Assessing bacterial magnetotactic behavior by using permanent magnet blocks

Invi ted Art ic le Materials Science

Assessing bacterial magnetotactic behavior by using permanentmagnet blocks

Tao Song • Hong-Miao Pan • Zheng Wang •

Tian Xiao • Long-Fei Wu

Received: 27 November 2013 / Accepted: 19 January 2014 / Published online: 28 March 2014

� Science China Press and Springer-Verlag Berlin Heidelberg 2014

Abstract Assessing the movement of magnetotactic

bacteria (MTB) under magnetic fields is a key to exploring

the function of the magnetotaxis. In this study, a simple

method was used to analyze the behavior of MTB, which

was based on the accumulation of cells on the walls of a

test tube when two permanent magnet blocks were applied

on the tube. Experimental results showed a significant

difference among the movements of the polar MTB, axial

MTB, and ferrofluid. The polar magnetotactic cells

aggregated as spots above or below the two magnet blocks

besides the aggregated spots underneath the magnet blocks.

By contrast, the axial magnetotactic cells aggregated only

as two round spots underneath the magnet blocks, and more

cells aggregated in the center than all around of the spot.

For the ferrofluid, two spots were also formed underneath

the magnet blocks, and the aggregated particles formed a

ring shape. Magnetic calculation by finite element method

was used to analyze the phenomenon, and the findings were

reasonably explained by the MTB features and magnetic

field theory. A scheme that differentiates polar MTB, axial

MTB, and magnetic impurity could be developed, which

would be beneficial to fieldworks involving MTB in the

future.

Keywords Magnetotactic bacteria � Permanent

magnet � Finite element method � Ferrofluid

1 Introduction

Magnetotactic bacteria (MTB) comprise a group of Gram-

negative aquatic prokaryotes found in both freshwater and

marine environments [1, 2]. These bacteria possess mag-

netosomes, which are intracellular magnetic iron nano-

crystals surrounded by a lipid bilayer membrane.

Magnetosomes form a chain (or chains) inside the cell, thus

enabling the bacterium to migrate along geomagnetic field

lines and to maintain its position within the boundary of the

oxic-anoxic transition zone (OATZ); this behavior is

known as magnetotaxis [3, 4].

According to an accepted viewpoint, the function of

magnetic fields on MTB is limited to passive orientation of

the bacterium that results from the torque exerted by the

ambient magnetic field on the biomagnetic compass (i.e.,

the magnetosome chains) of the bacterium as it swims [5].

Magnetotaxis in MTB actually exhibits either polar or axial

magneto-aerotaxis [6]. In several species of magnetotactic

spirilla, such as Magnetospirillum magnetotacticum, cells

are oriented by the ambient magnetic field. Thus, these

species swim parallel or antiparallel to the magnetic field to

form an aerotactic band, i.e., axial magneto-aerotaxis. By

contrast, bilophotrichously flagellated (i.e., with two fla-

gellar bundles on one hemisphere of the cell) magnetotactic

T. Song � Z. Wang

Beijing Key Laboratory of Bioelectromagnetism, Institute of

Electrical Engineering, Chinese Academy of Sciences,

Beijing 100190, China

T. Song (&) � H.-M. Pan � T. Xiao � L.-F. Wu

France-China Bio-Mineralization and Nano-Structures

Laboratory, Beijing 100193, China

e-mail: [email protected]

H.-M. Pan � T. Xiao

Key Laboratory of Marine Ecology & Environmental Sciences,

Institute of Oceanology, Chinese Academy of Sciences,

Qingdao 266071, China

L.-F. Wu

Laboratory of Bacterial Chemistry, UMR7283, Institute of

Mediterranean Microbiology, Aix-Marseille University, CNRS,

F-13402 Marseille Cedex 20, France

123

Chin. Sci. Bull. (2014) 59(17):1929–1935 csb.scichina.com

DOI 10.1007/s11434-014-0298-2 www.springer.com/scp

cocci swim persistently in a preferred direction that is

relative to the magnetic fields, i.e., polar magneto-aerotaxis

[6]. However, the relationships between magnetic fields

and aerotaxis have not yet been clearly elucidated. Several

reports even suggested that magnetotaxis is not a passive

progress but a magnetoreception [7, 8].

