9696

Martin ThanbichlerMax Planck Research Group / PUMa

Cellular differentiation and cell division in bacteria

Bacteria have evolved an abundance of different life cycles and cell shapes, but the mechanisms that under-lie this diversity are largely unknown. The aim of our group is to understand the molecular basis of cellular differentiation and cell division in bacteria and to elu-cidate how the one-dimensional information of the ge-netic code is translated into defi ned spatial and tempo-ral regulatory patterns. To address these questions, we are using a combination of cell biological, biochemical, biophysical, and genetic approaches. Our studies focus on the model organism Caulobacter crescentus, a Gram-negative α-proteobacterium that is characterized by its unique developmental cycle (Fig. 1). C. crescentus cells can easily be synchronized and then be observed as they progress synchronously through their developmental program, which greatly facilitates the study of cell cycle-dependent processes. Moreover, owing to their asym-metric cell division, they represent an attractive model for studying bacterial differentiation.

Division site placementA fundamental problem in cell biology is the proper tem-poral and spatial regulation of cell division. Our previous work has revealed the existence of a novel mechanism that is responsible for positioning of the cell division plane in C. crescentus. At its heart lies the Walker ATPase MipZ, which forms a dynamic complex with the chromosome partitioning protein ParB at the chromo-somal origin of replication (Fig. 2). Upon initiation of DNA replication, the two newly synthesized origin re-gions are immediately re-decorated with the MipZ/ParB complex and then positioned at the two opposite cell poles. This generates an intracellular gradient of MipZ, with its concentration being highest at the cell poles

and lowest at the cell center. Our analyses showed that MipZ has the ability to inhibit polymerization of the bac-terial tubulin homologue FtsZ and, thereby, to prevent the assembly of the so-called FtsZ ring, which forms the foundation of the ring-shaped cell division apparatus. As a consequence, cytokinesis is limited to the midcell re-gion and, furthermore, only initiates once chromosome segregation has started. MipZ is highly conserved among α-proteobacteria, which suggests that it represents the prototype of a new and widespread class of bacterial cell division regulators.

Fig. 1. The C. crescentus cell cycle. The C. crescentus swarmer cell bears a polar fl agellum and a single, replicationally quies-cent chromosome. At a certain point in the cell cycle, the fl agel-lum is shed and a stalk is formed at the previously fl agellated pole. Concomitantly, DNA replication and cell division are initi-ated. Cytokinesis then gives rise to a new G1-arrested swarmer cell and to a stalked cell that immediately enters the next divi-sion cycle.

Martin Thanbichler born 26.04.1973Diplom (Biology), Ludwig-Maximilans-Universität München, 1998Dr. rer. nat. (Microbiology), Ludwig-Maximilans-Universität München, 2002Postdoc (Prokaryotic Cell Biology), Stanford University, 2002-2006Head of the Max Planck Research Group „Prokaryotic Cell Biology“ at the MPI Marburg, since 2007Junior professor of Microbiology, Philipps-Universität Marburg, since 2008

9797

Martin Thanbichler Max Planck Research Group / PUMa

to the monomeric state. Losing its affi nity for DNA and FtsZ, it starts to diffuse rapidly within the cell until it returns to the ParB complex, thereby closing the cycle. MipZ thus shares similarity with small GTPases, using nucleotide binding and hydrolysis to transition between functionally distinct conformational states.

Cell divisionThe bacterial cell division apparatus (divisome) com-prises a variety of different proteins, whose precise func-tion and mode of cooperation is still poorly understood. Moreover, it is unclear to what extent the results ob-tained in the prototypic model system E. coli can be ap-plied to evolutionarily distinct organisms. Recent work suggests that a core of essential divisome components is widely conserved among bacteria, whereas a number of accessory factors are restricted to certain lineages or a few species only.

