer than the expression of the promoter activity in the native context.) Meanwhile, increased fluorescence was observed in stretches of vegetative cells that were neighbors to heterocysts, likely those that were most actively growing. Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 11 FIG 5 Spatiotemporal expression of the mreB gene promoter in Anabaena. (A) Genomic structure of strain CSCV3 (expressing [P mreB - gfp ] in comparison to PCC 7120 [WT]). (B) Filaments of strains PCC 7120 and CSCV3 grown in solid BG11 medium were transferred (at a cell density of 0.2 µg chlorophyll/ml) to BG11, BG110 + NH + , and BG11 media and incubated under culture conditions. At 24 h intervals, filaments were observed 4 0 under a fluorescence microscope and photographed. GFP fluorescence (left) and bright-field (right) images are shown. Arrowheads point to heterocysts: immature heterocysts (yellow); mature heterocysts, exhibiting polar refringent cyanophycin granules (white). Magnification is the same for all micrographs. After 72 h, the fluorescence in vegetative cells had decreased also (not shown). Thus, activity of the mreB promoter is maximum during the FEG phase, and upon N-stepdown expression transiently increases in the cells differentiating into heterocysts. mre mutants of Anabaena. To investigate the role of mreB, mreC, and mreD genes in Anabaena, derivatives carrying inactivated versions of these genes were created. Mutants were generated by substituting parts of each gene by an antibiotic-resistance- encoding gene cassette (Fig. 6A). Gene cassette C.S3 was inserted into the coding sequence of mreD. To avoid polar effects, gene cassette C.K1, which includes no transcriptional termination sequence, was inserted into genes mreB and mreC (see Materials and Methods for details). Strains CSCV1, CSCV4, and CSCV2 bear only deleted versions of mreB, mreC, and mreD genes, respectively (see Fig. S4). Strains CSCV1, CSCV4, and CSCV2 were characterized in terms of growth rate and morphology. The three mutants grew using either nitrate or ammonium, although growth  Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 12 4 0  FIG 6 Anabaena mreB, mreC, and mreD mutants. (A) Genomic structure in the mre region in strains CSCV1 ( mreB ), CSCV4 ( mreC ), and CSCV2 ( mreD ) in comparison to Anabaena (WT). (B) Filaments grown in BG110 + NH + or in BG11 (lacking combined nitrogen) medium were used to inoculate, at an initial cell density corresponding to 0.2 µg chlorophyll/ml, flasks containing the same medium, which were incubated under culture conditions. After 24 h, aliquots of each culture were taken and filaments photographed. Arrowheads point to heterocysts. Magnification is the same for all micrographs. was slower than in the wild type, especially with a regime of LC (the growth rate of the mutants was about 60% of that of the corresponding wild-type value) (Table 1). In the absence of combined nitrogen, the three mutants formed heterocysts (Fig. 6B). However, whereas strain CSCV2 was capable of sustained diazotrophic growth, CSCV1 and CSCV4 did not significantly increase mass and, indeed, got lysed after some days of incubation (Fig. 7). The diazotrophic growth rate of CSCV2 under LC was ca. 45% that of the wild type and 56% under HC (Table 1). To seek the cause of the impaired diazotrophic growth, nitrogenase activity was measured (Table 3). Whereas activity levels after 24 h of incubation in the absence of combined nitrogen were considerably lower in the three mutants than in the wild type, after 48 h, only strain CSCV4 was significantly impaired (activity was ca. 36% of the WT activity). Thus, the heterocysts formed in the mutants were largely active. The three mutants exhibited conspicuous morphological differences compared to the wild type, in both the presence and absence of combined nitrogen (Fig. 6). Figure 8 presents data on the determination of cell area (the significance of comparisons is shown in Data Set S1). Cells of the three mutants had similar sizes in medium supplemented with nitrate (LC or HC). Notably, the mean cell area values of CSCV2 (mreD), and especially of CSCV4 (mreC), incubated with ammonium were larger than those of nitrate-incubated cells of the same strain and larger than those of ammonium- Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 13 4 FIG 7 Growth of Anabaena and mutant strains CSCV1 (mreB), CSCV4 (mreC), and CSCV2 (mreD). Filaments grown in BG110 + NH + medium were used to inoculate, at an initial cell density corresponding to 0.2 µg chlorophyll/ml, flasks containing BG110 (lacking combined nitrogen) medium, which were incubated under culture conditions and photographed at the indicated times. incubated cells of CSCV1 ( mreB ). Moreover, cell size appeared heterogeneous in CSCV2, showing filament stretches of larger cells and others of smaller cells (see Fig. 6), which is indicative of the inability to regulate cell size. Under any growth condition, cells of any of the three mutants (vegetative cells for filaments incubated in the absence of combined nitrogen) were larger than those of the wild type (compare Fig. 8 and Fig. 1A). As in the wild type, cells growing diazotrophically (CSCV2) were smaller than those growing with combined nitrogen (Fig. 8). In filaments incubated in the absence of combined nitrogen, we also measured the size of heterocysts, which in the wild type were larger than vegetative cells and relatively constant during growth (compare Fig. 9A and Fig. 1A). Heterocysts of any of the mre mutants were larger than heterocysts of the wild type (Fig. 9A; Data Set S1). Regarding cell morphology, the aspect ratio of the cells of strains CSCV1, CSCV4, and CSCV2 was closer to 1 than the wild-type values (compare Fig. 10 and Fig. 3A; see Data Set S1), meaning that cells of the mutants are more rounded than cells of the wild type. Indeed, in the mutants, the aspect ratio values were smaller than 1, indicating that the cell axis parallel to the length of the filament is shorter than the axis perpendicular to the filament. Thus, the geometrical orientation of the cells in the filaments of the mutants is inverted compared to the orientation in the wild-type filaments (see Fig. 6B). Also, heterocysts of the three mutants were more rounded, and inverted, compared to the wild-type heterocysts, in which the aspect ratio was higher than in vegetative cells and showed a tendency to increase during the first 72 h of incubation (see Fig. 9B). Finally, in the three mre mutants, the filaments were generally longer than in the wild type (compare Fig. 11 and Fig. 2A; see Data Set S1). Thus, it appears that the intercellular septa of the mutants were more robust than those of the wild type. TABLE 3 Nitrogenase activity of mreB, mreC, and mreD mutants PCC 7120 (WT) 7.15 10.66 CSCV1 (mreB) 2.54 8.97 CSCV4 (mreC) 1.83 3.86 CSCV2 (mreD) 2.77 9.12 aNitrogenase activity was assayed in filaments of the indicated strain incubated in BG110 medium for 24 or 48 h. The figures represent the mean of the activities measured in two independent cultures that yielded similar results. Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Strain Nitrogenase activity (nmol · mg clorophyll - 1 · h - 1 ) a 24 h 48 h Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 14 FIG 8 Dynamics of cell area in mutant strains CSCV1 ( mreB ), CSCV4 ( mreC ), and CSCV2 ( mreD ) grown with different nitrogen and carbon supplies. Cells grown in the indicated conditions were used to inoculate, at an initial cell density corresponding to 0.2 µg chlorophyll/ml, cultures under the same conditions. At the indicated incubation times, aliquots of each culture were photographed and used for cell area determination. A total of 200 to 400 cells (vegetative cells in the diazotrophic cultures) from three different cultures of each time and condition were measured. Notched boxplot representations of the data are shown. The mean values are represented by black dots. Tukey tests were performed to assess significance of differences (Data Set S1). In order to identify possible factors contributing to the observed increased filament robustness in the mre mutants, we used Van-FL (fluorescent vancomycin) to estimate the septal width of vegetative cells. Van-FL labels sites of active PG synthesis (43), and in Anabaena, significant staining is found in the intercellular septa (e.g., 44). On average, septa were significantly wider in the mutants than in the wild type, and in CSCV1 and CSCV4, were wider than in CSCV2 (Fig. 12). Total Van-FL fluorescence detected at the septa was also higher in CSCV1 and CSCV4 than in CSCV2, which also rendered higher values than the wild type (Fig. 12). Increased septal width in the mutants can contribute to increased cell-to-cell cohesion and, thus, filament length. DISCUSSION Although many previous data on the growth of Anabaena under various experimental conditions have been reported, here we aimed at setting defined conditions Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 15 FIG 9 Morphological parameters of heterocysts in Anabaena and mutant strains CSCV1 (mreB), CSCV4 (mreC), and CSCV2 (mreD). Cells grown in BG110 medium and LC were used to inoculate, at an initial cell density corresponding to 0.2 µg chlorophyll/ml, flasks containing the same medium, which were incubated under culture conditions. (A and B) At the indicated times, aliquots of each culture were photographed and used for heterocyst cell area (A) and cell axis length (B) determination. The aspect ratio is the result of dividing the length of the axis parallel to the filament, by the length of the axis perpendicular to the filament. Ten heterocysts of each time and strain were measured. Boxplot representations of the data are shown. Black dots represent the mean values. Tukey tests were performed to assess significance of differences (Data Set S1). that would permit comparisons between growth rates and morphological features under different nutritional contexts. Indeed, we found that in this phototrophic bacterium, the nitrogen and carbon regime influence not only growth, but also cell size and shape and filament length. Regarding cell size, during the FEG phase, Anabaena using different nitrogen sources and carbon supplies would grossly meet the classical growth law, which entails that cells of rod-shaped bacteria exhibit larger volume when they grow with a nutrientimposed lower generation time (45), confirmed by recent studies at the population level (5). However, the positive linear correlation that was found between these two parameters in both E. coli and the slow-growing bacterium Sinorhizobium meliloti (46) could not be observed in Anabaena (Fig. S2). Moreover, in Anabaena, different relationships between growth and size were established during the growth cycle, resulting in a net cell size increase. These observations suggest that throughout active growth, the rate of cell division would progressively slow down more than the rate of total cell mass increase. In other words, the cell size at division would be increasing. Because Anabaena has multiple copies of the chromosome, and chromosome partitioning during cell division appears to be to a good extent aleatory (42), a large size would favor both daughter cells inheriting intact chromosomes during active growth (see reference 47). Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES  Velázquez-Suárez et al. September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 22 set bacterial cell size. Cell 165:1479 –1492. https://doi.org/10.1016/j.cell .2016.05.045. 7. Van Teeffelen S, Renner L. 2018. Recent advances in understanding how rod-like bacteria stably maintain their cell shapes. F1000Res 7:241. https://doi.org/10.12688/f1000research.12663.1. 8. Martins BMC, Tooke AK, Thomas P, Locke JCW. 2018. Cell size control driven by the circadian clock and environment in cyanobacteria. Proc Natl Acad Sci U S A 115:E11415–E11424. https://doi.org/10.1073/pnas .1811309115. 9. Chien A, Zareh S, Wang Y, Levin P. 2012. Changes in the oligomerization potential of the division inhibitor UgtP co-ordinate Bacillus subtilis cell size with nutrient availability. Mol Microbiol 86:594 – 610. https://doi.org/ 10.1111/mmi.12007. 10. Hill N, Buske P, Shi Y, Levin PA. 2013. A moonlighting enzyme links Escherichia coli cell size with central metabolism. PLoS Genet 9:e1003663. https://doi.org/10.1371/journal.pgen.1003663. 11. Monahan L, Hajduk I, Blaber S, Charles I, Harry E. 2014. Coordinating bacterial cell division with nutrient availability: a role for glycolysis. mBio 5:e00935–14. https://doi.org/10.1128/mBio.00935-14. 12. Randic A, Brun Y. 2015. Molecular mechanisms for the evolution of bacterial morphologies and growth modes. Front Microbiol 6:580. https://doi.org/10.3389/fmicb.2015.00580. 13. Egan AJF, Errington J, Vollmer W. 2020. Regulation of peptidoglycan synthesis and remodeling. Nat Rev Microbiol 18:446 – 460. https://doi .org/10.1038/s41579-020-0366-3. 14. Cava F, Kuru E, Brun Y, de Pedro MA. 2013. Modes of cell wall growth differentiation in rod-shaped bacteria. Curr Opin Microbiol 16:731–737. https://doi.org/10.1016/j.mib.2013.09.004. 15. Vollmer W, Bertsche U. 2008. Murein (peptidoglycan) structure, architecture and biosynthesis in Escherichia coli. Biochim Biophys Acta 1778: 1714 –1734. https://doi.org/10.1016/j.bbamem.2007.06.007. 16. Bendezú FO, Hale CA, Bernhardt TG, de Boer PAJ. 2009. RodZ (YfgA) is required for proper assembly of the MreB actin cytoskeleton and cell shape in E. coli . EMBO J 28:193–204. https://doi.org/10.1038/emboj.2008 .264. 17. Errington J. 2015. Bacterial morphogenesis and the enigmatic MreB helix. Nat Rev Microbiol 13:241–248. https://doi.org/10.1038/nrmicro3398. 18. Esue O, Cordero M, Wirtz D, Tseng Y. 2005. The assembly of MreB, a prokaryotic homolog of actin. J Biol Chem 280:2628 –2635. https://doi .org/10.1074/jbc.M410298200. 19. Dye NA, Pincus Z, Theriot J, Shapiro L, Gitai Z. 2005. Two independent spiral structures control cell shape in Caulobacter. Proc Natl Acad Sci U S A 102:18608 –18613. https://doi.org/10.1073/pnas.0507708102. 20. Leaver M, Errington J. 2005. Roles for MreC and MreD proteins in helical growth of the cylindrical cell wall in Bacillus subtilis. Mol Microbiol 57:1196 –1209. https://doi.org/10.1111/j.1365-2958.2005.04736.x. 21. Kruse T, Bork-Jensen J, Gerdes K. 2005. The morphogenetic MreBCD proteins of Escherichia coli form an essential membrane-bound complex. Mol Microbiol 55:78 – 89. https://doi.org/10.1111/j.1365-2958.2004.04367.x. 22. Bendezú FO, de Boer PAJ. 2008. Conditional lethality, division defects, membrane involution, and endocytosis in mre and mrd shape mutants of Escherichia coli . J Bacteriol 190:1792–1811. https://doi.org/10.1128/JB .01322-07. 