Early observations showed that polar MTB from the

Northern Hemisphere preferentially swim parallel to the

magnetic field, corresponding to a northward migration in

the geomagnetic field. Hence, such MTB are known as

north-seeking. Conversely, bacteria from the Southern

Hemisphere that preferentially swim antiparallel to the

magnetic field are known as south-seeking [3, 9]. The

relationships between bacteria and magnetic fields are

shown in Fig. 1. Interestingly, a significant number of

south-seeking MTB have been observed in the Northern

Hemisphere, thereby conflicting with previous models of

the adaptive value of magnetotaxis [10]. Such conflict is

attributed to the revelation of the selectivity of the redox

gradient configuration on the magnetotactic polarity of the

cells [11].

Several methods to assess the magnetotaxis of MTB

have been reported. A common method is to observe the

motion of MTB in a water droplet under an optical

microscope by applying the magnetic fields [12, 13].

However, the application of this method is limited by the

absence of controlling oxygen concentration and gradient.

Other methods include using microcapillary tubes [6] or

biochips [14]. A simpler method is to observe the accu-

mulation of cells on the wall of a bottle or a test tube when

two permanent magnet blocks were placed on opposite

sides of the bottle or test tube. Avoiding the effects of

magnetic force caused by the magnetic gradient is difficult

in this method. In fact, completely excluding magnetic

attraction to bacteria in all methods that involve permanent

magnet blocks remain a challenge. In the present study, we

used the aforementioned simple method to investigate the

magnetotaxic behavior of polar magnetotactic bacteria

collected from a seawater pond in an intertidal zone at

Huiquan Bay, the China Sea [15]. We compared the

behavior of these bacteria with those of the axial MTB M.

magnetotacticum and an artificial magnetic particle solu-

tion (ferrofluid). The validity and limit of the method in

exploring the magnetotaxic behavior were analyzed based

on the experimental results.

2 Materials and methods

2.1 Bacterial strains and ferrofluid

Polar MTB were collected from a seawater pond located at

Huiquan Bay, Qingdao, China. Sediments and seawater,

with a ratio of 1:2 were collected and stored in bottles. The

MTB in these samples were magnetically purified in Pas-

teur pipettes according to the racetrack-purification method

[16]. A virtual monoculture of ovoid-coccoid, bilopho-

trichously flagellated MTB was isolated and designated as

Qingdao Huiquan Low (QHL) tide MTB (Fig. 2a). These

cells ranged from 1.8 to 2.3 lm in width and from 2.0 to

2.8 lm in length. Most QHL cells had two magnetosome

chains. The average number of magnetosomes per cell was

18. Most of the magnetosome crystals (Fe3O4) in QHL

cells were rectangular and measured (101 ± 24) nm (mean

with standard deviation) in length and (83 ± 21) nm in

width. The suspension for the cells was concentrated to

108 cells/mL for the experiments.

M. magneticum AMB-1 (ATCC700264) was also used

in this study. Cells were grown microaerobically at 28 �C

in an enriched magnetic spirillum growth medium at pH

6.75 as described by Yang et al. [17]. M. magneticum

AMB-1 cells are helical with a diameter of approximately

Fig. 1 Definition of magnetotaxis. The long arrow indicates the direction of the magnetic field. The short arrows beside the bacteria indicate the

swimming direction. The short arrows inside the bacteria indicate the magnetized direction of the magnetosomes. When the swimming direction

of the bacteria is parallel to the magnetic field direction, the magnetotaxis is denoted as north-seeking. When the swimming direction is

antiparallel to the magnetic field direction, the magnetotaxis is denoted as south-seeking

1930 Chin. Sci. Bull. (2014) 59(17):1929–1935

123

0.5 lm and a length of approximately 3 to 10 lm [12]. The

average number of magnetosomes per cell was approxi-

mately 14. The magnetosomes of M. magneticum AMB-1

were roughly octahedral and 50 nm along each major axis

based on observations via transmission electron micros-

copy (TEM) (Fig. 2b). The bacterial concentration in the

experiment was 108 cells/mL.

