Cdvbased Cell Division and Cell Cycle Organization in the Thaumarchaeon

Molecular Microbiology (2011) ᭿ doi:10.1111/j.1365-2958.2011.07834.x

Cdv-based cell division and cell cycle organization in the

thaumarchaeon Nitrosopumilus maritimusmmi_7834 1..12

Erik A. Pelve,1† Ann-Christin Lindås,1† for evolutionary issues concerning the last archaeal Willm Martens-Habbena,2 José R. de la Torre,3 common ancestor and the relationship between David A. Stahl2 and Rolf Bernander1* archaea and eukaryotes. 1Department of Molecular Evolution, Evolutionary Biology Center, Uppsala University, Norbyvägen 18C, Introduction SE-752 36, Sweden. 2Department of Civil and Environmental Engineering, Two main evolutionary lineages, the Crenarchaeota and University of Washington, 201 More Hall, Seattle, WA Euryarchaeota phyla (kingdoms), were originally recog- 98195-2700, USA. nized within the Archaea domain of life (Woese et al., 3Department of Biology, San Francisco State University, 1990). The phylum Thaumarchaeota was proposed in 1600 Holloway Avenue, San Francisco, CA 94132-1722, 2008 (Brochier-Armanet et al., 2008) based on phyloge- USA. netic evidence and gene content analyses, a proposal that subsequently has gained further support (Spang et al., 2010; Gupta and Shami, 2011; Kelly et al., 2011). Summary Thaumarchaea are ubiquitous in both marine and terres- Cell division is mediated by different mechanisms in trial biotopes and are particularly abundant in the vast different evolutionary lineages. While bacteria and aphotic zone of ocean environments (Karner et al., 2001; euryarchaea utilize an FtsZ-based mechanism, most Leininger et al., 2006; Agogué et al., 2008), and thus rank crenarchaea divide using the Cdv system, related among the most numerous organisms on Earth. The to the eukaryotic ESCRT-III machinery. Intriguingly, Thaumarchaeota phylum comprises large and diverse thaumarchaeal genomes encode both FtsZ and Cdv clades of ammonia oxidizers (Francis et al., 2007; de la protein homologues, raising the question of their divi- Torre et al., 2008; Hatzenpichler et al., 2008; Zhang et al., sion mode. Here, we provide evidence indicating that 2010; Urakawa et al., 2011), including the first thaumar- Cdv is the primary division system in the thaumarchaeal species to be isolated and grown in pure culture, chaeon Nitrosopumilus maritimus. We also show that Nitrosopumilus maritimus (Könneke et al., 2005). Bio- the cell cycle is differently organized as compared to geochemical and physiological studies implicate thaumar- hyperthermophilic crenarchaea, with a longer pre- chaea as a key factor in global nitrogen cycling (Wuchter replication phase and a shorter post-replication et al., 2006; Beman et al., 2008), and N. maritimus dis- stage. In particular, the time required for chromosome plays a 200-fold higher ammonium affinity than any char- replication is remarkably extensive, 15–18 h, indicat- acterized bacterial ammonia oxidizer (Martens-Habbena ing a low replication rate. Further, replication did not et al., 2009). continue to termination in a significant fraction of In all crenarchaeal species so far analysed, the cell N. maritimus cell populations following substrate cycle is organized into a distinct pattern in which the depletion. Both the low replication speed and the pro- post-replicative stage (G2, mitosis and cell division) is the pensity for replication arrest are likely to represent longest phase during exponential growth (Bernander and adaptations to extremely oligotrophic environments. Poplawski, 1997; Poplawski and Bernander, 1997; Hjort The results demonstrate that thaumarchaea, crenar- and Bernander, 2001; Bernander, 2007; Lundgren et al., chaea and euryarchaea display differences not only 2008), possibly as an adaption to life in extreme envi- regarding phylogenetic affiliations and gene content, ronments (Andersson et al., 2010). In contrast, the eur- but also in fundamental cellular and physiological yarchaeal species that have been investigated do not characteristics. The findings also have implications display a coherent picture in terms of cell cycle organization (Malandrin et al., 1999; Maisnier-Patin et al., 2002; Majernik et al., 2005; Soppa, 2011). No information has, Accepted 31 August, 2011. *For correspondence. E-mail rolf. [email protected]; Tel. (+46) 18 4714698; Fax (+46) 18 to date, been reported about the cell cycle characteristics 4716404. †E.A.P. and A.-C.L. contributed equally to this work. of thaumarchaea.

Cell division is an integral part of the cell cycle in all organisms. A majority of the euryarchaea, as well as most bacteria, utilize a system based on the tubulin-like FtsZ protein (Bernander and Ettema, 2010). Recently a novel cell division mechanism, the Cdv machinery, was discov- ered in crenarchaea (Lindås et al., 2008; Samson et al., 2008). Two out of the three Cdv proteins display homology to the eukaryal ESCRT-III vesicle formation system, which also has been implicated in cytokinesis (Morita et al.,

