Introduction
Magnetothermal neurostimulation is emerging as a powerful technique for gaining insight into intricate brain circuits and performing therapeutic investigations of neurological disorders given its distinct advantages, such as minimal invasiveness, tether-free operation, deep brain accessibility, and high spatial resolution[1-2]. In this technique, biocompatible magnetic nanoparticles are utilized as nanoheaters to generate heat locally through the hysteresis loss process under alternating magnetic fields (AMFs)[3]. The increase in local temperature can activate the thermosensitive ion channels that are overexpressed on the cell membrane and trigger an influx of calcium (Ca2+) into neurons, thereby modulating neural activity[4]. To date, magnetothermal neurostimulation has been demonstrated to bidirectionally modulate neuronal activity[5-8], successfully regulate blood glucose[9-10] and adrenal hormone levels in vivo[11], and control the motor behaviors of worms[8, 12], flies[13] and freely moving mice[14]. However, current techniques still suffer from low activation efficiency and high required dosages of nanoheaters, leading to an unexpectedly long stimulus-response time and potential safety concerns[13].
According to the principles of magnetothermal neurostimulation, the precisely timed activation of a single neuron depends on both the expression level of thermosensitive ion channels and the heating efficiency of the nanoparticles under an AMF[11, 15-16]. Transient receptor potential cation channel subfamily V member 1 (TRPV1), the thermosensitive ion channel most commonly used in neuromodulation, can be activated by local heat (> 43 °C)-induced conformational changes in the channel pore[17-20]. Moreover, the rate of TRPV1 opening increases exponentially with increasing surrounding temperature[19, 21]. Because an increase in TRPV1 expression in vivo requires viral vector-mediated gene transfection, which comes with high safety risks, the key to quickly and safely activating TRPV1 lies in boosting the heat-generating performance of magnetic nanoparticles[3-4, 11]. Unfortunately, the superparamagnetic iron oxides (SPIOs) and ferritin protein widely used in magnetothermal neurostimulation show poor thermal conversion efficiency due to their superparamagnetic nature and minimal hysteresis loss[22-25]. In addition, an AMF with a high amplitude (H) and frequency (f ) is frequently needed to improve the heat-generating ability of the nanoheater[26-28]. Nevertheless, less improvement has been achieved because of the biologically acceptable limit (H ×f ≤ 5 × 109A/(m·s))[28-30]. Recognizing the above limitations, developing a high-performance magnetic nanoparticle-mediated stimulation technique is a valuable approach for realizing efficient and safe magnetothermal neuromodulation[31-33].
Significant efforts have been made to synthesize various magnetic nanoparticles with enhanced heat-generating performance given their wide application in magnetic hyperthermia, controlled drug delivery, cell signal transduction activation and magnetothermal neurostimulation[30, 34-39]. The thermal conversion efficiency of nanoparticles is commonly evaluated using the specific absorption rate (SAR), which mainly depends on the intrinsic magnetic properties of the nanoparticles[40-42]. Despite tuning their composition, shape, size and surface chemistry of various magnetic nanoparticles, most exhibit relatively low SAR values ranging from 250 W/g to 1000 W/g[30, 43-45]. As a result, these nanoparticles take more time to reach the threshold temperature and activate TRPV1, leading to long latency times (approximately 14.7 s) for behavior onset[13, 18, 46]. Applying a high dose of magnetic nanoparticles is an alternative method to reduce the heating time[26]. Nevertheless, using excessive amounts of nanoheaters increases the risk of nonnegligible safety issues in the brain, such as dopamine neuron damage, homeostatic disruption and inflammatory responses[47]. We previously reported that novel biocompatible ferromagnetic vortex-domain iron oxide nanorings (FVIOs) exhibit an ultrahigh SAR of greater than 3000 W/g because of their unique vortex-to-onion magnetization reversal phenomenon upon AMF exposure[48]. Notably, as our simulation results revealed, only 8.70 × 10-6 ng of FVIOs could effectively increase the local temperature from 37 °C to 43 °C, whereas 33.6 × 10-6 ng of SPIOs was needed under the same AMF conditions (Figure S1, Supporting Information ). In addition, relatively large size and biocompatible FVIOs are more stable in the brain, which is favorable for long-term and repeated magnetothermal neurostimulation in vivo[48]. As such, FVIOs are likely to be high-performance nanoheaters that facilitate quick activation of TRPV1, which might provide an opportunity to establish a highly efficient and safe magnetothermal neurostimulation technique.
In the present study, we comprehensively investigated FVIO-facilitated magnetothermal neurostimulation in vitro and in vivo, and compared the results with those of widely used SPIO-based approaches. The newly designed FVIOs with an anti-His antibody coating triggered Ca2+ influx in TRPV1-expressing HEK293T cells and cortical neurons with a minimum Fe concentration of 54 μg/mL, which was 20.27-fold lower than that of the SPIOs under the same AMF treatment. In vivo magnetothermal stimulation in the central nucleus of the amygdala (CeA) further demonstrated that FVIO treatment quickly evoked fear behaviors in mice, exhibiting a response time that was 2.3-fold faster than that in the SPIO-treated group. Overall, the lowest effective dose of FVIOs was 0.05 μg in vivo, 16.7 fold less than the effective dose of SPIOs. More importantly, the SPIO ( 0.8 μg)-treated mice exhibited significant histopathological changes (such as vacuolar degeneration) and upregulated proinflammatory cytokine (such as interleukin-6 (IL-6)) expression in the CeA. In contrast, the FVIO ( 0.05 μg)-treated mice showed negligible histopathological alterations and changes in proinflammatory cytokine expression. In addition, in the FVIO-treated groups, fear behaviors were successfully induced in the mice on the 60th day after injection. However, in the SPIO-treated group, fear behaviors could not be induced on the 30th day. Notably, only 0.28 μg of FVIOs activated the CeA endogenously expressing TRPV1 and elicited fear behaviors in transgene-free mice. The aim of this work was to establish an efficient and safe FVIO-facilitated magnetothermal neuromodulation technique, which is expected to be a powerful tool for future neuroscience investigations and therapeutic applications to neurological disorders.