COMMENTARY COMMENTARY Plasmid segregation by a moving ATPase gradient Daniela Kiekebuscha,b,c and Martin Thanbichlera,b,c,1 aMax Planck Research Group “Prokaryotic Cell Biology”, Max Planck Institute for Terrestrial Microbiology, 35043 Marburg, Germany; bFaculty of Biology, Philipps-Universität, 35043 Marburg, Germany; and cLOEWE Center for Synthetic Microbiology, 35043 Marburg, Germany

Unlike the mitotic segregation of eukaryotic called “partition complex” (8). This complex sister chromatids, DNA partitioning in bac- then dynamically interacts with ParA to drive teria is still not well understood. Bacterial the directed movement of the DNA cargo high–copy-number plasmids can be stably (Fig. 1). In the presence of ATP, ParA asso- maintained by random distribution of their ciates nonspecifically with DNA and exhibits copies during cell division. In contrast, the a weak intrinsic ATPase activity that is stim- faithful transmission of low–copy-number ulated synergistically by ParB and DNA (9– plasmids and many chromosomes depends 11). Moreover, several biochemical studies on an active process mediated by conserved, have shown that ParA can assemble into fil- tripartite segregation systems (1). A central amentous structures upon ATP binding (6, component of these machineries is a nucleo- 10, 12, 13). This finding, together with anal- side triphosphatase driving the partition- yses of ParA localization in vivo, has pro- ing reaction, which can be classified as an vided the basis for the filament-pulling actin-like ATPase (ParM), a tubulin-like model of DNA segregation (7). However, GTPase (TubZ), or a Walker-type ATPase the physiological relevance of ParA polymer- (ParA). Systems using an actin or tubulin ization is highly controversial (14, 15). In homolog function by means of a filament- particular, recent studies investigating the E. based pushing or pulling mechanism (2, coli P1 and F plasmid segregation systems – 3). Most low copy-number plasmids and have cast serious doubt on a role of filament Fig. 1. (A) Partitioning (sop)locusoftheE. coli F chromosomes are, however, segregated formation in the partitioning process (5, plasmid, comprising genes for an ATPase (SopA) and by Walker-type ATPases. Despite exten- 16–18). Instead, ParA was proposed to act an adaptor protein (SopB), as well as a centromere-like sive research, it is not yet unambiguously by a diffusion-ratchet mechanism, which region including multiple tandem repeats of the SopB binding site (sopC). (B)Thesop locusisessentialfor established how this third group of proteins includes the following steps: ParA-ATP the faithful segregation of F plasmids to the future harnesses the energy released during ATP dimers bind nonspecifically to chromo- daughter-cell compartments. Wild-type plasmids are reg- hydrolysis for plasmid movement. Based somal DNA and transiently tether plas- ularly positioned over the nucleoid (Upper), whereas on previous analyses, two competing mids to the nucleoid surface through plasmids lacking the sop locus are concentrated in –parS the polar regions of the cell (Lower). (C)Mechanism models have been put forward. In the fil- interaction with the ParB partition com- of sop-mediated plasmid movement. Nucleoid-asso- ament-pulling model, ParA is assumed to plex (17). In the resulting quaternary com- ciated SopA-ATP dimers bind to a SopB–sopC parti- form polymers that move DNA by repeated plex, ParB stimulates the ATPase activity of tion complex, thereby immobilizing a copy of the F polymerization/depolymerization cycles. ParA, thereby inducing its release from the plasmid on the nucleoid surface. After stimulation of In contrast, the diffusion-ratchet model DNA. Because reactivation of ParA involves its ATPase activity by SopB, SopA dissociates from the DNA, leaving behind a zone of low SopA con- proposes a concentration gradient of ParA a series of slow conformational changes, it is centration (Upper). As a consequence, the plasmid dimers on the nucleoid as the driving force unable to immediately reassociate with the starts to diffuse and then reattaches to the nucleoid for DNA segregation. In PNAS, Vecchiarelli nucleoid (17). This lag creates a ParA de- through association with neighboring SopA molecules Lower et al. (4) now provide direct evidence in pletion zone in the vicinity of the partition ( ). Iteration of these steps results in the directed movement of plasmid copies across the nucleoid support of the latter model by fully recon- complex, which ultimately results in the de- surface. stituting in vitro the segregation system of the tachment of the plasmid from the nucleoid Escherichia coli Fplasmid. surface. After its dissociation, the plasmid dif- Plasmid segregation systems based on fuses in a stochastically chosen direction. As central point of the model, the initial direc- Walker-type ATPases position plasmid cop- it reaches the edge of the depletion zone, it tion taken by the plasmid is reinforced by ies at regular distances over the nucleoid encounters an increasing number of nucleoid- (5–7). In addition to the ATPase compo- bound ParA dimers, which make new con- Author contributions: D.K. and M.T. wrote the paper. nent (ParA), they comprise a centromere-like tacts to the plasmid partition complex. Mov- The authors declare no conflict of interest. parS DNA sequence ( ) and an adaptor protein ing along this ParA gradient, the plasmid is See companion article 10.1073/pnas.1401025111. parS (ParB). Binding of ParB to motifs on the finally immobilized again and initiates the 1To whom correspondence should be addressed. E-mail: DNA cargo results in the formation of a so- formation of a new depletion zone. As a [email protected].