2010; Guizetti et al., 2011). Intriguingly, genome sequen- Nitrite [µM] cing of N. maritimus (Walker et al., 2010) and of other members of the recently recognized thaumarchaeal lineage, e.g. Caldiarchaeum subterraneum (Nunoura et al., 2011), Cenarchaeum symbiosum (Hallam et al., 2006), Nitrosoarchaeum limnia (Blainey et al., 2011) and Nitrososphaera gargensis (Spang et al., 2010), has revealed the presence of both FtsZ- and Cdv-encoding genes, bringing up the issue of which system is utilized for cytokinesis. Time [days] Here, we provide evidence indicating that N. maritimus Fig. 1. Growth of N. maritimus batch cultures. Growth of replicate employs the Cdv system for cell division, raising the ques- batch cultures measured by monitoring nitrite accumulation. Arrows tion of the function of FtsZ in this organism, and present indicate sampling for flow cytometry. the first cell cycle characterization of a thaumarchaeon. in which the exponential age distribution was taken into Results account, i.e. the fact that there are twice as many newborn as dividing cells. About two-fifths of the cell To investigate cell cycle characteristics and division fea- populations in actively growing cultures contained a single tures, N. maritimus cells were cultivated in batch cultures chromosome, resulting in a pre-replicative period length at 30°C (Martens-Habbena et al., 2009). Growth was of 19–29% of the cell cycle, depending on sample, or monitored by measuring nitrite accumulation (Martens- 6–10 h at a doubling time of about 33 h (calculated from Habbena et al., 2009), and cells were harvested at differ- Fig. 1). The proportion of replicating cells was 45–53%, ent time points to obtain samples representing actively corresponding to an S phase length of 15–18 h, and the growing cultures at different cell concentrations (Fig. 1). post-replicative period was estimated to 24–30% of the Cell cycle and stationary phase characteristics were cell cycle (8–10 h). A schematic illustration of the deduced determined by flow cytometry, while division features were cell cycle is shown in Fig. 3. investigated by immunofluorescence microscopy, using polyclonal antibodies raised against N. maritimus Cdv Growth cessation and rate of replication fork movement and FtsZ proteins. The overall organization of the cell cycle (above) con- trasted markedly with that observed in hyperthermophilic Relative lengths of cell cycle periods crenarchaea (Bernander and Poplawski, 1997; Poplawski The relative number of cells in different cell cycle phases and Bernander, 1997; Hjort and Bernander, 2001; in actively growing cultures was determined by flow Bernander, 2007; Lundgren et al., 2008), for which the cytometry analysis. The DNA content distributions (see relative peak heights in DNA content distributions are example in Fig. 2A) revealed a major peak corresponding opposite relative to N. maritimus, i.e. most cells reside in to a single copy of the chromosome (G1 stage), a propor- the post-replicative cell cycle stages (cells with two copies tion of about half the cell population containing between of the chromosome) during exponential growth, while the one and two chromosome equivalents (S phase), and a proportion of pre-replicative cells (single chromosome) is minor peak corresponding to two fully replicated chromo- small. An additional characteristic of hyperthermophilic somes (cells in the G2, mitosis and division stages). crenarchaea is a distinct tendency to rest in the post- To calculate the length of the cell cycle phases in the replicative cell cycle phase when growth has ceased (Ber- individual cell, all samples collected from actively growing nander and Poplawski, 1997; Hjort and Bernander, 1999; cultures (Fig. 2B; also from cultures not shown) were sub- Lundgren et al., 2008). Again, N. maritimus differed from jected to theoretical simulations (Michelsen et al., 2003), crenarchaea, in that the proportion of cells containing two

B Fig. 3. Cell cycle of N. maritimus. Explanation of designations: G1, gap 1 phase; S, DNA synthesis (replication) phase; G2, gap 2 phase (may not exist in N. maritimus); M, mitosis (genome segregation); D, division (cytokinesis). The micrographs were selected to illustrate different cell cycle stages although the precise phase of a cell with a single ﬂuorescence focus is unknown: an S stage cell is generally larger and displays a stronger ﬂuorescence signal (increased DNA content resulting from chromosome

replication) than a G1 cell. Cell perimeters were manually enhanced for clarity. Blue (DNA) and green (CdvA) ﬂuorescence signals were generated as explained in the legend to Fig. 5.

characteristic feature in the extremely oligotrophic envi- Relative number of cells Relative ronments in which marine thaumarchaea reside (see 1 2 DNA content Discussion). (number of chromosomes) With a genome size of 1.645 Mb (Walker et al., 2010) and assuming bidirectional replication from a single chro- Fig. 2. Flow cytometry analysis and cell cycle simulations of mosome replication origin, an S phase length of 15–18 h N. maritimus. corresponds to a replication rate of 13–15 bp s-1 for an A. Flow cytometry DNA content distribution corresponding to 9 day time point in Fig. 1. B. Winﬂow (Michelsen et al., 2003) simulation of DNA content distribution in A. Explanation of designations: diagonal stripes, cell populations AB containing a single chromosome (cells in pre-replicative stage; G1 9 days 12 days phase; see Fig. 3) or two chromosomes (post-replicative stage; G2, M and D; see Fig. 3); grey shading, cells containing between one and two genome equivalents (S phase); solid line, combined distribution. chromosomes decreased when nitrite accumulation (i.e. 12 1 2 cell multiplication) declined (Fig. 4), showing that cell divi- C D sion continued in the cultures during entry into stationary 14 days 19 days phase. Interestingly, despite continued incubation for several days after cessation of growth, the populations number of cells Relative never reached a state in which the cells exclusively contained fully replicated chromosomes. Many cells that were 12 1 2 in S phase at the time when ammonium was depleted and DNA content cellular growth ceased did, thus, not complete replication (number of chromosomes) and arrested with a genome content between one and two Fig. 4. DNA content distributions during substrate deprivation. chromosomes. This could be related to the extremely slow Flow cytometry DNA content distributions from samples of an rate of replication in N. maritimus (below), and might be a N. maritimus batch culture at different time points (see Fig. 1).