23. Pinho MG, Kjos M, Veening J-W. 2013. How to get (a)round: mechanisms controlling growth and division of coccoid bacteria. Nat Rev Microbiol 11:601– 614. https://doi.org/10.1038/nrmicro3088. 24. Massidda O, Nováková L, Vollmer W. 2013. From models to pathogens: how much have we learned about Streptococcus pneumoniae cell division? Environ Microbiol 15:3133–3157. https://doi .org/10.1111/1462-2920.12189. 25. Ting CS, Rocap G, King J, Chisholm SW. 2002. Cyanobacterial photosynthesis in the oceans: the origins and significance of divergent lightharvesting strategies. Trends Microbiol 10:134 –142. https://doi.org/10 .1016/s0966-842x(02)02319-3. 26. Flores E, Herrero A. 2014. The cyanobacteria: morphological diversity in a photoautotrophic lifestyle. pip 1:63–72. https://doi.org/10.1127/pip/ 2014/0008. 27. Zehr J, Capone D. 2020. Changing perspectives in marine nitrogen fixation. Science 368:eaay9514. https://doi.org/10.1126/science.aay9514. 28. Hahn A, Schleiff E. 2014. The cell envelope, p 29 – 87. In Flores E, Herrero A (ed), The cell biology of cyanobacteria. Caister Academic Press, Norfolk, UK. 29. Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY. 1979. Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111:1– 61. https://doi.org/10.1099/ 00221287-111-1-1. 30. Wilk L, Strauss M, Rudolf M, Nicolaisen K, Flores E, Kühlbrandt W, Schleiff E. 2011. Outer membrane continuity and septosome formation between vegetative cells in the filaments of Anabaena sp. PCC 7120. Cell Microbiol 13:1744 –1754. https://doi.org/10.1111/j.1462-5822.2011.01655.x. 31. Herrero A, Stavans J, Flores E. 2016. The multicellular nature of filamentous heterocyst-forming cyanobacteria. FEMS Microbiol Rev 40:831– 854. https://doi.org/10.1093/femsre/fuw029. 32. Wolk CP, Ernst A, Elhai J. 1994. Heterocyst metabolism and development, p 769 – 823. In Bryant DA (ed), The molecular biology of cyanobacteria. Kluwer Academic Publishers, The Netherlands. 33. Zhang J-Y, Lin G-M, Xing W-Y, Zhang C-C. 2018. Diversity of growth patterns probed in live cyanobacterial cells using a fluorescent analog of a peptidoglycan precursor. Front Microbiol 9:791. https://doi.org/10 .3389/fmicb.2018.00791. 34. Singh SP, Montgomery BL. 2014. Morphogenes bolA and mreB mediate the photoregulation of cellular morphology during complementary chromatic acclimation in Fremyella diplosiphon. Mol Microbiol 93: 167–182. https://doi.org/10.1111/mmi.12649. 35. Herrero A, Flores E. 2019. Genetic responses to carbon and nitrogen assimilation in Anabaena. Environ Microbiol 21:1–17. https://doi.org/10 .1111/1462-2920.14370. 36. Kim Y, Joachimiak G, Zi Y, Binkowski TA, Zhang R, Gornicki P, Callahan SM, Hess WR, Haselkorn R, Joachimiak A. 2011. Structure of transcription factor HetR required for heterocyst differentiation in cyanobacteria. Proc Natl Acad Sci U S A 108:10109 –10114. https://doi.org/10.1073/pnas .1106840108. 37. Flores E, Frías J, Rubio LM, Herrero A. 2005. Photosynthetic nitrate assimilation in cyanobacteria. Photosynth Res 83:117–133. https://doi .org/10.1007/s11120-004-5830-9. 38. Walsby AE. 2007. Cyanobacterial heterocysts: terminal pores proposed as sites of gas exchange. Trends Microbiol 15:340 –349. https://doi.org/ 10.1016/j.tim.2007.06.007. 39. Salje J, van den Ent F, de Boer P, Löwe J. 2011. Direct membrane binding by bacterial actin MreB. Mol Cell 43:478 – 487. https://doi.org/10.1016/j .molcel.2011.07.008. 40. van den Ent F, Leaver M, Bendezu F, Errington J, de Boer P, Löwe J. 2006. Dimeric structure of the cell shape protein MreC and its functional implications. Mol Microbiol 62:1631–1642. https://doi.org/ 10.1111/j.1365-2958.2006.05485.x. 41. Combet C, Blanchet C, Geourjon C, Deléage G. 2000. NPS@: Network Protein Sequence Analysis. Computer Corner 25:147–150. https://doi .org/10.1016/S0968-0004(99)01540-6. 42. Hu B, Yang G, Zhao W, Zhang Y, Zhao J. 2007. MreB is important for cell shape but not for chromosome segregation of the filamentous cyanobacterium Anabaena sp. PCC 7120. Mol Microbiol 63:1640 –1652. https:// doi.org/10.1111/j.1365-2958.2007.05618.x. 43. Daniel RA, Errington J. 2003. Control of cell morphogenesis in bacteria: two distinct ways to make a rod-shaped cell. Cell 113:767–776. https:// doi.org/10.1016/s0092-8674(03)00421-5. 44. Burnat M, Schleiff E, Flores E. 2014. Cell envelope components influencing filament length in the heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120. J Bacteriol 196:4026 – 4035. https://doi .org/10.1128/JB.02128-14. 45. Schaechter M, Maaloe O, Kjeldgaard NO. 1958. Dependency on medium and temperature of cell size and chemical composition during balanced growth of Salmonella typhimurium. J Gen Microbiol 19:592– 606. https:// doi.org/10.1099/00221287-19-3-592. 46. Dai X, Shen Z, Wang Y, Zhu M. 2018. Sinorhizobium meliloti, a slowgrowing bacterium, exhibits growth rate dependence of cell size under nutrient limitation. mSphere 3:e00567-18. https://doi.org/ 10.1128/mSphere.00567-18. 47. Männik J, Bailey MW. 2015. Spatial coordination between chromosomes and cell division proteins in Escherichia coli. Front Microbiol 6:306. https://doi.org/10.3389/fmicb.2015.00306. 48. Li Z, Kinosita Y, Rodriguez-Franco M, Nußbaum P, Braun F, Delpech F, Quax TEF, Albers S-V. 2019. Positioning of the motility machinery in halophilic Archaea. mBio 10:e00377-19. https://doi.org/10.1128/mBio .00377-19. 49. Lange R, Hengge-Aronis R. 1991. Growth-phase regulated expression of bolA and morphology of stationary-phase Escherichia coli cells are controlled by the novel sigma factor sigma s. J Bacteriol 173:4474 – 4481. https://doi.org/10.1128/jb.173.14.4474-4481.1991. Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Anabaena Morphology September/October 2020 Volume 5 Issue 5 e00747-20 msphere.asm.org 23 50. Wortinger MA, Quardokus EM, Brun YV. 1998. Morphological adaptation and inhibition of cell division during stationary phase in Caulobacter crescentus. Mol Microbiol 29:963–973. https://doi.org/10 .1046/j.1365-2958.1998.00959.x. 51. López-Igual R, Flores E, Herrero A. 2010. Inactivation of a heterocystspecific invertase indicates a principal role of sucrose catabolism in heterocysts of Anabaena sp. J Bacteriol 192:5526 –5533. https://doi.org/ 10.1128/JB.00776-10. 52. Mitschke J, Vioque A, Haas F, Hess WR, Muro-Pastor AM. 2011. Dynamics of transcriptional start site selection during nitrogen stress-induced cell differentiation in Anabaena sp. PCC7120. Proc Natl Acad Sci U S A 108:20130 –20135. https://doi.org/10.1073/pnas.1112724108. 53. Picossi S, Flores E, Herrero A. 2014. ChIP analysis unravels an exceptionally wide distribution of DNA binding sites for the NtcA transcription factor in a heterocyst-forming cyanobacterium. BMC Genomics 15:22. https://doi.org/10.1186/1471-2164-15-22. 54. Muro-Pastor AM, Valladares A, Flores E, Herrero A. 2002. Mutual dependence of the expression of the cell differentiation regulatory protein HetR and the global nitrogen regulator NtcA during heterocyst development. Mol Microbiol 44:1377–1385. https://doi.org/10 .1046/j.1365-2958.2002.02970.x. 55. Herrero A, Burnat M. 2014. Cyanophycin, a cellular nitrogen reserve material, p 211–219. In Flores E, Herrero A (ed), The cell biology of Cyanobacteria. Caister Academic Press, Norfolk, UK. 56. Lazaro S, Fernandez-Pinas F, Fernandez-Valiente E, Blanco-Rivero A, Leganes F. 2001. pbpB, a gene coding for a putative penicillin-binding protein, is required for aerobic nitrogen fixation in the cyanobacterium Anabaena sp. strain PCC7120. J Bacteriol 183:628 – 636. https://doi.org/ 10.1128/JB.183.2.628-636.2001. 57. Leganés F, Blanco-Rivero A, Fernández-Piñas F, Redondo M, Fernández- Valiente E, Fan Q, Lechno-Yossef S, Wolk PC. 2005. Wide variation in the cyanobacterial complement of presumptive penicillin-binding proteins. Arch Microbiol 184:234 –248. https://doi.org/10.1007/s00203-005-0046-8. 58. Berendt S, Lehner J, Zhang YV, Rasse TM, Forchhammer K, Maldener I. 2012. Cell wall amidase AmiC1 is required for cellular communication and heterocyst development in the cyanobacterium Anabaena PCC 7120 but not for filament integrity. J Bacteriol 194:5218 –5227. https://doi.org/ 10.1128/JB.00912-12. 59. Zheng Z, Omairi-Nasser A, Li X, Dong C, Lin Y, Haselkorn R, Zhao J. 2017. An amidase is required for proper intercellular communication in the filamentous cyanobacterium Anabaena sp. PCC 7120. Proc Natl Acad Sci U S A 114:E1405–E1412. https://doi.org/10.1073/pnas.1621424114. 60. Bornikoel J, Staiger J, Madlung J, Forchhammer K, Maldener I. 2018. LytM factor Alr3353 affects filament morphology and cell– cell communication in the multicellular cyanobacterium Anabaena sp. PCC 7120. Mol Microbiol 108:187–203. https://doi.org/10.1111/mmi.13929. 61. Frías JE, Flores E, Herrero A. 1994. Requirement of the regulatory protein NtcA for the expression of nitrogen assimilation and heterocyst development genes in the cyanobacterium Anabaena sp. PCC7120. Mol Microbiol 14:823– 832. https://doi.org/10.1111/j.1365-2958.1994.tb01318.x. 62. Camargo S, Valladares A, Flores E, Herrero A. 2012. Transcription activation by NtcA in the absence of consensus NtcA-binding sites in an Anabaena heterocyst differentiation gene promoter. J Bacteriol 194: 2939 –2948. https://doi.org/10.1128/JB.05994-11. 63. Mackinney G. 1941. Absorption of light by chlorophyll solutions. J Biol Chem 140:109 –112. 64. Cai Y, Wolk CP. 1990. Use of a conditionally lethal gene in Anabaena sp. strain PCC 7120 to select for double recombinants and to entrap insertion sequences. J Bacteriol 172:3138 –3145. https://doi.org/10.1128/jb .172.6.3138-3145.1990. 65. Olmedo-Verd E, Muro-Pastor AM, Flores E, Herrero A. 2006. Localized induction of the ntcA regulatory gene in developing heterocysts of Anabaena sp. strain PCC 7120. J Bacteriol 188:6694 – 6699. https://doi .org/10.1128/JB.00509-06. 66. Elhai J, Wolk CP. 1988. A versatile class of positive-selection vectors based on the nonviability of palindrome-containing plasmids that allows cloning into long polylinkers. Gene 68:119 –138. https://doi.org/10.1016/ 0378-1119(88)90605-1. 67. Merino-Puerto V, Herrero A, Flores E. 2013. Cluster of genes that encode positive and negative elements influencing filament length in a heterocyst-forming cyanobacterium. J Bacteriol 195:3957–3966. https:// doi.org/10.1128/JB.00181-13. 68. Black TA, Cai YP, Wolk CP. 1993. Spatial expression and autoregulation of hetR, a gene involved in the control of heterocyst development in Anabaena. Mol Microbiol 9:77– 84. https://doi.org/10.1111/j.1365-2958 .1993.tb01670.x. 69. Mohamed A, Jansson C. 1989. Influence of light on accumulation of photosynthesis-specific transcripts in the cyanobacterium Synechocystis 6803. Plant Mol Biol 13:693–700. https://doi.org/10.1007/BF00016024. 70. Paz-Yepes J, Herrero A, Flores E. 2007. The NtcA-regulated amtB gene is necessary for full methylammonium uptake activity in the cyanobacterium Synechococcus elongatus. J Bacteriol 189:7791–7798. https://doi .org/10.1128/JB.00404-07. 71. Vioque A. 1997. The RNase P RNA from cyanobacteria: short tandemly repeated repetitive (STRR) sequences are present within the RNase P RNA gene in heterocyst-forming cyanobacteria. Nucleic Acids Res 25: 3471–3477. https://doi.org/10.1093/nar/25.17.3471. 72. Valladares A, Maldener I, Muro-Pastor AM, Flores E, Herrero A. 2007. Heterocyst development and diazotrophic metabolism in terminal respiratory oxidase mutants of the cyanobacterium Anabaena sp. strain PCC 7120. J Bacteriol 189:4425– 4430. https://doi.org/10.1128/ JB.00220-07. Downloaded from http://msphere.asm.org/ on February 3, 2021 at CSIC-CENTRO DE INVESTIGACIONES Velázquez-Suárez et al., Fig. S1 WT N2 LC 4 4 3 3 2 2 1 0 0 20 0 40 0 60 0 80 0 Time (h) 4 3 2 1 0 0 20 0 40 0 60 0 80 0 Time (h) 4 3 2 1 1 0 5 4 3 2 1 0 3 2,5 2 1,5 1 0,5 0 0 20 0 40 0 60 0 80 0 Time (h) 0 20 0 40 0 60 0 80 0 Time (h) 0 0 20 0 40 0 60 0 80 0 Time (h) -0,5 0 20 40 60 80 10 0 Time (h) WT NO3- HC WT NO3- LC WT NH4+ LC WT N2 HC LN (At/A0) LN (At/A0) LN (At/A0) Velázquez-Suárez et al., Fig. S2 Week 2 NO3 HC - NO - LC 3 NH + LC 4 N2 HC N2 LC R2 = 0.8137 7,00 8,00 6,50 7,00 6,00 6,00 5,50 5,00 0 0,2 0,4 0,6 0,8 1 Growth rate constant (day-1) 5,00 4,00 0 0,2 0,4 0,6 0,8 Growth rate constant (day-1) 10, 00 9,00 8,00 8,00 7,00 7,00 6,00 6,00 5,00 5,00 4,00 0 0,05 0,1 0,15 0,2 Growth rate constant (day-1) 4,00 0 0,02 0,04 0,06 0,08 Growth rate constant (day-1) 8,00 7,00 6,00 5,00 4,00 -0, 04 -0, 02 0 0,02 0,04 0,06 Growth rate constant (day-1) Week 1 NH + LC 4 NO - HC 3 N2 LC NO - LC 3 N2 HC R2 = 0.5161 Fastest exponential growth NH LC 4 + NO - HC 3 NO - LC 3 N2 LC N2 HC R2 = 0.5375 Week 3 NO - LC 3 NO - HC 3 NH + LC 4 N2 LC N2 HC R2 = 0.229 Week 4 NO - LC 3 NH + LC NO - HC 3 4 N2 LC N2 HC R2 = 0.073 Mean cell area m2 Mean cell area m2 Mean cell area m2 Velázquez-Suárez et al., Fig. S3 10 m 24 h PCC 7120 CSCV3 alr0088 all0085 SmR SpR 1 kb CSCV3 (NH4+) 168 h 96 h 72 h 48 h all0087 all0086 gfpmut2 Velázquez-Suárez et al., Fig. S4 all0087 all0086 D alr0084-2 / all0086-5 all0085-5 / alr0086-5 PCC 7120 1404 bp 612 bp CSCV2 2989 bp - 1 2 3 4 5 6 WT 1 2 3 4 5 6 WT PCC 7120 alr0088 all0085 alr0084 all0086-2/alr0088-2 all0087-8/alr0088-2 all0088-2 CSCV1 all0087-8 all0086-2 1 kb 3 kb 1.5 kb 1 kb alr0088 all0085 alr0084 all0085-1/all0087-7 all0086-6/all0087-7 CSCV4 all0087-7 all0086-6 All0085-1 C.K1 all0085-1/ all0087-7 all0086-6/ all0087-7 PCC 7120 1743 pb 784 pb CSCV4 2386 pb - alr0088 all0085 alr0084 all0086-5 all0085-5 alr0084-2 CSCV2 all0086-2/alr0088-2 all0087-8/alr0088-2 all0087 all0086 1 2 3 4 5 6 WT 1 2 3 4 5 6 WT all0087 all0086 C.K1 C.S3 1 2 3 4 WT 1 2 3 4 WT all0086-2 / alr0088-2 all0087-8 / alr0088-2 PCC 7120 1660 bp 857 bp CSCV1 2249 bp - 88 Chapter I Chapter II CHAPTER II: The role of Mre factors and cell division in peptidoglycan growth in the multicellular cyanobacterium Anabaena 89 Peptidoglycan Growth in Anabaena mBio FIG 3 Van-FL staining of Anabaena mreB, mreC, and mreD mutants. Cultures of strains CSCV1 (mreB), CSCV4 (mreC), and CSCV2 (mreD) were set and, at the indicated times, samples were stained with Van-FL and observed under a fl uorescence microscope. (A) Filaments were photographed. Van-FL fl uorescence (green) and brightfi eld images are shown. Arrows point to tilted fl uorescent bands and brackets to cell compartments of disparate sizes. (B) Lateral and septal fl uorescence was quanti fi ed and compared to the v a lu es o b t ai ned f o r t h e wil d - t ype ( presen t e d i n F i g . 