Intrinsic functional neuron-type selectivity of transcranial focused ultrasound neuromodulation

Kai Yu, Xiaodan Niu, Esther Krook-Magnuson, Bin He

Nature Communications 2021, 12 · 10.1038/s41467-021-22743-7

rodenthealthyinvasive electrophysiologyhistology molecular

Abstract

Transcranial focused ultrasound (tFUS) is a promising neuromodulation technique, but its mechanisms remain unclear. We hypothesize that if tFUS parameters exhibit distinct modulation effects in different neuron populations, then the mechanism can be understood through identifying unique features in these neuron populations. In this work, we investigate the effect of tFUS stimulation on different functional neuron types in in vivo anesthetized rodent brains. Single neuron recordings were separated into regular-spiking and fast-spiking units based on their extracellular spike shapes acquired through intracranial electrophysiological recordings, and further validated in transgenic optogenetic mice models of light-excitable excitatory and inhibitory neurons. We show that excitatory and inhibitory neurons are intrinsically different in response to ultrasound pulse repetition frequency (PRF). The results suggest that we can preferentially target specific neuron types noninvasively by tuning the tFUS PRF. Chemically deafened rats and genetically deafened mice were further tested for validating the directly local neural effects induced by tFUS without potential auditory confounds.

Abstract via europepmc.

SpeciesWistar rat (Hsd:WI) and transgenic mouse (CaMKIIa-ChR2, PV-ChR2)

Subjects9, 10, 7, 3, 2swept animals

Sessions per subjectnot reported

Randomisedyes

Blindingnot reported

Sham / controlinactive transducer, active control site

Auditory controldeafened subjects

Readout timingonline

Anaesthesiaboth

Readoutsinvasive electrophysiology, histology molecular — Multi-unit and single-unit spiking activity classified into regular-spiking (RSU, presumed excitatory) and fast-spiking (FSU, presumed inhibitory) units from 32-channel intracranial electrode arrays; local field potentials; optogenetic ('optotagging') identification of CaMKIIa (excitatory) and parvalbumin (inhibitory) neurons in transgenic mice; hematoxylin and eosin histology for tissue damage.

Direction of effectexcitatoryRegular-spiking (presumed excitatory) units showed PRF-dependent increases in firing rate (significantly higher at 3000-4500 Hz vs 30 Hz), while fast-spiking (presumed inhibitory) units showed no significant change in firing rate across PRF levels, indicating cell-type-selective excitatory modulation.

Adverse eventsnone observedAll tFUS stimulation parameters used on the S1 cortices of rats and mice were maintained in brief exposures (i.e., 67-ms sonication per trial, with the total duty cycle of each trial being <3%) and with low intensities, which led to negligible calculated temperature rises (< 0.1 °C) at the targeted brain area. In addition, the mechanical index (MI) used in these experiments was less than 0.15, given the low peak negative pressure (i.e., <100 kPa).

Exposures

Exposure 1: tFUS PRF sweep at primary somatosensory cortex (S1)

Target: primary somatosensory cortex — “left primary somatosensory cortex (S1)”
Device: Olympus / Panametrics · Olympus Scientific Solutions Americas, Inc. / Blatek Industries, Inc. · V391-SU-F1.5IN-PTF / AT31529 ✓

Pulse timing
Waveform	pulsed
Fundamental frequency (kHz)	500	✓✓✓
Pulse duration (ms)	0.134, 0.2, 0.4, 2, 20swept	✓✓✓⚑
Pulse repetition frequency (Hz)	30, 300, 1,500, 3,000, 4,500swept	✓✓✓
Duty cycle (%)	0.6, 6, 30, 60, 90swept	✓✓✓
Sonication duration (s)	0.067	✓✓✓

Pressure and intensity, by domain
Free-field pressure (kPa)	148	✓✓✓
Free-field Isppa (W/cm²)	not reported
Free-field Ispta (W/cm²)	not reported
In-situ estimate	measurementsingle value
In-situ pressure (kPa)	79	✓✓✓
In-situ Isppa (W/cm²)	0.176	✓✓✓
In-situ Ispta (W/cm²)	0.00105, 0.0106, 0.0527, 0.105, 0.158swept	✓✓✓

Protocol, in the paper’s words

Rat S1 was stimulated with 500 kHz tFUS at 40-degree incidence at five PRF levels (30-4500 Hz), each sonication ('ultrasound duration', UD) lasting 67 ms, with inter-sonication interval (ISoI) of 2.5 s (10% jittered). In a first design (Table 1), tone-burst duration (TBD) was held constant at 200 µs while duty cycle varied with PRF (0.6-90%). In a second design (Table 2), TBD was instead adjusted per PRF (20, 2, 0.4, 0.2, 0.134 ms) to hold duty cycle constant at 60% across all five PRF levels, keeping Isppa, Isptp and Ispta constant. The same f0=500 kHz parameters (PRF 30 and 300 Hz only) were used in optogenetic mice. Three sham conditions were used: sham with flipped transducer (SFLP, aimed away from skull), sham at skull front (SSKF, a different skull site), and sham at electrode shank (SSHK, ultrasound delivered to the electrode shank away from the brain, at 100/43/28/13% derated energy levels).

Flags from extraction

exposures[0].timing.pulse_duration_ms — Combines pulse durations from two separate sub-experiments (Table 1: constant 0.2 ms TBD across PRFs; Table 2: TBD varied with PRF to hold duty cycle at 60%); list values are not positionally paired with the PRF/duty-cycle lists.
exposures[0].unspecified_domain.isppa_w_cm2 — Paper does not state whether Isppa/Ispta in Tables 1-2 are free-field or in-situ values; placed in unspecified_domain. Separately measured free-field (148 kPa) and through-skull (79 kPa) pressures are reported without explicit intensity equivalents.
n_subjects — Group sizes are reported for distinct sub-experiments (ketamine/xylazine rats N=9, isoflurane rats N=10, constant-duty-cycle isoflurane rats N=7, PV mice N=3, CaMKIIa mice N=2); possible overlap between the N=10 and N=7 rat cohorts is not stated. Additional chemically deafened rats and genetically deafened mice used in supplementary auditory-confound experiments are not enumerated in the main text.
sham_type — SSHK (ultrasound delivered to the electrode shank rather than a brain target) does not clearly fit a listed sham category and is recorded as 'other'.