Present address: School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, USA.
Loss of the actin-like protein MamK has pleiotropic effects on magnetosome formation and chain assembly in Magnetospirillum gryphiswaldense
Article first published online: 12 MAY 2010
© 2010 Blackwell Publishing Ltd
Volume 77, Issue 1, pages 208–224, July 2010
How to Cite
Katzmann, E., Scheffel, A., Gruska, M., Plitzko, J. M. and Schüler, D. (2010), Loss of the actin-like protein MamK has pleiotropic effects on magnetosome formation and chain assembly in Magnetospirillum gryphiswaldense. Molecular Microbiology, 77: 208–224. doi: 10.1111/j.1365-2958.2010.07202.x
- Issue published online: 24 JUN 2010
- Article first published online: 12 MAY 2010
- Accepted 3 May, 2010.
- Top of page
- Experimental procedures
- Supporting Information
Magnetotactic bacteria synthesize magnetosomes, which are unique organelles consisting of membrane-enclosed magnetite crystals. For magnetic orientation individual magnetosome particles are assembled into well-organized chains. The actin-like MamK and the acidic MamJ proteins were previously implicated in chain assembly. While MamK was suggested to form magnetosome-associated cytoskeletal filaments, MamJ is assumed to attach the magnetosome vesicles to these structures. Although the deletion of either mamK in Magnetospirillum magneticum, or mamJ in Magnetospirillum gryphiswaldense affected chain formation, the previously observed phenotypes were not fully consistent, suggesting different mechanisms of magnetosome chain assembly in both organisms. Here we show that in M. gryphiswaldense MamK is not absolutely required for chain formation. Straight chains, albeit shorter, fragmented and ectopic, were still formed in a mamK deletion mutant, although magnetosome filaments were absent as shown by cryo-electron tomography. Loss of MamK also resulted in reduced numbers of magnetite crystals and magnetosome vesicles and led to the mislocalization of MamJ. In addition, extensive analysis of wild type and mutant cells revealed previously unidentified ultrastructural characteristics in M. gryphiswaldense. Our results suggest that, despite of their functional equivalence, loss of MamK proteins in different bacteria may result in distinct phenotypes, which might be due to a species-specific genetic context.
- Top of page
- Experimental procedures
- Supporting Information
The ability of magnetic navigation in magnetotactic bacteria (MTB) is based on the synthesis of magnetosomes, which are complex intracellular organelles. In Magnetospirillum gryphiswaldense MSR-1 (in the following referred to as MSR) and related magnetospirilla magnetosomes comprise cubo-octahedral nanocrystals of magnetite (Fe3O4) that are enveloped by vesicles of the magnetosome membrane (MM) (Gorby et al., 1988; Schüler, 2004a). Previous studies revealed that the MM is a phospholipid bilayer, which is associated with a set of about 20 specific proteins (Gorby et al., 1988; Grünberg et al., 2001; 2004). In M. magneticum AMB-1 (in the following referred to as AMB) it was shown that the MM originates via invagination from the cytoplasmic membrane (CM) (Komeili et al., 2006).
For maximum sensitivity of magnetic orientation, magnetosomes are organized within single or multiple chains, which represent one of the highest structural levels found in a prokaryotic cell. However, a string of magnetic dipoles has an immanent tendency of collapsing to lower its magnetostatic energy unless it is properly stabilized (Kirschvink, 1982; Kobayashi et al., 2006). Recently, it has been found that the assembly and maintenance of magnetosome chains are governed by dedicated cellular structures. Two complementary studies investigated chain formation in the two related magnetic bacteria MSR and AMB by cryo-electron tomography (CET) (Komeili et al., 2006; Scheffel et al., 2006). CET has emerged as a powerful technology for bridging the gap between protein–protein interactions and cellular architecture (Lučićet al., 2005; Li and Jensen, 2009; Milne and Subramaniam, 2009). In CET, cells are embedded in vitreous ice in a close-to-native state, thereby avoiding artefacts typically associated with conventional electron microscopy. CET analysis of MSR and AMB demonstrated a cytoskeletal network of filaments, 3–4 nm in diameter, which traverse the cells adjacent to the CM. Magnetosomes were closely arranged along this magnetosome-associated cytoskeleton, which has been tentatively referred to as ‘magnetosome filament’ (MF) (Frankel and Bazylinski, 2006). It was speculated that MFs might be encoded by the mamK gene because of its sequence similarity to other cytoskeletal proteins (Grünberg et al., 2004; Schüler, 2004a), which together with other genes relevant for magnetosome formation is part of the large mamAB operon in all analysed MTB (Grünberg et al., 2001; Jogler and Schüler, 2009; Nakazawa et al., 2009; Jogler et al., 2009a; Murat et al., 2010). MamK proteins form a distinct and coherent branch within the large superfamily of prokaryotic actin-like proteins (Alps), which perform diverse functions in cell shape determination, establishment of polarity, cell division, chromosome segregation and plasmid partition (Carballido-Lopez, 2006; Derman et al., 2009).
Because of the difficulties with genetic manipulation of MTB, a mamK deletion mutant has so far been available only in AMB (Komeili et al., 2006). AMBΔmamK cells had lost their coherent chain-like structure. Instead, groups of few (two to three) neighbouring magnetosomes were separated by large gaps and appeared dispersed throughout the cell (Komeili et al., 2006). Cytoskeletal MFs could no longer be identified in tomograms of mutant cells, indicating that MamK in fact might be the structural element of the magnetosome-associated cytoskeleton. Similar to MreB and other Alps, MamK fusions to GFP displayed a filament-like organization in vivo and appeared as spiral or straight lines in cells of AMB (Dye et al., 2005; Pradel et al., 2006). MamK-GFP of AMB expressed in Escherichia coli formed straight filaments, which were structurally and functionally distinct from the known MreB and ParM filaments (Carballido-Lopez and Errington, 2003; Pradel et al., 2006). Recombinant MamK of Magnetospirillum magnetotacticum MS-1 (in the following referred as MS) in vitro polymerized into long straight filamentous bundles in the presence of a non-hydrolyzable ATP analogue (Taoka et al., 2007). In the related magnetic bacterium MSR MamK was shown to interact with the acidic repetitive MamJ protein, which was demonstrated to be another key player of magnetosome chain assembly (Scheffel et al., 2006; Scheffel and Schüler, 2007). A deletion mutant of mamJ did no longer produce straight magnetosome chains, but magnetite crystals were found arranged in compact clusters (Scheffel et al., 2006; Scheffel and Schüler, 2007), whereas empty vesicles and immature crystals are scattered throughout the cytoplasm and detached from the MFs, which were still present within MSRΔmamJ cells.
One obvious model that was suggested from these data is that MamJ connects magnetosomes to the cytoskeletal MF formed by MamK, which mechanically stabilizes the magnetosome chain and prevents it from collapsing (Komeili, 2007a,b; Scheffel and Schüler, 2007). According to this model, loss of either MamK or MamJ should result in an essentially identical phenotype, which is the abolishment of chain formation and agglomeration of magnetosome particles. However, the reported phenotypes of the MamJ deletion in MSR (agglomerated magnetosomes) and the MamK deletion in AMB (dispersed magnetosomes) were strikingly distinct. This raised the question, whether the MamJ and MamK proteins perform different or additional functions in different species of MTB, and whether the observed phenotypic differences are due to species-specific modes of chain formation caused by different genetic control in the two strains used in the two studies (Jogler and Schüler, 2007; Komeili, 2007b). In a very recent study, a second actin-like protein was discovered in the genome of AMB that shares 54.4% identity with MamK and is encoded within a genomic islet outside the MAI (Rioux et al., 2010). Like MamK, this MamK-like protein was also demonstrated to form filaments both in vivo and in vitro, and it was speculated that the presence of this second mamK-like gene might account for the variable phenotype of the ΔmamK mutant in AMB (Rioux et al., 2010).