Based on sequence similarity searches, most core com-ponents of the E. coli divisiome have previously been identifi ed in C. crescentus. One of the exceptions was FtsN, a peptidoglycan-binding protein that orchestrates the activity of cell wall synthases and hydrolases involved in cell wall invagination. Initially, the absence of this protein was not surprising, given that FtsN was thought to be a peculiarity of E. coli and its close relatives. How-ever, we have now identifi ed a C. crescentus protein that, despite the lack of sequence conservation, shows the structural and functional characteristics of FtsN and thus likely represents a highly diverged FtsN homologue (Möll & Thanbichler, 2009). Based on conserved fea-tures defi ned in our work, we were able to identify puta-tive FtsN homologues in a large variety of species cover-ing all proteobacterial lineages. Two of these candidate proteins were further investigated and verifi ed to be cell division proteins, indicating that FtsN-like proteins are widespread among bacteria, albeit highly variable at the sequence level. We found that the peptidoglycan-bind-ing (SPOR) domain of FtsN is dispensable for function but required for robust localization to the division site. Given that it is recruited to midcell in isolated form and retains its localization potential in heterologous hosts, it likely recognizes a conserved feature of the invaginating cell wall. Hence, the SPOR domain is likely to drive a positive feedback loop whereby only a small amount of FtsN is recruited to the division site by direct interaction with the divisome to initiate cell constriction. Synthesis of new cell wall material may then generate an increas-ing number of additional binding sites for FtsN, thus further accelerating the division process.

Fig. 2. Placement of the cell division plane by the spatial regu-lator MipZ. See text for explanations.

The MipZ system provides the fi rst example of a regula-tory protein gradient in bacteria. Our current research focuses on the molecular mechanisms responsible for gradient formation. Using a combination of crystallo-graphic, biochemical, and cell biological approaches, we were able to demonstrate that MipZ undergoes a dynamic localization cycle that is driven by differential affi nities of its ADP- and ATP-bound forms for ParB and chromosomal DNA (Kiekebusch et al., unpublished). In complex with ADP, MipZ exists as a monomer that specifi cally interacts with the polar ParB complex. Upon nucleotide exchange, MipZ dimerizes and becomes able to bind unspecifi cally to DNA. As a consequence, it leaves ParB and randomly attaches to nearby regions of the nucleoid, with the frequency of MipZ-DNA com-plexes decreasing as a function of distance from the cell pole. Only the ATP-bound form of MipZ is able to interact with FtsZ, thus inhibiting FtsZ ring assembly over the pole-proximal parts of the nucleoid. Owing to its intrinsic ATPase activity, the dimer eventually hydro-lyzes its nucleotide cofactor, and MipZ switches back

98

Martin ThanbichlerMax Planck Research Group / PUMa

tion. Yet a different situation is observed for Shewanella oneidensis, whose single bactofi lin homologue shows the typical localization pattern of a cell division protein. These fi ndings indicate that bactofi lins can adopt a wide range of functions in the cell. Biochemical analyses re-vealed that bactofi lins polymerize spontaneously in the absence of additional cofactors in vitro, forming stable ribbon- or rod-like fi lament bundles. These structures might have evolved as an alternative to intermediate fi la-ments, serving as versatile molecular scaffolds in a vari-ety of cellular pathways.

Fig. 4. Structure and function of bactofi lins in C. crescentus. Polymers formed by the purifi ed bactofi lin homologue BacA from C. crescentus were visualized by (A) DIC (bar: 10 µm) and (B) transmission electron microscopy (bar: 75 nm). (C) Model for the function of bactofi lin sheets in C. crescentus. A membrane-associated bactofi lin sheet interacts with the cytoplasmic tail of the cell wall synthase PbpC, thereby targeting PbpC to the stalked pole and promoting stalk biogenesis.

