Selective manipulation of excitatory and inhibitory neurons in top-down and bottom-up visual pathways using ultrasound stimulation
Yehhyun Jo, Xiaojia Liang, Hong Hanh Nguyen, Yeonseo Choi, Minji Choi, Ga-Eun Bae, Yakdol Cho, Jiwan Woo, Hyunjoo Jenny Lee
Brain Stimulation 2025, 18, 848-862 · 10.1016/j.brs.2025.04.008
Abstract
Introduction Techniques for precise manipulation of neurons in specific neural pathways are crucial for excitatory/inhibitory (E/I) balance and investigation of complex brain circuits. Low-intensity focused ultrasound stimulation (LIFUS) has emerged as a promising tool for noninvasive deep-brain targeting at high spatial resolution. However, there is a lack of studies that extensively investigate the modulation of top-down and bottom-up corticothalamic circuits via selective manipulation of excitatory and inhibitory neurons. Here, a comprehensive methodology using electrophysiological recording and c-Fos staining is employed to demonstrate pulse repetition frequency (PRF)-dependent E/I selectivity of ultrasound stimulation in the top-down and bottom-up corticothalamic pathways of the visual circuit in rodents. Materials and methods Ultrasound stimulation at various PRFs is applied to either the lateral posterior nucleus of the thalamus (LP) or the primary visual cortex (V1), and multi-channel single-unit activity is recorded from the V1 using a silicon probe. Results and conclusion Our results demonstrate that high-frequency PRFs, particularly at 3 kHz and 1 kHz, are effective at activating the bidirectional corticothalamic visual pathway. In addition, brain region-specific PRFs modulate E/I cortical signals, corticothalamic projections, and synaptic neurotransmission, which is imperative for circuit-specific applications and behavioral studies.
Abstract via europepmc.
Exposures
Exposure 1: LIFUS of the lateral posterior thalamic nucleus (LP), recording from V1 (LP-V1 group)
Target: thalamus — “lateral posterior nucleus of the thalamus (LP)”
Device: other named manufacturer · Hagisonic Inc. ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 5,000 | ✓✓✓ |
| Pulse duration (ms) | not reported | ⚑ |
| Pulse repetition frequency (Hz) | 3,000, 1,000, 80swept | ✓✓✓ |
| Duty cycle (%) | 50 | ✓✓✓ |
| Sonication duration (s) | 1 | ✓✓✓ |
| Free-field pressure (kPa) | 1,240 | ✓✓✓ |
|---|---|---|
| Free-field Isppa (W/cm²) | 51.25 | ✓✓✓ |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | not reported | |
| In-situ pressure (kPa) | not reported | |
| In-situ Isppa (W/cm²) | not reported | |
| In-situ Ispta (W/cm²) | not reported |
The first function generator delivered a square wave at PRFs of 3 kHz, 1 kHz, and 80 Hz, corresponding to 3000 cycles, 1000 cycles, and 80 cycles for a total pulse train of 1 s, all at a 50% duty cycle. The first function generator was controlled using a custom MATLAB program to generate 20-min stimulation sessions with an inter-stimulation interval (ISI) of 10 s. Three different PRF conditions of 3 kHz, 1 kHz, and 80 Hz were applied in a random sequence, each lasting 20 min. Sham conditions included SHAM-AG (air gap between collimator and brain), SHAM-OT (off-target transcranial site, to check bone-conduction auditory activation), and SHAM-PS (direct stimulation of the probe shank, to check vibration-induced activation).
Exposure 2: LIFUS of the primary visual cortex (V1), recording from V1 (V1-V1 group)
Target: primary visual cortex — “primary visual cortex (V1)”
Device: other named manufacturer · Hagisonic Inc. ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 5,000 | ✓✓✓ |
| Pulse duration (ms) | not reported | ⚑ |
| Pulse repetition frequency (Hz) | 3,000, 1,000, 80swept | ✓✓✓ |
| Duty cycle (%) | 50 | ✓✓✓ |
| Sonication duration (s) | 1 | ✓✓✓ |
| Free-field pressure (kPa) | 1,240 | ✓✓✓ |
|---|---|---|
| Free-field Isppa (W/cm²) | 51.25 | ✓✓✓ |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | not reported | |
| In-situ pressure (kPa) | not reported | |
| In-situ Isppa (W/cm²) | not reported | |
| In-situ Ispta (W/cm²) | not reported |
The first function generator delivered a square wave at PRFs of 3 kHz, 1 kHz, and 80 Hz, corresponding to 3000 cycles, 1000 cycles, and 80 cycles for a total pulse train of 1 s, all at a 50% duty cycle. The first function generator was controlled using a custom MATLAB program to generate 20-min stimulation sessions with an inter-stimulation interval (ISI) of 10 s. Three different PRF conditions of 3 kHz, 1 kHz, and 80 Hz were applied in a random sequence, each lasting 20 min. Sham conditions included SHAM-AG (air gap between collimator and brain), SHAM-OT (off-target transcranial site, to check bone-conduction auditory activation), and SHAM-PS (direct stimulation of the probe shank, to check vibration-induced activation).
Flags from extraction
n_subjects— 61 is the total mice used across the whole study (electrophysiology, c-Fos IHC, neurotransmitter co-staining, and sham/control groups); not all received active ultrasound. Primary electrophysiology cohorts were n=12 (LP-V1 group) and n=8 (V1-V1 group), reported in separate sentences.exposures[0].target.terms— The paper's target is the 'lateral posterior nucleus (LP) of the thalamus'; no exact match exists in the target vocabulary (closest concepts are pulvinar or posterior_thalamic_nucleus), so the parent term 'thalamus' is used.auditory_control— SHAM-OT (ultrasound aimed transcranially at an off-target site) was used specifically to test for bone-conduction auditory confounds, but does not match any listed auditory_control vocabulary term exactly; classified as other.exposures[0].timing.pulse_duration_ms— Duty cycle (50%) and PRF (3/1 kHz/80 Hz) are stated but individual pulse width in ms is not stated by the paper; left not_reported rather than computed.exposures[1].timing.pulse_duration_ms— Same as exposures[0]: pulse width not explicitly stated.exposures[0].in_situ.method— Craniotomy removed the skull over the target for the actual in vivo experiments, but the paper does not explicitly relabel the hydrophone-measured free-field values as in-situ brain values for this preparation.exposures[1].in_situ.method— Same as exposures[0].