To reconcile these conflicting observations and to further clarify the role of MamK, we characterized in detail a mamK deletion mutant of MSR by transmission electron microscopy (TEM) and CET. We show that in contrast to AMB, MamK of MSR is not required for chain formation. However, the absence of MFs results in a pleiotropic phenotype displaying shorter and fragmented chains that are displaced from their usual midcell localization. In addition, mutants are also impaired in magnetite formation. Our results suggest that MamK has a role in magnetosome chain positioning and MM vesicle formation, and loss of MamK may have distinct effects in different bacteria depending on the genetic context.
- Top of page
- Experimental procedures
- Supporting Information
Characterization of an unmarked, in-frame ΔmamK mutant of M. gryphiswaldense
We generated an unmarked, in-frame mamK deletion mutant of M. gryphiswaldense (MSR), in the following referred to as MSRΔmamK. Under standard conditions (microaerobic, 30°C) cells of MSRΔmamK strain exhibited morphology (Fig. 1), as well as growth and motility apparently identical to the wild type (WT). Whereas single crossover insertants of pEK32 were deficient in magnetite crystal formation, MSRΔmamK cells formed magnetosomes and aligned to magnetic fields. However, magnetic orientation of mutant cultures was markedly weaker than the WT as indicated by lower Cmag values (1.08 vs. 1.44 in the WT). Measurement of ∼590 magnetosomes from ∼30 cells revealed that the magnetite crystals were unaffected in shape and size, and the mean crystal diameter of 34 nm was essentially identical to that of the WT (33 nm). However, on average MSRΔmamK cells contained substantially fewer magnetosomes (19.7 crystals) per cell than WT cells (35.3 crystals) (Fig. 1D-iv) grown under identical conditions, which was also consistent with a reduced iron accumulation of mutant cells (30.4 µg mg-1 vs. 51.7 µg mg-1 dry weight in the WT, Fig. 1D-v).
Chain formation was assessed by TEM with respect to the average number of chains per cells, the average length (i.e. the number of particles), and the average distance between neighbouring particles. A chain was defined empirically by a minimum number of 10 magnetosomes that showed a linear alignment, and which were interspaced by not more than ∼50 nm from each other. Within this distance magnetic crystals are known to interact magnetically, whereas single particles spaced by about > 200 nm are magnetically uncoupled and behave as independent magnetic dipoles (Simpson, 2008; Li et al., 2009). Around 70% of WT magnetosome chains had between 19 and 44 particles, were 0.6–2.5 µm in length, tightly spaced (10.9 nm interparticle distance, 50.2 nm center-to-center distance), with smaller, apparently growing, and more widely spaced magnetite particles at the ends. In 92% of the cells straight, long and continuous chains were positioned at midcell (Fig. 1C-iii). Ectopic (e.g. terminal) chain localization or fragmented chains (6%) were found only occasionally. However, in about 17% of the WT cells 2–3 parallel chains were observed (Fig. 1A).
In contrast, MSRΔmamK cells displayed a distinct and much more inconsistent pattern of magnetosome chain configurations. A fraction of cells had lost their coherent, tightly spaced chain-like structure, but instead magnetosomes were dispersed along a linear axis throughout the cells and spaced by distances up to ∼920 nm (Fig. 1B and C). However, this pattern, which was somewhat reminiscent to the described phenotype of ΔmamK in AMB (Komeili et al., 2006) was found in only 18% of the cells (Fig. 1B-ii). Instead, the major fraction of cells still formed single or multiple tightly spaced chains. A minor fraction (5%) of the MSRΔmamK cells exhibited chains similar to those in the WT (i.e. one single, tightly spaced chain with > 10 magnetosome positioned at midcell; Fig. 1C-iii). In the majority of cells, however, chains were aberrant with respect to their number, length and position. For example, the occurrence of fragmented chains with multiple (up to 4) subchains was increased (average 1.38 chains per MSRΔmamK cell) compared with the WT (1.19 per cell). Individual fragmented chains on average were significantly shorter (14.2 particles in MSRΔmamK vs. 26.1 in the WT), and 29% of the cells had chains with fewer than 10 particles, compared with only 2% in the WT. Notably, parallel chains that were frequently present in WT cells, were never observed in MSRΔmamK cells. The position of magnetosome chains within cells was estimated by dividing the cells into three equal sectors to distinguish between midcell and terminal localization. A midcell localization of the chain was observed in only 43% of MSRΔmamK cells (WT: 92%), whereas in the majority of mutant cells the chains were located closely to the cell poles (Fig. 1C-i–ii).
We and others had observed previously that purified magnetosome particles from WT cells tend to form chains in vitro as long as the particles are enveloped by an intact MM (Grünberg et al., 2004; Kobayashi et al., 2006; Scheffel et al., 2006; Taoka et al., 2007; Li et al., 2009). To test whether the loss of MamK, which is associated with the MM of MSR (Fig. S1), had an effect on the integrity of the MM, and consequently, on chain formation in vitro, magnetosome particles purified from mutant and WT cells were investigated by TEM. Isolated mutant and WT magnetite crystals were surrounded by a MM-like organic layer of identical thickness (8–12 nm) and appearance as reported previously (Gorby et al., 1988; Schüler, 2004b; Taoka et al., 2006) with junctions that interconnected the individual particles (Fig. S2). Similar like WT magnetosomes, isolated mutant magnetosomes had a tendency to form chains as observed before (Grünberg et al., 2004) (Fig. S2-i–iv).
Complementation analysis: MamKMSR and MamKAMB are functionally equivalent
Immunodetection revealed that MamM and MamB, whose genes are located 338 and 7192 bp downstream of mamK within the mamAB operon, respectively, were expressed at WT levels in the mutant strains (data not shown), indicating that mamK deletion had no polar effects. To further preclude second-site mutations, we analysed cells transcomplemented with a functional mamK gene. Cloning and expression of the 1044 bp mamKMSR gene under control of the native PmamAB promoter (Lang et al., 2009) in pEK36 resulted in MamK expression comparable to WT levels (data not shown) upon transfer into MSRΔmamK. In addition, also magnetite formation, iron accumulation and chain formation were restored, albeit to a lower extent than in the WT (Figs 1D-iv–v and 4D).
To answer the question whether observed differences in WT and ΔmamK phenotypes between AMB and MSR might be due to sequence divergence of MamK orthologs, we also tested transcomplementation by mamKAMB, and in addition by mamKMS (M. magnetotacticum MS-1). Plasmids pEK37and pEK35 carrying mamKAMB and mamKMS, respectively, were conjugated into MSRΔmamK. Magnetosome numbers, iron accumulation, Cmag and chain formation were restored by MSRΔmamK + pEK35 (data not shown) and MSRΔmamK + pEK37 (Fig. 1D-iii–v) at comparable levels, indicating that mamKMS and mamKAMB are functionally equivalent to mamKMSR and can substitute its function in MSRΔmamK cells. We also investigated the effect of MamK overexpression by cloning of mamKMSR on pEK33 under control of the strong MSR promotor pmamDC (Lang et al., 2009). Transmission electron micrographs revealed long, straight WT-like magnetosome chains in MSR cells expressing mamK from pEK33, and filaments of similar abundance, length and thickness were visible within 3D maps. Likewise, cells displayed similar magnetosome numbers (Fig. 1D-i–v) and magnetic orientation (WT Cmag = 1.47, WT + pEK33 Cmag = 1.50), indicating that moderate overexpression of mamK has only minor effects on magnetosome chain formation (Fig. S3).
