Identification and functional characterization of ... - Wiley Online Library

Molecular Microbiology (2010) 76(2), 480–488 䊏

doi:10.1111/j.1365-2958.2010.07117.x First published online 31 March 2010

Identiﬁcation and functional characterization of liposome tubulation protein from magnetotactic bacteria mmi_7117 480..488

Masayoshi Tanaka, Atsushi Arakaki and Tadashi Matsunaga* Department of Biotechnology, Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei, Tokyo 184-8588, Japan.

Summary Magnetotactic bacteria synthesize intracellular magnetosomes that are comprised of membraneenveloped magnetic crystals. In this study, to identify the early stages of magnetosome formation, we isolated magnetosomes containing small magnetite crystals and those containing regular-sized magnetite crystals from Magnetospirillum magneticum AMB-1. This was achieved by using a novel size fractionation technique, resulting in the identification of a characteristic protein (Amb1018/MamY) from the small magnetite crystal fraction. The gene encoding MamY was located in the magnetosome island. Like the previously reported membrane deformation proteins, such as bin/amphiphysin/Rvs (BAR) and the dynamin family proteins, recombinant MamY protein bound directly to the liposomes, causing them to form long tubules. We established a mamY gene deletion mutant (DmamY) and analysed MamY protein localization in it for functional characterization of the protein in vivo. The DmamY mutant was found to have expanded magnetosome vesicles and a greater number of small magnetite crystals relative to the wild-type strain, suggesting that the function of the MamY protein is to constrict the magnetosome membrane during magnetosome vesicle formation, following which, the magnetite crystals grow to maturity within them.

Introduction Membrane constriction is an important process in organelle formation and cell/membrane division (Zhang and Hinshaw, 2001; Wang et al., 2009). Many endocytic proteins, such as dynamin and the bin/amphiphysin/Rvs (BAR) protein family, play a critical role during clathrinAccepted 24 February, 2010. *For correspondence. E-mail [email protected]; Tel. (+81) 42 388 7020; Fax (+81) 42 385 7713.

© 2010 Blackwell Publishing Ltd

mediated endocytosis in eukaryotic organelle formation (Dawson et al., 2006; Kuroiwa et al., 2008; Pucadyil and Schmid, 2009). On the other hand, the tubulin homologue, FtsZ, is the major cytoskeletal protein in bacterial cytokinesis and is related to cell division (Monahan et al., 2009; Osawa et al., 2009). These proteins have shown the potential for liposome tubulation activity in vitro (Danino et al., 2004; Peter et al., 2004). On the other hand, methyl-accepting chemotaxis protein (MCP) in Escherichia coli has been reported to deform the cell membrane (Khursigara et al., 2008). Magnetotactic bacteria synthesize highly crystalline intracellular magnetite (Fe3O4) surrounded by a lipid bilayer membrane (Blakemore, 1975). The membranous organelles, or magnetosomes, are derived from cytoplasmic membrane invagination (Komeili et al., 2006; Tanaka et al., 2006). Magnetosomes containing regular-sized (40–70 nm) magnetite crystals (RMs) and those containing small crystals (SMs), which are probably immature, are simultaneously present in the same cell (Fig. 1A). In addition, they are properly arranged as a linear chain. Recently, a magnetite crystallization mechanism was proposed whereby nucleation predominantly occurs within the vesicle, and subsequent crystal growth proceeds after the invaginated membranes are spatially separated from the cytoplasmic membrane (Scheffel et al., 2006; Faivre et al., 2007). The compartmentalization of the magnetosome vesicles from the cytoplasmic membrane may therefore be required to control crystal growth. Previous molecular studies have revealed that in magnetotactic bacteria, magnetosomes are formed by a unique set of proteins (Nakamura et al., 1995; Arakaki et al., 2003; Komeili et al., 2006; Scheffel et al., 2006). Although vesicle formation is believed to be a key process, limited knowledge regarding it has been obtained thus far. Mms16, a 16 kDa protein identified in Magnetospirillum magneticum AMB-1, is a small GTPase that was originally proposed to prime the invagination for vesicle formation (Okamura et al., 2001); however, a subsequent study on Magnetospirillum gryphiswaldense MSR-1 proposed that its function is poly(3hydroxybutyrate) depolymerization (Schultheiss et al., 2005). One of the dominant magnetosome proteins is a tetratricopeptide repeat protein designated as MamA (Okuda and Fukumori, 2001; Komeili et al., 2004). A

M. magneticum AMB-1 liposome tubulation protein 481

a mamY deletion mutant (DmamY) for in vivo study and a recombinant protein for in vitro study, with the aim of facilitating the understanding of the early stages of magnetosome formation.

