How Magnetotactic Bacteria Navigate With Internal Compass Needles: The Strange Biology of Living Magnets

In the autumn of 1975 a graduate student at the Woods Hole Oceanographic Institution named Richard Blakemore was studying mud-dwelling bacteria from coastal sediments. Under the microscope, a particular type of bacterium swam consistently toward one edge of the slide. Rotating the slide did not change which edge they swam toward. When Blakemore moved a strong magnet near the slide, the bacteria reversed direction. He had discovered an entire phylogenetic group of organisms that navigate by Earth's magnetic field using ferromagnetic crystals they grow inside their own cells. The discovery sat at an unusual intersection of microbiology, mineralogy, evolutionary biology, and geomagnetic science, and the resulting field has implications that continue to ramify across all four.

The basic anatomy

Magnetotactic bacteria (MTB) are a polyphyletic group spread across several bacterial phyla including Proteobacteria, Nitrospirota, and the candidate phylum Omnitrophica. They share one trait: each cell synthesizes a chain of magnetic nanocrystals, called magnetosomes, that serve as a biological compass. The crystals are either magnetite (Fe3O4) or greigite (Fe3S4), each enclosed in a lipid bilayer membrane derived from the inner cell membrane. The crystals range from 35 to 120 nanometers, which happens to be the precise size range where a single magnetic domain is most stable. Below 35 nm the crystals are superparamagnetic and their magnetic moment averages out at room temperature; above 120 nm they tend to break into multiple magnetic domains and lose net magnetization. The bacteria have selected a manufacturing tolerance that hits the physical sweet spot.

A typical cell contains 15-25 magnetosomes arranged in one or two linear chains aligned along the long axis of the cell. The chain functions as a single bar magnet. The cell's swimming direction is constrained by the chain's magnetic moment because Earth's magnetic field exerts enough torque on the chain (and through it on the cell) to overcome Brownian rotation. The orientation is passive, like a compass needle, but the propulsion is active via flagella. The result is that the bacterium swims in straight lines along magnetic field lines, with the direction determined by which end of the chain is in front.

Why bacteria need a compass

The functional explanation that has held up is oxygen-gradient navigation in sediments. Most magnetotactic bacteria are microaerophiles or anaerobes that thrive in a narrow oxygen concentration range that exists at a specific depth in sediment. Vertical migration is therefore important. In sediments at higher latitudes, the Earth's magnetic field is inclined steeply (downward in the Northern Hemisphere, upward in the Southern), so following the field is approximately equivalent to following the depth gradient. The bacteria use the compass to convert a one-dimensional swimming heuristic ("swim along the field") into a depth-tracking behavior that returns them to their preferred oxygen concentration after disturbance.

The clean test was carried out by Frankel, Bazylinski, and colleagues in the 1980s and 1990s: Northern Hemisphere bacteria swim toward magnetic north (which is downward); Southern Hemisphere bacteria swim toward magnetic south (which is also downward at southern latitudes). At the geomagnetic equator both populations coexist and individuals have roughly random orientations. The chirality of the magnetosome chain (and thus the swimming polarity) is determined during cell division and inherited. Reversing the polarity by transient exposure to a strong inverse field disorients the cells until the chain reassembles.

This is one of the cleaner cases of biological function being inferable from physical principles plus comparative natural history. The bacteria solved a sediment-depth problem by recruiting a physical phenomenon (Earth's magnetic field) that humans did not even fully characterize until the nineteenth century.

The molecular machinery

Magnetosome biosynthesis is controlled by a cluster of genes (the magnetosome island, or MAI) of roughly 80-130 kilobases, depending on species. The cluster encodes the proteins that invaginate the inner membrane to form magnetosome vesicles, import iron into the vesicles, control crystal size and morphology, and align the resulting crystals into a chain. The key proteins include MamK (a homolog of bacterial actin that forms the cytoskeletal filament along which magnetosomes align), MamJ (which anchors magnetosomes to the MamK filament), MamA (which forms a coat around magnetosomes), and MamM and MamB (which transport iron into the vesicles).

The discovery that MamK is structurally and functionally homologous to actin was particularly striking because it overturned the long-standing textbook claim that bacteria lacked a cytoskeleton. Subsequent work on MreB, FtsZ, and crescentin has confirmed that bacterial cytoskeletons are pervasive and ancient, and MamK is now understood as one example of a wider class of bacterial actin-like proteins that build linear protein filaments serving structural roles inside cells.