Intracellular localization of MamK
Plasmids harbouring various MamK-EGFP fusions were expressed in WT and several mutant backgrounds. Functionality of fusions was verified by partial restoration of WT-like phenotype upon expression of pAS_K, pAS_K1, pAS_K2 and pEK42 in MSRΔmamK. Fluorescence microscopy revealed distinct localization patterns from linear-to-helical filaments in all tested MamK-EGFP fusions. However, the respective patterns were dependent on the length of the linker between EGFP and MamK: Expression of pAS_K1 harbouring a C-terminal MamK-GSI-EGFP (Fig. 2A and B) fusion, and pAS_K2 harbouring an N-terminal EGFP-SAI-MamK fusion (data not shown) resulted in a linear fluorescence signal of about half the cell length that was restricted to midcell in E. coli and MSR, similar as observed in other studies (Pradel et al., 2006). In contrast, WT + pAS_K harbouring a N-terminal EGFP-LCLQGE-MamK displayed a linear-to-helical fluorescence pattern spanning from pole to pole throughout the entire cell, and occasionally forming loops (Fig. 2A). A similar pattern was also observed if pAS_K and pEK42 were expressed in MSRΔmamK, the non-magnetic mutant strain MSR-1B lacking most magnetosome genes by deletion (Schübbe et al., 2003; Ullrich et al., 2005), and E. coli (Fig. 2A–H). Previous experiments indicated that MamK interacts with the acidic MamJ protein (Scheffel et al., 2006; Scheffel and Schüler, 2007), but the filamentous localization of MamK was independent from MamJ (as demonstrated by expression in ΔmamJ strain) and other magnetosome genes (as demonstrated by expression in MSR-1B) (Scheffel et al., 2006). Therefore, we asked whether localization of MamJ was on the other hand dependent on the presence of MamK. Expression of pAS_J harbouring a MamJ-EGFP fusion in MSRΔmamK abolished its filamentous localization, but instead resulted in a punctual, cytoplasmic fluorescence signal, similar as the localization of MamJ-EGFP previously detected in MSR-1B (Scheffel et al., 2006)(Fig. 2A–J). This indicates that proper MamJ localization requires the presence of MamK. Because the linear-to-helical localization of EGFP-MamK (pAS_K, pEK42) was also observed in E. coli and M. gryphiswaldense MSR-1B backgrounds (Fig. 2E–H), this demonstrates that the filamentous localization is an intrinsic property of EGFP-MamKMSR, which does not require the presence of other magnetosome proteins. Expression of pAS_K in AMB resulted in the same filamentous signal pattern; however, spiral localization was absent (Fig. 2J).
Spontaneous formation of spheroplasts (‘coccoid bodies’) can be regularly observed in aging cells of M. gryphiswaldense and other magnetospirilla (Balkwill et al., 1980; Schüler and Köhler, 1992). Interestingly, in such spheroplast cells EGFP-MamK (pAS_K) did no longer localize within linear filaments, but instead yielded a ring-like fluorescence signal closely beneath the cell periphery, which is consistent with the peripheral ring-like, bent appearance of the magnetosome chains in such a cell as shown by fluorescence microscopy and TEM (Fig. 2J–M).
Previous CET analyses of magnetosome chain topology in AMB and MSR were limited to only single or a few cells (Komeili et al., 2006; Scheffel et al., 2006). Therefore, we analysed chain assembly and MFs in greater detail in a larger number of WT and MSRΔmamK cells of MSR. Filaments accompanying the magnetosome chains were identified in 16 of 28 of all analysed tomographic volumes from WT cells ( Fig. 3), and 2 of 10 analysed MSRΔmamK cells complemented with pEK42 (mamKMSR, Fig. 4A) and pEK34 (mamKAMB). As tomograms at the recorded magnification are limited in the field of view, they represent only a fraction of the whole cell. Moreover, some information is not accessible due to the limited tilt range and the resulting missing wedge. Therefore, it is possible that filaments in the other WT and transcomplemented cells escaped detection if localized perpendicular to the tilt-axis. Individual filaments, which were 3–6 nm in diameter and 0.5–1 µm in length, occasionally formed bundles of 2–4 filaments that were ∼20 nm in diameter and approximately 1 µm long (Movie S1). However, as filaments were not always within the same z-plane, their total dimensions can only be roughly estimated. Detected filaments were localized within the cytoplasm in close proximity to MM vesicles. Although the magnetosome chains were predominantly found at midcell, empty vesicles and filaments were also present in 100–300 nm vicinity to the inner membrane of the cell pole. However, we were unable to detect direct connections or discern distinct structures where filaments insert into the polar membrane. This was partially due to the presence of complex ordered structures resembling arrays of chemoreceptors (Briegel et al., 2009), which were located closely beneath the polar membrane (Fig. 3E–H). Analysis of different CET datasets determined an array length of ∼28 nm and a lattice distance of ∼11 nm (Fig. 3E, insets). Notably, in some cells filaments seemed to be connected laterally to the CM.
A couple of WT and transcomplemented cells were tomographed in state of early (Fig. 3A–E), and late division (Fig. 4A). According to those tomograms, the MF traverse the entire cell before septum formation, which remarkably might occur asymmetrically from only one side (Fig. 4A). During later division the magnetosome chain, and likely also the MF may be split into two parts (Figs 3A–E and 4D). However, due to the potential risk of technical artefacts these observations require future verification. Interestingly, two distinct filaments were detected in Fig. 4A within this cell. This was not observed in the analysed WT cells and might be the result of mamK expression from pEK42.
As in conventional micrographs, tomograms of WT cells frequently showed two parallel chains of particles that were aligned opposite to each other along the filaments (Figs 1A-i and 3A–E). In contrast, we failed to detect parallel chains in any of the 15 analysed MSRΔmamK tomograms, which always contained only one linear string of crystals per analysed section.
Sometimes filaments with association to the CM were detected, where magnetosomes were absent or sparse (Fig. 3A). Whereas the number of magnetite crystals was significantly lower in MSRΔmamK, the proportion of empty versus filled vesicles per tomographed area was not increased. Even though the absolute enumeration of vesicles per cell is not possible due the limited analysed area and technical limitations (e.g. limited tilt range, missing wedge), the total number of MM vesicles (empty or filled) per viewed section was also decreased in the mutant. If all tomograms of central sections at identical magnification were taken into account under the assumption that they may represent comparable, randomly chosen areas at midcell from the WT (15 cells) and MSRΔmamK (8 cells) cells, the average number of all vesicles were 21.6 (WT) versus 14.8 in the mutant, respectively, implying that the formation of MM vesicles is affected in the mutant.
Magnetosome membrane vesicles were predominantly found in close contact with the CM. In several cells, empty MM vesicles were detected that were invaginating from the CM, apparently with the MM forming a continuum with the CM (Fig. 3A, inset). The size of empty or partially filled vesicles varied between 13.4 and 33.5 nm. However, vesicle dimensions were not strictly correlated with the presence or absence of growing immature crystallites, as larger vesicles (18–33.5 nm) devoid of crystals as well as smaller vesicles harbouring small crystallites were apparent in the analysed tomograms.