Results Separation of SMs and RMs In previous studies, magnetosomes have been purified by magnetic separation or ultracentrifugation for proteomic analyses (Grunberg et al., 2004; Tanaka et al., 2006; Matsunaga et al., 2009). However, these methods do not allow fractionation of SMs and RMs. In this study, we developed a purification method using magnetic separation and centrifugation (Fig. 2A), which allowed comparative proteomic analysis of SMs and RMs (Fig. 2B). The most common and reliable techniques used for size distribution determination of the particles were transmission electron microscopy (TEM; Fig. 1B and C) and dynamic light scattering (DLS; data not shown), allowing optimization of the fractionation and purification conditions for obtaining the SMs. The average particle sizes were manually measured from the TEM images, by which small and regular magnetite crystals were found to be approximately 28 nm and 59 nm in size respectively.

Identification and bioinformatic characterization of protein in SMs

Fig. 1. Size distribution of magnetosomes from magnetotactic bacteria. A. Transmission electron micrographs of M. magneticum AMB-1 cells and the magnified image; scale bar = 200 nm. The SMs in the chain are encircled with dashed lines. B. Transmission electron micrographs of the fractionated magnetosomes from M. magneticum AMB-1. Purified RMs (i) and SMs (ii) were analysed; scale bar = 200 nm. The insets correspond to the magnified images. C. TEM for size distribution analysis of the fractionated SMs and RMs: dashed line, SMs; thin line, RMs; and thick line, magnetosomes in M. magneticum AMB-1.

mamA deletion mutant revealed the formation of a large number of empty vesicles, and the MamA protein has been proposed to function in the activation of magnetosome vesicles (Komeili et al., 2004). Here, we report the serendipitous discovery of a membrane tubulation protein, MamY, from an SM fraction. This report describes: (i) the development of a magnetosome size fractionation method, (ii) a comparative proteomic study and (iii) the successful identification of the novel MamY protein and its functional characterization by using

When membrane proteins (40 mg) were extracted from the SMs and RMs, many proteins were observed to be commonly distributed among the fractions and were similar to the profiles reported previously (Tanaka et al., 2006). Although successful separation of cell and magnetosome membranes was determined by the amount of the MamA protein in each fraction, one of the major outer membrane proteins (Msp1) was also found to be a contaminant (Tanaka et al., 2008). After comparing the protein profiles of the RM and SM fractions, the most distinct protein band appearing only in the SM profile was selected for the subsequent studies (Fig. 2B, band indicated with a black arrow). Although sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDSPAGE) revealed five protein bands specific to the SM fraction, only the gene product of mamY (amb1018) was successfully identified. The MamY protein has already been identified in M. gryphiswaldense MSR-1 by a proteomic study of magnetosomes containing both SMs and RMs (Grunberg et al., 2004); however, the protein could not be identified in the studies of RMs in M. magneticum AMB-1 (Tanaka et al., 2006) and Desulfovibrio magneticus RS-1 (Matsunaga et al., 2009), in which it is expected to be found, considering its localization to SMs

© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 76, 480–488

482 M. Tanaka, A. Arakaki and T. Matsunaga 䊏

Fig. 2. Size fractionation of SMs and RMs and identification of the MamY protein. A. Size fractionation of SMs and RMs: sup., supernatant; ppt., precipitate. B. SDS-PAGE protein profiles of various M. magneticum AMB-1 cell fractions. Protein (40 mg) from each fraction was loaded onto the gel. The bands corresponding to the MamY and MamA proteins are indicated by the black and white arrows respectively; Lane 1: cytoplasmic-periplasmic fraction; Lane 2: cell membrane fraction; Lane 3: proteins from SMs after washing with 100 mM NaCl; Lane 4: proteins from RMs after washing with 100 mM NaCl; M: marker.

(Fig. 2B). The mamY gene is coded within the magnetosome island (MAI), an important genome region for magnetosome formation in magnetotactic bacteria (Schubbe et al., 2003; Fukuda et al., 2006). The mamY gene operon is formed from four genes, and this gene cluster is conserved in other magnetotactic bacteria, such as M. gryphiswaldense MSR-1, Magnetospirillum magnetotacticum MS-1 and Magnetococcus MC-1 (Richter et al., 2007). Therefore, this operon is hypothesized to be involved in magnetosome formation. In order to predict the function of the mamY gene, a homology search was conducted against the DNA Data Bank of Japan (DDBJ) database (http://www.ddbj.nig.ac. jp/). The result showed that the MamY protein was annotated as an MCP. For further characterization of this protein, domain prediction using the SMART programme (http://smart.embl-heidelberg.de/) was conducted, and the MamY protein showed sequence similarity with BARrelated protein domains, such as the BAR domain (aa15– aa228, 4.45e+3) and the FER-CIP4 homology (FCH) domain (aa66–aa139, 1.16e+03). Furthermore, at the amino acid sequence level, the MamY protein had approximately 10% amino acid identity and 42% similarity to CentaurinBAR (BAR domain of Centaurinß2). The novel BAR proteins, RICH-1 and Tuba, have previously been reported in eukaryotic cells, showing 22% and 24% identity, respectively, and 44% and 39% similarity, respectively, with the classical BAR protein (Salazar et al., 2003;