Fig. 5. Distribution of the CdvA protein in actively growing N. maritimus cells. The cells were incubated with antibodies against the CdvA protein, followed by staining with secondary antibodies conjugated to Alexa Fluor 568 (red fluorescence). Nucleoids were visualized with the DNA-specific DAPI stain (blue). A. Overview of frequency of cells displaying CdvA-dependent fluorescence. B. Enlarged images of individual cells from A (rows 1–3) or from additional micrographs (rows 4–5). For clarity, cell perimeters have been manually enhanced in the stainings: actual cell perimeters may be observed in the phase contrast images (first column). Scale bar, 1 mm. individual replication fork. If the N. maritimus chromosome detected: most cells instead displayed evenly distributed would contain multiple replication origins, similar to a fluorescence across the interior. range of crenarchaeal species (Lundgren et al., 2004; Presence of the FtsZ protein was neither temporally nor Robinson et al., 2004; Robinson and Bell, 2007), the spatially correlated with nucleoid segregation and no speed of individual fork movement would be even lower. band structures were observed (Fig. 8). Instead, FtsZ, as Even assuming a single origin, this is a low replication rate well as the third CdvB paralogue, Nmar_0061 (Fig. 8C), compared to other, hyperthermophilic, archaeal species were evenly distributed in a majority of the cell population analysed (80–330 bp s-1) (Myllykallio et al., 2000; regardless of cell cycle stage, i.e. independent of cell size Lundgren et al., 2004), and more similar to rates observed or nucleoid distribution. in eukaryotic organisms (Jackson and Pombo, 1998; In addition to direct observation of division structures, Conti et al., 2007; de Moura et al., 2010). The low rate is the microscopy analyses provided information about likely to be an effect of the lower temperature at which overall cell cycle parameters. Since the fraction of cells DNA polymerization occurs, relative to hyperthermo- containing two chromosome equivalents in the flow philes, combined with adaptation to the low nutrient avail- cytometry analysis was in the order of 25–30%, the fact ability in the natural environment of Nitrosopumilus that 20–25% of the cells displayed segregated nucleoids species, with consequent limitations in terms of energy in fluorescence microscopy indicated that genome segre- and precursors for DNA synthesis. gation occurred within a short time interval after termina-

tion of chromosome replication, i.e. that the G2 stage is short, or absent, in N. maritimus. Finally, cells undergoing Cell division nucleoid segregation and cell division displayed a dis- Intriguingly, thaumarchaeal genomes contain genes asso- tinctly elongated morphology (e.g. Fig. 5B), showing that ciated with both main cell division systems used by bac- dynamic changes in cell organization occurred during cell teria and archaea. Antibodies against CdvA (Nmar_0700), cycle progression. three CdvB paralogues (Nmar_0029, Nmar_0061 and Nmar_0816), CdvC (Nmar_1088) and FtsZ (Nmar_1262) Discussion were used to elucidate which of the two systems is utilized by N. maritimus. Immunoﬂuorescence microscopy of cells We report evidence indicating that N. maritimus utilizes stained with antibodies against CdvA revealed sharp, cen- the Cdv machinery for cell division, and also show that the trally positioned, band-like structures in a subpopulation cell cycle displays strikingly differential characteristics consisting of 20–25% of the cells (Fig. 5A). The structures as compared to species belonging to the crenarchaeal correlated strongly with the presence of segregated lineage, with which thaumarchaea were previously nucleoids (Fig. 5B), similar to patterns observed in cre- grouped. narchaea from the genus Sulfolobus (Lindås et al., 2008; The CdvA and C proteins formed distinct centrally Samson et al., 2008). Conversely, in cells in which no positioned structures, presumably ring-shaped, between segregated nucleoids could be discerned, CdvA band segregated nucleoids, similar to the Cdv machinery of structures were rarely observed. Staining with antibodies Sulfolobus (Lindås et al., 2008; Samson et al., 2008). Two against CdvC resulted in a similar pattern (Fig. 6), indicat- out of three CdvB paralogues correlated with the pres- ing colocalization of the two proteins in a joint central cell ence of CdvA and CdvC, and also with segregated nucle- structure during the division stage of the cell cycle oids, indicating participation in the same process. The (Samson et al., 2011). absence of band-like structures for these could reﬂect Two of the three CdvB paralogues, Nmar_0029 and mechanistic differences, less stable association with the Nmar_0816, also correlated with the presence of segre- division structure, or a large protein excess that obscured gated nucleoids and, further, also with the presence of visualization. Regardless, we conclude that Cdv proteins, CdvA and CdvC in double stainings (Fig. 7A and B), i.e. rather than the FtsZ machinery as previously suggested the CdvB proteins were rarely observed in cells lacking (Makarova et al., 2010; Samson and Bell, 2011), are used either of the other two Cdv proteins. In contrast to CdvA for cell division in thaumarchaea, and that division is likely and CdvC, however, distinct band structures were not to occur in a similar mode as in crenarchaea. In terms

