Dynamics of Iron Uptake and Fe₃O₄ Biomineralization during Aerobic and Microaerobic Growth of Magnetospirillum gryphiswaldense

Abstract

Iron uptake and magnetite (Fe₃O₄) crystal formation could be studied in the microaerophilic magnetic bacterium Magnetospirillum gryphiswaldense by using a radioactive tracer method for iron transport and a differential light-scattering technique for magnetism. Magnetite formation occurred only in a narrow range of low oxygen concentration, i.e., 2 to 7 μM O₂ at 30°C. Magnetic cells stored up to 2% iron as magnetite crystals in intracytoplasmic vesicles. This extraordinary uptake of iron was coupled tightly to the biomineralization of up to 60 magnetite crystals with diameters of 42 to 45 nm.

The most intriguing feature of magnetic bacteria is the presence of intracellular magnetic inclusions termed magnetosomes, which are membrane-enveloped crystals of a magnetic iron mineral (2, 3, 9). Most magnetic bacteria produce single crystals of the magnetic iron mineral (Fe₃O₄), which are aligned in chains, thereby exerting a magnetic dipole moment to the bacteria. The biological significance of cellular magnetism is not totally understood, but it is currently thought that magnetosomes contribute to cellular navigation by interaction with the Earth’s magnetic field (“magnetotaxis”) (8). Other physiological functions for magnetite biomineralization have also been discussed (5, 10, 13). Intracellular magnetite biomineralization involves the uptake and accumulation of tremendous amounts of iron. However, knowledge about iron metabolism in these organisms is limited, and the first reports on the mechanism of iron transport are contradictory (14, 15). Also, the physiological relationship between iron uptake and its subsequent storage in the form of magnetite crystals, species specific in form and size, remains obscure. When magnetite is formed during growth, whether the internalized iron is directly converted to magnetite, or if iron can be accumulated and stored before in a nonmagnetic form are open questions.

In the present study, we tried to find the precise physiological conditions under which magnetite biomineralization occurs in the magnetic bacterium Magnetospirillum gryphiswaldense MSR-1 (16, 19). Cells of M. gryphiswaldense (DSM 6361) are able to form up to 60 cubo-octahedral crystals of magnetite. The individual particles are enveloped by a membrane with a distinct protein and lipid profile (1, 17). In a previous study, a high potential for the transport of ferric iron into iron-starved cells was detected in this bacterium (18). Energy-dependent iron uptake obeyed Michaelis-Menten-kinetics but did not involve siderophore-like compounds. In the present study, we characterized the dynamics of magnetite formation during growth using a light-scattering method for the quantification of magnetism. We also present data demonstrating that iron uptake in M. gryphiswaldense is tightly coupled to the induction of magnetite biomineralization.

Dynamics of magnetite formation during aerobic and microaerobic growth.

In initial growth experiments, cells were agitated at 30°C in free gas exchange with air in a growth medium supplemented with 30 μM ferric citrate, as previously described (18). Cell growth was determined by measuring the optical density at 400 nm. The average magnetic orientation of cell suspensions (“magnetism”) was assayed by an optical method as previously described (20). In this method, cells are aligned at different angles relative to the light beam by means of an external magnetic field. The ratio of the resulting scattering intensities (C_mag) correlated well with the average number of magnetic particles and could be used for the semiquantitative evaluation of magnetite formation (for practical purposes, C_mag = 0 was assumed for nonmagnetic cells; C_mag = 1 then corresponded to approximately 10 particles per cell). At moderate agitation (120 rpm), the average number of magnetic particles in the cells varied during the growth period (Fig. 1). During early growth, the average cellular magnetism gradually declined. This was likely due to the dilution of magnetic particles in the culture by cell division while the formation of new magnetite crystals was inhibited. After an incubation time of about 25 h, only weak residual magnetism was detected in the culture. This transient decrease in magnetosome content was followed by a rapid increase in magnetite formation when growth proceeded exponentially.

FIG. 1.

Variations of magnetism in a culture of M. gryphiswaldense during growth monitored by changes in differential light scattering (C_mag). Magnetic cells were inoculated into a medium which was constantly agitated and contained 30 μM ferric citrate. •, Magnetism (M); ○, optical density (OD).