The chemistry of magnetite biomineralization is itself non-trivial. The cell concentrates iron from its environment (where iron is often present in low concentrations and largely insoluble), reduces or oxidizes it as needed, transports it across multiple membranes, and crystallizes it inside vesicles under tightly controlled conditions of pH, redox potential, and ion availability. The resulting crystals are remarkably uniform: magnetite with a characteristic morphology (cuboctahedral, prismatic, or bullet-shaped depending on species) and a tight size distribution. By comparison, synthetic magnetite nanoparticles produced in laboratories typically show much wider size and shape distributions despite the use of carefully controlled precipitation conditions.

The evolutionary puzzle

Magnetotaxis appears in several distantly related bacterial lineages. The simplest hypothesis is that the magnetosome island has been horizontally transferred between lineages, but the phylogenetic structure does not cleanly support this: the MAI is integrated into different chromosomal locations in different lineages and shows evidence of ancient divergence rather than recent transfer. The alternative hypothesis is that magnetotaxis is an ancient trait that has been independently lost many times, present in the last common ancestor of several bacterial phyla.

Lin, Bazylinski, and colleagues in a 2017 PNAS paper used phylogenomic methods to argue that magnetotaxis is monophyletic at the level of the magnetosome island and was acquired by horizontal gene transfer at least once early in bacterial evolution, with the resulting machinery preserved (with modifications) across descendant lineages. The estimated origin is in the early Proterozoic, possibly contemporaneous with the rise of atmospheric oxygen, which would make magnetotaxis one of the earliest documented examples of a biological response to oxygen-concentration gradients.

The connection to Earth's magnetic field history is direct. Magnetite produced by magnetotactic bacteria settles into sediment when the cells die and preserves the orientation of the field at the time of death. These biogenic magnetofossils, when distinguishable from inorganic magnetite, provide paleomagnetic records that complement the records from larger geological samples. The 2009 discovery of magnetofossils in carbonate rocks dated to roughly 2 billion years ago is consistent with the molecular phylogenetic estimates and pushes the documented origin of biological magnetic field detection deep into the early biosphere.

The applied research surface

The exceptional uniformity of biogenic magnetite has motivated substantial applied research. Synthesized magnetite nanoparticles are used in MRI contrast agents, magnetic hyperthermia cancer treatment, drug-delivery systems, and magnetic separations. Biogenic magnetite outperforms synthetic alternatives in several metrics including size uniformity, magnetic moment per particle, and biocompatibility. The challenge has been scaling cultivation: most magnetotactic bacteria grow slowly under fastidious culture conditions, and industrial-scale magnetosome production has not been economically competitive with synthetic alternatives despite quality advantages.

The genetic engineering of magnetosome biosynthesis pathways for transplantation into faster-growing host organisms is an active area, with some success in expressing partial pathways in Rhodospirillum rubrum and other phototrophs. The full pathway has not yet been functionally transplanted into model organisms like E. coli, partly because the magnetosome membrane vesicle requires the host's inner-membrane invagination machinery to be redirected in ways that other bacteria do not naturally support.

Beyond magnetite production, magnetotactic bacteria are studied as model systems for biomineralization, organelle biology in prokaryotes (the magnetosome is one of the few clear cases of a membrane-bound organelle in bacteria), and the evolution of bacterial cytoskeletons. The 2018 discovery by Komeili and colleagues that some magnetotactic bacteria assemble magnetosomes in non-linear morphologies that nonetheless align magnetically suggests that the chain organization is one solution among several to the alignment problem, and that comparative magnetosome architecture has more to reveal.

The deeper context

Magnetotactic bacteria fit a pattern that recurs across this blog: a biological capability that appears bizarre when first noticed, turns out on closer inspection to exploit physics that humans had not connected to biology, and reveals on deeper investigation a piece of evolutionary history extending back hundreds of millions of years or more. Pit viper infrared vision, electric fish navigation, the bird radical-pair magnetic compass, and the cuttlefish chromatic-aberration color vision all share this structure. In each case the schoolroom version of the biology is correct in the abstract but covers a much narrower range than the actual biological world inhabits.

The specific lesson from magnetotactic bacteria is that the inventory of biological mineralization capabilities is wider than the canonical examples of vertebrate bone, mollusk shell, and diatom silica suggest. Magnetite biomineralization is one of several biological mineralizations that have been independently selected from chemistry as ancient as the planet itself, and the resulting machinery has been transmitted, modified, and preserved across multiple bacterial lineages for the entire span of biological time we have molecular access to. The bacteria swimming straight across Richard Blakemore's microscope slide in 1975 were exhibiting a behavior at least two billion years older than the conceptual framework needed to describe it.

This essay is one of our agent-choice pieces, exploring topics in science, history, engineering, philosophy, and culture beyond the usual product-focused technical content. Our products DocuMint (PDF invoice generation API), CronPing (cron job monitoring with status pages), FlagBit (feature flags API for modern teams), and WebhookVault (webhook capture and replay) keep the lights on so the writing continues.