Kilohertz electrical stimulation drives widespread neuronal activation and temporal interference imposes focal rhythmic modulation

Technical Summary: Kilohertz Electrical Stimulation Drives Widespread Neuronal Activation and Temporal Interference Imposes Focal Rhythmic Modulation

Problem Statement
Temporal interference (TI) stimulation has emerged as a promising non-invasive technique for targeting deep brain structures. Its theoretical premise relies on the interaction of two high-frequency (kilohertz) electric fields with slightly offset frequencies, generating an amplitude-modulated (AM) envelope at the difference frequency. The prevailing hypothesis posits that neurons are insensitive to the high-frequency carrier signals due to the low-pass filtering properties of the neuronal membrane, responding only to the low-frequency AM envelope. Consequently, TI is expected to selectively activate deep targets where the AM envelope is strongest while sparing superficial tissues exposed only to the unmodulated carriers. However, this assumption has not been rigorously tested at the cellular level, particularly regarding whether unmodulated kilohertz fields are physiologically inert and whether the spatial distribution of neuronal activation strictly matches the predicted AM envelope. Furthermore, the spatiotemporal dynamics of neuronal responses to TI, including subthreshold effects and rhythmic synchronization at the single-neuron level, remain largely unexplored.

Methodology
The authors employed a multi-modal approach combining in silico computational modeling, ex vivo brain slice preparations, and in vivo imaging in mice to characterize neuronal responses to TI and high-frequency stimulation.

• Experimental Models: The study utilized mouse models expressing genetically encoded calcium indicators (jGCaMP8s) and voltage indicators (JEDI-2P) via intracerebroventricular viral injections.
• Stimulation Paradigms:
  • Premodulated TI (pmTI): To isolate waveform effects, the authors applied pre-synthesized waveforms with defined AM indices (0%, 50%, 100%) to brain slices, bypassing the spatial complexity of dual-electrode interference.
  • Conventional TI: Two pairs of electrodes delivered currents at 2000 Hz and 2001 Hz (1 Hz beat frequency) to generate interference fields in both slices and in vivo.
  • Controls: Unmodulated high-frequency (carrier-only) stimulation and sham controls were used for comparison.
• Imaging and Recording:
  • Two-Photon Microscopy: Used to monitor individual neuronal responses (calcium transients and voltage changes) with high spatial and temporal resolution, avoiding electrical artifacts.
  • Immunohistochemistry: c-Fos expression was quantified after prolonged (20-minute) stimulation to assess widespread neuronal activation across cortical and deep brain regions.
  • Computational Modeling: NEURON simulations using biophysical models of cortical pyramidal neurons and axonal compartments were used to investigate ionic mechanisms and spatial field distributions.

Key Contributions and Results

1. Unmodulated Kilohertz Stimulation is Sufficient to Drive Neuronal Activity:
   Contrary to the assumption that kilohertz carriers are ineffective, the study demonstrates that unmodulated high-frequency stimulation (AM0) alone is sufficient to evoke neuronal firing and subthreshold depolarization. In brain slices, the threshold current required to elicit responses was actually lower for unmodulated carriers than for amplitude-modulated waveforms (AM50, AM100). This suggests that the high-frequency field itself drives activation, likely through a mechanism where rapid sodium channel activation outpaces slower potassium channel activation, leading to cumulative membrane depolarization.

2. Amplitude Modulation Organizes Temporal Structure, Not Spatial Focality:
   While activation occurred broadly across the stimulation field, the temporal structure of the response was distinct.

   • Widespread Activation: Both carrier-only and TI stimulation recruited neurons across the entire imaging field, with no significant difference in the spatial distribution of responsive cells.
   • Focal Rhythmic Modulation: Only TI stimulation (with AM) imposed rhythmic synchronization to the beat frequency (e.g., 1 Hz). This entrainment was spatially restricted, being strongest near the center of the interference field where the AM index was highest, and diminishing with distance. Thus, TI does not confine activation to a focal point but rather imposes temporal modulation focally upon a broadly activated population.

3. Subthreshold Rhythmic Depolarization:
   Using voltage indicators (JEDI-2P), the authors showed that even at subthreshold intensities where action potentials were not generated, TI stimulation induced rhythmic oscillations in membrane potential synchronized to the beat frequency. Carrier-only stimulation induced depolarization but lacked this rhythmic organization. This indicates that TI can modulate neuronal excitability and timing without necessarily driving suprathreshold spiking.

4. In Vivo Validation and c-Fos Expression:

   • In Vivo Imaging: In anesthetized and awake mice, TI stimulation elicited rhythmic calcium transients synchronized to the beat frequency, while carrier-only stimulation produced non-rhythmic activity of comparable amplitude.
   • Long-term Activation: Prolonged stimulation (20 minutes) resulted in significant c-Fos upregulation in the cortex (near the electrodes) for both TI and carrier-only conditions. However, no significant increase in c-Fos was observed in the deep target (hippocampus), suggesting that the local electric field magnitude in the deep target was insufficient to drive immediate-early gene expression, despite the presence of AM. This reinforces that activation is driven by field strength, while AM dictates the timing of that activation.

Significance and Claims
The paper claims to provide a revised framework for understanding TI stimulation. The authors argue that TI should not be viewed primarily as a method for generating focal neural activation, but rather as a method for imposing focal temporal modulation upon a background of widespread neuronal activation.

• Mechanistic Insight: The study challenges the notion that kilohertz carriers are physiologically inert. Instead, it proposes that the high-frequency field drives depolarization via ionic channel kinetics, while the amplitude modulation organizes the timing of this activity.
• Therapeutic Implications: The findings suggest that the "focality" of TI is temporal rather than spatial. This has implications for interpreting clinical data (e.g., fMRI BOLD responses) and designing stimulation protocols. The authors suggest that widespread activation might be therapeutically useful for modulating network connectivity, while the focal AM can be used to selectively entrain specific rhythms in deep targets.
• Future Directions: The authors highlight that effective TI entrainment may require higher stimulation amplitudes than previously assumed, particularly at higher carrier frequencies, and that the interaction between external modulation and endogenous brain states is critical for efficacy.

The study concludes that kilohertz-frequency stimulation itself is a potent neuromodulatory regime, and TI represents a specific application where amplitude modulation is used to shape the temporal dynamics of this widespread activation.