How Magnetotactic Bacteria Navigate by Earth's Field: The Strange Biomineralization of Magnetospirillum

In 1975, Richard Blakemore, then a graduate student in microbiology at the University of Massachusetts, was looking at sediment samples from a salt marsh under a microscope. He noticed that some of the bacteria in the samples were behaving oddly. They were not swimming randomly. They were all swimming in the same direction. When he rotated the microscope stage, they corrected their swimming direction to match. When he placed a bar magnet near the slide, the bacteria reoriented to align with the new field.

He had discovered a group of organisms that nobody had seriously expected to exist: bacteria that contain magnetic compasses built from crystals they grow themselves.

The basic biology

Magnetotactic bacteria are a polyphyletic group, meaning the trait appears in multiple bacterial lineages that are not closely related to each other. The genus Magnetospirillum is the best-studied, with species Magnetospirillum magneticum and Magnetospirillum gryphiswaldense being the model organisms for laboratory work. Other magnetotactic groups include various Proteobacteria, Nitrospirae, and some uncultured environmental lineages identified by DNA sequencing.

The defining feature is the magnetosome: a membrane-bound vesicle inside the cell that contains a single magnetic crystal. The crystals are typically magnetite (Fe3O4) or greigite (Fe3S4), depending on the species and the chemical environment. The crystals are tens of nanometers across, with a tight size distribution and a consistent shape that is species-specific.

The magnetosomes are arranged in chains that run along the long axis of the bacterial cell, held in position by a cytoskeletal filament made of a protein called MamK. The total chain is long enough and the magnetic moments are aligned enough that the chain acts as a single magnetic dipole large enough to overcome Brownian motion at body temperature. The cell rotates passively to align with whatever external magnetic field it experiences.

Why bacteria evolved magnetic navigation

The standard interpretation is that magnetotaxis is an aid to chemotaxis in environments where the geometry of useful chemistry is vertical. Magnetotactic bacteria are typically microaerophilic: they prefer specific oxygen concentrations that are lower than atmospheric but higher than zero. In a sediment-water column, the optimal oxygen concentration is found at a specific depth, with too much oxygen above and too little below.

Earth's magnetic field is inclined relative to the surface in most places: in the Northern Hemisphere it points northward and downward, in the Southern Hemisphere northward and upward, with the angle of inclination varying by latitude. A bacterium that swims in the direction of the magnetic field is moving in a roughly known direction relative to vertical. Combined with chemotactic sensing of the local oxygen concentration, the magnetic compass turns a slow random search for the right depth into a fast directed search.

The hypothesis predicts that magnetotactic bacteria in the Northern Hemisphere should be predominantly north-seeking and those in the Southern Hemisphere should be predominantly south-seeking. This was confirmed by Blakemore's followup work in Australia and New Zealand in the late 1970s. The transition zone near the magnetic equator contains roughly equal populations of north-seeking and south-seeking variants, with neither having a clear advantage when the field is parallel to the surface.

The biomineralization pathway

The magnetosome synthesis pathway is one of the most studied biomineralization systems in biology. The process begins with the formation of an invagination from the inner cell membrane, which is then pinched off to form a closed vesicle. Iron is transported into the vesicle by specialized membrane proteins. Inside the vesicle, the iron is processed through several oxidation and reduction steps that ultimately produce magnetite crystals through a controlled biomineralization process.

The proteins involved are encoded in a cluster of genes called the magnetosome island, which can be hundreds of kilobases long and contains dozens of genes. The Mam proteins (Mam stands for magnetosome membrane) handle the iron transport, membrane organization, crystal nucleation, and crystal shape control. Different Mam proteins are responsible for different aspects of the synthesis: MamA appears to organize the magnetosome membrane, MamK forms the cytoskeletal filament that aligns the chain, MamJ links the magnetosomes to the filament, and MamN handles iron transport. Many other genes contribute to the overall process.

The crystal shape is genetically controlled with remarkable precision. Different species produce crystals of different shapes (cuboctahedral, hexagonal prismatic, elongated bullet-shaped) and different sizes (typically in the 35 to 120 nanometer range). The shape and size are species-specific and stable across many generations of laboratory culture, which means the genetic program that controls crystal morphology operates at the nanometer scale on the bulk synthesis chemistry.

The biomimetic engineering interest

Magnetite nanoparticles have many industrial applications: magnetic data storage, contrast agents for medical imaging, drug delivery vehicles, environmental remediation, and a growing list of biotechnology uses. Synthetic magnetite nanoparticles are routinely produced at industrial scale, but the synthetic particles have substantially worse properties than the biological ones for several reasons.

Synthetic magnetite tends to have polydisperse size distribution, irregular shape, and surface defects. Biological magnetite from magnetosomes has narrow size distribution, consistent shape, and clean surfaces. For applications where particle uniformity matters, biological magnetite outperforms synthetic by a significant margin. Several biotechnology companies have attempted to produce magnetosomes at commercial scale by growing Magnetospirillum in large fermenters and harvesting the magnetosomes for industrial use.

The commercial scaling has been more difficult than expected. The bacteria grow slowly, the magnetosome yield per cell is moderate, and the purification process is expensive. As of 2026, magnetosomes are produced commercially in small quantities for specific high-value applications, but they have not displaced synthetic magnetite for the bulk market.

The synthetic-biology approach of transferring the magnetosome synthesis genes to a faster-growing host organism (such as Escherichia coli or yeast) has been attempted by multiple research groups. Partial success has been achieved: recombinant hosts can produce magnetite-like material, but the crystals are typically smaller, less uniform, and less magnetic than the native bacterial product. The mechanism by which Magnetospirillum achieves such tight control over crystal properties is still not fully understood at the level required to replicate it in a different cellular context.

Three observations

The first observation is that biological nanotechnology consistently outperforms human nanotechnology when the relevant scale is small enough to be entirely below human engineering tools. Cells routinely build structures with nanometer precision through self-assembly processes that have been refined over evolutionary timescales. Human nanotechnology operates at the same scale but uses top-down fabrication tools that are inherently less precise. The gap is largest for chemistry-driven self-assembly problems and smallest for problems that can be solved by lithography.

The second observation is that organisms with unusual capabilities are often phylogenetically scattered. Magnetotaxis appears in multiple bacterial lineages that are not closely related, which means the trait has either evolved independently several times or has been transferred horizontally between lineages. The current evidence suggests both have happened: the magnetosome island has been transferred horizontally between species, and convergent evolution has produced similar systems independently in other lineages.

The third observation is that scientific discovery in microbiology often depends on the specific moment when an unusual organism happens to be visible to a researcher who is trained to notice. Blakemore's discovery in 1975 required exactly the right combination of sample, microscope, and curious observer. The bacteria were presumably present in similar sediment samples that had been examined by many earlier researchers; none of them happened to notice the directional swimming because they were not looking for it.

The deeper observation is that the inventory of strange biological capabilities is much larger than the inventory of capabilities currently characterized. Magnetotactic bacteria are one example of an organism whose existence requires capabilities biology textbooks did not anticipate before the discovery. Cuttlefish color vision, octopus chemotactile receptors, bowhead whale longevity, and electric eel voltage generation are other examples. The pattern suggests that future biology will continue to find capabilities that current biology does not anticipate, and that the bottleneck is not whether such capabilities exist but whether someone happens to be looking in the right place at the right time.

Read more essays and technical writing at anethoth.com — a notebook on databases, distributed systems, biology, and the engineering that holds the world together.