1 B ) . Fo r t h e rat i o L / S , P val u es we r e , 0 . 01 for comparisons of the WT with CSCV1 or CSCV2 at any time, and with CSCV4 at 216 h, and P value was , 0 . 05 between t h e W T a n d CSC V 4 a t 48 h ( D a t a S et S1 ) . M a g n i fi c a t i on is t h e sa m e f o r all mic r o gr a p h s . construction in the dividing cells, could be detected. It is worth noting that some cultures, especially of CSCV8 after 24 h of incubation, presented cells bigger and rounder than the wild-type cells. This could indicate that GFP-MreD might interfere with MreD function. Nonetheless, the specificity of the GFP fluorescence localization in CSCV8 and its consistency with the signals produced by GFP-MreB (in CSCV6) and GFP-MreC (in CSCV7) support that GFP-MreD is reporting the physiological localization of MreD. In the three strains, the peripheral fluorescence became more diffuse after 48 h of incubation, and the septal signal also diminished (Fig. 2A). In summary, during active growth, MreB, MreC, and MreD localize throughout the cell periphery. Notably, they also localize to the septa that are under construction during cell division, remaining localized in the intercellular septal regions after compartmentalization of the resulting daughter cells. Influence of MreB, MreC, and MreD on PG growth. Van-FL staining was also performed in fi laments of strains containing only inactivated versions of mreB (CSCV1), mreC (CSCV4), or mreD (CSCV2), previously generated (30). Despite conspicuous alterations in fi lament and cell size and morphology (30), strong septal bands of alternating intensity that resemble the septal labeling pattern in the WT, could be observed in fi laments of the three mutants (Fig. 3A and Fig. S1). Indeed, the average septal fl uorescence intensity in the mutants was somewhat higher than in the wild type (compare Fig. 3B and Fig. 1B). As in the wild type, fl uorescence matching the progression of septum formation during cell division could be detected in some cells. However, in the mutants, the septal July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 6 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio FIG 4 Localization of FtsZ in Anabaena mreB, mreC, and mreD mutants. (A) Cultures of strains CSSC19 (ftsZ-gfp-mut2 in WT background), CSCV20 (ftsZ-gfp-mut2 in mreB background), CSCV21 (ftsZ- gfp-mut2 in mreC background) and CSCV22 (ftsZ-gfp-mut2 in mreD background) were set and, after 48 h, fi laments were visualized by confocal microscopy and photographed. GFP fl uorescence (green), merged GFP and cyanobacterial auto fl uorescence (red), and brightfi eld images are shown. To improve visibility, contrast is higher for the green image of CSCV22. (B) Immunolocalization with antibodies against Anabaena FtsZ in the wild-type and strains CSCV1 (mreB), CSCV4 (mreC), and CSCV2 (mreD) incubated for 48 h in liquid BG11 medium (initial cell density, 1 mg chlorophyll/mL). Arrows point to tilted fl uorescence bands and brackets to cell compartments of disparate sizes. Magni fi cation is the same for all micrographs in (A or B). bands frequently appeared tilted or separating compartments of different sizes. These alterations could explain the observed local deviations in the filament plane and heterogeneity of cell size in the mutants (30) (Fig. 3A and Fig. S1). Thus, MreB, MreC, and MreD are not needed for septal PG synthesis during cell division, but they appear to be needed for the correct positioning of the septal PG growth-plane. On the other hand, peripheral labeling of the mutant round cells was also detected in strains CSCV1, CSCV2, and CSCV4 (Fig. 3 and Fig. S1). Localization of FtsZ and ZipN in mre mutants. Given the observed instances of misplacement of the medial and septal bands of Van-FL staining in the mre mutants, we asked whether the absence of Mre proteins affected the localization of the cell division plane. We then studied the localization of FtsZ-rings and ZipN in the absence of MreB, MreC, or MreD. To study FtsZ, a gene construct, including an ftsZ-gfp reporter expressed from the P ftsZ promoter was transferred to strains CSCV1 (mreB), CSCV4 (mreC), and CSCV2 (mreD), generating strains CSCV20, CSCV21, and CSCV22, respectively. Strain CSSC19 expresses the ftsZ-gfp reporter in the WT background (33). As in CSSC19, fl uorescent FtsZ-rings were readily detected in strains CSCV20, CSCV21, and CSCV22. However, in the mutant backgrounds some rings were not parallel to those of neighboring cells, and the distance between two consecutive Z-rings was irregular (Fig. 4A). In addition, FtsZ was detected by immunolocalization with antibodies against Anabaena FtsZ in the mutants CSCV1, CSCV4, and CSCV2 in comparison to the wild type. Consistent with the results obtained with GFP fusions, some aberrantly located rings could be detected in the mutants (Fig. 4B). To study the localization of ZipN in the absence of MreB, MreC, or MreD, an sfgfpzipN reporter gene expressed from the PzipN promoter was transferred to strains CSCV1, CSCV4, and CSCV2, generating strains CSCV14, CSCV15 and CSCV16, respectively. In filaments of strain CSAV39 (sfgfp-zipN in the wild-type background, (40)), fluorescence July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 7 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio FIG 5 Localization of ZipN in Anabaena mreB, mreC and mreD mutants. Cultures of strains CSAV39 (sfgfpzipN in WT background), CSCV14 (sfgfp-zipN in mreB background), CSCV15 (sfgfp-zipN in mreC background) and CSCV16 (sfgfp-zipN in mreD background) were set and, after 24 h, filaments were visualized by confocal microscopy and photographed. GFP fluorescence (green), merged GFP and cyanobacterial autofluorescence (red), and bright-field images are shown. Arrows point to tilted fluorescence bands and brackets to cell compartments of disparate sizes. Magnification is the same for all micrographs. was detected in midcell bands and as bands or spots in recently matured septa. In fresh cultures of strains CSCV14, CSCV15, and CSCV16, midcell and septal fl uorescent bands or spots could be detected frequently. However, especially in CSCV15, cells exhibiting diffuse fl uorescence over the cell periphery were also found (Fig. 5). As with FtsZ-GFP, some neighboring cells exhibited tilted GFP-ZipN bands, and at some places, the distances between consecutive bands in the fi lament varied. These results suggest that the misplaced Van-FL-labeled septal bands observed in the mre mutants corresponded to places where cell division took place at a plane tilted with regard to that of the previous division or deviated from the midcell position. PG growth in Anabaena mutants was impaired in FtsZ or ZipN. We have previously described Anabaena derivatives that conditionally down-express ftsZ (strain CSFR18; (35)) or zipN (strain CSL109; (34)) from PND, a synthetic promoter that directs low expression levels in the presence of ammonium. Upon transfer of filaments grown with nitrate (permissive conditions for PND-ftsZ or PND-zipN expression) to a medium with ammonium (restrictive conditions), strains CSFR18 and CSL109 underwent drastic alterations in cell morphology and extensive cell lysis (see representative images in Fig. 6 and 7, respectively). In the two mutants, morphological alterations preceding cell lysis included cell elongation, bulging of the central cell region, and drastic cell swelling. In CSFR18, long bulged cells could be observed, initially forming filaments, and later as single cells (which likely results from lysis of accompanying cells in the filament). Strain CSL109 presented giant spherical-like cells or ellipsoidal cells in which the axis perpendicular to the filament is longer than that parallel to the filament. These and previous observations (33–35) show that in Anabaena impairment of cell division leads to unchecked cell enlargement with restricted elongation producing giant cells with aberrant shapes. Van-FL staining was performed in strains CSFR18 and CSL109. In filaments incubated with nitrate (permissive conditions), Van-FL staining produced weak peripheral, and strong septal fluorescence signals similar to those observed in the WT (not shown). Upon transfer of strain CSFR18 to medium with ammonium (restrictive conditions for PND- ftsZ expression), it showed weak peripheral labeling in the still slightly enlarged cells, and strong dispersed peripheral fluorescence, including occasional discrete spots, in the aberrantly more enlarged cells, which were increasingly abounding (see some July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 8 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio 4 FIG 6 Van-FL staining of strain CSFR18 impaired in FtsZ expression. Filaments of strain CSFR18 (P ND -ftsZ) (see the text for details) grown in BG11 medium (permissive conditions) were transferred to BG11 0 1 NH 1 medium (restrictive conditions for ftsZ expression) and incubated under culture conditions. At the indicated times, samples were stained with Van-FL, observed under a fl uorescence microscope, and photographed. Van-FL fl uorescence (green) and brightfi eld images are shown. Magni fi cation is the same for all micrographs. examples in Fig. 6). After prolonged incubation under these conditions, most of the remaining cells appeared very deformed and had lost fl uorescence, although some very aberrant, mostly isolated cells with conspicuous peripheral fl uorescence could also be observed (Fig. 6). In strain CSL109 transferred to ammonium-containing medium (restrictive conditions for P ND -zipN expression), some cells that began to adopt aberrant engrossed shapes exhibited strong peripheral labeling. However, upon prolonged incubation, most cells showed very weak, progressively fading labeling (Fig. 7). Regarding midcell staining, no Van-FL labeling was detected in the enlarged cells of CSFR18 or CSL109. Regarding septal labeling, Van-FL bands could be initially detected in the intercellular regions of slightly elongated cells, but they did not increase in width in parallel to the observed increase in cell volume (note enlarged cells connected by tiny septal fl uorescent spots in Fig. 6, 48 and 168 h, and in Fig. 7, 168 and 216 h). In CSL109, septal fl uorescence was promptly lost, and in CSFR18 many aberrant cells appeared also devoid of septal fl uorescence. Thus, the observed septal Van-FL signals appear to correspond to cell division events initiated before restrictive conditions for ftsZ, or zipN expression were imposed. In summary, in the two strains, the transfer to restrictive conditions leads to the cessation of divisome-associated midcell PG incorporation, but also decreased PG incorporation at the septa and the cell periphery, although a phase of strong peripheral PG incorporation concomitant with severe morphological alterations took place. BACTH analysis of protein interactions. Given the alterations of Z-ring localization in mreB, mreC, and mreD mutants, and the localization of Mre proteins also to division rings in dividing cells, we tested direct interactions between Mre and cell division proteins. We used a bacterial two-hybrid (BACTH) assay with MreB, MreC, MreD, and the divisome proteins FtsZ, ZipN, ZipS (a cyanobacterium-speci fi c cell division factor, (41)), Alr0487 (putative SepF), Alr1706 (putative FtsE), All1757 (putative FtsX), FtsQ, All7666 (putative FtsK), Alr0718 (putative divisome transpeptidase FtsI), and All0154 (putative divisome glycosyl-transferase FtsW). Besides MreB self-interactions, fully signi fi cant, although weak, positive interactions were detected with the pairs: T18-MreB/T25-FtsQ and MreD-T18/T25-FtsI (Student ’ s t tests , 0.02 for comparisons with all the three July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 9 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio 4 FIG 7 Van-FL staining of strain CSL109 impaired ZipN expression. Filaments of strain CSL109 (P ND - zipN) (see the text for details) grown in BG11 medium (permissive conditions) were transferred to BG11 0 1 NH 1 medium (restrictive conditions for zipN expression) and incubated under culture conditions. At the indicated times, samples were stained with Van-FL, observed under a fl uorescence microscope, and photographed. Van-FL fl uorescence (green), cyanobacterial auto fl uorescence (red), and brightfi eld images are shown. Magni fi cation is the same for all micrographs. controls). In addition, the results obtained suggested interactions between MreD-T18/ T25-FtsW; and MreD-T18/T25-FtsQ (Fig. 8A). DISCUSSION In the rod-shaped cells of Anabaena sp. strain PCC 7120, Van-FL staining produced weak continuous peripheral fluorescence in the cylindrical part of the cells (Fig. 1; see also reference (32)). This pattern differs from that of weak tilted bands and strong peripheral dots detected in other rod-shaped bacteria such as Bacillus subtilis (25) or C. crescentus (9), which is indicative of a disperse mode of PG incorporation in the cylindrical part of the cell. In B. subtilis, this peripheral PG incorporation is independent of MreB but dependent on Mbl (a MreB homolog), MreC, and MreD (42). Both MreB (8) and Mbl (43) form filaments that rotate around the cell's long axis, and MreC and MreD localize in a similar banded pattern (42). In E. coli, MreB also forms peripheral filaments, whereas MreC and MreD were uniformly localized through the cell periphery (44). In C. crescentus, MreB and MreC form nonoverlapping periplasmic spirals along the perimeter of the cell and, in contrast to predivisional MreB localization at midcell, MreC is not present at the division site (45). Here, we show that in Anabaena, MreB, MreC, and MreD localize throughout the cell periphery (Fig. 2). In the presence of combined nitrogen none of the three mre genes is essential (30), but mutants lacking any of them form rounded cells or even cells in which the longer axis is perpendicular to the filament. These results indicate an involvement of MreB, MreC, and MreD in cell July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 10 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio FIG 8 BACTH assay of interactions of Mre and divisome proteins. (A) Interactions of protein pairs produced in E. coli were assayed by measurements of b-galactosidase activity in liquid cultures incubated at 30°C. The topology of each fusion is indicated by the order of components (T18-protein and T25-protein denote the corresponding adenylate cyclase domain fused to the N terminus of the tested protein, whereas protein-T18 and protein-T25 denote fusions to the C terminus). Data are the mean and standard deviation of the indicated number of determinations of the activity (nmol ONP min 2 1 mg protein 2 1 ) with the indicated protein fused to T25 (or the empty vectors pKNT25 or pKT25) and the indicated protein fused to T18 (or the empty vectors pUT18C or pUT18). Signi fi cance of differences was assessed by Student ’ s t tests: p refers to the test against the strain containing both empty vectors; p a and p b denote tests against the strains expressing each fused protein and the complementary empty vector. (B) Schematic of the positive interactions detected: colored solid lines d e no t e in tera ctio n s wi t h p , p a , a n d p b , 0. 