Richnau et al., 2004). In addition, the BAR domain and MCP possess a coiled-coil repeat structure for stable binding to biological membranes (Shimada et al., 2007; Khursigara et al., 2008). PSIPRED programme (http:// bioinf.cs.ucl.ac.uk/psipred/) predicted this characteristic secondary structure to be present in the MamY protein, suggesting that this protein is functionally similar to these membrane deformation proteins (Fig. S1). Liposome sedimentation assay with CentaurinBAR and MamY To analyse the protein function in vitro, the CentaurinBAR gene fragment and the mamY gene were expressed and purified (purity > 98%; Fig. S2). The BAR domain is known to exhibit direct binding activity with liposomes that are composed of phospholipids (Peter et al., 2004; Richnau et al., 2004). The positive charge of the basic amino acid, positioned at the inner centre of the concave shape of the BAR domain, induces membrane deformation (Peter et al., 2004; Ren et al., 2006; Shimada et al., 2007). In order to evaluate whether the MamY protein interacts directly with the liposomes, a liposome sedimentation assay was conducted. As shown in Fig. 3, the CentaurinBAR and MamY proteins coprecipitated with the liposomes, showing that both bind directly; however, the control (GST) did not coprecipitate. © 2010 Blackwell Publishing Ltd, Molecular Microbiology, 76, 480–488


proteins respectively. At a high concentration of the MamY protein (120 mM), membrane deformation occurred much more readily with many liposome tubules and buds; however, the deformation ratio hardly showed a change with a low concentration of the protein (Fig. 4). GST and magnetosome protein Mms6 were evaluated as negative controls for liposome tubulation activity: no liposome structural changes were observed following the addition of either of these proteins. In addition, we examined and verified membrane tubulation activity by using a lipid extract of the magnetosome membrane (Bligh and Dyer, 1959), and tubulation of the magnetosome membrane was successfully observed on using the MamY protein (30 mM; Fig. 4). Fig. 3. Liposome sedimentation assay with the CentaurinBAR and MamY proteins. Coomassie-stained gels of the sedimentation assay of CentaurinBAR, MamY and GST (negative control) proteins at 6 mM each with or without 0.6 mg ml-1 of liver lipid liposomes; P, pellet; S, supernatant.

Liposome tubulation assay with CentaurinBAR and MamY Purified recombinant MamY protein (30 mM) efficiently deformed the lipid bilayer of the liposomes in the liposome-based assay that supported the generation of coated intermediates of the invaginated membranes, leading to the formation of tubules with outer diameters of 20 nm (Fig. 4) (Ren et al., 2006). The CentaurinBAR protein, used as a positive control, showed similar liposome tubulation activity (20 nm diameter; Fig. 4). Under these conditions, approximately 7% and 4% of the liposomes were tubulated by the CentaurinBAR and MamY

mamY gene deletion and functional analysis in M. magneticum AMB-1 The mamY gene is located in an operon, consisting of four genes, within the MAI: mamY (amb1018), amb1017, amb1016 and amb1015. An in frame gene deletion (to avoid polar effects) was carried out for mamY gene mutation. The growth curves of the obtained strains (the DmamY strain, the complementation strain harbouring pUMPmamY_mamY_gfp, and the control wild-type strain harbouring pUMPmamY_Gmr) were charted. As shown in Fig. S3, gene deletion or complementation barely affected cell growth, and 90% of both the wild-type and DmamY strains yielded a positive magnetic response, turning in a rotating magnetic field. Because the MamY protein was annotated as the MCP protein on BLAST search, MamY

Fig. 4. Liposome tubulation assay with the CentaurinBAR and MamY proteins. Electron micrographs of the liver lipid liposomes incubated with the CentaurinBAR and MamY proteins at 30 mM each. The use of a higher concentration of the MamY protein (120 mM) and liposome tubulation using lipids from the magnetosome membrane were also examined; scale bar = 200 nm. The Mms6 and GST proteins were used as the negative controls. The inset shows liposomes at a higher magnification. © 2010 Blackwell Publishing Ltd, Molecular Microbiology, 76, 480–488


Fig. 5. Transmission electron micrographs and distributions of the diameters of the magnetosome vesicles in the DmamY, complementation and wild-type strains. A. Micrographs of ultrathin sections of the DmamY (i) and wild-type (ii) strains. The arrows indicate the magnetosome vesicles; scale bar = 100 nm. Note the presence of a large magnetosome vesicle and small magnetite crystals in the DmamY strain. B. Size distribution of magnetosome vesicles (i) and magnetite crystals (ii) in various strains; thick line, DmamY strain; thin line, complementation strain; dashed line, wild-type strain.