www.pnas.org/cgi/doi/10.1073/pnas.1402867111 PNAS Early Edition | 1of2 Downloaded by guest on September 28, 2021 low ParA concentrations in the wake of the narrow space between the nucleoid surface In doing so, they tracked along a moving partition complex, thereby giving rise to ro- and the cytoplasmic membrane, a situation SopA gradient, leaving behind a region of bust, unidirectional movement of plasmid that greatly facilitates repeated interactions low SopA concentration that was slowly molecules (Fig. 1C). of ParB–parS partition complexes with nucle- refilled with SopA molecules from solution. As a first step toward verifying the diffu- oid-bound ParA molecules. Therefore, in line It still needs to be determined if, analogous to sion-ratchet model, the P1 and F plasmid with the experimental results, spatial con- P1 ParA, a time-delay between ATP binding segregation systems have recently been re- finement was proposed to be another key and reassociation with DNA is involved in constituted in a cell-free set-up (16, 18). To requirement for ParA-mediated plasmid the emergence of the SopA gradient. Interest- mimic the nucleoid surface, the bottom of ingly, not all beads exhibited directional a 25-μm-thick microfluidic flow cell was The study by Vecchiarelli movement. Whereas directed beads consis- densely coated with nonspecific DNA frag- et al. significantly tently stayed in close contact with the sur- ments. The flow cell was then loaded with a face, others diffused freely along the surface mixture of ParA-GFP, ParB, and a fluores- advances our under- and “bounced” in and out of the TIRFM cently labeled parS-bearing plasmid prepared standing of bacterial illumination area, ruling out a significant in an ATP-containing buffer. Subsequently, DNA segregation. contribution of the magnetic force to the the dynamics of the components were visu- straight motion of directed beads. alized by total internal reflection microscopy segregation by a diffusion-ratchet mechanism Although the characteristics of a magnetic (TIRFM), a technique that specifically re- (16, 18). bead differ from those of a DNA molecule, solves surface-associated processes. The two In the present study, Vecchiarelli et al. (4) the study by Vecchiarelli et al. (4) signifi- systems exhibited similar dynamics, suggest- have verified this hypothesis by reinvesti- cantly advances our understanding of bacte- ing a comparable segregation mechanism. gating the dynamics of F plasmid segregation rial DNA segregation. The work identifies the Initially, plasmids were stably tethered to using an advanced version of their cell-free partitioning reaction driven by (at least a sub- a layer of ParA-GFP associated with the system. The plasmids used in the previous set of) Walker-type ATPases as a diffusion- DNA carpet. After some time, they started to studies were replaced by magnetic beads regulated process, and thus adds to the growing move around their anchor point in a circular coated with fluorescently labeled sopC- body of evidence suggesting that the physio- Brownian motion, leading to the displacement containing DNA. Spatial confinement was logical relevance of filament formation by of ParA-GFP in their surroundings. Eventu- then simulated by application of a magnetic these proteins may have been overestimated. ally, the plasmids dissociated from the sur- force perpendicular to the DNA-coated flow It will be interesting to perform the same face and diffused away, leaving behind cell surface. This experimental set-up not kind of analysis on other ParA-dependent a zone depleted of ParA-GFP molecules. only reproduced the previously observed P1 segregation systems that have been proposed These observations strikingly mirrored the and F plasmid dynamics, but also made it to use a polymerization-based mechanism in vivo dynamics of the P1 plasmid (5) and possible to follow the cargo during its mo- (19). Moreover, modeling studies will be re- were consistent with key aspects of the dif- bile phase. Strikingly, beads migrated in quired to understand how the movement of fusion-ratchet model described above. Im- a directed manner over distances of sev- plasmids along ParA gradients finally leads to portantly, ParA molecules were evenly eral microns, driven by repeated cycles of the faithful distribution of sister copies to the distributed over the DNA carpet and ex- ParA depletion and surface detachment. two daughter cells. changed rapidly with molecules in solution (16, 18), supporting the idea that the active, nucleoid-bound species of ParA consists of 1 Gerdes K, Howard M, Szardenings F (2010) Pushing and pulling 10 Leonard TA, Butler PJ, Löwe J (2005) Bacterial chromosome in prokaryotic DNA segregation. Cell 141(6):927–942. segregation: Structure and DNA binding of the Soj dimer—A dimeric complexes rather than filaments. 