A water-soluble ferrofluid was purchased from Anhui

Jinke Magnetic Liquids Co. Ltd. (China). The concentration

of the ferrofluid was 9 % (v:v), and its specific gravity was

1.34 g/cm3. The size of the magnetic particles (Fe3O4) in the

ferrofluid was measured by TEM. The diameters of the

particles ranged from 10 to 20 nm (Fig. 2c). The particles in

the ferrofluid were superparamagnetic. The saturation

magnetization of the ferrofluid was 40.34 kA/m, whereas

the saturation magnetization of the magnetic particle was

approximately 448 kA/m. The ferrofluid was diluted to 1010

particles/mL for the experiment.

2.2 Permanent magnet blocks

Permanent magnet blocks measuring U6 mm 9 5 mm,

with a magnetic flux density of approximately 0.37 T at the

surface, were used to detect magnetotactic behavior. The

magnet blocks were made of Nd-Fe-B with a magnetic

energy product of 240 kJ/m3.

2.3 Experimental procedures

A 30 mL QHL cell suspension at 108 cells/mL was placed

in two 25 mL test tubes of U15 mm 9 170 mm averagely.

One test tube was supplemented with 300 lL (1 lmol/L)

carbonyl cyanide m-chlorophenyl-hydrazone (CCCP,

Sigma Chemical Co., USA) to disable the rotation of the

flagella of bacteria. A 30 mL of AMB-1 culture was placed

in two 25 mL test tubes averagely. One test tube was also

supplemented with 300 lL CCCP. 15 mL ferrofluid at 1010

particles/mL was placed in a 25 mL test tube.

Two magnet blocks were placed on the opposite sides of

each test tube and oriented toward the same magnetization

direction (Fig. 3). The accumulation of cells on the walls of

the test tubes was observed.

2.4 Magnetic field analysis

The magnetic fields of the permanent magnet blocks were

analyzed by using finite element method software (Ansoft

Co., USA). The magnetic field is rotationally symmetrical

around the axes of the permanent magnet blocks. Thus, the

3-dimensional calculation of the magnetic fields can be

simplified into a 2-dimensional calculation.

3 Results

For the QHL culture, six bacterial spots were observed on

the wall of the test tube after two magnet blocks were

placed on the test tube for 30 min (Fig. 4). Two spots (S1

and N1) appeared underneath the magnet blocks, two spots

(N2 and S2) showed up above the magnet blocks, and two

spots (N3 and S3) were found below the magnet blocks.

Bacteria were more abundant at N1, N2, and N3 than at S1,

Fig. 2 TEM photos of the samples. (a) Morphotypes of QHL cells

and the arrangement of the magnetosome chains. (b) Magnetospiril-

lum magneticum strain AMB-1. (c) The ferrofluid used in the

experiment

Chin. Sci. Bull. (2014) 59(17):1929–1935 1931

123

S2 and S3. The distances between the upper/lower spots and

the rotational symmetrical axis of the magnet blocks (i.e.,

the X axis) were almost the same (18 mm). When two

magnet blocks were placed on each side of the test tube

(i.e., the thickness of the block on each side was increased

to 10 mm), six bacterial spots appeared on the wall of the

test tube, but the distance between the upper/lower spots

and the X axis was increased to 21 mm. For the QHL

culture with CCCP, no bacterial spot appeared on the wall

of the test tube after two magnet blocks were placed on the

test tube for 30 min.

For M. magneticum AMB-1, wild-type cells aggregated

as two small spots on the inner wall of the test tube

underneath the two magnet blocks after the magnet blocks

were placed for 5 min. As time went on, the bacterial

aggregation underneath the two magnet blocks increased

gradually, and only two spots appeared during the entire

period. The aggregated cells formed a round shape, and

most cells aggregated in the center of the round spot

(Fig. 5a). The CCCP-treated bacteria aggregated in two

small spots on the inner wall of the test tube underneath the

two magnet blocks after the magnet blocks were placed for

1 h. As time went on, the bacterial aggregation underneath

the two magnet blocks also increased gradually. The

aggregated cells formed a circular ring shape, and few cells

aggregated in the center of the circle (Fig. 5b).

