Glutamate indicators with increased sensitivity and tailored deactivation rates

Abhi Aggarwal, Adrian Negrean, Yang Chen, Rishyashring R. Iyer, Daniel Reep, Anyi Liu, Anirudh Palutla, Michael E. Xie, Bryan MacLennan, Kenta M. Hagihara, Lucas W. Kinsey, Julianna L. Sun, Pantong Yao, Jihong Zheng, Arthur Tsang, Getahun Tsegaye, Yonghai Zhang, Ronak Patel, Benjamin J. Arthur, Julien Hiblot, Philipp Leippe, Mirosław Tarnawski, Jonathan S. Marvin, Jason Vevea, Srinivas C. Turaga, Alison G. Tebo, Matteo Carandini, L. Federico Rossi, David Kleinfeld, Arthur Konnerth, Karel Svoboda, Glenn Turner, Jeremy P. Hasseman, Kaspar PodgorskiNature Methods 2025 · DOI

Introduction

Glutamate is the primary excitatory neurotransmitter in the vertebrate brain, serving as the fundamental "go" signal between neurons. To decode how neural circuits process information, researchers must observe this chemical signaling in real-time across thousands of individual synapses. However, imaging glutamate in vivo presents significant technical hurdles. The neurotransmitter is released in minute quantities into the synaptic cleft and cleared within milliseconds. While previous indicators, such as iGluSnFR3, represented a major advancement, they often lacked the signal-to-noise ratio (SNR) required to record from large populations of synapses simultaneously. Furthermore, different experimental setups require different temporal resolutions: some need extreme speed to resolve high-frequency bursts, while others benefit from slower signals that are easier to capture with standard video-rate microscopy.

Engineering a New Generation of Indicators

To address these limitations, researchers developed a massive high-throughput engineering and screening pipeline to optimize the iGluSnFR scaffold. The team generated and screened a library of 3,365 variants of the iGluSnFR3 protein.

The screening process was conducted in two stages. Initial screening took place in rat primary neuron cultures using field stimulation to evoke action potentials. This was followed by an "optical mini" screen to observe spontaneous, single-vesicle glutamate release—a much more stringent test of sensitivity and kinetics than bulk stimulation.

A key innovation in this study was the use of a Generalized Linear Model (GLM) to analyze how 12 specific mutations interacted with one another, a phenomenon known as epistasis. By mapping the "sequence-to-function" landscape, the researchers identified non-obvious combinations of mutations that yielded superior performance. This computational approach was complemented by structural biology, including the resolution of a high-resolution crystal structure of iGluSnFR3 (PDB: 9FBU) to physically map the effects of these mutations.

Tailored Kinetics for Diverse Applications

The study's most significant insight was that "faster" is not always "better" for every imaging application. By optimizing the deactivation rates, the researchers produced two distinct variants:

  • iGluSnFR4s (Slow): With a decay time of approximately 153 ms, this variant is optimized for high sensitivity and large-scale imaging. Its slower kinetics make it ideal for recording from massive populations of synapses using standard frame rates (e.g., 30 Hz).
  • iGluSnFR4f (Fast): With a decay time of approximately 26 ms, this variant is essential for capturing rapid dynamics and resolving high-frequency inputs.

Validation was performed in vivo using two-photon microscopy in the mouse visual (V1) and somatosensory (barrel) cortices, as well as fiber photometry in the midbrain.

Performance and Key Results

The new iGluSnFR4 variants demonstrated substantial improvements over their predecessors:

  • Enhanced Sensitivity: Both variants showed significantly larger signal amplitudes for single action potentials compared to iGluSnFR3. They are sensitive enough to detect the release of a single synaptic vesicle in vivo.
  • High-Frequency Resolution: In the barrel cortex, iGluSnFR4f successfully resolved glutamate transients during whisking at frequencies up to 20 Hz, a feat that outperformed previous indicators.
  • Deep-Brain Photometry: In fiber photometry experiments targeting GABAergic neurons in the ventral tegmental area (VTA), iGluSnFR4s produced signals approximately 3.6 times larger than iGluSnFR3, facilitating the detection of subtle neurotransmitter changes.

Limitations and Future Directions

The study also revealed intriguing biological complexities. The new variants showed a reduced dynamic range when tested as purified soluble proteins, despite their superior performance on the cell surface. This suggests that the environment of the cell membrane fundamentally alters the protein's behavior, highlighting the necessity of screening directly in neurons rather than using purified protein shortcuts.

Additionally, the slower, higher-affinity variant (4s) exhibits slightly more spatial "crosstalk" between nearby synapses because the signal persists longer and can diffuse further.

Looking forward, the researchers suggest that the massive dataset generated during this study could be used to train machine learning models. Such models could predict even more effective protein sequences, moving beyond the 12 mutation sites explored in this work to further refine the tools available for neuroscientific discovery.

Summary updated

Abstract

Understanding how neurons integrate signals from thousands of input synapses requires methods to monitor neurotransmission across many sites simultaneously. The fluorescent protein glutamate indicator iGluSnFR enables visualization of synaptic signaling, but the sensitivity, scale and speed of such measurements are limited by existing variants. Here we developed two highly sensitive fourth-generation iGluSnFR variants with fast activation and tailored deactivation rates: iGluSnFR4f for tracking rapid dynamics, and iGluSnFR4s for recording from large populations of synapses. These indicators detect glutamate with high spatial specificity and single-vesicle sensitivity in vivo. We used them to record natural patterns of synaptic transmission across multiple experimental contexts in mice, including two-photon imaging in cortical layers 1-4 and hippocampal CA1, and photometry in the midbrain. The iGluSnFR4 variants extend the speed, sensitivity and scalability of glutamate imaging, enabling direct observation of information flow through neural networks in the intact brain.
BibTeX
@inproceedings{aggarwal2025glutamate,
  author = {Abhi Aggarwal and Adrian Negrean and Yang Chen and Rishyashring R. Iyer and Daniel Reep and Anyi Liu and Anirudh Palutla and Michael E. Xie and Bryan MacLennan and Kenta M. Hagihara and Lucas W. Kinsey and Julianna L. Sun and Pantong Yao and Jihong Zheng and Arthur Tsang and Getahun Tsegaye and Yonghai Zhang and Ronak Patel and Benjamin J. Arthur and Julien Hiblot and Philipp Leippe and Mirosław Tarnawski and Jonathan S. Marvin and Jason Vevea and Srinivas C. Turaga and Alison G. Tebo and Matteo Carandini and L. Federico Rossi and David Kleinfeld and Arthur Konnerth and Karel Svoboda and Glenn Turner and Jeremy P. Hasseman and Kaspar Podgorski},
  title = {{Glutamate indicators with increased sensitivity and tailored deactivation rates}},
  booktitle = {Nature Methods},
  year = {2025},
  doi = {10.1038/s41592-025-02965-z},
}
Figure