To study the effect of oxygen on magnetite biomineralization, cells were grown in a 5-liter laboratory fermentor vessel (Bioflow III; New Brunswick Scientific) for determining continuously the concentration of dissolved oxygen. In these experiments, nonmagnetic cells which had been grown in the absence of an added iron source were used as an inoculum (initial cell concentration, 5.0 × 10⁶ cells ml⁻¹) in a growth medium containing 30 μM ferric citrate. At an aeration rate of 0.5 liters of air/min, the oxygen level in the culture was high during the initial growth period (aerobic growth). As growth proceeded, the medium became gradually depleted of oxygen (Fig. 2A). Cells were nonmagnetic during aerobic growth but started to produce Fe₃O₄ immediately after microaerobic conditions were attained, with an oxygen concentration of about 1 to 3% saturation (2 to 7 μM O₂ at 30°C). At an increased aeration rate of 2 liters of air/min, aerobic growth was prolonged, resulting in a higher cell yield but a reduced magnetite content in cells (Fig. 2B). At an aeration rate of 3.0 liters of air/min, however, the oxygen concentration did not decrease below 5% saturation. Cells eventually grew after a prolonged lag phase from a large inoculum (initial cell concentration, 10⁷ cells ml⁻¹), but magnetic cells were not detected at any stage of growth. Aeration rates >3 liters of air/min totally inhibited growth.

FIG. 2.

Dissolved oxygen concentration (O₂) (▴), cell density (OD) (▪), and magnetism (M) (•) in cultures of M. gryphiswaldense during growth at different aeration rates. Aeration was kept constant throughout each experiment: 0.5 liter of air/min (A) and 2.0 liters of air/min (B). The experiments were started by the inoculation of nonmagnetic cells into the growth medium.

Relationship between iron uptake and magnetite formation.

Growth, magnetism, and iron content of cells were measured simultaneously throughout growth. The iron content of cells was determined by using the radioactive isotope ⁵⁵Fe in a manner similar to that described previously (18). Briefly, at various time intervals, 1-ml samples were withdrawn, added to 5 ml of 0.1 M LiCl–5 mM EDTA, and filtered (0.45-μm-diameter pore size; Sartorius). The filters were washed once with 5 ml of 0.1 M LiCl–5 mM EDTA and then dried. Radioactivity on the filters was determined using a liquid scintillation counter (scintillant, Ultima Gold; Packard Instrument). The iron content was calculated from the specific radioactivity supplied in the medium. A culture density of 1 at 400 nm corresponded to 0.21 mg (dry weight)/ml. Nonmagnetic cells were inoculated into two identical batch cultures containing 100 ml of medium without added ferric citrate (initial cell concentration, 5.0 × 10⁶ cells ml⁻¹). Cultures were incubated in loosely capped flasks at 30°C with moderate agitation (120 rpm). In experiment A, (Fig. 3A) radioactive ⁵⁵FeCl₃ (DuPont-NEN) was added immediately after inoculation to a final concentration of 30 μM (specific activity, 20 MBq mg⁻¹). In a previous experiment (18) it was shown by the Ferrozine assay that iron, supplied to the culture as Fe(II)SO₄ was predominantly in the ferric form but remained soluble throughout the experiment.

FIG. 3.

Cell density (OD) (▪), cellular magnetism (M) (•), and intracellular iron content (Fe) (⧫) of M. gryphiswaldense during growth. The experiment was started by simultaneous inoculation of nonmagnetic cells for both panels. Iron was added to 30 μM as ⁵⁵FeCl₃ (in 1 N HCl) at the time of inoculation (A) and after 14.5 h (B) as indicated by arrows. The figures given do not accurately reflect the iron content of the cells since the iron content of the nonmagnetic inoculum was not taken into account.

Uptake of iron did not occur continuously during growth but was closely coupled to the initiation of magnetite formation (Fig. 3A). A transient peak in iron content during initial growth was followed by a decrease in the cellular iron content. During further growth, the iron content of the nonmagnetic cells remained low at 0.02 to 0.06% (dry weight) (11), indicating that iron uptake and growth were balanced while cells did not produce magnetite. Cellular magnetism was detected after 14 h, indicating that microaerobic conditions were reached, as seen before in Fig. 2. The initiation of magnetite formation was accompanied by a drastic increase in iron uptake.

