DEVELOPMENT OF HIGH-SPEED PHOTOMETRY FOR IMAGING MEMORY-RELATED NEURAL OSCILLATIONS IN FREELY BEHAVING ANIMALS

Memory-related oscillations such as theta and sharp wave–ripples (SWRs) are typically studied with local field potential (LFP) recordings, but electrophysiology alone cannot assign these dynamics to specific cell types. Genetically encoded voltage indicators (GEVIs) enable millisecond-scale readouts of subthreshold activity and spike timing, yet their low photon budget and rapid photobleaching impose stringent requirements on detection sensitivity and sampling speed.

We developed a high-speed fiber photometry system based on single-photon avalanche diode (SPAD) sensors, enabling kHz sampling with low noise under photon-limited conditions. The system is compatible with simultaneous in vivo electrophysiology and is readily scalable to multi-fiber imaging across multiple brain regions. In freely behaving and sleeping mice, SPAD photometry resolved theta- and gamma-band oscillations and SWR-associated voltage dynamics. In hippocampal CA1, population membrane potential in pyramidal neurons preferentially depolarised near theta troughs, was modulated by gamma phase, and peaked during SWRs. In contrast, parvalbumin-expressing interneurons depolarised across multiple theta phases and exhibited significant theta–gamma coupling.

We further demonstrate extensibility to multi-region recordings by imaging pyramidal-cell membrane potential from left CA3 and bilateral CA1. Notably, even without an LFP reference, voltage signals alone were sufficient to quantify theta-phase and gamma-phase coupling across hippocampal subregions.

These findings support SPAD-based photometry as a scalable platform for studying fast, cell-type-specific circuit dynamics during behaviour and sleep.