Electromagnetic brain stimulation is an increasingly used therapeutic approach in neurology and psychiatry, aimed at harnessing the brain's intrinsic repair mechanisms. Its clinical outcomes nevertheless remain variable from one patient to another and poorly reproducible, largely because the underlying mechanisms are still poorly understood. The development of therapeutic protocols therefore relies more on trial and error than on a rational adjustment targeting the pathology. This study set out to elucidate the action of weak magnetic fields by distinguishing two key questions: are the dose and the rhythm of the pulses decisive, and does the magnetic field act directly on neurons, independently of any electrical discharge?
To address this, the authors used low-intensity repetitive transcranial magnetic stimulation (LI-rTMS, in the millitesla range) delivered focally, in an experimental model allowing a biological effect to be readily quantified: the repair of the mouse olivocerebellar pathway after injury. Following unilateral section of the cerebellar peduncle, which deprives one cerebellar hemisphere of its climbing fibers, stimulation was applied for ten minutes per day over two weeks, both in vivo and ex vivo on organotypic cultures. Various stimulation patterns drawn from human clinical practice were compared — single pulses at 1 and 10 Hz, continuous, intermittent, or random theta-burst patterns, and a complex high-frequency biomimetic pattern (BHFS). The role of the magnetic field itself was tested in the presence or absence of cryptochrome, a putative cellular magnetoreceptor, using knockout mice. Reinnervation was assessed by immunohistochemistry, cellular activation by c-fos labeling, and changes in gene expression by quantitative PCR.
This work shows that LI-rTMS induces the axonal regrowth and synaptogenesis required for cerebellar reinnervation. This reparative effect depends closely on the stimulation pattern: complex biomimetic rhythms proved more effective than the regular rhythms typically used in human rTMS. Repair further required the presence of cryptochrome. Only the patterns promoting repair increased the expression of genes involved in neuronal regeneration, nearly 40% of which are cryptochrome targets.
These findings indicate that weak magnetic fields can modify the brain without directly triggering neuronal activity. Rather than an activation of neurons by induced electrical currents, the authors propose that these fields act through cryptochrome to activate cellular signaling cascades. This new understanding of the mechanisms underlying structural neuroplasticity opens avenues for optimizing electromagnetic stimulation and designing treatments tailored to different neurological diseases.