02 , a n d d as h e d li n es d enote i n t e r acti on s wit h p an d eit h e r p a or p b , 0.02. Previously deduced interactions between divisome proteins (33, 34) are depicted in black. elongation ((30); Fig. 3). However, the detection of Van-FL peripheral labeling in the round cells of the mutants suggests that some PG incorporation, likely serving maintenance functions, can take place in them. It is worth noting that the pattern of Van-FL staining shown here for Anabaena, with weak peripheral and strong septal labeling, is similar to that observed with the fluorescent amino acid HADA that can be incorporated into PG (46), thus precluding that our observations were due to poor access of Van- FL to the Anabaena periplasm or Van-FL staining artifacts. Anabaena also showed Van-FL staining in the progressing septum under construction during cell division, and intense bands in the intercellular regions along the fi lament, even under phases of slow growth (Fig. 1), suggesting that septal growth is maintained during the cell cycle. In addition, the observed alternance of septal Van-FL band intensities (stronger in the older septa and weaker in the more recent ones) is indicative of the persistence of septal PG growth or remodeling after cell division has been completed. During the phases of less active growth, the intensity of fl uorescence in septal bands homogenize through the fi lament and remains strong, also suggesting septal PG growth/remodeling after cell division. This contrasts with the pattern in Bacillus in which little polar Van-FL staining is observed in mature cells (25), which is consistent with a narrow period of septal PG growth after which the cell poles remain inert during most of the cell cycle. In E. coli, preseptal elongation from the midcell is observed only at the initiation of July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 11 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio cell division (12), and after completion, the cell poles are inert for PG synthesis (47). In C. crescentus zonal PG synthesis near the midcell takes place between a period of dispersed lateral growth and cell constriction (13). In S. pneumoniae, PG synthesis is halted after completion of cell division and the resulting cell poles become inert (25). In contrast, in the case of C. glutamicum, PG growth directed by the divisome remains active after cell division has completed, so that growth continues at the cell poles leading to a rod-like shape (25). These models differ from the situation in Anabaena, in which the persistent and strong septal labeling appears restricted to the intercellular regions of the filament, consistent with the thickening of the PG layer observed in these regions. In Anabaena, our observations with strain CSCV6 (sfgfp-mreB) suggest the localization of MreB throughout the cell periphery, including the septal regions as well as in the divisome (Fig. 2). These observations could be compared with two previous reports: one with a gfp-mreBCD construct overexpressed from a heterologous promoter in a replicative plasmid, which gave conspicuous fluorescent spots at the cell poles (31), and another with gfp-mreB expressed from PmreB in a replicative plasmid, which showed filaments dispersed through the cell without any directional preference (48). On the other hand, strains CSCV7 (sfgfp-mreC) and CSCV8 (sfgfp-mreD) allowed the localization of MreC and MreD, respectively, at midcell matching the growing septum in dividing cells, even at advanced stages of septum formation, and in the intercellular regions between adjacent cells in the filament after septum closure as well as in nondividing cells (Fig. 2). Indeed, in mreB, mreC and mreD mutants, intercellular septa are wider and septal Van-FL labeling signal is stronger than in the wild type, likely contributing to increase filament length and rigidity (compare Fig. 1 and 3; (30)). These results suggest that, although MreB, MreC, and MreD are not required for septal PG synthesis, they participate in septum constriction and intercellular PG maintenance at all stages of the cell cycle. In contrast to the wild type, in which the new septa localize at midcell in parallel to that of the previous division and perpendicular to the fi lament long axis, in a mreB, mreC, or mreD mutant septal PG bands are frequently found tilted with regard to those of previous divisions, a disposition never observed in the wild type, or separating compartments with different sizes (Fig. 3). These alterations appear to cause the observed variability of cell size and distortions in the fi lament plane, even leading to apparent branches, in the mutants (30) (Fig. 3 and Fig. S1). Furthermore, we have shown here that during cell division, whereas in the wild-type the proteins FtsZ and ZipN localized to parallel Z-rings in consecutive cells, in the mreB, mreC, and mreD mutants some tilted and diverted FtsZ (Fig. 4) and ZipN (Fig. 5) rings are observed. Also, especially in the mreC mutant, fl uorescence from GFP-ZipN appeared somewhat dispersed (Fig. 5). Thus, in the mutants the alterations in the localization of the septal PG incorporation likely originate from misplacement of the Z-ring. Therefore, although MreB, MreC, and MreD are not required for divisome-associated PG growth, they are needed for its correct geometric localization. The altered localization of the division plane might result, at least in part, from the increased cell volume and rounded cell morphology of the mre mutants. However, a more direct effect of the Mre proteins localized at the division site is possible, as supported also by our BACTH results suggesting direct interactions involving MreB and MreD and the divisome components FtsQ, FtsW, and FtsI (Fig. 8). In Anabaena, the morphological alterations observed when the initial steps of cell division are inhibited differ from those reported for other bacteria. Hence, in E. coli FtsZ depletion (49) or mutation (50) leads to the formation of long fi laments; in B. subtilis, FtsZ depletion leads to cell elongation but does not affect growth rate (25), and in C. crescentus FtsZ depletion leads to cell elongation by a disperse mode of PG synthesis over the sidewalls (13). Thus, in E. coli, B. subtilis, and C. crescentus, lateral PG growth to elongate the cells can be maintained in the absence of cell division. In contrast, Anabaena derivatives that conditionally down-express FtsZ or ZipN show strong deformations of the rod-shape morphology normally exhibited by the wild type, or by the mutants under permissive conditions, forming enlarged cells with irregular width (larger in the central region than in the polar regions), more abundant under FtsZ July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 12 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio x shortage, and giant cells with inverted polarity in the filaments that progress to lysis under ZipN shortage (34, 35) (Fig. 6 and 7). Upon transfer to repressive conditions, Van- FL staining of strains CSFR18 (PND-ftsZ) and CSL109 (PND-zipN) revealed a phase of strong abnormal PG incorporation through the periphery of aberrant cells, apparently followed by cessation of PG incorporation. Interestingly, it has been previously described that treatment of Anabaena with aztreonam (an inhibitor of FtsI) leads, in the short term, to strong peripheral HADA incorporation (46). Moreover, PG-growth in intercellular septa also ceased leading to the presence of minuscule septa, in comparison to the increased cell volume, joining the aberrant enlarged cells. In summary, in Anabaena, inhibition of the initial steps of cell division also inhibits the incorporation of PG at mature filament septa that are observed in the wild type and affects the peripheral PG growth preventing sustained cell elongation without widening. All present and previous results are consistent with a model in which three modes of PG growth operate in Anabaena, namely, peripheral, divisional, and intercellular septal. The peripheral mode requires MreB, MreC, and MreD. The divisional and septal PG growth could proceed in the absence of MreB, MreC, and MreD, although these factors have a role in the correct localization of divisional PG growth at midcell and the correct septal PG dimensions. As was the case for many septal proteins of Anabaena, we propose that the localization of Mre proteins to division rings serves to place them in the mature intercellular septa, where the proteins remain after cell division has completed. Septal Mre proteins might represent a topological determinant and regulate the extent of peripheral PG growth. Finally, continuous activity of PG growth and remodeling in the mature intercellular septa could contribute to the maintenance of the inserted septal junction arrays that mediate intercellular communication. MATERIALS AND METHODS Strains and growth conditions. Anabaena sp. Strain PCC 7120 and mutant strains were grown in a BG11 medium (containing NaNO 3 as a nitrogen source) (51). Cultures were incubated at 30°C with illumination (12 mE m 2 2 s 2 1 white light emitted from Osram LED lamps 16.4 W/4000K) in Erlenmeyer fl asks with shaking or in plates of medium solidi fi ed with 1% Difco agar. For the mutants, media were supplemented with antibiotics: spectinomycin dihydrochloride pentahydrate (Sp) and streptomycin sulfate (Sm) at 5 mg mL 2 1 each in solid medium or 2.5 mg mL 2 1 each in liquid medium (CSCV2, CSCV6, CSCV7, CSCV8, CSFR18, CSSC19, CSL109, CSAV39), with neomycin sulfate (Nm) at 25 mg mL 2 1 in solid medium or 5 mg mL 2 1 in liquid medium (CSCV1, CSCV4), or with Sm, Sp, and Nm (CSCV14, CSCV15, CSCV16, CSCV20, CSCV21, CSCV22). Unless otherwise speci fi ed, for the experiments described in this work BG11- grown fi laments of the indicated strains were transferred to (at a cell density of 0.5 mg chlorophyll/mL) and incubated in a BG11-based medium lacking NaNO 3 and supplemented with 4 mM NH 4 Cl and 8 mM TES-NaOH buffer, pH 7.5. It should be noted that, except for experiments with strains CSFR18 and CSL109, similar results were obtained with fi laments incubated with nitrate instead of ammonium (unpublished data). The growth of cultures was monitored by measuring the absorbance at 750 nm in aliquots withdrawn at the indicated times. The chlorophyll content of the cultures was determined after extraction with methanol (52). In Anabaena, 1 mg chlorophyll corresponds to ca. 3.3 10 6 cells (53). Plasmid and strain constructions. Strains CSCV6, CSCV7, and CSCV8 bear an sfgfp-mreB, sfgfp-mreC, or sfgfp-mreD fusion gene, respectively, expressed from the native mreBCD operon promoter (PmreB), in a heterologous genomic locus, thrS2, which is dispensable under normal growth conditions (54). These strains were generated in two steps. In a first step, the construct PmreB-sfgfp was introduced into the coding sequence of thrS2, generating strain CSCV5. For that, the sequence of PmreB was amplified from Anabaena genomic DNA with primers all0087-20/all0087-21 (all oligodeoxynucleotide primers used in this work are listed in Table S2), and the sfgfp from plasmid pCSAL39 (55) with primers SF-GFP-F2/SFGFP- R2 (plasmids used in this work are listed in Table S1). Both DNA fragments were joined together by overlapping PCR and introduced into previously cloned thrS2 sequence in plasmid pMBLT (Canvas), generating plasmid pCSCV9. (To generate a site for introducing the PmreB-sfgfp construct, the plasmid containing thrS2 was amplified with primers all4723-3/all4723-4, both corresponding to internal thrS2 sequences and ending with NsiI restriction sites.) The insert of pCSCV9 was transferred to the conjugative vector pRL277 including genes for SmR/SpR and sacB, encoding susceptibility to sucrose for positive selection (56), generating plasmid pCSCV10. Finally, pCSCV10 was transferred to Anabaena by conjugation selecting first for resistance against Sm and Sp, and later for sensitivity to sucrose. One clone that bore the PmreB-sfgfp construct inserted on thrS2 by double crossover (verified by PCR) was selected (strain CSCV5). In a second step, a DNA fragment encoding sfgfp lacking the stop codon (amplified with primers SFGFP-F/SF-GFP-R) was fused by overlapping PCR to fragments encoding mreB (amplified with all0087- 22/all0087-23), mreC (all0086-13/all0086-14) or mreD (all0085-8/all0085-9), preceded by a 5-Gly encoding sequence. The resulting fusions were cloned into the conjugative vector pCSV3 including determinants for SmR/SpR (57), generating plasmids pCSCV11, pCSCV12, and pCSCV13, which were transferred to strain July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 13 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 14 x x CSCV5. Clones bearing plasmids pCSCV11, pCSCV12, or pCSCV13 inserted by single crossover into thrS2 were selected and veri fi ed by PCR. Strains CSCV14, CSCV15, and CSCV16 bear the fusion sfgfp-zipN expressed from the P zipN promoter, together with a native copy of P zipN -zipN, in a CSCV1, CSCV4, or CSCV2 background, respectively. To generate them, the pCSV3-based plasmid pCSAV285, which includes the P zipN -sfgfp-zipN construct (40), was transferred to strains CSCV1 and CSCV4 with selection for Sm R /Sp R . On the other hand, the insert of pCSAV285 was cloned into the KpnI site of pRL424 (58) generating plasmid pCSCV37, which was transferred to strain CSCV2 with selection for Nm R . One clone resulting from each conjugation that had inserted the transferred plasmid by a single crossover into the zipN locus