Cell polarityIn cooperation with Patrick H. Viollier (Case Western Reserve University, Cleveland, Ohio), we have identifi ed a novel protein, SpmX, that plays a vital role in control-ling the differential fate of stalked and swarmer cells in C. crescentus (Radhakrishnan et al, 2008). SpmX is a membrane protein that is targeted to the stalked pole at the early predivisional stage of the cell cycle. Polar localization is dependent on a periplasmic domain with

Fig. 3. Midcell localization of FtsN-like proteins in different bacteria. Fluorescent protein fusions of FtsN-like proteins from C. crescentus (α−proteobacteria), M. xanthus (δ-proteobacteria) and B. thailandensis (β-proteobacteria) were visualized by dif-ferential interference contrast (DIC) and fl uorescence micros-copy.

Searching for interaction partners of FtsN, we have re-cently identifi ed a peptidoglycan hydrolase, designated DipM, that is required for proper invagination of the cell wall during cytokinesis (Möll et al., unpublished). On deletion of the dipM gene, the outer layers of the cell envelope fail to follow the inner membrane during constriction, thus uncoupling compartmentalization of the cytoplasm from daughter cell separation. As a conse-quence, cells tend to form fi laments or grow in chains, with branches emerging from the poles of individual cytoplasmic compartments. These results suggest that constriction of the cell may occur by deposition of ad-ditional peptidoglycan layers at the inner face of the cell wall and concomitant removel of the outer layers by lytic enzymes such as DipM. Moreover, they underscore the key role of FtsN in in the organization of peptidoglycan remodeling at the division site.

Cytoskeletal elementsThe cytoskeleton has a vital function in the temporal and spatial organization of both prokaryotic and eukaryotic cells. Our work has revealed the existence of a new class of polymer-forming proteins, designated bactofi lins, that are widely conserved among bacteria (Kühn et al, 2009). In C. crescentus, two bactofi lin paralogues cooperate to form a sheet-like structure that lines the cytoplasmic membrane in proximity of the stalked cell pole. These assemblies mediate polar localization of a peptidoglycan synthase involved in stalk formation, thus complement-ing the function of the actin-like cytoskeleton and the cell division machinery in the regulation of cell wall bio-genesis. Myxococcus xanthus, by contrast, synthesizes four bactofi lin paralogues. Three of them are encoded in an operon located in a cluster of motility-related genes, assembling into a massive, rod-shaped structure that oc-cupies the mid-third of the cell. Deletion of this operon resulted in a severe defect in pilus-mediated gliding mo-tility, suggesting direct or indirect involvement in locomo-

99

Martin Thanbichler Max Planck Research Group / PUMa

Finished theses

Diploma/MSc thesesAndrea Möll (2008) Auf den Spuren von FtsN – Studie eines Zellteilungsproteins in Caulobacter crescentus.

Katja Leser (2008) CC1873 and CC3022 – Charak-terisierung mutmaßlicher neuer Cytoskelett-Elemente in Caulobacter crescentus.

Lin Yang Zhang (2008) Analyse der Rolle von ParB bei der Regulation der Zellteilung in Caulobacter crescen-tus.

Kathrin E. Klein (2008) Die Rolle der Klasse A Peni-cillin-Bindeproteine bei der Morphogenese von Cau-lobacter crescentus.

Anne Raßbach (2009) Untersuchungen zur Biologie des Planctomyceten Gemmata obscuriglobus.

BSc thesesAljona Gutschmidt (2009) Biochemische Charak-terisierung des Zellteilungsregulators MipZ aus Cau-lobacter crescentus.

Nicole Schnaß (2009) Funktionelle Analyse von Pro-teinen der Zellwandbiosynthese in Caulobacter crescen-tus.

Structure of the group (12/2009)

Group leader: Jun.-Prof. Dr. Martin Thanbichler

PhD students: Andrea Möll, Daniela Kiekebusch, Juliane Kühn, Anne Raßbach, Susan Schlimpert

MSc student: Alexandra Jung

BSc students: Oliver Leicht, Wolfgang Strobel

Technical assistant: Stephanie Wick

similarity to phage lysozymes, suggesting that the pro-tein recognizes specifi c structural features of the polar cell wall. SpmX acts as a targeting factor that tethers and thereby activates the histidine kinase DivJ. Activated DivJ, in turn, launches the stalked-cell-specifi c develop-mental program in the stalked sibling once cell division has occurred. The swarmer cell, by contrast, lacks the SpmX/DivJ complex and thus expresses a different set of genes which encodes swarmer-specifi c functions.