Effect of MSRΔmamK on dynamic assembly of magnetosome chains
Deletion of mamK did not result in agglomeration of magnetosomes, but mutant cells still maintained a chain-like configuration. We therefore reasoned that instead merely providing a backbone or scaffold that mechanically stabilizes the magnetosome chain, the primary function of MamK might be rather in the control of the dynamic positioning and concatenating the individual particles during chain assembly and cell division as speculated before (Frankel and Bazylinski, 2006; Pradel et al., 2006). For clarification, we undertook a growth experiment in which magnetite synthesis was induced in growing MSRΔmamK and WT cells by the addition of iron to previously non-magnetic, iron-starved cells. As shown in Fig. 5, both WT and MSRΔmamK cells started to form magnetite after addition of 50 µM Fe(III)-citrate, and small (10–20 nm) crystallites could be first detected after 30 min. As in previous induction experiments (Scheffel et al., 2006; Faivre et al., 2007), initiation of magnetosome formation took place usually at few locations close to the cell periphery and scattered throughout the cell in WT and MSRΔmamK. During the first 80 min magnetite biomineralization proceeded at similar rates in both strains. After about 100 min, magnetite formation of the mutant then lagged behind the WT, resulting in a lower magnetic orientation for the MSRΔmamK over the remaining 200 min (Fig. 5). After 90 min irregular chains of magnetosomes became apparent at different loci in both mutant and WT cells. However, while WT cells formed the typical linear, and sometimes parallel chain fragments already located at midcell, chains in induced MSRΔmamK mutant cells were displaced, short and more distorted than in those WT cells. After 120 and 300 min, further development of WT chains resulted in gradual extension and midcell localization, whereas in MSRΔmamK cells fragmented, displaced chains were formed.
- Top of page
- Experimental procedures
- Supporting Information
Previous studies had suggested that the actin-like protein MamK is required for magnetosome chain assembly of AMB by aligning these organelles within the cell (Komeili et al., 2006; Pradel et al., 2006). However, the phenotype of a ΔmamK mutant of MSR generated in this study argues against an essential role in chain formation in this organism, as in spite of the absence of filamentous structures mutant cells are still able to form linear and coherent chains, albeit with reduced lengths and ectopic positions. For the assembly of a functional magnetosome chain, individual magnetite crystals have to be (i) aligned along a common linear axis, (ii) concatenated (i.e. by bringing newly synthesized particles to the nascent, tightly spaced magnetosome chain), and finally (iii) chains have to be positioned properly at their usual midcell localization. Instead merely providing a backbone or scaffold that mechanically stabilizes magnetosome chain, our data support speculations that MamK might be involved in sorting, concatenating and intracellular positioning of the magnetosome chain (Pradel et al., 2006; Jogler and Schüler, 2009). On the other hand, the fact that a small percentage of ΔmamK cells still form long chains does not necessarily indicate that MamK has no scaffolding role at all, but just could mean that it cannot be the only factor promoting chains.
Our induction experiments show that MSRΔmamK cells are also still able to assemble short, but coherent chains from dispersed magnetosome particles if challenged with the de novo synthesis of magnetite particles in non-magnetic, iron-starved cells undergoing cell division. This raises the question of how these chains are maintained in MSRΔmamK mutants against their immanent tendency to agglomerate. Several mechanisms for chain formation have been suggested. In AMB chain initiation starts with the simultaneous formation of multiple, adjacent magnetosomes (Komeili et al., 2004; Li et al., 2009). In marked difference, magnetite biomineralization in MSR is initiated at multiple sites dispersed over the entire cell (Scheffel et al., 2006) before eventually becoming concatenated into the mature coherent chains located at midcell, which suggests a control over dynamic localization and intracellular positioning of magnetosomes in this organism. When magnetic crystals continue to grow, increasing magnetostatic interactions between adjacent particles force them into close contact, which is then further stabilized by interactions through MM constituents. In the absence of the ‘bead-on-a-string’ like alignment along the MamK filaments, connections between the invaginating MM vesicles and the CM may be sufficient to maintain the linear arrangement of subchains in MSRΔmamK. In addition to the cytoskeletal MF, an elusive sheath-like structure has been postulated that may hold magnetosome chains together and in place within the cell (Kobayashi et al., 2006; Taoka et al., 2006). However, in our extensive CET studies we failed to detect any indications for the existence of such an intracellular structure.
Like in AMB, deletion of the mamK gene resulted in the complete absence of MFs. In addition, we were unable to detect any other filament-like structures in MSRΔmamK cells. This is interesting, as MSRΔmamK cells are likely to contain further cytoskeletal structures formed by other filamentous proteins, such as MreB and FtsZ, which are encoded in the genome of MSR. In a study on Caulobacter crescentus, in which the formation of MreB filaments was inhibited by A22 treatment, similar filaments, which were speculated to be FtsZ, were still detectable by CET (Li et al., 2007). Our failure to detect them thus might indicate that other cytoskeletal structures are way less abundant than MamK filaments within the cell.
Complementation of the MSRΔmamK mutant not only with mamKMSR, but also with mamKAMB and mamKMS did restore chain formation and the presence of magnetosomal filaments, indicating that these mamK alleles have equivalent functions and can substitute each other. However, despite the morphological and ultrastructural resemblance of the two organisms, the phenotypes of mamK mutants are distinct between MSR and AMB. Whereas the effect on chain assembly is less pronounced in MSRΔmamK than in AMB1ΔmamK, in which the mutant lacked the long, highly organized chains seen in WT (Komeili et al., 2006), loss of MamK in MSR has several additional, pleiotropic effects. These differences do not result from different growth conditions, as MSR grown in magnetic spirillum growth medium (i.e. the genuine AMB medium) displayed the same morphology and magnetosome organization as in FSM (Flask Standard Medium, see Experimental procedures) (data not shown). On the other hand, WT AMB cells grown in FSM under identical conditions as MSR display a distinct chain configuration that is characterized by fairly loose, widely spaced chains that may extend through the entire cell (Fig. S4), compared with the midcell localization of densely spaced MSR chains (Fig. 1A), and it has been recently argued by Rioux et al. that the fraction of misaligned magnetosome chains in AMB1ΔmamK may be well within the variability usually observed in the WT strain (Rioux et al., 2010). This relatively weak phenotype and the differences between MSR and AMB might be explained by the fact that only 52% of all genes are shared between the two strains (Richter et al., 2007). In fact, in the recent study by Rioux et al. a second mamK-like gene was discovered outside the MAI in an islet together with six additional magnetotaxis-related genes (Rioux et al., 2010). These findings support the idea that the observed differences between AMB and MSR and their mamK mutants are due to a different genomic context. Therefore, only a double deletion mutant of the mamK-like gene in addition to ΔmamK will reveal a more conclusive picture about MamK function in AMB.