1 1 1 1 2 2 2 2

3 3 3 3

B Cell shape Nucleoids CdvA Merged 1

2 II

III 3

Fig. 6. Distribution of the CdvC protein in Cell shape Nucleoids CdvC Merged actively growing N. maritimus cells. The cells were incubated with antibodies against the CdvC protein, followed by staining with secondary antibodies conjugated to Alexa Fluor 488 (green ﬂuorescence). Nucleoid staining and manual enhancement of cell perimeters is described in the legend to Fig. 5. Scale bar, 1 mm.

of regulatory features, the cdv genes are cell-cycle- evolved a different function in thaumarchaea. In agree- specifically transcribed in S. acidocaldarius, such that ment, thaumarchaeal FtsZ sequences form a well- strong induction occurs shortly before the division stage supported distinct branch in phylogenetic analyses (Lundgren and Bernander, 2007). The specific correlation (Makarova and Koonin, 2010), separate from bacterial between Cdv fluorescence foci, elongated cells and seg- and euryarchaeal clades containing FtsZ proteins shown regated nucleoids indicate similar regulation of cdv to be involved in cell division. In further support of an expression in N. maritimus. alternate role, atypical FtsZ homologues suggested to The FtsZ protein, as well as one CdvB paralogue, function in other membrane-remodelling processes than showed no correlation to nucleoid distribution or cell cycle cytokinesis have been reported in Sulfolobus (Makarova stage, no band-like structures, and a largely uniform intra- et al., 2010), in which division is Cdv-mediated. In this cellular distribution. In Sulfolobus, involvement in viral context, it is also of interest that FtsZ homologues (TubZ) release (Ortmann et al., 2008; Snyder and Young, 2011) in Bacillus species form axial filaments that contribute to and vesicle budding (Ellen et al., 2009) for Cdv proteins plasmid segregation (Ni et al., 2010), suggesting that an has been discussed, and one or more of the Nitrosopumi- involvement in chromosome segregation for the N. mar- lus CdvB homologues could also be involved in processes itimus homologue is conceivable. other than cytokinesis, in line with their disperse localiza- The N. maritimus DNA content distributions represent a tion in the genome. Since we find no support in our data mirror image of those observed in crenarchaea, thus con- for a cytokinesis function for FtsZ, even if a subtle role stituting a differential feature between organisms from the cannot be ruled out, we propose that the protein has Thaumarchaeota and Crenarchaeota phyla. The fact that

Fig. 7. Distribution of CdvB proteins in actively growing N. maritimus cells. A. Cells incubated with antibodies against the CdvA and CdvB (Nmar_0029) proteins. B. Cells incubated with antibodies against the CdvA and CdvB (Nmar_0816) proteins. Secondary antibody staining, nucleoid staining and manual enhancement of cell perimeters is described in the legends to Figs 5 and 6. Scale bar, 1 mm.

A CdvB CdvA+ CdvB+ Cell shape Nucleoids CdvA (Nmar_0029) Nucleoids Nucleoids

B CdvB CdvA+ CdvB+ Cell shape Nucleoids CdvA (Nmar_0816) Nucleoids Nucleoids

CdvA+ FtsZ+ Nucleoids Nucleoids

B CdvA+ FtsZ+ Cell shape Nucleoids CdvA FtsZ Nucleoids Nucleoids

C CdvB Cell shape Nucleoids (Nmar_0061) FtsZ

Fig. 8. Distribution of FtsZ and Cdv proteins in actively growing N. maritimus cells. A. Overview of frequency of cells displaying CdvA- and FtsZ-dependent fluorescence. B. Enlarged images of individual cells displaying CdvA- and FtsZ-dependent fluorescence. C. Enlarged images of individual cells displaying CdvB (Nmar_0061)- and FtsZ-dependent fluorescence. Secondary antibody staining, nucleoid staining and manual enhancement of cell perimeters is described in the legends to Figs 5 and 6. Scale bar, 1 mm.