was overexpressed in the M. magneticum AMB-1 and E. coli cells. However, no significant change was observed in their membrane structure under cryo-electron microscopy (data not shown). To compare the DmamY strain with the wild-type strain of M. magneticum AMB-1, the magnetosome vesicles in the DmamY strain were evaluated by TEM [Fig. 5A and B(i)]. Size determination from ultrathin-sectioned cell images carries the risk of underestimation because magnetosome vesicles may not be sliced accurately along their maximum vesicle sizes. However, this risk was reduced by measuring many samples (> 300 magnetosome membranes) to determine the influence of the mamY gene mutation (Scheffel et al., 2008). Overall, the magnetosome vesicles appeared larger on average in the DmamY strain (67.9 nm) than vesicles in the wild-type cells (60.1 nm) [Fig. 5B(i)]. The Mann–Whitney probability (P-value) determined for the diameter of the magnetosome vesicles of the wild-type and DmamY strains was < 1E-04, indicating a highly statistically significant difference (Mann and Whitney, 1947). Conversely, while the average number of crystals per cell was similar in the DmamY strain and wild-type (~20), the number of small crystals relative to the large ones was greater in the DmamY strain [Fig. 5B(ii); P < 1E-05].

Subcellular localization analysis of MamY in M. magneticum AMB-1 We used the bioinformatic programmes, HMMTOP (Tusnady and Simon, 1998, Tusnady and Simon, 2001) and TMHMM (Sonnhammer et al., 1998; Moller et al., 2001), to predict two transmembrane segments in the MamY protein. To further understand MamY localization and function, a C-terminal fusion of MamY to the green fluorescent protein (GFP) was established in M. magneticum AMB-1 (Fig. 6A). MamY-GFP was visualized as a linear clumped structure in the DmamY strain, which appears to correspond to a clumped area of SMs within the chain (Fig. 1A). Localization of the MamY-6H protein was further confirmed by Western blot analysis (Fig. 6B) using a transformant harbouring the pUMPmamY_mamY_ 6H plasmid. The analysis confirmed that the MamY protein was abundantly localized to the SMs, but rarely to the RMs or the cell membrane.

Discussion In this study, we used a technique involving simple centrifugation and magnetic separation to successfully purify magnetosomes containing magnetite crystals of two average-sized populations, SMs (28 nm) and RMs © 2010 Blackwell Publishing Ltd, Molecular Microbiology, 76, 480–488


Fig. 6. Localization analysis of the MamY protein in M. magneticum AMB-1. A. Analysis of MamY-GFP within the DmamY strain by fluorescent microscopy. The images to the left show the nuclei stained with Hoechst 33342, the central images show GFP localization, and the images to the right are an overlay of the Hoechst 33342 and GFP images; scale bar = 1 mm. B. Western blot analysis of the MamY-6H protein. The obtained samples from the transformant harbouring pUMPmamY_mamY_6H were loaded for SDS-PAGE and detected using the anti-hexa histidine antibody as described under Experimental procedures; Lane 1: cytoplasmic-periplasmic fraction; Lane 2: cell membrane fraction; Lane 3: proteins from SMs after washing with 100 mM NaCl; Lane 4: proteins from RMs after washing with 100 mM NaCl.

(59 nm), from M. magneticum AMB-1 (Fig. 1B and C). A comparative study was conducted to identify the protein in the early stages of magnetosome formation. The novel protein, MamY, in M. magneticum AMB-1 was identified from the SM fraction. Magnetotactic bacteria synthesize magnetite crystal chains with clumps of large and small particles distributed throughout their cells (Fig. 1A). The GFP-tagged MamY protein showed a linear clumped structure, and MamY-6H was mainly detected in the SM fraction (Fig. 6B). These data clearly showed that MamY localizes in the SMs in M. magneticum AMB-1. The eukaryotic membrane compartmentalization process is known to involve many membrane proteins (Dawson et al., 2006; Tsujita et al., 2006). A similar membrane constriction process may be required for magnetosome vesicle formation. Some coiled-coil repeat structure within the protein functions in stable binding of the membrane to other biological membranes and its deformation (Shimada et al., 2007; Khursigara et al., 2008). This characteristic secondary structure was predicted to be present in the MamY protein (Fig. S1), and the activities of binding to liposome and its tubulation suggest that it is responsible for membrane constriction leading to magnetosome vesicle formation (Figs 3 and 4). However, whether the interaction of the MamY protein and liposome is based on the coiled-coil structure, the predicted transmembrane regions, or the other positively charged amino acids remains unknown. Magnetosomes are derived from cytoplasmic membrane invagination (Komeili et al., 2006; Tanaka et al.,