Publications

1. Kühn, J., Briegel, A., Mörschel, E., Kahnt, J., Leser, K., Wick, S., Jensen, G.J., and Thanbichler, M. (2010). Bactofi lins, a ubiquitous class of cytoskeletal proteins mediating polar localization of a cell wall synthase in Caulobacter crescentus. EMBO J. 29, 327-339.

2. Thanbichler, M. (2009). Synchronization of chromo-some dynamics and cell division in bacteria. Cold Spring Harb. Perspect. Biol. 2, a000331.

3. Thanbichler, M. (2009). Spatial regulation in Cau-lobacter crescentus. Curr. Opin. Microbiol. 12, 715-721.

4. Thanbichler, M. (2009). Closing the ring: a new twist to bacterial chromosome condensation. Cell 137, 598-600.

5. Möll, A., and Thanbichler, M. (2009). FtsN-like proteins are conserved components of the cell division machinery in proteobacteria. Mol. Microbiol. 72, 1037-1053.

6. Radhakrishnan, S.K., Thanbichler, M., and Viollier P.H. (2008) The dynamic interplay between a cell fate determinant and a lysozyme homolog drives the asym-metric division cycle of Caulobacter crescentus. Genes Dev. 22, 212-225.

7. Thanbichler, M., and Shapiro, L. (2008) Getting orga-nized – how bacteria move proteins and DNA. Nat. Rev. Microbiol. 6, 28-40.

100

Martin ThanbichlerMax Planck Research Group / PUMa

Address

Jun.-Prof. Dr. Martin ThanbichlerMax-Planck-Institut für terrestrische MikrobiologieKarl-von-Frisch-Straße 1035043 Marburg/Germany

Phone: +49 6421 178-300Fax: +49 6421 178-209E-mail: thanbichler@mpi-marburg.mpg.de

External funding

Human Frontier Science Program: Young Investigator Award (RGY 69/2008)

Invited lectures

Jahrestagung der Vereinigung für Allgemeine und An-gewandte Mikrobiologie, Frankfurt, Germany, Special Group: Structure and Microscopy: „From micro- to microscopical imaging in microbiology“ (9–11/3/2008)

162nd Meeting of the Society for General Microbiol-ogy, Edinburgh, Great Britain, Physiology, Biochemistry & Molecular Genetics Group Session: “Prokaryotic cell biology” (31/3–3/4/2008)

108th General Meeting of the American Society for Microbiology, Boston, USA, Session 085: “Spatial Regulation of Cytokinesis: From Bacteria to Eukary-otes” (1–5/6/2008)

8th Graduate Retreat of the MPI for Biochemistry, Tegernsee, Germany (1–2/7/2008)

Fakultät für Biology, Bereich Mikrobiologie, Ludwig-Maximilians-Universität München (8/7/2008)

Departement für Chemie und Biochemie, Universität Bern (8/12/2008)

Instituto de Biochimica Vegetal y Fotosintesis, CSIC/Universidad de Sevilla, Spain (16/4/2009)

Lehrstuhl für Mikrobiologie/Organismische Interak-tionen, Eberhard-Karls-Universität Tübingen, Germany (28/5/2009)

Centre de Genetique Moleculaire, CNRS, Gif-sur-Yvettes, France (20/11/2009)

101

Max Planck Research Group / PUMaMartin Thanbichler

The Max Planck Research Group “Prokaryotic Cell Biology“. From left to right: Stephanie Wick, Oliver Leicht, Wolfgang Strobel, Alexandra Jung, Martin Thanbichler, Juliane Kühn, Anne Raßbach, Andrea Möll, Susan Schlimpert, Daniela Kiekebusch