the proportion of cells with segregated nucleoids in the chaea and many bacteria, N. maritimus cells may not microscopy analyses was similar to the fraction containing contain sufficient energy reserves to complete replication two genome equivalents in flow cytometry, indicates that after extended periods of substrate depletion. In the genome segregation occurs rapidly after termination of marine environment, N. maritimus-like thaumarchaea replication in N. maritimus, and that the G2 phase, conse- may, however, rarely become completely deprived of sub- quently, is short or absent, also in contrast to the situation strate due to their extremely high affinity for ammonium in crenarchaea. It should, however, be noted that these (Martens-Habbena et al., 2009) and, thus, seldom, if ever, are relative numbers: at a doubling time around 33 h, experience comparable arrest of growth and replication. even a G2 period of only about 3% would last a full hour. The suggestion that Thaumarchaeota constitutes a The general tendency towards a genome content of a separate phylum is supported not only through phyloge- single chromosome could reflect the different environ- netic inferences but also by comparative genomics, ment the organism is adapted to, relative to Sulfolobus showing that thaumarchaeal chromosomes contain a species. While multiple copies of the chromosome may mixture of genes previously assumed to be specific for constitute an advantage under the highly mutagenic either Crenarchaeota or Euryarchaeota (Brochier- growth conditions of hyperthermophilic crenarchaea Armanet et al., 2008; 2011; Spang et al., 2010; Walker (Andersson et al., 2010), a single copy of the genome et al., 2010; Kelly et al., 2011). We demonstrate that also may suffice in the environmentally stable deep-sea con- at the cellular and physiological levels, the properties of ditions characteristic of marine thaumarchaea. We specu- the thaumarchaeon N. maritimus are mixed. The cell divi- late that a low DNA content even might provide an sion process is, thus, similar to that of several crenar- advantage, and thus be positively selected for, by reduc- chaeal clades while the cell cycle organization differs ing cellular costs for chromosome maintenance and DNA considerably and, although the FtsZ protein is shared with repair (cf. Giovannoni et al., 2005), in line with the fact that euryarchaea, it does not play the same pivotal role, if any, the N. maritimus genome is one of the smallest among in cytokinesis. Yet another differential feature concerns free-living organisms (Walker et al., 2010). the mode of interaction between CdvA and CdvB, since The most remarkable aspect of the cell cycle organiza- protein domains important for binding in crenarchaea tion is the extensive time, in absolute terms, required for are missing in N. maritimus (Samson et al., 2011). The replication: an N. maritimus cell, thus, spends 15–18 h absence of CdvA in Caldiarchaeum subterraneum replicating a genome of only 1.645 Mb (Walker et al., (Nunoura et al., 2011), in which both CdvB and CdvC, as 2010). This requires continual maintenance of all repli- well as FtsZ, are present, indicates further cell division some components, and supply of DNA precursors, over variation among deeply branching archaea. extended time periods, and also results in long-term gene Phylogenomic evidence suggests that the last common dosage effects, with genes located in the origin-proximal ancestor of archaea and eukaryotes contained a combi- part of the chromosome being present in two copies for nation of cdv and ftsZ genes (Makarova et al., 2010), and many hours, relative to a single copy for genes closer to that Thaumarchaeota may be among the archaeal lin- the terminus. eages most closely related to Eukarya (Foster et al., The fact that chromosome replication did not continue to 2009; Gribaldo et al., 2010; Kelly et al., 2011), further completion in many cells after growth arrest could be strengthened by our observation of similarities in DNA related to the slow replication rate. If the time required to synthesis rates. Characterization of the physiological, cel- complete replication after substrate depletion exceeded lular and genomic properties of additional thaumarchaeal that during which the cells could supply sufficient metabolic species, in particular soil isolates (Schleper et al., 2005; energy and precursors for DNA synthesis, a fraction of the Schleper and Nicol, 2010), and of other recently discov- cell population would arrest during replication. Notably, ered deep archaeal lineages (de la Torre et al., 2008; genome sequence analysis, as well as electron micros- Spang et al., 2010; Blainey et al., 2011; Nunoura et al., copy, has suggested that N. maritimus may not produce 2011) will, thus, be necessary to increase our understand- significant amounts of energy reserves such as glycogen, ing of issues related to the origin of the eukaryotic lineage, polyhydroxybutyrate or polyphosphate (Walker et al., and of the fundamental evolutionary relationships 2010; Urakawa et al., 2011). Thus, in contrast to crenar- between the three domains of life.

