Patterned photostimulation via visible-wavelength photonic probes for deep brain optogenetics

Eran Segev, Jacob Reimer, Laurent Moreaux, Trevor M. Fowler, Derrick Chi, Wesley D. Sacher, Maisie Lo, Karl Deisseroth, Andreas S. Tolias, Andrei Faraon, M. L. RoukesNeurophotonics 2016 · DOI

TL;DR

Researchers have developed novel, implantable photonic probes that overcome the limitations of light scattering in brain tissue, enabling precise, patterned optogenetic stimulation deep within the brain. These probes, fabricated using integrated nanophotonics and MEMS, offer a scalable and minimally invasive solution for controlling neural circuits with unprecedented spatial and temporal resolution.

What Problem Does This Paper Address?

Optogenetics, a powerful technique for controlling genetically modified neurons with light, has revolutionized neuroscience. However, a major hurdle for its application in deep brain regions is the scattering and absorption of light by neural tissue. Traditional methods using free-space optics or optical fibers struggle to deliver light effectively beyond a few millimeters, limiting the ability to precisely stimulate or inhibit neural circuits deep within the brain. This paper addresses this challenge by proposing and demonstrating a new class of implantable photonic probes designed to deliver patterned light with cellular-scale precision to these deeper brain areas.

Methods

The core innovation lies in the development of implantable, ultranarrow silicon-based photonic probes. These probes integrate several key technologies:

  • Integrated Nanophotonics: The probes utilize silicon nitride (Si₃N₄) photonic circuitry, including arrayed waveguide gratings (AWGs) for wavelength division multiplexing (WDM) and grating couplers (termed "E-pixels") for efficient light emission.
  • Wavelength Division Multiplexing (WDM): This technique allows multiple independent light signals, each encoded on a different wavelength, to be transmitted through a single optical fiber. On the probe, AWGs demultiplex these signals, routing each wavelength to a specific E-pixel. This enables independent control of numerous illumination points along the probe.
  • Microelectromechanical Systems (MEMS): MEMS fabrication processes are used to create the ultranarrow, needle-like probe shanks, minimizing tissue displacement upon implantation.
  • Grating Couplers (E-pixels): These are designed to efficiently couple light off the probe shank into the surrounding neural tissue with low divergence, creating small, localized illumination spots comparable in size to neuronal cell bodies.
  • Probe Architecture: The probes feature thin, multi-layered shanks (18 µm thick) with minimal cross-sections (down to 20 µm at the tip) to reduce tissue damage. The probe head houses the AWG for WDM.

Datasets

Not discussed in the paper. The study primarily relies on in vivo experimental validation using animal models.

Key Insights and Ideas

  • WDM as a Scalable Solution: The paper demonstrates that WDM is a highly effective strategy for multiplexing many independent optical channels onto a single fiber, enabling a massive increase in the number of addressable illumination points (E-pixels) on a probe.
  • Minimally Invasive Design: The use of MEMS fabrication for ultranarrow shanks is crucial for reducing tissue damage and enabling deeper penetration with less perturbation.
  • Precise Illumination Control: The combination of WDM and precisely engineered grating couplers allows for highly localized and patterned photostimulation, with beam profiles comparable to the size of individual neurons.
  • Passive Components for Heat Reduction: The use of passive photonic components (AWGs, waveguides) for WDM and demultiplexing avoids the heat generation associated with active components like micro-LEDs, which is critical for avoiding thermal artifacts in neural tissue.
  • Bridging Optical Communications and Neuroscience: The work successfully adapts mature optical communication technologies (WDM, AWGs) to the visible light spectrum relevant for optogenetics, opening new avenues for neural interface development.

Key Results

  • Functional Validation in vivo: The probes were successfully implanted in mice and demonstrated the ability to optogenetically activate neurons.
  • Hippocampal Stimulation: Extracellular electrical recordings in the hippocampus of mice expressing channelrhodopsin-2 (ChR2) showed robust multi-unit spiking activity in response to light pulses delivered by the probe.
  • Cortical Stimulation: Two-photon functional imaging in the cortex of mice coexpressing ChR2 and GCaMP6s showed localized neural activation, with minimal overlap between adjacent E-pixels, demonstrating cellular-scale precision.
  • Low Divergence Beams: E-pixels produced beams with low divergence angles (2-3.5 degrees), allowing for illumination of neural tissue up to ~200 µm away with a beam width comparable to neuronal cell bodies.
  • Scalable E-pixel Density: The design allows for E-pixel pitches as low as 50 µm, with future designs aiming for even smaller spacing.
  • Insertion Loss: Initial prototypes had an insertion loss of approximately 20 dB, with the majority attributed to coupling losses. Future generations are projected to achieve <10 dB.

Limitations and Future Work

  • Insertion Loss: The current insertion loss is significant, primarily due to coupling inefficiencies between the optical fiber and the on-chip circuitry. Future work will focus on optimizing these coupling interfaces, potentially using angle-polished fibers and advanced foundry fabrication.
  • Number of Channels: While WDM offers scalability, the number of addressable E-pixels per AWG is limited by the spectral bandwidth of the optogenetic effector and the spectral channel width. Further improvements in laser sources and fabrication precision are needed to increase channel density.
  • Fabrication Complexity: While the process is designed for scalability, current fabrication relies on shared university facilities, leading to some loss and variability. Transitioning to commercial foundry processes is a key future direction.
  • Future Designs: The authors are actively developing next-generation probes with denser E-pixel arrays (100 µm spacing, aiming for 50 µm) and improved optical performance. They also envision hierarchical on-probe photonic circuits to further reduce the number of external fiber connections. The ultimate goal is widespread deployment for neuroscience research and neuromedical applications.

Summary updated

Abstract

Optogenetic methods developed over the past decade enable unprecedented optical activation and silencing of specific neuronal cell types. However, light scattering in neural tissue precludes illuminating areas deep within the brain via free-space optics; this has impeded employing optogenetics universally. Here, we report an approach surmounting this significant limitation. We realize implantable, ultranarrow, silicon-based photonic probes enabling the delivery of complex illumination patterns deep within brain tissue. Our approach combines methods from integrated nanophotonics and microelectromechanical systems, to yield photonic probes that are robust, scalable, and readily producible <i>en masse</i>. Their minute cross sections minimize tissue displacement upon probe implantation. We functionally validate one probe design <i>in vivo</i> with mice expressing channelrhodopsin-2. Highly local optogenetic neural activation is demonstrated by recording the induced response-both by extracellular electrical recordings in the hippocampus and by two-photon functional imaging in the cortex of mice coexpressing GCaMP6.
BibTeX
@inproceedings{segev2016patterned,
  author = {Eran Segev and Jacob Reimer and Laurent Moreaux and Trevor M. Fowler and Derrick Chi and Wesley D. Sacher and Maisie Lo and Karl Deisseroth and Andreas S. Tolias and Andrei Faraon and M. L. Roukes},
  title = {{Patterned photostimulation via visible-wavelength photonic probes for deep brain optogenetics}},
  booktitle = {Neurophotonics},
  year = {2016},
  doi = {10.1117/1.nph.4.1.011002},
}
Figure