For the ferrofluid, two spots appeared underneath the

two magnet blocks. Aggregation in the ferrofluid was

remarkable after the magnet blocks were placed for 3 min.

Aggregation increased significantly with time. The aggre-

gated particles formed a circular shape similar to that of the

CCCP-treated AMB-1 bacterial spot. The color of the

solution gradually became pale until the upper layer of the

solution turned completely colorless after 50 min. After

1 h, the color of the solution 10 mm away beneath the

Fig.4 Distribution of QHL bacterial spots, as indicated by N2, N3, S2,

and S3, in a test tube with an artificial magnetic field generated by two

magnets. (a) Photogram of the distribution of QHL bacterial spots.

(b) Schematic illustration of the result

Fig. 5 Distribution of Magnetospirillum magneticum AMB-1 bacte-

rial spots. (a) Aggregated round spot of wild-type cells. (b) Aggre-

gated circularity of CCCP-treated bacteria

Fig. 3 Diagrammatic sketch of the experimental setup. Two magnets

were placed on the opposite sides of the test tube. The frame of the

axes is also shown in the figure, wherein the X-axis is at the rotational

symmetrical axis of the cylindrical magnet blocks

1932 Chin. Sci. Bull. (2014) 59(17):1929–1935

123

magnet blocks started to become pale. With the elapsed

time, the lower part of the test tube become pale after 2 h

and 40 min when almost all the ferrofluid aggregated at the

location attracted by magnet blocks.

4 Discussion

To analyze the aggregation results, we calculated the

magnetic fields of the two permanent magnet blocks. Fig-

ure 6 shows the distribution of the magnetic fields.

Based on the distribution of the magnetic direction

shown in Fig. 6b, we found that the directions of the

magnetic field were perpendicular to the wall at points N2c,

N3c, S2c, and S3c besides the points N1c and S1c. So the

north-seeking bacteria would aggregate at N1c, N2c, and

N3c, while the south-seeking bacteria would aggregate at

S1c, S2c, and S3c. An implicit assumption is that the

swimming direction of MTB is parallel to or antiparallel to

applied magnetic field direction, which is the definition of

magnetotaxis. The distance of N2c (N3c, S2c, and S3c) to the

X axis was calculated as 17 mm when the thickness of the

magnet block on each side of the test tube was 5 mm.

When the thickness of the magnet block was increased to

10 mm, the distance of N2c (N3c, S2c, and S3c) to the X axis

became 21 mm. It is noticed that the aggregation is not

related with the shape and the magnetic moment of MTB

but with the distribution of magnetic fields and the mag-

netotaxis of MTB.

The experiment on QHL cells showed that six aggre-

gated spots were formed and that the distances of the four

aggregated spots (N2, N3, S2, and S3) to the X axis was 18

or 21 mm when the thickness of the magnet blocks was 5

or 10 mm, respectively. The locations of the six spots are

the same as those in aforementioned analysis, demon-

strating that the QHL cells include north-seeking and

south-seeking bacteria. The greater abundance of bacteria

at N2 and N3 than at S2 and S3 confirmed that MTB from

the Northern Hemisphere preferentially swim to the North

Pole. As expected, QHL cells exhibited polar magneto-

tactic behavior when they were examined in a water droplet

under an optic microscope [15]. The distance of N2 (S2) to

the X axis was the same as that of N3 (S3), implying that

magnetotactic behavior of QHL cells was not influenced by

oxygen concentration. Of course, a possible reason is that

there might be no significant difference of the oxygen

concentration at these positions in the test tubes.