leaving an intact P zipN -zipN gene (veri fi ed by PCR) was selected. Strains CSCV20, CSCV21, and CSCV22 bear the fusion ftsZ-gfpmut2 expressed from the PftsZ promoter, together with a native copy of PftsZ-ftsZ, in a CSCV1, CSCV4, or CSCV2 background, respectively. To generate them, the pRL277-based plasmid pCSSC39, which includes the PftsZ-ftsZ-mut2gfp construct (33) was transferred to strains CSCV1 and CSCV4 with selection for SmR/SpR. On the other hand, the insert of pCSSC39 was cloned into the SacI/XhoI sites of pRL278 (59) generating plasmid pCSCV39, which was transferred to strain CSCV2 with selection for NmR. One clone resulting from each conjugation that had inserted the transferred plasmid by a single crossover into the ftsZ locus leaving an intact PftsZ-ftsZ (verified by PCR) was selected. Van-FL staining and quanti fi cation. For Van-FL staining, fi laments were suspended in a liquid medium supplemented with 2 mg/mL Vancomycin-FL (Bodipy-FL conjugate, Invitrogen) and incubated for 1 h in the dark with shaking at 30°C. Filaments were washed twice with liquid medium and spotted in agar. Lateral and septal fl uorescence was quanti fi ed with ImageJ (60) processing of fl uorescence images by collecting total fl uorescence in manually de fi ned equal square sections (0.41 mm in length) at the periphery and the intercellular septa of each counted cell. For each cell, lateral fl uorescence was calculated as the mean of the values of four sections, and septal fl uorescence as the mean of two sections, one at each cell pole. Twenty to thirty cells were counted for each strain and condition, and the average values were calculated. ImageJ was used also for recording fl uorescence along determined fi lament areas (Fig. 1C) and for quanti fi cation of septal width and total septal fl uorescence using manually de fi ned sections including the whole septa (Fig. 1D). BACTH analysis. BACTH assays based on the reconstitution of adenylate cyclase from Bordetella pertussis (61) were performed. Genes were amplified by PCR using Anabaena DNA as the template and oligonucleotide pairs: all0087-10/all0087-11 (T25-MreB), all0087-12/all0087-11 (T18-MreB), all0087-12/ all0087-13 (MreB-T25, MreB-T18), all0086-17/all0086-18 (T25-MreC), all0086-19/all0086-18 (T18-MreC), all0085-12/all0085-13 (MreD-T25, MreD-T18), all0154-11b/all0154-13 (FtsW-T18), all1616-4/all1616-5 (T25-ZipS, T18-ZipS), all7666-3/all7666-2 (T18-FtsK) and all7666-3/all7666-4 (FtsK-T25, FtsK-T18), alr1706- 4/alr1706-2 (FtsE-T18). The resulting PCR products, which were fl anked by PstI/EcoRI ends, were cloned in pUT18, pUT18C, pKNT25 or pKT25 digested with the same enzymes, producing fusions to the 5 9 or 3 9 ends of the genes encoding the adenylate cyclase T18 and T25 fragments. All the resulting plasmids were veri fi ed by sequencing. The fusions encoding MreC-T18, MreC-T25, T18-MreD, and T25-MreD could not be cloned. Other fusions to the genes ftsZ, zipN, sepF, zipS, ftsX, ftsI, ftsQ, ftsW, and ftsE were as previously described (33 – 35). Plasmids were transformed into E. coli XL1-Blue for ampli fi cation. Isolated plasmids were cotransformed into strain BTH101 (cya-99), and the transformants were plated on a solid LB medium containing selective antibiotics and 1% glucose. Cotransformants were grown in a liquid medium in the presence of IPTG plus antibiotics supplemented with o-nitrophenol-b-galactoside. The onitrophenol produced per mg of protein versus time was represented, and the b-galactosidase activity was deduced from the slope of the linear function. Microscopy. Van-FL fl uorescence was visualized with a Leica DM6000B fl uorescence microscope and a FITCL5 fi lter (excitation band-pass, 480/40; emission band-pass, 527/30), and photographed with an ORCA-ER camera (Hamamatsu). For immunolocalization with antibodies against Anabaena FtsZ (35), fi laments were treated as speci fi ed in reference (34) and visualized by fl uorescence microscopy as above. GFP fl uorescence was monitored with an Olympus TCS SP2 confocal laser-scanning microscope equipped with an HCX PLAN-APO 63 1.4 NA oil immersion objective (excitation, 488-nm; collection, 500 to 540 nm for GFP or 630 to 700 nm for cyanobacterial auto fl uorescence) or with an Olympus FLUOVIEW FV3000 (hyper-resolution) confocal laser-scanning microscope equipped with a UPlanApo 60 1.5 NA oil immersion objective (excitation, 488-nm; collection 500 to 540 nm for GFP or excitation, 640 nm; collection, 650 to 750 nm for cyanobacterial autofluorescence). SUPPLEMENTAL MATERIAL Supplemental material is available online only. DATA SET S1, XLSX file, 0.02 MB. FIG S1, PDF file, 0.2 MB. TABLE S1, DOCX file, 0.02 MB. TABLE S2, DOCX file, 0.02 MB. ACKNOWLEDGMENTS We thank Enrique Flores for their suggestions and critical reading of the manuscript. The work was supported by grant BFU2016-77097-P funded by MCIN/AEI/10.13039/ 501100011033/and grant P20-00032 funded by Junta de Andalucía and FEDER. C.V.-S. Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Peptidoglycan Growth in Anabaena mBio was the recipient of a Formación de Personal Investigador (FPI) contract from the Spanish Government. Author contributions were as follows: conceptualization-A.H.; investigation-C.V.-S., A. V., and I.L.; writing (original draft)-A.H.; writing (review and editing)-A.H., I.L., C.V.-S., and A.V.; funding acquisition-A.H. and I.L.; and supervision-A.H. and I.L. The authors have no conflict of interest to declare. REFERENCES 1. Lovering AL, Safadi SS, Strynadka NCJ. 2012. Structural perspective of peptidoglycan biosynthesis and assembly. Annu Rev Biochem 81:451–478. https://doi.org/10.1146/annurev-biochem-061809-112742. 2. Egan AJF, Errington J, Vollmer W. 2020. Regulation of peptidoglycan synthesis and remodelling. Nat Rev Microbiol 18:446–460. https://doi.org/10 .1038/s41579-020-0366-3. 3. Bisson-Filho A, Hsu Y, Squyres G, Kuru E, Wu F, Jukes C, Sun Y, Dekker C, Holden S, VanNieuwenhze MS, Brun YV, Garner EC. 2017. Treadmilling by FtsZ filaments drives peptidoglycan synthesis and bacterial cell division. Science 355:739–743. https://doi.org/10.1126/science.aak9973. 4. Yang X, Lyu Z, Miguel A, McQuillen R, Huang KC, Xiao J. 2017. GTPase activity– coupled treadmilling of the bacterial tubulin FtsZ organizes septal cell wall synthesis. Science 355:744–747. https://doi.org/10.1126/science.aak9995. 5. Whitley KD, Jukes C, Tregidgo N, Karinou E, Almada P, Cesbron Y, Henriques R, Dekker C, Holden S. 2021. FtsZ treadmilling is essential for Z-ring condensation and septal constriction initiation in Bacillus subtilis cell division. Nat Commun 12:2448. https://doi.org/10.1038/s41467-021-22526-0. 6. Sun Y-J, Bai F, Luo A-C, Zhuang X-Y, Lin T-S, Sung Y-C, Shih Y-L, Lo C-J. 2021. Probing bacterial cell wall growth by tracing wall-anchored protein complexes. Nat Commun 12:2160. https://doi.org/10.1038/s41467-021-22483-8. 7. Martins A, Contreras-Martel C, Janet-Maitre M, Miyachiro MM, Estrozi LF, Trindade DM, Malospirito CC, Rodrigues-Costa F, Imbert L, Job V, Schoehn G, Attrée I, Dessen A. 2021. Self-association of MreC as a regulatory signal in bacterial cell wall elongation. Nat Commun 12:2987. https://doi.org/10 .1038/s41467-021-22957-9. 8. Errington J. 2015. Bacterial morphogenesis and the enigmatic MreB helix. Nat Rev Microbiol 13:241–248. https://doi.org/10.1038/nrmicro3398. 9. Divakaruni AV, Baida C, White CL, Gober JW. 2007. The cell shape proteins MreB and MreC control cell morphogenesis by positioning cell wall synthetic complexes. Mol Microbiol 66:174–188. https://doi.org/10.1111/j .1365-2958.2007.05910.x. 10. Hussain S, Wivagg CN, Szwedziak P, Wong F, Schaefer K, Izoré T, Renner LD, Holmes MJ, Sun Y, Bisson-Filho AW, Walker S, Amir A, Löwe J, Garner EC. 2018. MreB filaments align along greatest principal membrane curvature to orient cell wall synthesis. Elife 7:e32471. https://doi.org/10.7554/eLife.32471. 11. Laddomada F, Miyachiro MM, Dessen A. 2016. Structural insights into protein-protein interactions involved in bacterial cell wall biogenesis. Antibiotics 5:14. https://doi.org/10.3390/antibiotics5020014. 12. de Pedro MA, Quintela J, Höltje J, Schwarz H. 1997. Murein segregation in Escherichia coli. J Bacteriol 179:2823–2834. https://doi.org/10.1128/jb.179 .9.2823-2834.1997. 13. Aaron M, Charbon G, Lam H, Schwarz H, Vollmer W, Jacobs-Wagner C. 2007. The tubulin homologue FtsZ contributes to cell elongation by guiding cell wall precursor synthesis in Caulobacter crescentus. Mol Microbiol 64:938–952. https://doi.org/10.1111/j.1365-2958.2007.05720.x. 14. Pazos M, Peters K, Casanova M, Palacios P, VanNieuwenhze M, Breukink E, Vicente M, Vollmer W. 2018. Z-ring membrane anchors associate with cell wall synthases to initiate bacterial cell division. Nat Commun 9:5090. https://doi.org/10.1038/s41467-018-07559-2. 15. Figge RM, Divakaruni AV, Gober JW. 2004. MreB the cell shape determining bacterial actin homologue coordinates cell wall morphogenesis in Caulobacter crescentus. Mol Microbiol 51:1321–1332. https://doi.org/10 .1111/j.1365-2958.2003.03936.x. 16. Vats P, Rothfield L. 2007. Duplication and segregation of the actin (MreB) cytoskeleton during the prokaryotic cell cycle. Proc Natl Acad Sci U S A 104:17795–17800. https://doi.org/10.1073/pnas.0708739104. 17. R van der P, Verheul J, Vischer NO, Alexeeva S, Hoogendoorn E, Postma M, Banzhaf M, Vollmer W, den Blaauwen T. 2013. Colocalization and interaction between elongasome and divisome during a preparative cell division phase in Escherichia coli. Mol Microbiol 87:1074–1087. https://doi .org/10.1111/mmi.12150. 18. Vats P, Shih Y-L, Rothfield L. 2009. Assembly of the MreB-associated cytoskeletal ring of Escherichia coli. Mol Microbiol 72:170–182. https://doi.org/ 10.1111/j.1365-2958.2009.06632.x. 19. Fenton AK, Gerdes K. 2013. Direct interaction of FtsZ and MreB is required for septum synthesis and cell division in Escherichia coli. EMBO J 32: 1953– 1965. https://doi.org/10.1038/emboj.2013.129. 20. Yoshii Y, Niki H, Shiomi D. 2019. Division-site localization of RodZ is required for efficient Z ring formation in Escherichia coli. Mol Microbiol 111:1229–1244. https://doi.org/10.1111/mmi.14217. 21. van Teeseling MCF. 2021. Elongation at midcell in preparation of cell division requires FtsZ, but not MreB nor PBP2 in Caulobacter crescentus. Front Microbiol 12:732031. https://doi.org/10.3389/fmicb.2021.732031. 22. Monteiro JM, Fernandes PB, Vaz F, Pereira AR, Tavares AC, Ferreira MT, Pereira PM, Veiga H, Kuru E, VanNieuwenhze MS, Brun YV, Filipe SR, Pinho MG. 2015. Cell shape dynamics during the staphylococcal cell cycle. Nat Commun 6:8055. https://doi.org/10.1038/ncomms9055. 23. Pinho M, Kjos M, Veening J-W. 2013. How to get (a)round: mechanisms controlling growth and division of coccoid bacteria. Nat Rev Microbiol 11: 601–614. https://doi.org/10.1038/nrmicro3088. 24. Massidda O, Nováková L, Vollmer W. 2013. From models to pathogens: how much have we learned about Streptococcus pneumoniae cell division? Environ Microbiol 15:3133–3157. https://doi.org/10.1111/1462-2920.12189. 25. Daniel RA, Errington J. 2003. Control of cell morphogenesis in bacteria: two distinct ways to make a rod-shaped cell. Cell 113:767–776. https://doi .org/10.1016/s0092-8674(03)00421-5. 26. Herrero A, Stavans J, Flores E. 2016. The multicellular nature of filamentous heterocyst-forming cyanobacteria. FEMS Microbiol Rev 40:831–854. https://doi.org/10.1093/femsre/fuw029. 27. Arbel-Goren R, Buonfiglio V, Di Patti F, Camargo S, Zhitnitsky A, Valladares A, Flores E, Herrero A, Fanelli D, Stavans J. 2021. Robust, coherent, and synchronized circadian clock-controlled oscillations along Anabaena filaments. Elife 10:e64348. https://doi.org/10.7554/eLife.64348. 28. Arévalo S, Nenninger A, Nieves-Morión M, Herrero A, Mullineaux CV, Flores E. 2021. Coexistence of communicating and noncommunicating cells in the filamentous cyanobacterium Anabaena. mSphere 6:e01091- 20. https://doi.org/10.1128/mSphere.01091-20. 29. Lehner J, Berendt S, Dörsam B, Pérez R, Forchhammer K, Maldener I. 2013. Prokaryotic multicellularity: a nanopore array for bacterial cell communication. FASEB J 27:2293–2300. https://doi.org/10.1096/fj.12-225854. 30. Velázquez-Suárez C, Luque I, Herrero A. 2020. The inorganic nutrient regime and the mre genes regulate cell and filament size and morphology in the phototrophic multicellular bacterium Anabaena. mSphere 5:e00747-20. https://doi.org/10.1128/mSphere.00747-20. 31. Hu B, Yang G, Zhao W, Zhang Y, Zhao J. 2007. MreB is important for cell shape but not for chromosome segregation of the filamentous cyanobacterium Anabaena sp. PCC 7120. Mol Microbiol 63:1640–1652. https://doi .org/10.1111/j.1365-2958.2007.05618.x. 32. Burnat M, Schleiff E, Flores E. 2014. Cell envelope components influencing filament length in the heterocyst-forming cyanobacterium Anabaena sp. Strain PCC 7120. J Bacteriol 196:4026–4035. https://doi.org/10.1128/JB .02128-14. 33. Corrales-Guerrero L, Camargo S, Valladares A, Picossi S, Luque I, Ochoa de Alda JAG, Herrero A. 2018. FtsZ of filamentous, heterocyst-forming cyanobacteria has a conserved N-terminal peptide required for normal FtsZ polymerization and cell division. Frontiers Microbiol 9:2260. https://doi.org/ 10.3389/fmicb.2018.02260. 34. Camargo S, Picossi S, Corrales-Guerrero L, Valladares A, Arévalo S, Herrero A. 2019. ZipN is an essential FtsZ membrane tether and contributes to the septal localization of SepJ in the filamentous cyanobacterium. Anabaena Sci Rep 9:2744. https://doi.org/10.1038/s41598-019-39336-6. 