One of the various effects of the mamK deletion in MSR is the formation of multiple, fragmented chains, which are strongly reminiscent to the subchains observed by Li et al. in experiments, in which magnetite formation was induced in aerobically grown non-magnetic cells of AMB by a shift to microaerobic incubation (Li et al., 2009). Each of these subchains behaved as an ideal uniaxial single-domain particle with extremely weak magnetostatic interactions between subchains (Li et al., 2009). Another distinctive feature of the MSR mutant is that the chains are displaced from their usual midcell location, which indicates that MamK is involved in positioning of magnetosomes by an as yet unknown mechanism. It has been speculated that positional information might be provided possibly by interaction with the divisome machinery that regulates proper cell division and determines septum formation (Schüler, 2008; Adams and Errington, 2009). This hypothesis has been further stimulated by the fact that a second ftsZ-like gene has been identified in the mamXY operon of MSR (Richter et al., 2007). A putative chimeric protein was found within a metagenomic clone, in which a MamK-like domain is coupled to a FtsZ-like domain (Jogler et al., 2009b). A recent study has shown that the second FtsZ-like protein in MSR is not involved in cell division but had rather an effect on biomineralization and, apparently, also on chain assembly (Ding et al., 2009). So far, it still remains to be analysed whether this or the genuine FtsZ has an effect in providing information about midcell position.
Previous studies of MamKAMB localization by Komeili et al. revealed that a C-terminal GFP fusion appeared in straight lines extending across most of the cell approximately along its inner curvature, consistent with the magnetosome-associated filaments in both localization and extent (Komeili et al., 2006). CET revealed networks of long filaments 200 to 250 nm in length running parallel along the chain of an AMB WT cell. At any position within the chain, up to seven of these filaments flanked the magnetosome with no obvious spatial pattern (Komeili et al., 2006). In contrast, fluorescence microscopy in our study revealed a more variable localization pattern of MamKMSR depending on the particular GFP fusion and linker length. Variable fluorescence patterns were also detected for other actin-like proteins like ParM and MreB (Jones et al., 2001; Møller-Jensen et al., 2002), as incorporation of EGFP into a polymeric structure might affect MamK tertiary structure. Thus, a longer linker might provide sterical freedom required for its native localization. In addition, different linkers might affect the stability of EGFP fusions, and therefore, the distinct localization of the MamK fusions might also be function-unrelated. Although MamK-EGFP fusions (pAS_K1/2), which show midcell fluorescence, also seem to complement the deletion of mamK, the pole-to-pole localization observed for EGFP-MamK (pAS_K, pEK42) is more consistent with the polar localization of MF bundles observed by CET, and also with the observation by Pradel et al. that one extremity of MamKAMB was located at the pole if expressed in E. coli (Pradel et al., 2006; 2007). It has been also suggested by these authors that MamK might somehow be involved in magnetoreception, possibly by interaction with other polar components (IcsA) (Pradel et al., 2006). Our finding that extremities of MamK in some MSR cells extend towards arrays of chemoreceptors would support a possible interaction between mamK and these structures in signal transduction.
It has been demonstrated previously that MamK interacts with MamJ, but MamJ is not required for MamK filament formation (Scheffel et al., 2006; Scheffel and Schüler, 2007). We found that on the other hand MamK is required for the filamentous localization of MamJ-EGFP (Fig. 2A–J), which is abolished in the MSRΔmamK mutant. Previous studies have suggested that MamJ attaches magnetosomes to filaments formed by MamK by direct interaction (Scheffel and Schüler, 2007). If this would be the only MamJ function, then loss of either MamJ or MamK would be expected to result in essentially identical phenotypes, which is the collapse of chains and agglomeration of magnetosomes. As chain configuration, however, is maintained even in the absence of MamK, but not MamJ, the mode of chain stabilization by MamJ cannot exclusively be accomplished by MamK interaction, but has to involve other, so far unknown mechanisms as well.
MSRΔmamK chains are also shorter, presumably as a consequence of the decreased numbers of magnetite crystals in the mutants. At this point, the reason for this unexpected effect on magnetosome formation is not clear. It has been suggested that MamK could act through establishment of magnetosome biogenesis factors (Komeili, 2007b). This might be, for example, by mediating the recruitment of other proteins required for biomineralization to the MM. A role in the recruitment and positioning of particular proteins has been demonstrated for other Alps and tubulin-like proteins, such as FtsZ, ParM and MreB, in Bacillus subtilis and E. coli (Salje and Löwe, 2008; Adams and Errington, 2009; Gamba et al., 2009; Vats et al., 2009). However, a preliminary comparative analysis of the MM between the mutant and the WT revealed virtually the same band patterns in SDS-PAGE experiments (data not shown). Intriguingly, our data imply that the number of MM vesicles is higher (21.6) in tomograms of the WT compared with 14.8 in the mutant, and the reduced number of MM vesicles in the mutant might be the reason why less magnetosome crystals are formed in cells devoid of MamK.
Our extensive CET analysis of > 40 WT and mutant cells also revealed several previously unrecognized ultrastructural features with relevance for magnetosomal organelle formation. For example, we have shown that, like in AMB the MM in MSR invaginates from the CM and at least transiently forms a continuum between the two membranes (Komeili et al., 2006), indicating that the mechanisms of intracellular differentiations are similar between these organisms. In addition, we found that in the WT formation of two parallel chains is rather common, which was never observed in MSRΔmamK cells. The reason for the formation of multiple chains in the WT is not entirely clear, but it can be speculated that MamK filaments may have a strong affinity for binding MM vesicles, possibly by its interaction with MamJ and other proteins, and thus tend to gather magnetosomes within their vicinity.
In conclusion, our data argue for a function of MamK in positioning and concatenating magnetosome chains rather than merely providing a rigid scaffold for chain alignment. Our results further suggest that the role of MamK is likely to be more complex and somewhat distinct from previously reported models that were mostly inferred from the single mamK deletion in AMB (Komeili et al., 2006; Rioux et al., 2010). At this point, the precise mechanisms by which magnetosome chains are positioned by MamK remains elusive and will require further investigation of interaction with and localization of further constituents of the magnetosome assembly and vesicle formation machinery.
- Top of page
- Experimental procedures
- Supporting Information
Bacterial strains, media and magnetosome isolation
Strains M. magneticum AMB-1 (AMB) and M. gryphiswaldense MSR-1 (MSR) were grown in modified liquid FSM and LIM (Low iron media, modified FSM) (Heyen and Schüler, 2003; Faivre et al., 2007). Modified magnetic spirillum growth medium (Blakemore et al., 1979; Komeili et al., 2004) (50 µM ferric citrate instead of ferric malate) was alternatively used where indicated. Growth of E. coli strain BW29427 (Table 1) was accomplished in lysogeny broth supplemented with 1 mM DL-αε-diaminopimelic acid (Sigma-Aldrich, Switzerland). Culture conditions for E. coli strains were as described (Sambrook and Russel, 2001). Magnetosomes were isolated from microaerobically grown 5l cultures as described elsewhere (Grünberg et al., 2004; Ohuchi and Schüler, 2009). Optical densities and Cmag values of MSR cultures were measured turbidimetrically at 565 nm with immotile cells inactivated by the addition of formaldehyde (Fluka, Switzerland) to a final concentration of 0.1% prior to the measurement (Scheffel et al., 2006).