The accumulation of MTB on the walls of a bottle or a

test tube under permanent magnet blocks indicated the

magnetotactic behavior. However, such accumulation

Fig. 6 (Color online) Distribution of the magnetic flux density of two permanent magnet blocks. (a) Magnetic lines of force. (b) Magnetic

direction of the magnetic fields (short arrows). (c) Magnetic flux density along the X axis. (d) Gradient of the magnetic field along the X axis

Chin. Sci. Bull. (2014) 59(17):1929–1935 1933

123

cannot completely exclude the possible magnetic attraction

of bacteria to these positions. For polar MTB, we observed

four additional bacterial spots above and below the mag-

nets. Nevertheless, CCCP-treated bacteria did not accu-

mulate on these additional positions. Therefore, the

localization of bacteria on these positions reflects their

magnetotactic behavior.

M. magneticum AMB-1 exhibits axial magneto-aero-

taxis. These bacteria are oriented by the ambient magnetic

field, and they swim parallel or antiparallel to the magnetic

field. Thus, only two spots were formed on the inner wall

of the test tube underneath the two magnet blocks.

Although two spots were observed for the CCCP-treated

AMB-1 cells, their differences were apparent (Fig. 5).

Another difference was aggregation time. The aggregation

of active bacteria caused by magnetotactic behavior was

faster than that of CCCP-treated bacteria with restrained

flagella. The aggregation of CCCP-treated bacteria resulted

from passive magnetic attraction.

The magnitude of the magnetic force on a magnetotactic

bacterium is Fmag ¼ nM dB

dsV , where Fmag is the magnitude

of the magnetic force, M is the magnetization of the

magnetosome (M is approximately 480 kA/m if the mag-

netosome is a single-domain Fe3O4), B is the magnetic flux

density of the applied magnetic field, dB

dsis a gradient of B,

V is the volume of one magnetosome, and n is the number

of magnetosomes. Therefore, the magnetic force linearly

correlated with the gradient of the magnetic field. For the

permanent magnet block used in our experiment, a high

magnetic gradient region was located near the fringe of the

round spot. This phenomenon explains why the aggregated

CCCP-treated bacteria formed a circular shape, which is

different from the shape formed by the aggregated active

bacteria. As the active axial MTB swam parallel to or

antiparallel to the magnetic field, the aggregated cells

formed a round shape, and most cells aggregated in the

center of the round spot.

The aggregation of ferrofluid also resulted from the

passive magnetic attraction, as showed by the CCCP-

treated AMB-1 cells. Thus the aggregated particles formed

a circular shape similar to the CCCP-treated bacterial spot.

The magnetic particles of ferrofluid form a large magnetic

bead under an applied magnetic field, thereby increasing

the magnetic force [18]. Consequently, the aggregation

speed of ferrofluid is faster than that of CCCP-treated

bacteria.

According to the experimental and theoretical analysis,

we could find a scheme to differentiate polar MTB, axial

MTB and magnetic impurity in environmental samples. The

scheme is based on observing the accumulation on the walls

of a test tube after applying two cylindrical permanent

magnet blocks. If there are aggregated spots above or below

the two magnet blocks besides the aggregated spots under-

neath the permanent magnet blocks, the aggregation should

be polar MTB. According to the relationship between mag-

netic direction of the magnet block and the location of

aggregated spots, the seeking direction of the polar MTB

could be confirmed. If there are only two aggregated round

spots, and the color in the center of the round is darker than

the color around it, the aggregation should be axial MTB. If

there are two circular aggregated spots, the aggregation

should be magnetic impurity or dead MTB.

5 Conclusions

A simple method, i.e., observing the accumulation of cells

on the walls of a test tube when two permanent magnet

blocks were placed on the tube, was used to explore the

magnetotaxis of MTB. For the polar MTB QHL, aggre-

gated spots were formed above or below the two magnet

blocks besides the aggregated spots underneath the per-

manent magnet blocks. For the axial MTB AMB-1, only

two aggregated round spots were formed, and the color in

the center of the round spot was darker than the color

around it. For the ferrofluid, two spots were also formed

underneath the two magnet blocks, and the aggregated

particles formed a circular shape similar to the CCCP-

treated AMB-1 bacterial spot. The findings were reason-

ably explained by the MTB features and magnetic field

theory. A scheme that can differentiate polar MTB from

axial MTB and magnetic impurity could be developed

based on these results, which could be a useful tool for

fieldworks involving MTB.