35. Ramos-León F, Mariscal V, Frías JE, Flores E, Herrero A. 2015. Divisome-de- pendent subcellular localization of cell-cell joining protein SepJ in the July/August 2022 Volume 13 Issue 4 10.1128/mbio.01165-22 15 Downloaded from https://journals.asm.org/journal/mbio on 12 September 2022 by 161.111.154.117. Chapter III 112 life Article The Role of MreB, MreC and MreD in the Morphology of the Diazotrophic Filament of Anabaena sp. PCC 7120 Cristina Velázquez-Suárez, Ignacio Luque and Antonia Herrero * Instituto de Bioquímica Vegetal y Fotosíntesis, CSIC and Universidad de Sevilla, 41092 Seville, Spain * Correspondence: herre[email protected]ic.es Citation: Velázquez-Suárez, C.; Luque, I.; Herrero, A. The Role of MreB, MreC and MreD in the Morphology of the Diazotrophic Filament of Anabaena sp. PCC 7120. Life 2022, 12, 1437. https://doi.org/ 10.3390/life12091437 Academic Editors: Asunción Contreras and Jose I. Labella Received: 28 July 2022 Abstract: The cyanobacterium Anabaena sp. PCC 7120 forms filaments of communicating cells. Under conditions of nitrogen scarcity, some cells differentiate into heterocysts, allowing the oxygen-sensitive N2-reduction system to be expressed and operated in oxic environments. The key to diazotrophic growth is the exchange of molecules with nutritional and signaling functions between the two types of cells of the filament. During heterocyst differentiation, the peptidoglycan sacculus grows to allow cell enlargement, and the intercellular septa are rebuilt to narrow the contact surface with neighboring cells and to hold specific transport systems, including the septal junction complexes for intercellular molecular transfer, which traverse the periplasm between heterocysts and neighboring vegetative cells through peptidoglycan nanopores. Here we have followed the spatiotemporal pattern of peptidoglycan incorporation during heterocyst differentiation by Van-FL labeling and the localization and role of proteins MreB, MreC and MreD. We observed strong transitory incorporation of peptidoglycan in the periphery and septa of proheterocysts and a maintained focal activity in the center of mature septa. During differentiation, MreB, MreC and MreD localized throughout the cell periphery and at the cell poles. In mreB, mreC or mreD mutants, instances of strongly increased peripheral and septal peptidoglycan incorporation were detected, as were also heterocysts with aberrant polar morphology, even producing filament breakage, frequently lacking the septal protein SepJ. These results suggest a role of Mre proteins in the regulation of peptidoglycan growth and the formation of the heterocyst neck during differentiation, as well as in the maintenance of polar structures for intercellular communication in the mature heterocyst. Finally, as previously observed in filaments growing with combined nitrogen, in the vegetative cells of diazotrophic filaments, the lack of MreB, MreC or MreD led to altered localization of septal peptidoglycan-growth bands reproducing an altered localization of FtsZ and ZipN rings during cell division. Keywords: heterocyst differentiation; heterocyst neck; intercellular communication; Mre proteins; peripheral peptidoglycan growth; polar peptidoglycan growth Accepted: 9 September 2022 Published: 15 September 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1. Introduction Filamentous heterocyst-forming cyanobacteria, such as the model strain Anabaena sp. PCC 7120 (hereafter Anabaena) represents genuine pluricellular bacteria in which the unit for selection has shifted from the single cell to the filament [1,2]. The Anabaena filament is composed of hundreds of cells, each of which is surrounded by an individual cytoplasmic membrane, whereas the outer membrane is continuous along the filament, defining a commonly shared periplasm. Indeed, the continuous periplasm has been considered a path that allows intercellular molecular exchange through the filament [3]. In addition, neighboring cells are communicated through septal junction complexes composed of polytopic cytoplasmic-membrane anchored proteins [4,5] that traverse the septal peptidoglycan (PG) by perforations termed nanopores, which are drilled by PG amidases [6–8]. Regarding the PG sacculus, the continuous filament periplasm contains a single molecule that completely encircles each cell and is engrossed in the septal location between contiguous cells [9]. In relation to its multicellular structure, Anabaena presents distinct Life 2022, 12, 1437. https://doi.org/10.3390/life12091437 https://www.mdpi.com/journal/life Life 2022, 12, 1437 2 of 15 characteristics in the pattern of PG incorporation. During cell division, septal incorporation occurs in the divisome to build the new septa of the resulting daughter cells. In addition, weak peripheral PG growth elongates the cells during cell growth, and continuous PG incorporation occurs in the intercellular locations along the filament, which persists after the division has been completed [10–12]. In model unicellular rod-shaped bacteria, the proteins MreB, MreC and MreD are involved in the spatial localization and regulation of the enzymatic machinery for lateral PG growth organized in the multiprotein complex termed the elongasome. MreB is an actin structural homolog that forms dynamic polymers associated with the inner face of the cytoplasmic membrane. MreC is a cytoplasmic membrane-anchored protein with most of its extension residing in the periplasm, and MreD is an integral membrane protein [13]. Notably, in Anabaena, MreB, MreC and MreD have been localized all around the cell periphery, including the mature intercellular septa, and in the divisome through all stages of cell division, having been implicated in the regulation of peripheral, medial (divisome-associated) and septal PG incorporation [10]. Depending on environmental conditions, the Anabaena filament can include different cell types specialized in specific functions. Thus, heterocysts are cells specialized in the fixation of atmospheric nitrogen that differentiates at semi-regular intervals along the filament under nitrogen deprivation conditions. Heterocysts supply vegetative cells with organic nitrogen. Vegetative cells fix CO2 through oxygenic photosynthesis and provide the heterocysts with organic carbon to be utilized as a source of reductants and as carbon skeletons for the incorporation of the ammonium resulting from N2 fixation [14,15]. Because the enzymatic machinery for N2 reduction to ammonium is extremely O2 sensitive, a main issue in heterocyst performance is the maintenance of a microoxic internal milieu, to which the presence of extra glycolipid and polysaccharide envelope layers, deposited during the differentiation process, contribute (see [16]). However, the entrance of N2 into the heterocysts is needed. Key to regulating gas exchanges is the differentiation of the heterocyst cell poles to form the specialized structure known as the heterocyst neck, which narrows the contact space between the heterocyst and the adjacent vegetative cells and represents the main path for gas exchange [14,17]. Indeed, intercellular molecular exchange involving heterocysts is thought to take place to a good extent through the heterocyst neck. Thus, although some membrane transporters of different protein families are localized through the heterocyst periphery, some others, such as the ABC-type exporters HetC, possibly involved in the export of morphogens from the heterocyst [18,19], and DevBCD-HgdD, for heterocyst glycolipid export [20], preferentially localize at the heterocyst poles. Additionally, the pentapeptide-repeat protein HglK [21] and the septal-junction associated proteins SepJ [22] and FraCD [23] are localized to the heterocyst-vegetative cell septa. Finally, the ABC-permease protein FraE [24] and the TonB-like protein SjdR [25] have been identified as required for the constriction and arrangement of the heterocyst polar structures. Besides peripheral PG growth required for cell enlargement, the differentiation of the heterocyst neck also appears to involve PG remodeling/turnover see [11]. Consistent with this, mutations in genes encoding enzymes involved in different steps of PG synthesis lead to impaired heterocyst differentiation [26–29]. Moreover, PG septal disk diameter, nanopore diameter and nanopore number are larger in septa between vegetative cells and heterocysts than in septa between two vegetative cells [24], and mutants deficient in the PG amidases AmiC1 [7] or AmiC3 [8] are unable of diazotrophic growth. In this work, we have monitored the incorporation of PG and the localization of the MreB, MreC and MreD proteins in Anabaena during heterocyst differentiation and in mature diazotrophic filaments. We have also addressed the influence of MreB, MreC and MreD on peripheral, medial and septal PG growth, the filament geometry and the formation of heterocyst-vegetative cell connections. Life 2022, 12, 1437 3 of 15 × × 2. Materials and Methods 2.1. Strains and Growth Conditions Anabaena sp. strain PCC 7120 and mutant strains (strains and plasmids used in this work are listed in Table S1) were grown in BG11 medium (containing NaNO3 as a nitrogen source) [30] containing ferric citrate instead of ferric-ammonium citrate. Cultures were incubated at 30 ◦C with illumination (12 µE m−2 s−1 white light from Osram LED lamps 16.4 W/4000 K), in Erlenmeyer flasks with shaking (100 rpm) or in plates of medium solidified with 1% Difco agar. For the mutants, media were supplemented with antibiotics: spectinomycin dihydrochloride pentahydrate (Sp) and streptomycin sulfate (Sm) at 5 µg mL−1 each in solid medium or 2.5 µg mL−1 each in liquid medium (CSCV2, CSCV6, CSCV7, CSCV8, CSSC19, CSAV39), with neomycin sulfate (Nm) at 25 µg mL−1 in solid medium or 5 µg mL−1 in liquid medium (CSS89, CSCV1, CSCV4), or with Sm, Sp and Nm (CSCV14, CSCV15, CSCV16, CSCV17, CSCV18, CSCV19, CSCV20, CSCV21, CSCV22). For the experiments described in this work, BG11-grown filaments of the indicated strains were transferred to (at a cell density corresponding to 0.5 µg chlorophyll mL−1) and incubated in BG110 medium [30] lacking NaNO3 and antibiotics. Chlorophyll a (Chl) content of the cultures was determined after extraction with methanol [31]. In Anabaena, 1 µg Chl corresponds to ca. 3.3 × 106 cells [32]. 2.2. Strain Constructions Strains CSCV17, CSCV18 and CSCV19 bear a sepJ-gfpmut2 fusion gene in a mreB, mreC or mreD mutant background, respectively. To generate strain CSCV19, the conjugative plasmid pCSVT22 (encoding Nm resistance and including sequences of the 3’ part of Anabaena sepJ fused to gfpmut2) [23] was transferred to strain CSCV2 (mreD) by conjugation, and one clone in which the plasmid was inserted by single cross-over into sepJ was selected and verified by PCR. To generate strains CSCV17 and CSCV18, the insert of pCSVT22 was amplified with oligonucleotides alr2338-BamH1 and gfp-BamH1 (listed in Table S1) and cloned in the conjugative vector pCSV3 encoding Sm/Sp resistance [33] generating plasmid pCSCV38, which was transferred to strains CSCV1 (mreB) and CSCV4 (mreC) by conjugation. Clones in which pCSCV38 was inserted by single cross-over into sepJ were selected and verified by PCR. 2.3. Microscopy For Van-FL labeling and detection, filaments were treated as described [10] and visualized with a Leica DM6000B fluorescence microscope and a FITCL5 filter (excitation band-pass, 480/40; emission band-pass, 527/30), and photographed with an ORCA-ER camera (Hamamatsu). GFP fluorescence was monitored with an Olympus TCS SP2 confocal laser-scanning microscope equipped with an HCX PLAN-APO 63 1.4 NA oil immersion objective (excitation, 488-nm; collection, 500–540 nm for GFP or 630–700 nm for cyanobacterial autofluorescence) or with an Olympus FLUOVIEW FV3000 (hyper-resolution) confocal laser-scanning microscope equipped with a UPlanApo 60 1.5 NA oil immersion objective (excitation, 488-nm; collection 500–540 nm for GFP or excitation, 640 nm; collection, 650–750 nm for cyanobacterial autofluorescence). 3. Results 3.1. PG Growth during Heterocyst Differentiation in Anabaena Van-FL is a fluorescent derivative of the antibiotic vancomycin that binds to the PG precursor Lipid II in the periplasm, either bound or not to the sacculus but not involved in crosslinking, thus providing a probe for nascent PG synthesis [34]. We used Van-FL to label Anabaena filaments during heterocyst differentiation upon transfer of filaments grown in BG11 medium (containing nitrate) to BG110 medium lacking combined nitrogen. The differentiation process can be followed through a succession of cellular signals that include progressive loss of red fluorescence due to photosynthetic pigment rearrangement, cell elongation and enlargement, deposition of extracellular envelope layers, and finally, Life 2022, 12, 1437 4 of 15 5 m accumulation in refringent polar granules of cyanophycin resulting from N2 fixation. In the differentiating cells, the localization of Van-FL fluorescence appeared to follow a sequence through the different stages of the differentiation process (Figure 1). Thus, at early stages (cells already enlarged but with no conspicuous change in morphology and still detectable red fluorescence), Van-FL fluorescence was detected as a homogenous peripheral signal and in foci localized in the heterocyst neck. In the neighboring vegetative cells, fluorescence was detected at the pole contiguous to the differentiating heterocyst (see the second row in Figure 1; see also [35]). Afterward, concomitant with morphological reshaping and red-fluorescence loss, the peripheral signal in the differentiating cells appeared more pronounced, especially in the polar regions. Moreover, a strong label was detected in the pole proximal to the heterocyst of the neighboring vegetative cells (see the third row in Figure 1). Later, in mature heterocysts (showing conspicuous polar granules), the peripheral signal faded, and fluorescence was more restricted to foci in the center of the heterocyst poles and the neighboring vegetative cell pole (see the fourth row in Figure 1). These observations suggest that during heterocyst differentiation, a phase of peripheral PG growth occurs at intermediate stages of differentiation (as previously noted [11]), being this a transient event. In addition, continuous activity of PG remodeling appears to be maintained at the poles of heterocysts and neighboring vegetative cell poles. Figure 1. Van-FL staining of filaments of Anabaena sp. strain PCC 7120 during heterocyst differentiation. Filaments grown in BG11 medium were transferred to BG110 (lacking combined nitrogen) medium and incubated under culture conditions. After 24 (row 2) or 48 h (rows 1, 3 and 4), samples were stained with Van-FL and observed under a fluorescence microscope and photographed. Van-FL fluorescence (green), cyanobacterial autofluorescence (red), and bright-field images are shown. Cells in early stages of differentiation (S), immature heterocysts (P) and mature heterocysts exhibiting polar refringent cyanophycin granules (H) (see the text for details) are indicated. Green arrows in row 1 point to fluorescence matching the divisome complex. 