2006). A study conducted on M. gryphiswaldense MSR-1 suggested that the process of magnetite crystal growth within magnetosome vesicles has two distinct stages: (i) small magnetite crystallization within an immature magnetosome vesicle that is still attached to and has invaginated from the cytoplasmic membrane and (ii) crystal growth within the mature dissociated magnetosome vesicle (Scheffel et al., 2006; Faivre et al., 2007). On the other hand, magnetite crystal growth within the invaginated magnetosome (not pinched off from the cytoplasmic membrane) and many magnetosome invaginations (~34%) of the cytoplasmic membrane were observed in M. magneticum AMB-1 (Komeili et al., 2006). However, it remains unclear whether the magnetosome membranes are spatially connected or separated from the cytoplasmic membrane when the crystals mature. The stage of magnetite crystal growth in which the separation event occurs (iron accumulation, nucleation, or maturation) also remains unknown. In this study, we propose that MamY is involved in the constriction of the cytoplasmic membrane to form magnetosome vesicles. The DmamY mutant was found to have expanded magnetosome vesicles and a greater number of SMs relative to the wild-type strain (Fig. 5B); however, the number of magnetite crystals had barely changed (~20). The size defects of the vesicles in the DmamY mutant accounts for the incomplete membrane constriction process. Because magnetite formation is highly sensitive to the conditions (Arakaki et al., 2008; Faivre and Schüler, 2008), unregulated sizes of the vesicles and/or the constriction spaces connecting the vesicles and cytoplasm might affect magnetite crystal growth, resulting in a greater number of small crystals formed in the deletion mutant. However, whether the MamY protein directly encourages crystal growth or not remains unclear. In addition, the absence of the MamY protein did not crucially affect magnetosome formation in the magnetotactic bacterium (Fig. 5). Probably, some proteins besides the MamY protein play a role in the complicated process of magnetosome membrane invagination as shown in eukaryotic endocytosis. Based on our results and observations, we propose that the function of the MamY protein is to constrict the magnetosome membrane during magnetosome vesicle formation, and then, the magnetite crystals grow to maturity within them. Although these data are inconclusive with regard to the membrane deformation mechanism, this study provides new information on the evolution of the vesicle formation process from prokaryotes to eukaryotes. Further studies, including structural analysis, characterization of the interaction with the MamY protein, and direct observation of protein localization at the molecular level in vivo, are necessary to comprehend the function of this protein and its similarity to the other membrane deforma-



tion protein. The identification of a liposome tubulation protein should also facilitate molecular-level investigations into the similarity of the membrane deformation process and dynamics in eukaryotes and prokaryotes. Furthermore, the size-optimized magnetosomes obtained by the novel simple fractionation method could have practical applications in established technologies.

Experimental procedures Strains, plasmids, primers, double-stranded DNA and growth conditions The strains, plasmids, primers and double-stranded DNA used in the study are described in detail in Table S1. M. magneticum AMB-1 (ATCC700264) was cultured anaerobically in magnetic spirillum growth medium (MSGM) in an 8-L fermentor as described previously (Yang et al., 2001).

Isolation of SM and RM fractions Figure 2A presents a schematic representation of the entire procedure used for separating the SMs; all the experimental steps were performed at 4°C. Approximately 5.0 ¥ 1013 cells in 100 ml of HEPES buffer (10 mM, pH 7.4) were disrupted five times using a French press (1500 kg cm-2, UR-200P; Tomy Seiko). The magnetosomes were separated magnetically from the disrupted cells using cylindrical Nd–B magnets (15 mm in diameter, 10 mm in height). Because magnetosomes align as a chain within the cells, a magnetosome chain, comprising SMs and RMs, was collected during this separation step. The magnetosome chain was then disrupted by weak sonication for 30 s with W-170ST (Honda Electronics) in HEPES buffer (10 mM, pH 7.4). The dispersed magnetosomes were collected magnetically to obtain a precipitate comprising RMs and a supernatant containing a mixture of SMs and non-precipitated RMs. During these magnetic separation steps, the empty magnetosome membranes were eliminated from the samples. The precipitate was washed 18 times by weak sonication with HEPES buffer containing 100 mM NaCl (NaCl–HEPES buffer). RMs were removed from the supernatant by centrifugation (10 000 g, 2 ¥ 10 min). After centrifugation, NaCl (100 mM, final concentration) was added to the supernatant containing the SMs. Finally, the SMs were collected with an Nd–B magnet and washed 18 times with the same NaCl–HEPES buffer. The other cellular fractions were collected by previously described methods (Okamura et al., 2001; Tanaka et al., 2006). The SMs and RMs were verified by TEM under a Hitachi H-700 and DLS on a DLS-6000 spectrophotometer (Otsuka Electronics; Fig. 1B and C). The TEM samples of the fractionated magnetosomes were prepared by drop coating onto a carbon-coated copper grid, and the TEM measurements were carried out at an accelerating voltage of 150 kV.