© 2011 Blackwell Publishing Ltd, Molecular Microbiology 10 E. A. Pelve et al. ᭿ Experimental procedures fluorescent nucleic acid stain (Invitrogen) in TE buffer, and analysed in a customized Apogee A40 Analyzer flow cytom- Growth, harvest and DNA extraction eter using a 488 nm wavelength solid-state laser. Alterna- tively, the cells were stained with a combination of ethidium Nitrosopumilus maritimus strain SCM1 was grown in liquid bromide and mithramycin A and analysed with a 405 nm batch culture and growth was monitored by accumulation of solid-state laser in the same instrument, as described previ- nitrite as described previously (Martens-Habbena et al., ously (Lundgren et al., 2008). 2009). For flow cytometry, 50–100 ml samples from four dif- The relative number of cells in the pre-replicative, replica- ferent time points in two parallel cultures were deposited by tive and post-replicative stages in the DNA content histo- filtration onto 47 mm polycarbonate membranes (0.1 mm pore grams were simulated with Winflow software (Michelsen size, Whatman) at 4°C by vacuum filtration. Filters were et al., 2003), until good fits between the theoretical simula- immediately transferred to 50 ml tubes (BD Falcon) contain- tions and the experimental data were obtained. ing 70% ethanol, and kept on ice. Cells were resuspended by gently rinsing off the filter. The same protocol was used for sampling for microscopy, except that 450 ml duplicate Acknowledgements samples were used. DNA was extracted as described earlier (Urakawa et al., 2010). We thank Anna Knöppel for artwork in Fig. 3. This work was supported by grant 621-2010-5551 from the Swedish Research Council, by a Strong Research Environment grant Cloning and protein purification (Uppsala Microbiomics Center) from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Genes encoding Cdv and FtsZ proteins were amplified using a Planning to R.B., by NSF award MCB-0604448 to D.A.S. and BIO-X-ACT short PCR kit (BIOLINE), and cloned into the J.R.T., by NSF award OCE-0623174 to A. E. Ingalls, D.A.S., pET-45b(+) vector (Novagen) using T4 DNA ligase (BIOLINE), and by A. H. Devol, and NSF Award MCB-0920741 to D.A.S. generating recombinant proteins with an N-terminal hexa- and M. Hackett. histidine tag. All constructs were verified by DNA sequencing (Big Dye Terminator v3.1 Cycle Sequencing kit, Applied Bio- systems) before electroporation into Escherichia coli Rosetta References DE3 cells. The histidine-tagged proteins were affinity-purified Agogué, H., Brink, M., Dinasquet, J., and Herndl, G.J. (2008) on 5 ml HiTrap NiChelating HP columns (GE Healthcare) and Major gradients in putatively nitrifying and non-nitrifying eluted with a 0–500 mM imidazole gradient in 20 mM sodium- Archaea in the deep North Atlantic. Nature 456: 788–791. phosphate buffer, pH 7.4, containing 0.5 M NaCl. The imida- Andersson, A.F., Pelve, E.A., Lindeberg, S., Lundgren, M., zole was removed using size exclusion chromatography Nilsson, P., and Bernander, R. (2010) Replication-biased (PD-10 columns; GE Healthcare). genome organisation in the crenarchaeon Sulfolobus. BMC Genomics 11: 454. Antibody generation Beman, J.M., Popp, B.N., and Francis, C.A. (2008) Molecular and biogeochemical evidence for ammonia oxidation by Polyclonal antibodies against purified proteins were exter- marine Crenarchaeota in the Gulf of California. ISME J 2: nally produced (Innovagen, Lund, Sweden). For Nmar_0029, 429–441. Nmar_0816 and Nmar_1262 antibodies were raised in rabbit, Bernander, R. (2007) The cell cycle of Sulfolobus. Mol Micro- and for Nmar_0061, Nmar_0700 and Nmar_1088 antibodies biol 66: 557–562. were raised in chicken. Bernander, R., and Ettema, T.J.G. (2010) FtsZ-less cell division in archaea and bacteria. Curr Opin Microbiol 13: 747– 752. Immunostaining and fluorescence microscopy Bernander, R., and Poplawski, A. (1997) Cell cycle charac- Ethanol-fixed N. maritimus cells were washed with PBST teristics of thermophilic archaea. J Bacteriol 179: 4963– 4969. buffer (phosphate-buffered saline containing 0.05% Tween Blainey, P.C., Mosier, A.C., Potanina, A., Francis, C.A., and 20) and stained with primary and secondary antibodies as Quake, S.R. (2011) Genome of a low-salinity ammonia- previously described (Lindås et al., 2008). Immunostained oxidizing archaeon determined by single-cell and metage- cells were added to a thin layer of 1% agarose containing nomic analysis. PLoS ONE 6: e16626. 5 mgml-1 of the DNA-specific DAPI (4′,6-diamidino-2- Brochier-Armanet, C., Boussau, B., Gribaldo, S., and Fort- phenylindole) stain on a glass slide, and visualized using a erre, P. (2008) Mesophilic crenarchaeota: proposal for a DMRXE epifluorescence microscope (Leica) at 1000-fold third archaeal phylum, the Thaumarchaeota. Nat Rev magnification, also as described previously (Lindås et al., Microbiol 6: 245–252. 2008). Brochier-Armanet, C., Forterre, P., and Gribaldo, S. (2011) Phylogeny and evolution of the Archaea: one hundred Flow cytometry and determination of relative lengths genomes later. Curr Opin Microbiol 14: 274–281. of cell cycle periods Conti, C., Saccà, B., Herrick, J., Lalou, C., Pommier, Y., and Bensimon, A. (2007) Replication fork velocities at adjacent Nitrosopumilus maritimus cells fixed in ethanol were washed replication origins are coordinately modified during DNA with TE buffer (pH 7.9), stained with 2 mM SYTO 13 green replication in human cells. Mol Biol Cell 18: 3059–3067.