5 m H H H 5 m P P P S S S 5 m Life 2022, 12, 1437 5 of 15 Regarding the vegetative cells of filaments incubated in BG110 medium, continuous peripheral Van-FL fluorescence and strong fluorescence bands in the intercellular regions, reminiscent of the pattern found in filaments growing with combined nitrogen [10], were observed (first row in Figure 1). As in the latter, in actively-growing diazotrophic filaments, the septal bands alternate in intensity, with the intensity detected in the recently formed septa being lower than in the older septa. Additionally, midcell fluorescence progressing inwards from the cell surface, matching the progressing septum under construction, could be detected in dividing cells. 3.2. Localization of MreB, MreC and MreD during Heterocyst Differentiation We studied the localization of MreB, MreC and MreD in filaments of Anabaena subjected to N-stepdown using protein fusions to GFP. For that, we used the previously generated strains CSCV6, CSCV7 and CSCV8 that express the genes sfgfp-mreB, sfgfp-mreC or sfgfp-mreD (encoding superfolder GFP fused to MreB, MreC or MreD, respectively) preceded by the native promoter of the mreBCD operon, located on an ectopic chromosomal locus, keeping an intact mreBCD operon in its native genomic location [10]. GFP fluorescence was monitored in filaments of CSCV6, CSCV7 and CSCV8 grown with nitrate and transferred to a medium lacking combined nitrogen. In the three strains, fluorescence was localized over the periphery of cells in the early stages of differentiation, as well as of immature and mature heterocysts (exhibiting complete loss of red fluorescence and conspicuous polar granules) (Figure 2), and tended to become more diffuse in older heterocysts (not shown). GFP fluorescence could also be detected frequently as foci in the heterocyst poles (Figure 2A; see CSCV8 in Figure 2B). A CSCV6 CSCV7 CSCV8 B WT CSCV6 CSCV7 CSCV8 Figure 2. Localization of MreB, MreC and MreD in Anabaena. Strain PCC 7120 and their derivatives CSCV6 (sfGFP-MreB), CSCV7 (sfGFP-MreC) and CSCV8 (sf-GFP-MreD) were grown in BG11 medium, transferred to BG110 medium and incubated under culture conditions. After 24 h, filaments were visualized by confocal microscopy with TCS (A) or FLUOVIEW (B) equipment. GFP fluorescence (green), cyanobacterial autofluorescence (red), merged GFP and cyanobacterial autofluorescence, and bright-field images are shown. Cells in early stages of differentiation (S), immature heterocysts (P) and mature heterocysts (H) are indicated. Green arrows point to GFP fluorescence matching divisome complexes. Magnification is the same for all micrographs in (A) or (B). 10  m H H H H H H H H H H H H H H H 10  m H H H P P P S S S P P P Life 2022, 12, 1437 6 of 15 In the vegetative cells of strains CSCV6, CSCV7 and CSCV8, GFP fluorescence could also be detected in the cell periphery during active growth, and at midcell in signals progressing inwards from the periphery in the dividing cells (Figure 2). As previously noted [10], especially cells of CSCV7 and CSCV8 appeared bigger and rounder than those of the wild type. This could indicate some interference of the fusion proteins with the function of the corresponding native proteins. Nonetheless, the GFP fluorescence localization’s specificity and consistency with the signals produced by GFP-MreB (in CSCV6) support that GFP-MreC and GFP-MreD are reporting the physiological localization of MreC and MreD, respectively. 3.3. Influence of MreB, MreC and MreD in PG Growth during Heterocyst Differentiation Strains CSCV1, CSCV4 and CSCV2 contain only inactivated versions of mreB, mreC, or mreD, respectively. These mutants show strong morphological alterations, forming cells, including heterocysts, that are larger and rounder than wild-type cells [36]. Van-FL staining was performed in filaments of strains CSCV1, CSCV4 and CSCV2 grown with nitrate and incubated in the absence of combined nitrogen. Fluorescence localization was compared to the wild-type (shown in Figure 1). Regarding fluorescence distribution in differentiating cells, some heterocysts with peripheral fluorescence and septal foci could be detected in CSCV1 (see the second row in Figure 3), CSCV4 (see first and the second row in Figure 4) and CSCV2 (see the second row in Figure 5). However, instances of deformed septal fluorescence distribution or aberrantly increased labeling were also found (see the third row in Figure 3, for CSCV1, and the third row in Figure 5, for CSCV2). In the three mutants, but especially in CSCV4, few mature heterocysts with conspicuous polar granules were detected after 48 h of incubation in the absence of combined nitrogen, consistent with the previous description of retarded expression of nitrogenase activity [36]. Frequently, heterocysts with conspicuous peripheral labeling could still be detected at 48 h (see the fourth row in Figure 3 for CSCV1 and Figure 5 for CSCV2). CSCV1 24 h 24 h 24 h 48 h Figure 3. Van-FL staining of Anabaena mreB mutant. Strain CSCV1 (mreB) was treated under the same conditions described for the WT (PCC 7120) in Figure 1. Van-FL fluorescence (green), cyanobacterial autofluorescence (red) and bright-field images are shown. To improve visibility, contrast/brightness is different for the image at left in row 1. The bracket indicates a cell compartment with disparate size; the blue arrow points to a tilted fluorescent band and the green arrow to fluorescence matching the divisome. Cells in the early stages of differentiation (S) and mature heterocysts (H) are indicated. 5 m S S S 5 m S S S 5 m H H H 5 m Life 2022, 12, 1437 7 of 15 CSCV4 24 h 24 h 48 h Figure 4. Van-FL staining of Anabaena mreC mutant strain CSCV4. See the legend in Figure 3 for details. To improve visibility, contrast/brightness is different for the image at left in row 1. CSCV2 24 h 24 h 24 h 48 h Figure 5. Van-FL staining of Anabaena mreD mutant strain CSCV2. See the legend in Figure 3 for details. Immature heterocysts (P) and mature heterocysts (H) are indicated. To improve visibility, contrast/brightness differs for the images in row 1. In vegetative cells (see the first row in Figures 3–5), the three mutants showed peripheral, septal and midcell Van-FL labeling. However, the consecutive septal bands appeared frequently tilted with regard to each other or at irregular distances from each other (compare first rows in Figures 1 and 3–5). Inspection of 100 to 120 septal bands of each PCC 7120, CSCV1, CSCV4 and CSCV2 incubated for 24 and 48 h in BG110 medium showed 5 m S S S 5 m 5 m 5 m P P P 5 m P P P H H H Life 2022, 12, 1437 8 of 15 that, in the mutants, 29–40% septal bands appeared conspicuously tilted with regard to a consecutive band. In the wild type, 95–98% appeared parallel to its consecutive band (Mann–Whitney tests indicated fully significant differences, p < 0.001, for any comparison between a mutant and the WT, and no significant differences between the mutants). CSCV4 even showed some instances where the division plane appeared inverted (see row 3 in Figure 4). In addition, abundant cell debris from cell lysis was observed especially in CSCV1 and CSCV4. 3.4. Localization of FtsZ and ZipN in Mre Mutants Because heterocyst differentiation is linked to inhibition of cell division (see [15]) and because, as shown above, MreB, MreC and MreD are also localized to the divisome, we asked whether the Mre proteins affected the localization of FtsZ and ZipN during the differentiation process. FtsZ initiates the polymerization of the division ring (see [37]), and ZipN is the main tether for binding FtsZ to the cytoplasmic membrane and organizer of the divisome in Anabaena [38]. We monitored FtsZ localization using strains that express an ftsZ-gfp reporter from the PftsZ promoter in the WT background (strain CSSC19 [37]) and in the mreB (strain CSCV20), mreC (strain CSCV21) and mreD (strain CSCV22) mutant backgrounds [10]. Upon N-stepdown, fluorescent FtsZ-rings could be detected in vegetative cells of filaments of all strains (Figure 6). However, in the mutant backgrounds, some rings were tilted compared to those of neighboring cells, and in some places, the distance between two consecutive Z-rings was irregular. In the differentiating cells described for strain CSSC19 (39), Z-rings were detected during differentiation but not in mature heterocysts. CSSC19 mostly exhibited mature heterocysts 48 h after N-stepdown, whereas in CSCV20, CSCV21 and CSCV22, cells exhibiting fluorescent rings were still abundant at that time, consistent with retardation of heterocyst differentiation. Of note is the observation of some aberrant events in CSCV20, such as the differentiation of consecutive cells (see third and fourth rows in Figure 6), which is very rare in the wild type, or the presence of a Z-ring (see the seventh row in Figure 6) or apparent septation (see the third row in Figure 6) in mature heterocysts. The effect of the lack of MreB, MreC or MreD on the localization of ZipN was studied in strains that express an sfgfp-zipN reporter gene from the PzipN promoter in the WT background (strain CSAV39 [39]) and the mreB (strain CSCV14), mreC (strain CSCV15) and mreD (strain CSCV16) mutant backgrounds [10]. Upon N-stepdown, strain CSAV39 presented midcell fluorescent bands in vegetative cells, and this signal was progressively lost in the differentiating cells, being absent in mature heterocysts (Figure 7; [39]). In CSCV14, CSCV15 and CSCV16, the midcell GFP- ZipN fluorescence was also lost in most mature heterocysts, although in some of them, some signals could still be detected (see the heterocyst in the seventh row in Figure 7 for CSCV16). In the three mutant backgrounds, but especially in CSCV15, fluorescence appeared more dispersed than in the wild-type background (strain CSAV39). Of note is the observation that in the three mutant backgrounds some heterocysts deviated from the filament longitudinal plane. Moreover, in CSCV14 and CSCV16, heterocysts with aberrant polar morphology even provoking filament breakage were observed (Figure 7). Life 2022, 12, 1437 9 of 15 CSSC19 CSCV20 CSCV21 CSCV22 Figure 6. Localization of FtsZ in mreB, mreC and mreD mutants. Strains CSSC19 (ftsZ-gfpmut2 in WT background), CSCV20 (ftsZ-gfpmut2 in CSCV1 background), CSCV21 (ftsZ-gfpmut2 in CSCV4 background) and CSCV22 (ftsZ-gfpmut2 in CSCV2 background) were treated as above described for Figure 2. After 48 h, filaments were visualized by confocal microscopy with FLUOVIEW equipment. GFP fluorescence (green), merged GFP and cyanobacterial autofluorescence (red), and bright-field images are shown. To improve visibility, contrast is higher for the green image of CSCV22. Blue arrows point to tilted fluorescence bands, and blue brackets to cell compartments with disparate sizes. Immature (P) and mature heterocysts (H) are indicated. Blue arrowheads point to an apparent instance of septation in mature heterocysts; orange arrowheads to a mature heterocyst with Z-ring fluorescence; and blue asterisks to contiguous heterocysts. Magnification is the same for all micrographs. H H 10  m H H H H H H H P P P P P P P P P H P P H P P H P P H H H H H H P P P H H H P P P H H H P P P P P P H H H Table S1. Cyanobacterial strains, plasmids and oligonucleotides used in this work Strain Genotype Resistance Source Anabaena sp. PCC 7120 WT Pasteur Culture Collection CSCV1 mreB::C.K1 Nm 36 CSCV2 mreD::C.S3 Sm, Sp 36 CSCV4 mreC::C.K1 Nm 36 CSCV6 thrS2::PmreB-sfgfp-mreB Sm, Sp 10 CSCV7 thrS2::PmreB-sfgfp-mreC Sm, Sp 10 CSCV8 thrS2::PmreB-sfgfp-mreD Sm, Sp 10 CSSC19 PftsZ-ftsZ-gfpmut2 Sm, Sp 37 CSCV20 PftsZ-ftsZ-gfpmut2, mreB Nm, Sm, Sp 10 CSCV21 PftsZ-ftsZ-gfpmut2, mreC Nm, Sm, Sp 10 CSCV22 PftsZ-ftsZ-gfpmut2, mreD Nm, Sm, Sp 10 CSAV39 PzipN-sfgfp-zipN Sm, Sp 39 CSCV14 PzipN-sfgfp-zipN, mreB Nm, Sm, Sp 10 CSCV15 PzipN-sfgfp-zipN, mreC Nm, Sm, Sp 10 CSCV16 PzipN-sfgfp-zipN, mreD Nm, Sm, Sp 10 CSS89 sepJ-gfpmut2 Nm 38 CSCV17 sepJ-gfpmut2, mreB Nm, Sm, Sp This study CSCV18 sepJ-gfpmut2, mreC Nm, Sm, Sp This study CSCV19 sepJ-gfpmut2, mreD Nm, Sm, Sp This study Plasmid Description Resistance marker Source pCSCV38 pCSV3 carrying sepJ-gfpmut2 Sm, Sp This study pCSVT22 pRL424 carrying sepJ-gfpmut2 Nm 23 Oligodeoxynucleotide primers1 Sequence (5’-3’) alr2338-BamHI CGTGGGATCCTTTTCTGTGGTGAGGTGC gfp-BamHI AAGCGGATCCTTATTTGTATAGTTCATCCATGCC 1The underlined letters indicate a restriction site. General discussion 129 GENERAL DISCUSSION General discussion 130 General discussion 131 The influence of the environment in the Anabaena morphology The results of this work showed that environmental conditions, like the carbon and nitrogen availability, influence cell size and shape and filament length in Anabaena. In diverse bacteria, cell morphology has been reported to be important for interaction with the environment and for survival (Young, 2006; van Teeseling et al., 2017). For instance, the spiral morphology of H. pylori and the curved shape of V. cholerae allow these strains to colonize the gastric and intestinal epithelia, respectively, and favor their movement through the mucus, which facilitates infection (Young, 2006; van Teeseling et al., 2017). Other pathogens like M. tuberculosis and uropathogen E. coli adopt an elongated morphology that helps them to avoid phagocytosis by macrophages (Young, 2006). In some other cases, environmental conditions impact the morphology of different bacteria although the possible advantage of this effect is not evident. This is the case of Ancalomicrobium, which in poor environments undergo filamentation and branching, or the cyanobacterium F. diplosiphon that changes morphology in response to the quality of light (Young, 2006; Singh and Montgomery, 2011). Regarding cell size, in Anabaena we have observed differences according to the nutritional regime of carbon and nitrogen and the intensity of the incident light (Velázquez-Suárez et al., in preparation). Also, different relationships were found between the growth rate and cell size through the growth cycle, especially in the presence of combined nitrogen (nitrate or ammonium). In the phase of fastest growth, Anabaena cells were larger in the presence of combined nitrogen than in diazotrophic cultures, an effect that would meet the classical growth law entailing that cells are larger when growing with a nutrient-imposed lower generation time (Schaechter et al., 1958; reviewed in Morcinek-Orlowska et al., 2019). Larger cell size during active growth would facilitate chromosome inheritance by a random mechanism, as proposed for Anabaena (Hu et al., 2007), during cell division. In the absence of combined nitrogen, a small cell size could facilitate intercellular molecular exchanges through the filament, which take place by diffusion and are more stringently