1. Transcranial Magnetic Stimulator for Focal Stimulation of Rodent Brain Transcranial magnetic stimulation (TMS) is emerging as a therapeutic tool for treating many neuropsychiatric disorders, but its underlying mechanisms remain poorly understood. Preclinical rodent studies are critically important in this regard. TMS systems, in their most basic form, are made of a treatment coil and a driving circuit that sends high current pulses to the coil to generate a Tesla-level magnetic field. The field penetrates brain tissue and induces eddy current when the coil is located over the skull. This stimulates the neurons causing the release of neurotransmitters and clinical effects. However, commercial rodent TMS coils lack the spatial focality to mimic human TMS treatment conditions - a large volume of brain tissue is stimulated even with the best available rodent TMS coils which prevents spatially relevant mechanistic understanding of preclinical models and its translation to human studies. Conventional TMS coil design is limited by coil-to-brain size ratio; very high electric current is needed if a human TMS coil is scaled to rodent brain size, causing excessive heating and demanding power supply requirements. We have developed an innovative concept to significantly improve the focality for rodent brain stimulation. In our coil design, a stack of long silicon steel sheets is used as a magnetic core due to high permeability and high magnetic saturation. Individual sheets are insulated to minimize eddy current generated in the core during stimulation. Litz wires are used for coil wrapping, eliminating the skin effect of the current in the coil. The coil is wrapped with a tilted angle and induces a tight focal spot in its asymmetric electric field distribution. The design has been tested in both computer simulations and pilot experiments. We have also developed an impedance matched TMS circuit to drive this coil, providing high current pulses. In in vivo experiments on mice, motor-evoked potential was recorded while TMS targeted the hindlimb motor cortex. Here, TMS induced limb twitches only on the contralateral hindlimb, but not any other body part, suggesting a much better focality than commercial rodent coils. This is the first TMS system that can induce unilateral movements on rodents, providing important access to spatially relevant mechanistic understanding of preclinical models and great possibilities for translation to human studies. 2. Functional MRI models of awake rodents performing goal-directed tasks Although fMRI has been routinely conducted in human participants while performing a goal-directed task, such a study in rodents is extremely challenging due to the difficulty in limiting head motion while performing an operant task. As a result, rodent fMRI studies have only been performed under anesthesia or with severe restraint, both of which precludes goal-directed tasks. We have developed a novel fMRI platform to overcome this long-lasting problem in the field, such that fMRI based dynamic brain activity can be measured in real time while a rodent performs a goal-directed task. Mice were transfected with channelorhodopsin-2 into the infralimbic cortex (ILc) and implanted with an optical fiber and an MRI-compatible head-mount. After recovery, mice were trained to remain head-fixed and to walk on a customized MRI-compatible treadmill, an activity that helped to alleviate stress associated with head-fixation. Animals were initially trained to lick a spout, which resulted in a photobeam break and the delivery of sucrose reward only when a light cue was present (ON period); reward was not available in the absence of the cue (OFF period). After mice successfully learned to lick during the ON period, sucrose was replaced with optogenetic stimulation delivered to the ILc, which consisted of a train of optogenetic pulses. FMRI data were acquired with a paradigm including the cue ON and OFF periods. During the cue ON period, optogenetic self-stimulation induced BOLD activation in regions consistent with the underlying projections of the medial PFC including cingulate cortex (CG1/CG2), prelimbic and ILc, piriform cortex, nucleus accumbens (NAc), septum, and midline/mediodorsal thalamic nuclei. Notably, NAc activation reflects the known rewarding, goal-directed properties of this pathway, and the activation of cingulate cortex suggests the engagement of executive control in this goal-directed task. In contrast, during the cue OFF period, the conditioned lick responses were accompanied by negative BOLD signals in the anterior olfactory nucleus (AO), anterior NAc, lateral septum and dorsal hippocampus. These negative BOLD signals are novel and of interest in light of previous research findings suggesting that negative BOLD signals are related to negative emotion. This novel awake imaging platform permits dynamic fMRI acquisition from a mouse actively engaged in a goal-directed, reinforced behavior. 3. Neural pathways of anterior and posterior insula revealed by optogenetic stimulation and fMRI The insula is known to receive interoceptive signals and plays a critical role in affective functions. A functional distinction between the anterior and posterior insular has been proposed. While human studies of insular functions are mainly based on insular lesions, animal models allow us to investigate the functions and processes of the insula more directly. Here, using optogenetic stimulation and fMRI, we examine the downstream brain activation associated with stimulation at the anterior or posterior insula, with the aim to understand the functional profile of the insula and to further explore its role in addiction related behaviors. Preliminary results show that optogenetic stimulation at the posterior insula resulted in robust activation along the entire ipsilateral insular strip, extending to the very anterior. This is unusual because optogenetic stimulation usually results in the highest activation at the targeted area and decreases with distance from the target. Stimulation at the anterior insula resulted in the typical activation pattern with the highest activation at the target site and stops at around the mid-insula. In the control groups, no activation was shown at the insula. We also found that the two groups did not show preference or aversion when the anterior insula was stimulated, but only the active group showed aversion when the posterior insula was stimulated. Further analyses of the imaging and behavioral data are underway.
Showing the most recent 10 out of 19 publications