Ellen, A.F., Albers, S.V., Huibers, W., Pitcher, A., Hobel, Nicol, G.W., et al. (2006) Archaea predominate among C.F.V., Schwarz, H., et al. (2009) Proteomic analysis ammonia-oxidizing prokaryotes in soils. Nature 442: 806– of secreted membrane vesicles of archaeal Sulfolobus 809. species reveals the presence of endosome sorting Lindås, A.-C., Karlsson, E.A., Lindgren, M.T., Ettema, T.J.G., complex components. Extremophiles 13: 67–79. and Bernander, R. (2008) A unique cell division machinery Foster, P.G., Cox, C.J., and Embley, T.M. (2009) The primary in the Archaea. Proc Natl Acad Sci USA 105: 18942– divisions of life: a phylogenomic approach employing 18946. composition-heterogeneous methods. Philos Trans R Soc Lundgren, M., and Bernander, R. (2007) Genome-wide tran- Lond B Biol Sci 364: 2197–2207. scription map of an archaeal cell cycle. Proc Natl Acad Sci Francis, C.A., Beman, J.M., and Kuypers, M.M.M. (2007) USA 104: 2939–2944. New processes and players in the nitrogen cycle: the Lundgren, M., Andersson, A., Chen, L., Nilsson, P., and Ber- microbial ecology of anaerobic and archaeal ammonia nander, R. (2004) Three replication origins in Sulfolobus oxidation. ISME J 1: 19–27. species: synchronous initiation of chromosome replication Giovannoni, S.J., Tripp, H.J., Givan, S., Podar, M., Vergin, and asynchronous termination. Proc Natl Acad Sci USA K.L., Baptista, D., et al. (2005) Genome streamlining in a 101: 7046–7051. cosmopolitan oceanic bacterium. Science 309: 1242– Lundgren, M., Malandrin, L., Eriksson, S., Huber, H., and 1245. Bernander, R. (2008) Cell cycle characteristics of Crenar- Gribaldo, S., Poole, A.M., Daubin, V., Forterre, P., and chaeota: unity among diversity. J Bacteriol 190: 5362– Brochier-Armanet, C. (2010) The origin of eukaryotes and 5367. their relationship with the Archaea: are we at a phyloge- Maisnier-Patin, S., Malandrin, L., Birkeland, N.K., and Ber- nomic impasse? Nat Rev Microbiol 8: 743–752. nander, R. (2002) Chromosome replication patterns in the Guizetti, J., Schermelleh, L., Mäntler, J., Maar, S., Poser, I., hyperthermophilic euryarchaea Archaeoglobus fulgidus Leonhardt, H., et al. (2011) Cortical constriction during and Methanocaldococcus (Methanococcus) jannaschii. abscission involves helices of ESCRT-III-dependent Mol Microbiol 45: 1443–1450. filaments. Science 331: 1616–1620. Majernik, A.I., Lundgren, M., McDermott, P., Bernander, R., Gupta, R.S., and Shami, A. (2011) Molecular signatures for and Chong, J.P.J. (2005) DNA content and nucleoid dis- the Crenarchaeota and the Thaumarchaeota. Antonie Van tribution in Methanothermobacter thermautotrophicus. Leeuwenhoek 99: 133–157. J Bacteriol 187: 1856–1858. Hallam, S.J., Konstantinidis, K.T., Putnam, N., Schleper, C., Makarova, K.S., and Koonin, E.V. (2010) Two new families of Watanabe, Y., Sugahara, J., et al. (2006) Genomic analy- the FtsZ-tubulin protein superfamily implicated in mem- sis of the uncultivated marine crenarchaeote Cenar- brane remodeling in diverse bacteria and archaea. Biol chaeum symbiosum. Proc Natl Acad Sci USA 103: Direct 5: 33. 18296–18301. Makarova, K.S., Yutin, N., Bell, S.D., and Koonin, E.V. (2010) Hatzenpichler, R., Lebedeva, E.V., Spieck, E., Stoecker, K., Evolution of diverse cell division and vesicle formation Richter, A., Daims, H., and Wagner, M. (2008) A moder- systems in Archaea. Nat Rev Microbiol 8: 731–741. ately thermophilic ammonia-oxidizing crenarchaeote from Malandrin, L., Huber, H., and Bernander, R. (1999) Nucleoid a hot spring. Proc Natl Acad Sci USA 105: 2134–2139. structure and partition in Methanococcus jannaschii:an Hjort, K., and Bernander, R. (1999) Changes in cell size archaeon with multiple copies of the chromosome. Genet- and DNA content in Sulfolobus cultures during dilution ics 152: 1315–1323. and temperature shift experiments. J Bacteriol 181: Martens-Habbena, W., Berube, P.M., Urakawa, H., de la 5669–5675. Torre, J.R., and Stahl, D.A. (2009) Ammonia oxidation Hjort, K., and Bernander, R. (2001) Cell cycle regulation in kinetics determine niche separation of nitrifying Archaea the hyperthermophilic crenarchaeon Sulfolobus acidocal- and Bacteria. Nature 461: 976–979. darius. Mol Microbiol 40: 225–234. Michelsen, O., Teixeira de Mattos, M.J., Jensen, P.R., and Jackson, D.A., and Pombo, A. (1998) Replicon clusters are Hansen, F.G. (2003) Precise determinations of C and D stable units of chromosome structure: evidence that periods by flow cytometry in Escherichia coli K-12 and B/r. nuclear organization contributes to the efficient activation Microbiology 149: 1001–1010. and propagation of S phase in human cells. J Cell Biol 140: Morita, E., Colf, L.A., Karren, M.A., Sandrin, V., Rodesch, 1285–1295. C.K., and Sundquist, W.I. (2010) Human ESCRT-III and Karner, M.B., DeLong, E.F., and Karl, D.M. (2001) Archaeal VPS4 proteins are required for centrosome and spindle dominance in the mesopelagic zone of the Pacific Ocean. maintenance. Proc Natl Acad Sci USA 107: 12889– Nature 409: 507–510. 12894. Kelly, S., Wickstead, B., and Gull, K. (2011) Archaeal phylo- de Moura, A.P.S., Retkute, R., Hawkins, M., and Nieduszyn- genomics provides evidence in support of a methanogenic ski, C.A. (2010) Mathematical modelling of whole chromo- origin of the Archaea and a thaumarchaeal origin for the some replication. Nucleic Acids Res 38: 5623–5633. eukaryotes. Proc Biol Sci 278: 1009–1018. Myllykallio, H., Lopez, P., López-García, P., Heilig, R., Saurin, Könneke, M., Bernhard, A.E., de la Torre, J.R., Walker, C.B., W., Zivanovic, Y., et al. (2000) Bacterial mode of replication Waterbury, J.B., and Stahl, D.A. (2005) Isolation of an with eukaryotic-like machinery in a hyperthermophilic autotrophic ammonia-oxidizing marine archaeon. Nature archaeon. Science 288: 2212–2215. 437: 543–546. Ni, L., Xu, W., Kumaraswami, M., and Schumacher, M.A. Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., (2010) Plasmid protein TubR uses a distinct mode of HTH-