|pEK32||4 kb flanking sequences of mamK||This work|||
|pAS_J||mamJ_egfp||(Scheffel et al., 2006)|||
|pK19mobGII||–||(Katzen et al., 1999)|||
|pBBR1MCS-2||–||(Kovach et al., 1994)|||
|pBBRPmamDC||–||(Lang et al., 2009)|||
|pBBRPmamAB||–||(Lang et al., 2009)|||
|Primer||Sequence (5′-3′)||Restriction site|
|Name||Relevant genotype or phenotype||Construction|
|Magnetospirillum gryphiswaldense,||MSR-1 E10, ΔmamK, Rifr,||This work|
|Magnetospirillum gryphiswaldense MSR-1||Rifr, Strepr derivative of DSM6361 (WT)||(Schultheiss et al., 2004)|
|Magnetospirillum gryphiswaldense MSR-1,||Rifr, Strepr, ΔmamJ||(Scheffel et al., 2006)|
|Magnetospirillum gryphiswaldense MSR-1B||Rifr, Strepr, spontaneous deletion mutant, lacking 40.4 kb within MAI, mag-||(Schübbe et al., 2003; Ullrich et al., 2005)|
|Magnetospirillum magneticum AMB-1||WT ATCC700264||(Kawaguchi et al., 1992)|
|E. coli DH5α||F-supE44 DacU169 (80lacZDλacU169(Φ 80lacZDM15) hsdR17recA1endA1gyrA96thi-1relA1||(Bethesda Research Laboratories, 1986)|
|E. coli BW29427||thrB1004 pro thi rpsL hsdS lacZΔM15 RP4-1360Δ(araBAD)567ΔdapA 1341::[erm pir(Wildtyp)]tra||K.A. Datsenko and B.L. Wanner (Purdue University, IN, USA)|
DNA techniques and southern blot
Total DNA from all strains used in this study was isolated as described previously (Marmur, 1961; Grünberg et al., 2001). Genetic constructs used to generate the MSRΔmamK deletion were amplified using standard polymerase chain reaction (PCR) procedures. The primers and plasmids used in this study are shown in Table 1. Primer sequences for amplification of DNA fragments from MSR were deduced from GenBank sequence deposition BX571797. The primer pairs used for the amplification of mamK from Magnetospirillum strains AMB and MS were deduced from sequence deposition NC_007626 and NZ_AAAP01003824.1 respectively. For sequencing, BigDye terminators v3.1 (Applied Biosystems, Darmstadt, Germany) were used. Sequence data were analysed with Lasergene 6 (DNAstar, Madison, WI) and MacVector 7.2.3 (Oxford Molecular, Oxford, UK) programs. Southern blots were performed by standard procedures as described previously (Ullrich et al., 2005) (Fig. S5).
Construction of an unmarked in-frame mamK deletion mutant
Gene deletion was accomplished via homologous recombination of up- and downstream mamK flanking sequence in pEK32 with the MSR chromosome. A 2003 bp upstream and a 2009 bp downstream mamK flanking region were amplified with primer pairs EK2_K_u_f /EK_K_u_r and EK_K_d_f/EK2_K_d_r. The fragments were fused by cloning into the XbaI and BamHI restriction sites of pK19mobGII vector, resulting in pEK32. Plasmids were transferred into M. gryphiswaldense by biparental conjugation and screening was performed as previously described (Schultheiss et al., 2004). Screening for the gusA marker (Katzen et al., 1999) revealed 8 KanR insertion mutants occurring with a frequency of 1.7 × 10−7 on X-Gluc (5-Bromo-4-Chloro-3-Indolyl-β-D-glucuronic acid, 0.5 mM) containing solid FSM plates. After 2 passages for 3 days in 300 µl liquid culture and subsequent 1 day in 10 ml (∼5 generations), an incubation for 10 days on solid FSM at 30°C under microoxic conditions was performed. Screening by PCR of 24 white colonies for putative double crossover resolvants revealed eight positive candidates that were gusA- and KanS. Double crossover mutants were obtained at a final frequency of 2.6 × 10−3. Successful unmarked in-frame deletion of mamK was confirmed by PCR and southern blotting using a 816 bp DNA probe amplifying parts of mamJ and mamK gene of MSR with primer pairs EK_K_f/ EK_probeK02_r (Fig. S5). One clone named MSRΔmamK was selected for further analysis.
A cloned fragment comprising an open-reading frame of 1083 bp that starts with a CTG codon according to the previously deposited database sequence [CAJ30118 (Ullrich et al., 2005)] was not functionally expressed (data not shown). Inspection of the aligned sequences of mamK nucleotide sequences of mamK orthologs from MSR, AMB (1044 bp, 33 nt mismatches, 90.8% aa identity) and MS (1044 bp, 34 nt mismatches, 90.6% aa identity) revealed a conserved ATG in all three orthologs 39 bp downstream of the predicted CTG start codon in the older database version of mamKMSR, which hereafter was considered as the correct start codon, resulting in an open-reading frame of 1044 bp as in the other strains.
The mamK gene was either amplified with primer pair EKmamK_F_kurz/EKmamK_R02 and the 1056 bp fragment subsequently cloned into NdeI and BamHI restriction sites of pBBRPmamDC generating pEK36, or with primer pair EKmamK_F02/EKmamK_R02 resulting in a 1083 bp fragment in which the start codon CTG was changed into an ATG. Likewise, the ATG defines the recognition site of 5′ NdeI restriction enzyme. Together with the 3′ BamHI recognition site the fragment was cloned into pBBR1pmamAB (Lang et al., 2009) yielding pEK36 and subsequently transformed into MSRΔmamK via biparental mating as described (Schultheiss and Schüler, 2003; Schultheiss et al., 2004). The mamK of AMB was PCR amplified with primer pairs EK_mamK_AMB1_F/EK_mamK_AMB1_R2 yielding a 1056 bp fragment, which was cloned into NdeI/ BamHI restriction site of pBBR1pmamAB or pBBRpmamDC, resulting in pEK37 and pEK34 respectively. Amplification and subsequent cloning of MS mamK into pBBR1pmamDC resulted in pEK35.
Construction of Mam-EGFP fusions
Several different mamK-egfp (enhanced GFP) expression fusions were constructed via fusion PCR (Ho et al., 1989). The mamK gene of MSR (accession number CAJ30118) and egfp (pEGFP-N2, Clontech) were amplified using primers as described, resulting in the C-terminal MamK fusion pAS_K1 (Scheffel, 2007). For construction of the N-terminal fusion with a 6 amino acid linker sequence, primer pairs for egfp-gene ASEGFP_f10/ ASEGFP_r11, and for mamK ASmamKs_f3/ ASmamKe_r2, were used and resulted in pAS_K vector (Scheffel, 2007).
Gel electrophoresis and Western blot experiments
Protein concentrations were determined with a BCA-Protein Micro assay kit (Pierce) according to the manufacturer's instructions. For one-dimensional SDS-PAGE of magnetosome-associated proteins we used the procedure of Laemmli (Laemmli, 1970). An amount of magnetosome particles or solubilisate equivalent to 6.5 µg of protein was mixed with electrophoresis sample buffer containing 2% (wt/wt) SDS and 5% (wt/vol) 2-mercaptoethanol. After boiling for 5 min, samples were centrifuged for 3 min. The supernatants were loaded onto polyacrylamide gels containing various concentrations of polyacrylamide (8% to 16%). The non-magnetic fraction of whole cells was further processed as described elsewhere (Grünberg et al., 2001), and 6.5 µg protein was used for analysis on SDS gels. Western blot experiments were performed as explained in previous work (Schübbe et al., 2006). The anti-MamK primary antibody was raised by S. Schübbe as described (Schübbe, 2005).