Acknowledgments This work was supported by the National Basic

Research Program of China (51037006) and the National Natural

Science Foundation of China (41276170). We thank Jing Lv and

Qiufeng Ma for assistance in experiments.

References

1. Komeili A (2012) Molecular mechanisms of compartmentaliza-

tion and biomineralization in magnetotactic bacteria. FEMS

Microbiol Rev 36:232–255

2. Lin W, Li B, Pan YX (2011) DMTB: a comprehensive online

resource of 16S rRNA genes, ecological metadata, oligonucleo-

tides, and magnetic properties of magnetotactic bacteria. Chin Sci

Bull 56:476–478

3. Blackmore RP (1975) Magnetotactic bacteria. Science 190:

377–379

4. Bazylinski DA, Frankel RB (2004) Magnetosome formation in

prokaryotes. Nat Rev Microbiol 2:217–230

5. Frankel RB, Blakemore RP (1980) Navigational compass in

magnetic bacteria. J Magn Magn Mater 15:1562–1564

6. Frankel RB, Bazylinski DA, Johnson MS et al (1997) Magneto-

aerotaxis in marine coccoid bacteria. Biophys J 73:994–1000

1934 Chin. Sci. Bull. (2014) 59(17):1929–1935

123

7. Greenberg M, Canter K, Mahler I et al (2005) Observation of

magnetoreceptive behavior in a multicellular magnetotactic pro-

karyote in higher than geomagnetic fields. Biophys J 88:1496–1499

8. Philippe N, Wu LF (2010) An MCP-like protein interacts with the

MamK cytoskeleton and is involved in magnetotaxis in Mag-

netospirillum magneticum AMB-1. J Mol Biol 400:309–322

9. Blakemore RP, Frankel RB, Kalmijn AJ (1980) South-seeking

magnetotactic bacteria in the Southern Hemisphere. Nature

286:384–385

10. Simmons SL, Bazylinski DA, Edwards KJ (2006) South-seeking

magnetotactic bacteria in the Northern Hemisphere. Science

311:371–374

11. Zhang WJ, Chen CF, Li Y et al (2010) Configuration of redox

gradient determines magnetotactic polarity of the marine bacteria

MO-1. Environ Microbiol Rep 2:646–650

12. Seong S, Park TH (2001) Swimming characteristics of magnetic

bacterium, Magnetospirillum sp. AMB-1, and implications as

toxicity measurement. Biotechnol Bioeng 76:11–16

13. Steinberger B, Petermann H, Weiss DG (1994) Movement of

magnetic bacteria in time-varing magnetic fields. J Fluid Mech

273:189–211

14. Li N, Luo CX, Zhu XJ et al (2011) Microfluidic generation and

dynamically switching of oxygen gradients applied to the

observation of cell aerotactic behaviour. Microelectron Eng

88:1698–1701

15. Pan HM, Zhu KL, Song T et al (2008) Characterization of a

homogeneous taxonomic group of marine magnetotactic cocci

within a low tide zone in the China Sea. Environ Microbiol

10:1158–1164

16. Wolfe R, Thauer R, Pfennig N (1987) A ‘capillary racetrack’

method for isolation of magnetotactic bacteria. FEMS Microbiol

Lett 45:31–35

17. Yang C, Takeyama H, Tanaka T (2001) Effects of growth med-

ium composition, iron sources and atmospheric oxygen concen-

trations on production of luciferase-bacterial magnetic particle

complex by a recombinant Magnetospirillum magneticum AMB-

1. Enzyme Microbiol Technol 29:13–19

18. Xu H, Song T, Bao XQ et al (2005) Site-directed research of

magnetic nanoparticles in magnetic drug targeting. J Magn Magn

Mater 293:514–519

Chin. Sci. Bull. (2014) 59(17):1929–1935 1935

123