Liposome sedimentation and tubulation assays with CentaurinBAR and MamY Liposomes from the total liver lipids (Avanti) were extruded 15 times through polycarbonate membranes (Avanti) to achieve

the desired size. For the sedimentation assay, the CentaurinBAR and MamY proteins were bound at a concentration of 6 mM each to 400 nm of liposomes at a concentration of 0.6 mg ml-1. The liposomes were sedimented at 45 000 r. p.m. in a Beckman Coulter TLA 45 rotor for 30 min. The supernatant was decanted completely, and the sedimented liposomes were diluted quickly in liposome buffer (150 mM NaCl, 1 mM DTT and 20 mM HEPES; pH 7.4). The liposome sample was then analysed by SDS-PAGE (Fig. 3). The liposome tubulation assay was performed as described previously (Peter et al., 2004; Masuda et al., 2006; see also McMahon’s laboratory protocols: http://www.endocytosis.org/ techniqs/Liposome.html). For the tubulation assay, 1 mg ml-1 of 400 nm liposomes was incubated for 20 min at 37°C with 30 mM (low concentration) or 120 mM (high concentration) of each protein. The sample was spotted onto 150-mesh copper grids (Nisshin EM) and analysed under a Hitachi H-7100 operated at an accelerating voltage of 100 kV (Fig. 4).

Characterization of the DmamY, complementation and wild-type strains of M. magneticum AMB-1 The cultured magnetotactic bacteria in the stationary phase were fixed with 2% glutaraldehyde in 0.05 M sodium cacodylate buffer overnight. After washing with 0.05 M sodium cacodylate buffer, the samples were post-fixed with 2% osmium tetroxide for 2 h at 4°C, washed with MilliQ at 4°C, dehydrated with ethanol and embedded in EPON812. Ultrathin sections were obtained from several blocks, stained with lead citrate and uranyl acetate and observed under a JEOL JEM1200EX operated at 80 kV. Over 300 magnetosome vesicles in the cells (WT strain, 320; DmamY strain, 302; complementation strain, 308) were manually counted and measured. Mann–Whitney test was conducted to show the significance of the differences in the magnetosome vesicle sizes of the wild-type and the DmamY strains (Mann and Whitney, 1947). At a P-value of < 1E-03, the difference was considered to be highly statistically significant in this test. For characterizing the magnetite crystals, the cultured bacteria were directly spotted onto the grids, and the crystals within the cells (WT strain, 305; DmamY strain, 300; complementation strain, 302) were observed under a Hitachi H-7100.

Localization analysis of MamY by fluorescent observation and Western blotting The DmamY strain was transformed with pUMPmamY_mamY_gfp, harvested and stained with Hoechst 33342 (Invitrogen) according to the manufacturer’s protocol. The cells were placed on thin agarose pads on microscope slides. Microscopy was carried out with an Olympus BX51 and a charge-coupled device camera (DP70; Olympus). Hoechst 33342 and GFP were visualized with a standard WU filter and WIBA filter respectively. The images were acquired with the ¥100 oil objective, overlaid into the DP Controller imaging software and cropped using Adobe Photoshop 7.0. To verify the localization of the MamY protein, the DmamY strain was also transformed with pUMPmamY_mamY_6H. The obtained cells were fractionated to the cytoplasmicperiplasmic, cell membrane and SM and RM fractions by a © 2010 Blackwell Publishing Ltd, Molecular Microbiology, 76, 480–488


previously described method (Yoshino and Matsunaga, 2006). Twenty micrograms of the proteins from these fractions were subjected to SDS-PAGE on a 15% (wt/vol) polyacrylamide gel. The gel was electroblotted onto a polyvinylidene difluoride membrane. For immunostaining of the Western blots, His-probe (H-3) mouse monoclonal antibodies (1:2000 dilution; Santa Cruz Biotechnology) and goat anti-mouse IgG alkaline phospatase conjugates (1:2000 dilution; Calbiochem) were used. The membrane was developed with BCIP/NBT-Blue (substrate solution; Sigma).

Acknowledgements This work was funded in part by a Grant-in-Aid for Scientific Research (A) (No. 18206084) from the Ministry of Education, Science, Sports and Culture of Japan. M. Tanaka thanks the Japan Society for the Promotion of Science (JSPS) for financial support. We are grateful to Tetsushi Mori for assistance in preparing the manuscript in English.