MSR WT and MSRΔmamK bearing the plasmids pAS_J, pAS_K(1,2), were grown in 15 ml polypropylene tubes with sealed screw caps and a culture volume of 10 ml to stationary phase. The cell membranes were stained with the membrane stain FM4-64 (Invitrogen, Karlsruhe, Germany) at a final concentration of 16.4 µM, immobilized on agarose pads (FSM salts in H2O, supplemented with 1% agarose), and imaged with an Olympus BX81 microscope equipped with a 100 UPLSAPO100XO objective (numerical aperture of 1.40) and a Hamamatsu Orca AG camera. Images were captured and analysed using Olympus cell software.
TEM and CET
For TEM analysis, unstained cells were adsorbed on carbon coated copper grids and air-dried (Plano, Wetzlar). Bright field TEM was performed on a FEI Tecnai F20 transmission electron microscope (FEI; Eindhoven, the Netherlands) at an accelerating voltage of 200 kV. Images were captured with a FEI Eagle 4096 × 4096 pixel CCD camera using EMMenue 4.0 and FEI's Explore 3D.
For CET, a FEI Tecnai F30 Polara transmission electron microscope (FEI; Eindhoven, the Netherlands), equipped with 300 kV field emission gun, a Gatan GIF 2002 Post-Column Energy Filter, and a 2048 × 2048 pixel Gatan CCD Camera (Gatan; Pleasanton, CA) were used. All data collection was performed at 300 kV, with the energy filter operated in the zero-loss mode (slit width of 20 eV). Tilt series were acquired using Serial EM (Mastronarde, 2005) and FEI's Explore 3D software.
Quantifoil copper grids (Quantifoil Micro Tools GmbH, Jena) were prepared by placing a 5 µl droplet of 10–15 nm colloidal gold clusters (Sigma) on each grid for subsequent alignment purposes. A 5 µl droplet of a fresh MSR culture was added onto the prepared grid, and after blotting was embedded in vitreous ice by plunge freezing into liquid ethane (temperature c. −170°C). Single-axis tilt series for tomography were typically recorded with 2 increments over an angular range of ±65. To minimize the electron dose applied to the ice-embedded specimen, data were recorded at low-dose conditions by using automated data acquisition software. The total dose accumulated during the tilt series was kept below 100 e/Å2. To account for the increased specimen thickness at high tilt angles, the exposure time was multiplied by a factor of 1/cos α. The object pixel size in unbinned images was 0.661 at a magnification of 34 000×, 0.805 at 27 500× and 0.979 at 22 500×. Images were recorded at nominal defocus values of −8 µm or −4 µm.
Three-dimensional reconstructions from tilt series were performed with the weighted back-projection method and further analysis of the tomograms was done using the TOM toolbox (Nickell et al., 2005). Visualizations of the tomograms were done with Amira (http://www.amiravis.com) on 2 times binned volumes.
Determination of iron content by atomic absorption spectroscopy
Cells were grown to an optical density of 0.2 and aliquots of 1 ml were taken and pelleted. After a subsequent pellet washing step in 0.5 ml 20 mM HEPES + 5 mM EDTA both the supernatant (900 µl) and the pellet were supplemented with 10 µl and 100 µl 65% nitric acid respectively. Incubation at 98°C for 2 h dissolved all organic material and magnetosomes, and dilutions of 1:100 and 1:1000 in H2O were analysed by atomic absorption spectroscopy (Varian AA240) using SpectrAA 240FS software version 5.01 (Varian, Australia) with the following parameters set: wavelength 248.3 nm, slit width 0.2 nm, cathode lamp current 10.0 mA.
- Top of page
- Experimental procedures
- Supporting Information
We are grateful to Günter Pfeifer (MPI of Biochemistry) for help with the TEM. This work was supported by the Max Planck Society and the Deutsche Forschungsgemeinschaft (Schu1080/9-1 and 10-1).
- Top of page
- Experimental procedures
- Supporting Information
- 2009) Bacterial cell division: assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7: 642–653. , and (
- 1980) Ultrastructure of a magnetotactic spirillum. J Bacteriol 141: 1399–1408. , , and (
- Bethesda Research Laboratories (1986) BRL pUC host: E. coli DH5 alpha competent cells. Focus 8: 1–19.
- 1979) Isolation and pure culture of freshwater magnetic spirillum in chemically defined medium. J Bacteriol 140: 720–729. , , and (
- 2009) Universal architecture of bacterial chemoreceptor arrays. Proc Natl Acad Sci USA 106: 17181–17186. , , , , , , et al. (
- 2006) The bacterial actin-like cytoskeleton. Microbiol Mol Biol Rev 70: 888–909. (
- 2003) A dynamic bacterial cytoskeleton. Trends Cell Biol 13: 577–583. , and (
- 2009) Phylogenetic analysis identifies many uncharacterized actin-like proteins (Alps) in bacteria: regulated polymerization, dynamic instability and treadmilling in Alp7A. Mol Microbiol 73: 534–552. , , , , , , et al. (
- 2009) Deletion of the ftsZ-like gene results in the production of superparamagnetic magnetite magnetosomes in Magnetospirillum gryphiswaldense. J Bacteriol 192: 1097–1105. , , , , , , et al. (
- 2005) Two independent spiral structures control cell shape in Caulobacter. Proc Natl Acad Sci USA 102: 18608–18613. , , , , and (
- 2007) Intracellular magnetite biomineralization in bacteria proceeds by a distinct pathway involving membrane-bound ferritin and an iron(II) species. Angew Chem Int Ed Engl 46: 8495–8499. , , , and (
- 2006) How magnetotactic bacteria make magnetosomes queue up. Trends Microbiol 14: 329–331. , and (
- 2009) Two-step assembly dynamics of the Bacillus subtilis divisome. J Bacteriol 191: 4186–4194. , , , , and (
- 1988) Characterization of the bacterial magnetosome membrane. J Bacteriol 170: 834–841. , , and (
- 2001) A large gene cluster encoding several magnetosome proteins is conserved in different species of magnetotactic bacteria. Appl Environ Microbiol 67: 4573–4582. , , , and (
- 2004) Biochemical and proteomic analysis of the magnetosome membrane in Magnetospirillum gryphiswaldense. Appl Environ Microbiol 70: 1040–1050. , , , , and (
- 2003) Growth and magnetosome formation by microaerophilic Magnetospirillum strains in an oxygen-controlled fermentor. Appl Microbiol Biotechnol 61: 536–544. , and (
- 1989) Site-directed mutagenesis by overlap extension using the polymerase chain reation. Gene 77: 51–59. , , , , and (
- 2007) Genetic analysis of magnetosome biomineralization. In Magnetoreception and Magnetosomes in Bacteria. SchülerD. (ed.). Berlin, Heidelberg: Springer-Verlag Berlin Heidelberg, pp. 133–161. , and (
- 2009) Genomics, genetics, and cell biology of magnetosome formation. Annu Rev Microbiol 63: 501–521. , and (
- 2009a) Comparative analysis of magnetosome gene clusters in magnetotactic bacteria provides further evidence for horizontal gene transfer. Environ Microbiol 11: 1267–1277. , , , , , , et al. (
- 2009b) Toward cloning of the magnetotactic metagenome: identification of magnetosome island gene clusters in uncultivated magnetotactic bacteria from different aquatic sediments. Appl Environ Microbiol 75: 3972–3979. , , , , , , et al. (