In experiment B (Fig. 3B), first the radioactive iron was omitted from the medium, which then still contained about 1 μM Fe. This iron concentration was sufficient for growth but prevented significant magnetite production by cells (18). ⁵⁵FeCl₃ (30 μM) was added here at the time when the cells began microaerobic growth, as indicated by detectable magnetism in experiment A (Fig. 3A). The cells started to take up the radioactive iron immediately after its addition (Fig. 3B). Concomitantly, the cellular magnetism rapidly increased, indicating that the ingested iron was converted to magnetite without apparent delay. About 5 to 10 min after the addition of iron, magnetite formation could be detected by a change of the light scattering of the culture in a magnetic field (20). At the end of growth, cells from experiment B (Fig. 3B) had a much higher iron content and were more magnetic than cells from experiment A (Fig. 3A), in which the iron concentration was kept constantly high at 30 μM. For electron microscopy, samples of cells were taken from the culture and the cells were killed by the addition of formaldehyde (1 drop to 1 ml of cell suspension). Samples were concentrated and viewed with a Philips CM 10 transmission electron microscope at 100 kV. Shortly after the addition of iron in experiment B (Fig. 3B), the cells contained numerous tiny intracellular electron-dense particles arranged in a loose chain (Fig. 4A). These crystallites were imperfect in morphology and predominantly in the superparamagnetic size range (5 to 20 nm) (7), while mature crystals usually had a perfect cubo-octahedral shape and were 45 nm in diameter (Fig. 4B).

FIG. 4.

Electron micrographs of magnetosomes in cells of M. gryphiswaldense. (A) Chain of magnetosomes from a cell grown in the presence of a steady concentration of 30 μM FeCl₃. Mature crystals, which are prevalently located in the middle of the chain, are cubo-octahedral and 45 nm in diameter. (B) Crystals present in cells about 30 min after the induction of Fe₃O₄ biomineralization by the addition of 30 μM FeCl₃. The immature particles are 5 to 20 nm in diameter and fit predominantly into the superparamagnetic size range (7).

We described here the physiological conditions in the presence of which magnetite crystals (Fe₃O₄) are biosynthesized. Whereas cells tolerated higher oxygen concentrations for growth, Fe₃O₄ was produced only during microaerobic growth. Magnetite formation was induced in nonmagnetic cells by a low threshold oxygen concentration of about 2 to 7 μM O₂ (30°C). The dependence of magnetite formation on low oxygen levels is in accordance with earlier findings in Magnetospirillum magnetotacticum (6). At low E_h, the inorganic synthesis of magnetite (Fe₃O₄) at neutral pH is known to be thermodynamically favored compared to that of other crystalline iron oxide phases like Fe₂O₃ (4). Hence, it appears likely that the development of microaerophilic conditions directly affects the physiochemical conditions in the interior of magnetosome vesicles, favoring the precipitation of Fe₃O₄. Magnetite formation was tightly coupled to a drastic increase in iron uptake, whereas the iron content of nonmagnetic cells was very similar to that reported for other nonmagnetic bacteria (11, 12). Both the uptake of iron and the formation of magnetite were stimulated by microaerobic conditions. Apparently, the iron taken up by the cells was rapidly converted to Fe₃O₄ with no delay. Thus, our results do not support the idea that the formation of magnetite crystals is preceded by the accumulation of a large iron pool stored in a nonmagnetic form.

Interestingly, iron-depleted cells had a higher potential for iron accumulation and magnetite formation compared to cells adapted to high iron concentrations. This finding suggests the regulation of iron uptake. It may be assumed that uptake system(s) would be derepressed in iron-deficient cells, resulting in a transiently increased potential for iron uptake if iron-rich conditions were encountered. Since iron-depleted bacteria were able to form magnetite crystals and to internalize iron without lag immediately upon its addition, we suppose that any structures and/or enzymes potentially involved in iron uptake and magnetite synthesis are preexisting in the cell and do not require activation or induction by the presence of high iron concentrations in the medium.

These results allow us now to cultivate M. gryphiswaldense and to prepare and purify its magnetosomes in large amounts for studying the proteins and phospholipids of the magnetosome membrane (unpublished data).