References Arakaki, A., Webb, J., and Matsunaga, T. (2003) A novel protein tightly bound to bacterial magnetic particles in Magnetospirillum magneticum strain AMB-1. J Biol Chem 278: 8745–8750. Arakaki, A., Nakazawa, H., Nemoto, M., Mori, T., and Matsunaga, T. (2008) Formation of magnetite by bacteria and its application. J R Soc Interface 5: 977–999. Blakemore, R.P. (1975) Magnetotactic bacteria. Science 190: 377–379. Bligh, E., and Dyer, W. (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37: 911–917. Danino, D., Moon, K.H., and Hinshaw, J.E. (2004) Rapid constriction of lipid bilayers by the mechanochemical enzyme dynamin. J Struct Biol 147: 259–267. Dawson, J.C., Legg, J.A., and Machesky, L.M. (2006) Bar domain proteins: a role in tubulation, scission and actin assembly in clathrin-mediated endocytosis. Trends Cell Biol 16: 493–498. Faivre, D., and Schüler, D. (2008) Magnetotactic bacteria and magnetosomes. Chem Rev 108: 4875–4898. Faivre, D., Bottger, L.H., Matzanke, B.F., and Schuler, D. (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. Fukuda, Y., Okamura, Y., Takeyama, H., and Matsunaga, T. (2006) Dynamic analysis of a genomic island in Magnetospirillum sp. strain AMB-1 reveals how magnetosome synthesis developed. FEBS Lett 580: 801–812. Grunberg, K., Muller, E.C., Otto, A., Reszka, R., Linder, D., Kube, M., et al. (2004) Biochemical and proteomic analysis of the magnetosome membrane in Magnetospirillum gryphiswaldense. Appl Environ Microbiol 70: 1040–1050. Khursigara, C.M., Wu, X., Zhang, P., Lefman, J., and Subramaniam, S. (2008) Role of HAMP domains in chemotaxis signaling by bacterial chemoreceptors. Proc Natl Acad Sci USA 105: 16555–16560. Komeili, A., Vali, H., Beveridge, T.J., and Newman, D.K.

(2004) Magnetosome vesicles are present before magnetite formation, and MamA is required for their activation. Proc Natl Acad Sci USA 101: 3839–3844. Komeili, A., Li, Z., Newman, D.K., and Jensen, G.J. (2006) Magnetosomes are cell membrane invaginations organized by the actin-like protein MamK. Science 311: 242–245. Kuroiwa, T., Misumi, O., Nishida, K., Yagisawa, F., Yoshida, Y., Fujiwara, T., and Kuroiwa, H. (2008) Vesicle, mitochondrial, and plastid division machineries with emphasis on dynamin and electron-dense rings. Int Rev Cell Mol Biol 271: 97–152. Mann, H.B., and Whitney, D.R. (1947) On a test of whether one of two random variables is stochastically larger than the other. Ann Math Stat 18: 50–60. Masuda, M., Takeda, S., Sone, M., Ohki, T., Mori, H., Kamioka, Y., and Mochizuki, N. (2006) Endophilin BAR domain drives membrane curvature by two newly identified structure-based mechanisms. EMBO J 25: 2889–2897. Matsunaga, T., Nemoto, M., Arakaki, A., and Tanaka, M. (2009) Proteomic analysis of irregular, bullet-shaped magnetosomes in the sulphate-reducing magnetotactic bacterium Desulfovibrio magneticus RS-1. Proteomics 9: 3341– 3352. Moller, S., Croning, M.D., and Apweiler, R. (2001) Evaluation of methods for the prediction of membrane spanning regions. Bioinformatics 17: 646–653. Monahan, L.G., Robinson, A., and Harry, E.J. (2009) Lateral FtsZ association and the assembly of the cytokinetic Z ring in bacteria. Mol Microbiol 74: 1004–1017. Nakamura, C., Burgess, J.G., Sode, K., and Matsunaga, T. (1995) An iron-regulated gene, magA, encoding an iron transport protein of Magnetospirillum sp. strain AMB-1. J Biol Chem 270: 28392–28396. Okamura, Y., Takeyama, H., and Matsunaga, T. (2001) A magnetosome-specific GTPase from the magnetic bacterium Magnetospirillum magneticum AMB-1. J Biol Chem 276: 48183–48188. Okuda, Y., and Fukumori, Y. (2001) Expression and characterization of a magnetosome-associated protein, TPRcontaining MAM22, in Escherichia coli. FEBS Lett 491: 169–173. Osawa, M., Anderson, D.E., and Erickson, H.P. (2009) Curved FtsZ protofilaments generate bending forces on liposome membranes. EMBO J 28: 3476–3484. Peter, B.J., Kent, H.M., Mills, I.G., Vallis, Y., Butler, P.J., Evans, P.R., and McMahon, H.T. (2004) BAR domains as sensors of membrane curvature: the amphiphysin BAR structure. Science 303: 495–499. Pucadyil, T.J., and Schmid, S.L. (2009) Conserved functions of membrane active GTPases in coated vesicle formation. Science 325: 1217–1220. Ren, G., Vajjhala, P., Lee, J.S., Winsor, B., and Munn, A.L. (2006) The BAR domain proteins: molding membranes in fission, fusion, and phagy. Microbiol Mol Biol Rev 70: 37–120. Richnau, N., Fransson, A., Farsad, K., and Aspenstrom, P. (2004) RICH-1 has a BIN/Amphiphysin/Rvsp domain responsible for binding to membrane lipids and tubulation of liposomes. Biochem Biophys Res Commun 320: 1034– 1042. Richter, M., Kube, M., Bazylinski, D.A., Lombardot, T., Glock-



