Magnetogenetics has just emerged as a way to noninvasively activate neurons. The technology involves fusing or targeting the iron-binding protein ferritin to TRP channels in the neuronal membrane. A magnetic field is then thought either to exert a magnetic force on ferritin complexes or to heat the iron-containing complexes, both of which would open up the associated TRP channels and result in neuronal activation.

After reading a report on a molecular magnetic compass and later the reports on magnetogenetics, Markus Meister from the California Institute of Technology in Pasadena questioned the findings “because of the incredibly minuscule amount of iron involved and the relatively tiny magnetic fields they were using,” he says. His intuition arose from his early work on bacterial taxis, where he also encountered magnetotactic bacteria. “That's why I had kind of an intuition for how big a compass needle needs to be in order to orient itself in the earth field,” explains Meister.

According to him, the physical limitations of magnetogenetics stem from the paramagnetic nature of ferritin complexes as well as the small size of their iron-containing core. The smallest known iron particles that are permanently magnetic (i.e., ferromagnetic) contain about a million iron atoms within a diameter of 30 nanometers. In contrast, the iron in ferritin complexes is paramagnetic, because the iron core is too small to be ferromagnetic. Paramagnetic materials are only weakly magnetic and require the application of an external magnetic field to exhibit magnetism.

On the basis of physics principles, Meister assessed the plausibility of different scenarios that could potentially explain the mechanisms underlying magnetogenetics. He put forward several hypotheses for how magnetic force could gate an ion channel. For one, a magnetic field applied to paramagnetic particles such as ferritin complexes could be pulled by a force that is proportional to the magnetic field gradient. Alternatively, multiple ferritin complexes targeted to the same ion channel could tug in different directions. Another mechanism could be a rotational force on the channel that arises from a possible asymmetric magnetizability of ferritin. Finally, the effect on the ion channel could be more indirect through stress on the membrane that could be generated cumulatively by many ferritins. According to Meister's calculations, the forces that could be achieved by these mechanisms are five to ten orders of magnitude too small to open ion channels.

Similarly, the proposed alternative mechanism of magnetically induced heat, which acts on temperature-sensitive TRP channels, is physically implausible according to Meister. Ferromagnetic nanoparticles have been shown to heat up in a size-dependent manner. However, these nanoparticles are more susceptible to heating than the paramagnetic ferritin complexes are. In addition, the size of the iron core in ferritin complexes is smaller than the particle size that is considered useful for heating applications in the nanoparticle field. Even under the most benevolent assumptions, Meister found that the achievable temperature change would be in the range of nano-Kelvins, which is not large enough to open temperature-sensitive TRP channels.

While Meister questions the current magnetogenetics approaches, further modifications or alternative approaches may emerge. But some basic experimental standards should be applied in such studies. Meister finds that the magnetosensation literature is full of reports of single, isolated responses without indication of repeated stimulation. “The literature is just filled with accidental coincidences like that,” Meister cautions. Considering the stakes for people trying to repeat experiments, Meister thinks that “there is a moral obligation to go beyond chasing the spectacular story and spending more time on checking things out.”