- 2001) Control of cell shape in bacteria: helical, actin-like filaments in Bacillus subtilis. Cell 104: 913–922. , , and (
- 1999) New mobilizable vectors suitable for gene replacement in gram-negative bacteria and their use in mapping of the 3′ end of the Xanthomonas campestris pv. campestris gum operon. Appl Environ Microbiol 65: 278–282. , , , , and (
- 1992) Phylogeny and 16S rRNA sequence of Magnetospirillum sp. AMB-1, an aerobic magnetic bacterium. Nucleic Acids Res 20: 1140. , , and (
- 1982) Paleomagnetic evidence for fossil biogenic magnetite in western Crete. Earth Planet Sci Lett 59: 388–392. (
- 2006) Experimental observation of magnetosome chain collapse in magnetotactic bacteria: sedimentological, paleomagnetic, and evolutionary implications. Earth Planet Sci Lett 245: 538–550. , , , , , , et al. (
- 2007a) Cell biology of magnetosome formation. In Magnetoreception and Magnetosomes in Bacteria. Schüler, D. (ed.). Berlin, Heidelberg: Springer-Verlag Berlin Heidelberg, pp. 163–173. (
- 2007b) Molecular mechanisms of magnetosome formation. Annu Rev Biochem 76: 351–366. (
- 2004) Magnetosome vesicles are present before magnetite formation, and MamA is required for their activation. Proc Natl Acad Sci USA 101: 3839–3844. , , , and (
- 2006) Magnetosomes are cell membrane invaginations organized by the actin-like protein MamK. Science 311: 242–245. , , , and (
- 1994) pBBR1MCS: a broad-host-range cloning vector. BioTechniques 16: 800–802. , , , , and (
- 1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685. (
- 2009) Identification of promoters for efficient gene expression in Magnetospirillum gryphiswaldense. Appl Environ Microbiol 75: 4206–4210. , , and (
- 2009) Electron cryotomography: a new view into microbial ultrastructure. Curr Opin Microbiol 12: 1–8. , and (
- 2009) Magnetite magnetosome and fragmental chain formation of Magnetospirillum magneticum AMB-1: transmission electron microscopy and magnetic observations. Geophys J Int 177: 33–42. , , , , , and (
- 2007) The structure of FtsZ filaments in vivo suggests a force-generating role in cell division. EMBO J 26: 4694–4708. , , , and (
- 2005) Structural studies by electron tomography: from cells to molecules. Annu Rev Biochem 74: 833–865. , , and (
- 1961) A procedure for the isolation of deoxiribonucleic acid from microorganisms. J Mol Biol 3: 208–218. (
- 2005) Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol 152: 36–51. (
- 2009) Cryo-electron tomography of bacteria: progress, challenges and future prospects. Nat Rev Microbiol 7: 666–675. , and (
- 2002) Prokaryotic DNA segregation by an actin-like filament. EMBO J 21: 3119–3127. , , , and (
- 2010) Comprehensive genetic dissection of the magnetosome gene island reveals the step-wise assembly of a prokaryotic organelle. Proc Natl Acad Sci USA 107: 5593–5598. , , , and (
- 2009) Whole genome sequence of Desulfovibrio magneticus strain RS-1 revealed common gene clusters in magnetotactic bacteria. Genome Res 19: 1801–1808. , , , , , , et al. (
- 2005) TOM software toolbox: acquisition and analysis for electron tomography. J Struct Biol 149: 227–234. , , , , , , et al. (
- 2009) In vivo display of a multisubunit enzyme complex on biogenic magnetic nanoparticles. Appl Environ Microbiol 75: 7734–7738. , and (
- 2006) Biogenesis of actin-like bacterial cytoskeletal filaments destined for positioning prokaryotic magnetic organelles. Proc Natl Acad Sci USA 103: 17485–17489. , , , , and (
- 2007) Polar positional information in Escherichia coli spherical cells. Biochem Biophys Res Commun 353: 493–500. , , , , , and (
- 2007) Comparative genome analysis of four magnetotactic bacteria reveals a complex set of group-specific genes implicated in magnetosome biomineralization and function. J Bacteriol 189: 4899–4910. , , , , , , and (
- 2010) A second actin-like MamK protein in Magnetospirillum magneticum AMB-1 encoded outside the genomic magnetosome island. PLoS ONE 5: e9151. , , , , , and (
- 2008) Bacterial actin: architecture of the ParMRC plasmid DNA partitioning complex. EMBO J 27: 2230–2238. , and (
- 2001) Molecular Cloning: A Laboratory Manual. New York: Cold Spring Harbor Laboratory Press, p. 999. , and (
- 2007) Molekulare und strukturelle Untersuchung zur Bildung von Magnetosomen und zur Assemblierung von Magnetosomenketten. Doctoral thesis. University of Bremen, 1–194. (
- 2007) The acidic repetitive domain of the Magnetospirillum gryphiswaldense MamJ protein displays hypervariability but is not required for magnetosome chain assembly. J Bacteriol 189: 6437–6446. , and (
- 2006) An acidic protein aligns magnetosomes along a filamentous structure in magnetotactic bacteria. Nature 440: 110–114. , , , , , , and (
- 2005) Untersuchung zur molekularen Regulation und Organisation der mam-Gene in Magnetospirillum gryphiswaldense MSR-1. Doctoral thesis. University of Bremen, 1–167. (
- 2003) Characterization of a spontaneous nonmagnetic mutant of Magnetospirillum gryphiswaldense reveals a large deletion comprising a putative magnetosome island. J Bacteriol 185: 5779–5790. , , , , , , et al. (
- 2006) Transcriptional organization and regulation of magnetosome operons in Magnetospirillum gryphiswaldense. Appl Environ Microbiol 72: 5757–5765. , , , , , , and (
- 2004a) Molecular analysis of a subcellular compartment: The magnetosome membrane in Magnetospirillum gryphiswaldense. Arch Microbiol 181: 1–7. (
- 2004b) Characterization of the magnetosome membrane in Magnetospirillum gryphiswaldense. In Biominerlization. Baeuerlein, E. (ed.). Weinheim: Wiley-VCH Verlag GmbH, pp. 109–118. (
- 2008) Genetics and cell biology of magnetosome formation in magnetotactic bacteria. FEMS Microbiol Rev 32: 654–672. (
- 1992) The isolation of a new magnetic spirillum. Zentralbl Mikrobiol 147: 150–151. , and (
- 2003) Development of a genetic system for Magnetospirillum gryphiswaldense. Arch Microbiol 179: 89–94. , and (
- 2004) Inactivation of the flagellin gene flaA in Magnetospirillum gryphiswaldense results in nonmagnetotactic mutants lacking flagellar filaments. Appl Environ Microbiol 70: 3624–3631. , , and (
- 2008) Electron holography of isolated and interacting magnetic nanocrystals. Doctoral thesis. University of Cambridge, 1–197. (
- 2006) Spatial localizations of Mam22 and Mam12 in the magnetosomes of Magnetospirillum magnetotacticum. J Bacteriol 188: 3805–3812. , , , , , and (
- 2007) Polymerization of the actin-like protein MamK, which is associated with magnetosomes. J Bacteriol 189: 8737–8740. , , , and (
- 2005) A hypervariable 130-kilobase genomic region of Magnetospirillum gryphiswaldense comprises a magnetosome island which undergoes frequent rearrangements during stationary growth. J Bacteriol 187: 7176–7184. , , , , and (
- 2009) Assembly of the MreB-associated cytoskeletal ring of Escherichia coli. Mol Microbiol 72: 170–182. , , and (
- Top of page
- Experimental procedures
- Supporting Information
|MMI_7202_sm_FigS1-5-MovieS1.pdf||454K||Supporting info item|
|MMI_7202_sm_MovieS1.mpg||45837K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.