Ultrasound pulse repetition frequency preferentially activates different neuron populations independent of cell type
Jack Sherman, Emma Bortz, Erynne San Antonio, Hua-an Tseng, Laura Raiff, Xue Han
Journal of Neural Engineering 2024, 21, 056008 · 10.1088/1741-2552/ad731c
Abstract
Objective . Transcranial ultrasound (US) stimulation serves as an external input to a neuron, and thus the evoked response relies on neurons' intrinsic properties. Neural activity is limited to a couple hundred hertz and often exhibits preference to input frequencies. Accordingly, US pulsed at specific physiologic pulse repetition frequencies (PRFs) may selectively engage neurons with the corresponding input frequency preference. However, most US parametric studies examine the effects of supraphysiologic PRFs. It remains unclear whether pulsing US at different physiologic PRFs could activate distinct neurons in the awake mammalian brain. Approach . We recorded cellular calcium responses of individual motor cortex neurons to US pulsed at PRFs of 10, 40, and 140 Hz in awake mice. We compared the evoked responses across these PRFs in the same neurons. To further understand the cell-type dependent effects, we categorized the recorded neurons as parvalbumin positive fast spiking interneurons or putative excitatory neurons and analyzed single-cell mechanosensitive channel expression in mice and humans using the Allen Brain Institute's RNA-sequencing databases. Main results . We discovered that many neurons were preferentially activated by only one PRF and different PRFs selectively engaged distinct neuronal populations. US-evoked cellular calcium responses exhibited the same characteristics as those naturally occurring during spiking, suggesting that US increases intrinsic neuronal activity. Furthermore, evoked responses were similar between fast-spiking inhibitory neurons and putative excitatory neurons. Thus, variation in individual neuron's cellular properties dominates US-evoked response heterogeneity, consistent with our observed cell-type independent expression patterns of mechanosensitive channels across individual neurons in mice and humans. Finally, US transiently increased network synchrony without producing prolonged over-synchronization that could be detrimental to neural circuit functions. Significance . These results highlight the feasibility of activating distinct neuronal subgroups by varying PRF and the potential to improve neuromodulation effects by combining physiologic PRFs.
Abstract via europepmc.
Exposures
Exposure 1: Motor cortex US at PRFs of 10, 40, 140 Hz (physiologic) and 2 kHz (supraphysiologic)
Target: motor cortex — “motor cortex”
Device: Ultran · Ultran · GS350-D13 ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 350 | ✓✓✓ |
| Pulse duration (ms) | not reported | ⚑ |
| Pulse repetition frequency (Hz) | 10, 40, 140, 2,000swept | ✓✓✓ |
| Duty cycle (%) | 20, 42swept | ✓✓✓⚑ |
| Sonication duration (s) | 1 | ✓✓✓ |
| Free-field pressure (kPa) | 522 | ✓✓✓ |
|---|---|---|
| Free-field Isppa (W/cm²) | 9.11 | ✓✓✓ |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | simulationmean or range across subjects | |
| In-situ pressure (kPa) | 13, 95swept | ✓✓✓ |
| In-situ Isppa (W/cm²) | 0.2, 3.16swept | ✓✓✓ |
| In-situ Ispta (W/cm²) | not reported |
350 kHz US was delivered for 1 s per trial pulsed at PRFs of 10, 40, or 140 Hz at 20% duty cycle, or at a supraphysiologic 2 kHz PRF at 42% duty cycle. In single-PRF sessions, each of 20 trials lasted 50 s (10 s pre-US, 1 s US, 39 s recovery). In alternating-PRF sessions, 25 blocks of 3 trials (10, 40, 140 Hz, randomly ordered) were run, each trial with 10 s pre-US, 1 s US, and 10 s post-US. Sham stimulation used the same transducer setup but omitted the US gel coupling the transducer to the mouse's chin, preventing acoustic propagation into the brain while the mouse still experienced any transducer-associated sound. k-Wave simulation estimated an overall in-situ peak pressure of 289.5 kPa (~0.29 MPa) resulting from the 522 kPa free-field pressure, with a lower simulated pressure/intensity range specifically at the calcium-imaging site (13-95 kPa; 0.2-3.16 W/cm2, average 1.5 W/cm2); a phantom-skull hydrophone measurement of 125 kPa validated the upper end of this estimate.
Flags from extraction
exposures[0].in_situ.reported_as— The 13-95 kPa / 0.2-3.16 W/cm2 values are a spatial range across the calcium-imaging site from a single simulation, not a range or mean across subjects; the paper separately states a single overall simulated in-situ peak pressure of 289.5 kPa (~0.29 MPa), which is noted in the protocol description rather than as a separate field value.exposures[0].timing.pulse_duration_ms— Not stated directly; would require dividing duty cycle by PRF, which is arithmetic this extraction avoids.sham_type— Sham omitted the US gel coupling medium rather than powering off the transducer; classified as inactive_transducer (blocked coupling) but this is an inference from the described method.exposures[0].timing.duty_cycle_pct— Duty cycle values (20% and 42%) correspond to different PRF subsets (10/40/140 Hz vs 2 kHz respectively) but are listed together without pairing to specific PRFs.