ner, F.O., Reinhardt, R., and Schuler, D. (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. Salazar, M.A., Kwiatkowski, A.V., Pellegrini, L., Cestra, G., Butler, M.H., Rossman, K.L., et al. (2003) Tuba, a novel protein containing bin/amphiphysin/Rvs and Dbl homology domains, links dynamin to regulation of the actin cytoskeleton. J Biol Chem 278: 49031–49043. Scheffel, A., Gruska, M., Faivre, D., Linaroudis, A., Plitzko, J.M., and Schuler, D. (2006) An acidic protein aligns magnetosomes along a filamentous structure in magnetotactic bacteria. Nature 440: 110–114. Scheffel, A., Gardes, A., Grunberg, K., Wanner, G., and Schuler, D. (2008) The major magnetosome proteins MamGFDC are not essential for magnetite biomineralization in Magnetospirillum gryphiswaldense but regulate the size of magnetosome crystals. J Bacteriol 190: 377–386. Schubbe, S., Kube, M., Scheffel, A., Wawer, C., Heyen, U., Meyerdierks, A., et al. (2003) Characterization of a spontaneous nonmagnetic mutant of Magnetospirillum gryphiswaldense reveals a large deletion comprising a putative magnetosome island. J Bacteriol 185: 5779–5790. Schultheiss, D., Handrick, R., Jendrossek, D., Hanzlik, M., and Schüler, D. (2005) The presumptive magnetosome protein Mms16 is a poly(3-hydroxybutyrate) granule-bound protein (phasin) in Magnetospirillum gryphiswaldense. J Bacteriol 187: 2416–2425. Shimada, A., Niwa, H., Tsujita, K., Suetsugu, S., Nitta, K., Hanawa-Suetsugu, K., et al. (2007) Curved EFC/F-BARdomain dimers are joined end to end into a filament for membrane invagination in endocytosis. Cell 129: 761–772. Sonnhammer, E.L., von Heijne, G., and Krogh, A. (1998) A hidden Markov model for predicting transmembrane helices in protein sequences. Proc Int Conf Intell Syst Mol Biol 6: 175–182. Tanaka, M., Okamura, Y., Arakaki, A., Tanaka, T., Takeyama, H., and Matsunaga, T. (2006) Origin of magnetosome membrane: proteomic analysis of magnetosome membrane and comparison with cytoplasmic membrane. Proteomics 6: 5234–5247. Tanaka, M., Nakata, Y., Mori, T., Okamura, Y., Miyasaka, H.,

Takeyama, H., and Matsunaga, T. (2008) Development of a cell surface display system in a magnetotactic bacterium, ‘Magnetospirillum magneticum’ AMB-1. Appl Environ Microbiol 74: 3342–3348. Tsujita, K., Suetsugu, S., Sasaki, N., Furutani, M., Oikawa, T., and Takenawa, T. (2006) Coordination between the actin cytoskeleton and membrane deformation by a novel membrane tubulation domain of PCH proteins is involved in endocytosis. J Cell Biol 172: 269–279. Tusnady, G.E., and Simon, I. (1998) Principles governing amino acid composition of integral membrane proteins: application to topology prediction. J Mol Biol 283: 489–506. Tusnady, G.E., and Simon, I. (2001) The HMMTOP transmembrane topology prediction server. Bioinformatics 17: 849–850. Wang, Q., Navarro, M.V., Peng, G., Molinelli, E., Lin Goh, S., Judson, B.L., et al. (2009) Molecular mechanism of membrane constriction and tubulation mediated by the F-BAR protein Pacsin/Syndapin. Proc Natl Acad Sci USA 106: 12700–12705. Yang, C., Takeyama, H., Tanaka, T., and Matsunaga, T. (2001) Effects of growth medium composition, iron sources and atmospheric oxygen concentrations on production of luciferase-bacterial magnetic particle complex by a recombinant Magnetospirillum magneticum AMB-1. Enzyme Microb Technol 29: 13–19. Yoshino, T., and Matsunaga, T. (2006) Efficient and stable display of functional proteins on bacterial magnetic particles using mms13 as a novel anchor molecule. Appl Environ Microbiol 72: 465–471. Zhang, P., and Hinshaw, J.E. (2001) Three-dimensional reconstruction of dynamin in the constricted state. Nat Cell Biol 3: 922–926.

Supporting information Additional supporting information may be found in the online version of this article. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