required under these diazotrophic conditions. In any case, differences in cell size report on differences in the balance between cell growth and division through the cell cycle, with an impact of the environment. Recent reports have shown that in Anabaena, the level of the signal molecule c-di-GMP plays a role in determining cell size. Cell size control is dependent on the interaction of the CdgR protein with the DevH transcription factor, an interaction that only occurs when CdgR is not bound to c-di-GMP (Zeng et al., 2023). It would be worth to investigate whether c-di-GMP is involved in cell size determination in response to the availability of carbon and nitrogen. Regarding cell shape, in the presence of combined nitrogen, Anabaena cells appeared to elongate through the phases of active growth, and to decrease in length thereafter. Cells of cultures supplemented with nitrate or ammonium were longer than cells of diazotrophic cultures, in which cell morphology appeared to be less variable. In the presence of nitrate, a high C supply also had a positive effect on cell elongation. In rod-shaped cells in general, an increase in length will increase the total surface area, which represents an advantage when nutrients General discussion 132 (nitrate, ammonium, bicarbonate) are to be taken from the external medium. In contrast, under diazotrophic conditions, nitrogen nutrients are exchanged between vegetative cells and heterocysts mainly by septal connections, thus making less clear the advantage of increasing the cell surface. Finally, in ntcA and hetR mutants growing with ammonium, cells were longer than in the wild type, also supporting an impact of the C-to-N balance on cell morphology (see below). Regarding the length of the filaments (i.e., the number of cells they contain), diazotrophic filaments showed a shorter length than those in cultures supplemented with combined nitrogen, which can be related to the assumption that the septa between heterocysts and vegetative cells are weaker than those between two vetative cells. PG growth and the role of MreB, MreC and MreD in Anabaena In model unicellular rod-shaped bacteria, the elongasome directs a disperse mode of PG synthesis in the cylindrical part of the cell, whereas during cell division the divisome directs septal PG synthesis to form the new poles of the daughter cells. In Anabaena, both Van-FL (this work) and HADA (Zhang et al., 2018), produces a homogeneous, although rather weak, labeling in the cell periphery. Consistently, in this work GFP fusions to MreB, MreC and MreD were localized through the cell periphery. These observations, together with the round shape of the Anabaena mreB, mreC and mreD mutans, which is similar to the phenotype observed upon Mre depletion in E. coli and B. subtilis (Dye et al., 2005; Kruse et al., 2005; Leaver and Errington, 2005), indicate that lateral PG growth dependent on Mre factors also take place in Anabaena enabling cell elongation. Consistently also, expression of the mreBCD operon is higher during the phases of stronger growth, when cells are longer, and is negatively regulated by NtcA, which occurs at higher levels in diazotrophic filaments, where the cells are shorter. Instead of occurring with a patchy distribution excluding the cell poles, like e.g., in B. subtilis (Daniel and Errington, 2003), in Anabaena peripheral PG growth appears to take place all around the cell surface, an issue worth of further study. Van-FL staining in Anabaena also indicated that PG synthesis linked to the divisome and, remarkably, in the mature intercellular septa takes place (see Table 3). This distinct labeling is indicative of a permanent growth or remodeling of the intercellular septal PG during the phases of active growth, in contrast to the situation in common model rod-shaped unicellular bacteria, like E. coli, B. subtilis and C. crescentus, where the cell poles become inert for PG synthesis once division is completed (Daniel and Errington, 2003; Aaron et al., 2007; Kawazura et al., 2017). These results are consistent with those obtained by pulse-chase experiments with HADA in Anabaena, which show a fast turnover of this analog in mature septa (Zhang et al., 2018). It is possible that septal PG remodeling in Anabaena were required for the assembly or maintenance of the SJ complexes that expand the intercellular septa through PG nanopores. Indeed, MreB, MreC and MreD were observed also in the intercellular septa and, although inactivation of mre genes did not preclude PG incorporation at such location, in the mutants the intercellular septa were wider and more robust than in the WT. Thus, Mre proteins are not required for PG General discussion 133 incorporation at mature intercellular septa, but they appear required for the formation of dimensioned septa. It has been described that under conditions of rapid growth, limitation for PG precursors in E. coli could lead to a competition between the incorporation of PG at the lateral and midcell locations. Also, increased or decreased incorporation of PG at the division site provokes the inverse effect at the cell periphery (Navarro et al., 2022). In Anabaena, increased PG incorporation at the intercellular septa in the absence of elongation in mre mutants could suggest that a competition between septal and peripheral PG incorporation could also operate. Table 3. Modes of PG incorporation in Anabaena. Mode of PG incorporation Role of Mre proteins Divisional Topologycal regulation Lateral Elongasome dependent Intercellular septal Septal dimension regulation Interactions between the elongasome and the divisome in Anabaena In Anabaena, MreB, MreC and MreD match the localization of the divisome during all stages of the cell division process. Also, by BACTH tests, we have identified interactions between MreB or MreD and some divisome components, suggesting contacts between the divisome and the elongasome during cell division. This is consistent with the observation of a direct interaction between MreB and FtsZ in other filamentous cyanobacteria (Springstein et al., 2020b). In E. coli, MreB was shown to interact and co-localize with FtsZ in Z-rings at the initiation of cell division, leading to the proposal that MreB guides some components of the elongasome PG synthesis General discussion 134 machinery be passed on to the divisome (Fenton and Gerdes, 2013). We cannot rule out that in Anabaena Mre factors do play a similar role, but our observations are indicative of additional functions connected to the divisome. In this work we have observed that in mreB, mreC or mreD mutants midcell PG growth bands along the filament, as well as, FtsZ and ZipN rings were frequently non-equidistant and not parallel in consecutive cells and, in many cases, tilted with respect to the filament long axis. This suggests a role of MreB, MreC and MreD in the correct localization of the divisome, which is unprecedented in the literature. The work presented here evidences that the multicellular character of Anabaena makes this organism a good model for the study of septa positioning in successive generations of cells. Evidence for an interplay between the divisome and the elongasome in Anabaena also derive from the morphological alterations observed in strains depleted of FtsZ or ZipN (Corrales- Guerrero et al., 2018; Camargo et al., 2019). Model rod-shaped unicellular bacteria depleted of FtsZ maintain PG growth over the sidewalls leading to the classical filamentation phenotype (Dai and Lutkenhaus, 1991; Daniel and Errington, 2003; Redick et al., 2005; Aaron et al., 2007). However, depletion of FtsZ or ZipN in Anabaena led to a very aberrant cell and filament morphology, including giant cells with irregular width, or exhibiting central bulged areas, connected by minuscule septa in comparison to the increased cell volume. Also frequent were events of cell detachment from the filament and cell lysis, all this indicative of unchecked cell enlargement with restricted elongation. Van-FL staining revealed either very strong, frequently patched, or very weak, progressively fading, peripheral signals in the aberrant cells. Septal signal did not increase in width during cell enlargement, and frequently was undetected. Thus, lowering the expression of Z-ring components not only impair divisional, but also peripheral and septal PG growth. PG growth during heterocyst differentiation In Anabaena none of the mre genes are essential, although their mutants were impaired in growth, more drastically under diazotrophic conditions. Diazotrophic growth in Anabaena depends on heterocyst differentiation to establish a filament with two cell types: the vegetative cells, which fix CO2 and transfer carbon compounds, and the heterocysts, which fix N2 and donate nitrogen-rich compounds to the vegetative cells. Heterocysts differentiate from certain vegetative cells of the filament under conditions of nitrogen deficiency. The differentiation process involve extensive morphological changes, including increase in cell size, elongation, remodeling of the cell envelope and remodeling of the cell poles, all involving PG synthesis (Nicolaisen et al., 2009a; Zeng and Zhang, 2022). Van-FL labeling has revealed different events of PG incorporation during differentiation. Regarding peripheral labeling, strong signals were detected at intermediate stages of differentiation, which became weaker in mature heterocysts. Ours and others’ (e.g., Zhang et al., 2018) studies are indicative of the occurrence of transient lateral PG growth effecting cell enlargement and elongation. The involvement of MreB, MreC and MreD in this transient General discussion 135 peripheral PG incorporation is supported by the expression of mreB, mreC and mreD in the differentiating cells and the localization of these proteins to the cell periphery, as well as by the morphological alterations in heterocysts from mre mutants that showed increased size, in comparison to the wild-type heterocysts, and a rounder shape. The heterocyst polar neck represents a differentiated structure for intercellular communication with the neighboring vegetative cell. Indeed, differences in the array of PG nanopores holding septal junction complexes have been detected between vegetative cell-vegetative cell and vegetative cell-heterocyst connections (Arévalo et al., 2021), indicative of remodeling of septal communication structures during differentiation. Van-FL and HADA labeling revealed strong PG incorporation in the heterocyst neck and the pole of the neighboring vegetative cell during heterocyst differentiation. In mreB, mreC and mreD mutants, alterations in the heterocyst polar morphology, including events of displacement of the heterocyst with regard to the filament plane, filament breakage at heterocyst-vegetative cell connections and lack of SepJ at some heterocyst poles were detected. MreB, MreC and MreD were conspicuously detected at the cell poles in differentiating and mature heterocysts. These results involve MreB, MreC and MreD in the remodeling of PG during differentiation and maintenance of the heterocyst septal structures. In future, it would be interesting to further investigate the septal structures in mre mutants, including the features of the nanopore array and the activity of molecular exchanges with neighboring vegetative cells. Heterocysts are terminally differentiated cells that have lost the capacity for cell division, an important step in setting the commitment to differentiation. Indeed, the Z-ring is dismantled early, and ftsZ and zipN expression is arrested during differentiation (e.g., Valladares et al., 2020). Two factors, PatA (Valladares et al., 2020) and PatD (Wang et al., 2021), whose expression increases transiently in the differentiating cells, have been proposed to participate in the destabilization of the Z-ring during differentiation through a direct interaction with divisome components. An important observation of this work was that proheterocysts of mreB, mreC and mreD mutants frequently showed midcell FtsZ and ZipN rings at times when they are already absent in the wild type. It is tempting to speculate that the Mre proteins could play a role in the inhibition of Z-ring formation during heterocyst differentiation, a proposal consistent also with our observation of MreB and MreD interactions with divisome components. It is worth noting that although Z-rings were visible in some proheterocysts of mre mutants, midcell PG incorporation was never observed. This suggests that a functional divisome is not assembled, which is consistent with the existence of additional inhibitory mechanisms, as described above (see section 8.2). General discussion 136 In summary, this work has shown the importance of environmental parameters, such as the carbon and nitrogen sources, and the growth phase in the determination of cell size and morphology in the filamentous cyanobacterium Anabaena. In this cyanobacterium, the expression of the mreB, mreC and mreD genes is regulated by the growth phase and the environment through the transcriptional regulator NtcA that responds to the C-to-N balance. Anabaena presents three main modes of PG incorporation: a peripheral synthesis that elongate the cells during growth; a midcell synthesis dependent on the divisome that forms the poles of the daughter cells during cell division and a persistent incorporation in the mature intercellular septa along the filament. Peripheral PG growth requires MreB, MreC and MreD. Although divisional and septal PG incorporation do not require MreB, MreC and MreD, these factors are needed for the correct positioning of the divisome and the correct geometry and dimensions of the intercellular septa. Similarly, in the differentiating heterocysts, MreB, MreC and MreD are localized to the cell periphery and the cell poles, regulating the dimensions and shape of the PG wall. MreB, MreC and MreD are also localized to the divisome, which can represent a way for their localization at mature septa once division has completed. During cell division and in the mature septa, as well as in the heterocyst neck, MreB, MreC and MreD could contribute to establishment and maintenance of septal structures involved in intercellular communication. Thus, this work has uncovered unprecedented roles for MreB, MreC and MreD in relation to the multicellular character of Anabaena, underscoring the importance of studying basic bacterial mechanisms outside of a few classical models. Conclusions 137 CONCLUSIONS