DNA binding and recruits the prokaryotic tubulin homolog Snyder, J.C., and Young, M.J. (2011) Potential role of cellular TubZ to effect DNA partition. Proc Natl Acad Sci USA 107: ESCRT proteins in the STIV life cycle. Biochem Soc Trans 11763–11768. 39: 107–110. Nunoura, T., Takaki, Y., Kakuta, J., Nishi, S., Sugahara, J., Soppa, J. (2011) Ploidy and gene conversion in Archaea. Kazama, H., et al. (2011) Insights into the evolution of Biochem Soc Trans 39: 150–154. Archaea and eukaryotic protein modifier systems revealed Spang, A., Hatzenpichler, R., Brochier-Armanet, C., Rattei, by the genome of a novel archaeal group. Nucleic Acids T., Tischler, P., Spieck, E., et al. (2010) Distinct gene set in Res 39: 3204–3223. two different lineages of ammonia-oxidizing archaea sup- Ortmann, A.C., Brumfield, S.K., Walther, J., McInnerney, K., ports the phylum Thaumarchaeota. Trends Microbiol 18: Brouns, S.J.J., van de Werken, H.J.G., et al. (2008) Tran- 331–340. scriptome analysis of infection of the archaeon Sulfolobus de la Torre, J.R., Walker, C.B., Ingalls, A.E., Könneke, M., solfataricus with Sulfolobus turreted icosahedral virus. and Stahl, D.A. (2008) Cultivation of a thermophilic J Virol 82: 4874–4883. ammonia oxidizing archaeon synthesizing crenarchaeol. Poplawski, A., and Bernander, R. (1997) Nucleoid structure Environ Microbiol 10: 810–818. and distribution in thermophilic Archaea. J Bacteriol 179: Urakawa, H., Martens-Habbena, W., and Stahl, D.A. (2010) 7625–7630. High abundance of ammonia-oxidizing Archaea in coastal Robinson, N.P., and Bell, S.D. (2007) Extrachromosomal waters, determined using a modified DNA extraction element capture and the evolution of multiple replication method. Appl Environ Microbiol 76: 2129–2135. origins in archaeal chromosomes. Proc Natl Acad Sci USA Urakawa, H., Martens-Habbena, W., and Stahl, D.A. (2011) 104: 5806–5811. Physiology and genomics of ammonia-oxidizing archaea. Robinson, N.P., Dionne, I., Lundgren, M., Marsh, V.L., Ber- In Nitrification. Ward, B.B., Klotz, M.G., and Arp, D.J. (eds). nander, R., and Bell, S.D. (2004) Identification of two Washington, DC: ASM Press, pp. 117–155. origins of replication in the single chromosome of the Walker, C.B., de la Torre, J.R., Klotz, M.G., Urakawa, H., archaeon Sulfolobus solfataricus. Cell 116: 25–38. Pinel, N., Arp, D.J., et al. (2010) Nitrosopumilus maritimus Samson, R.Y., and Bell, S.D. (2011) Cell cycles and cell genome reveals unique mechanisms for nitrification and division in the archaea. Curr Opin Microbiol 14: 350–356. autotrophy in globally distributed marine crenarchaea. Proc Samson, R.Y., Obita, T., Freund, S.M., Williams, R.L., and Natl Acad Sci USA 107: 8818–8823. Bell, S.D. (2008) A role for the ESCRT system in cell Woese, C.R., Kandler, O., and Wheelis, M.L. (1990) Towards division in archaea. Science 322: 1710–1713. a natural system of organisms: proposal for the domains Samson, R.Y., Obita, T., Hodgson, B., Shaw, M.K., Chong, Archaea, Bacteria, and Eucarya. Proc Natl Acad Sci USA P.L.-G., Williams, R.L., and Bell, S.D. (2011) Molecular and 87: 4576–4579. structural basis of ESCRT-III recruitment to membranes Wuchter, C., Abbas, B., Coolen, M.J.L., Herfort, L., van during archaeal cell division. Mol Cell 41: 186–196. Bleijswijk, J., Timmers, P., et al. (2006) Archaeal nitrifica- Schleper, C., and Nicol, G.W. (2010) Ammonia-oxidising tion in the ocean. Proc Natl Acad Sci USA 103: 12317– archaea–physiology, ecology and evolution. Adv Microb 12322. Physiol 57: 1–41. Zhang, L.-M., Offre, P.R., He, J.-Z., Verhamme, D.T., Nicol, Schleper, C., Jurgens, G., and Jonuscheit, M. (2005) G.W., and Prosser, J.I. (2010) Autotrophic ammonia oxida- Genomic studies of uncultivated archaea. Nat Rev Micro- tion by soil thaumarchaea. Proc Natl Acad Sci USA 107: biol 3: 479–488. 17240–17245.