2 Garner EC, Campbell CS, Weibel DB, Mullins RD (2007) conserved biological switch. EMBO J 24(2):270–282. The visualized dynamics were dependent on Reconstitution of DNA segregation driven by assembly 11 Barillà D, Carmelo E, Hayes F (2007) The tail of the ParG DNA thepresenceofbothParAandParB,aswell of a prokaryotic actin homolog. Science 315(5816): segregation protein remodels ParF polymers and enhances ATP 1270–1274. hydrolysis via an arginine finger-like motif. Proc Natl Acad Sci USA as on the ATPase activity of ParA. When 3 Montabana EA, Agard DA (2014) Bacterial tubulin TubZ-Bt 104(6):1811–1816. either of the Par proteins was omitted from transitions between a two-stranded intermediate and a four-stranded 12 Hui MP, et al. (2010) ParA2, a Vibrio cholerae chromosome the reaction, plasmids did not associate with filament upon GTP hydrolysis. Proc Natl Acad Sci USA 111(9): partitioning protein, forms left-handed helical filaments on DNA. 3407–3412. Proc Natl Acad Sci USA 107(10):4590–4595. the flow cell surface. In contrast, in the ab- 4 Vecchiarelli AG, Neuman KC, Mizuuchi K (2014) A propagating 13 Lim GE, Derman AI, Pogliano J (2005) Bacterial DNA segregation by sence of ATP hydrolysis, the plasmids re- ATPase gradient drives transport of surface-confined cellular cargo. dynamic SopA polymers. Proc Natl Acad Sci USA 102(49):17658–17663. Proc Natl Acad Sci USA, 10.1073/pnas.1401025111. 14 Vecchiarelli AG, Mizuuchi K, Funnell BE (2012) Surfing biological mained irreversibly attached to the surface, 5 Hatano T, Niki H (2010) Partitioning of P1 plasmids by gradual surfaces: Exploiting the nucleoid for partition and transport in highlighting the importance of transient distributionoftheATPaseParA.Mol Microbiol 78(5): bacteria. Mol Microbiol 86(3):513–523. ParA–DNA interactions. A shortcoming of 1182–1198. 15 Szardenings F, Guymer D, Gerdes K (2011) ParA ATPases can 6 Niki H, Hiraga S (1999) Subcellular localization of plasmids move and position DNA and subcellular structures. Curr Opin the in vitro system used in these studies was containing the oriC region of the Escherichia coli chromosome, with Microbiol 14(6):712–718. that it could only recapitulate the initial stages or without the sopABC partitioning system. Mol Microbiol 34(3): 16 Hwang LC, et al. (2013) ParA-mediated plasmid partition driven of the diffusion-ratchet model because the 498–503. by protein pattern self-organization. EMBO J 32(9):1238–1249. 7 Ringgaard S, van Zon J, Howard M, Gerdes K (2009) Movement 17 Vecchiarelli AG, et al. (2010) ATP control of dynamic P1 ParA- large size of the flow cells allowed plasmids and equipositioning of plasmids by ParA filament disassembly. DNA interactions: A key role for the nucleoid in plasmid partition. to diffuse away from the surface after detach- Proc Natl Acad Sci USA 106(46):19369–19374. Mol Microbiol 78(1):78–91. 8 Havey JC, Vecchiarelli AG, Funnell BE (2012) ATP-regulated 18 Vecchiarelli AG, Hwang LC, Mizuuchi K (2013) Cell-free study ment. As a consequence, it was not possible interactions between P1 ParA, ParB and non-specific DNA that are of F plasmid partition provides evidence for cargo transport by a to observe the iterative diffusion/association stabilized by the plasmid partition site, parS. Nucleic Acids Res 40(2): diffusion-ratchet mechanism. Proc Natl Acad Sci USA 110(15): cycles that were proposed to drive directional 801–812. E1390–E1397. 9 Davis MA, Martin KA, Austin SJ (1992) Biochemical activities of 19 Barillà D, Rosenberg MF, Nobbmann U, Hayes F (2005) Bacterial plasmid movement. In the bacterial cell, the the parA partition protein of the P1 plasmid. Mol Microbiol 6(9): DNA segregation dynamics mediated by the polymerizing protein diffusion of plasmids is constrained by the 1141–1147. ParF. EMBO J 24(7):1453–1464.

2of2 | www.pnas.org/cgi/doi/10.1073/pnas.1402867111 Kiekebusch and Thanbichler Downloaded by guest on September 28, 2021