Lawrence Livermore National Laboratory (LLNL) scientists, working with researchers at University of California, San Francisco (UCSF), have developed a novel system for recording widespread brain activity, using high-density implantable devices to collect real-time data over longer timescales and across multiple areas of the brain.
The new platform, as recently reported in the journal Neuron, is capable of continuously measuring the activity of nearly 400 single neurons over a period of at least five months from devices distributed in multiple regions of rodent brains. Using an automated spike sorting system developed at UCSF, the researchers concluded the recordings are stable enough to track large numbers of individual neurons for more than a week. The research is part of a project aimed at gaining a better understanding of the mechanisms of learning that occur over long periods of time.
"Being able to record from all of these regions for long periods of time with high-fidelity, you can start looking at patterns of learning and how memory changes over time," said LLNL researcher Angela Tooker, a co-author on the paper. "This is a huge step. It’s taking a lot of different things and putting it altogether into one package. Now we can see how that picture is changing from day to day in response to training, retraining, new stimuli. It opens up a whole new field."
The team said the polymer, array-based system provides a new tool for understanding brain activity and makes it possible to examine how different regions of the brain interact. UCSF scientists implanted probes in four distant regions of the rodent’s brains — the hippocampus, ventral striatum, orbitofrontal cortex and medial prefrontal cortex. The recordings showed how hippocampal sharp-wave ripples, which play a role in memory, also activate the orbitofrontal cortex, the medial pre-frontal cortex and nucleus accumbens.
Researchers said the platform represents a "substantial advance" over previous technologies, the main advantage being that they can last longer in the brain than most high-density silicon-based devices. The researchers hope the platform can provide key insights into how the brain functions, particularly how it stores memories, and allow them to ask questions such as how neurons change following sleep, how brain activity changes within an area over time, or how brain regions evolve as information is stored, consolidated and retrieved.
"Important activity in the brain doesn’t occur in just one place, it happens across many areas all connected together; it’s a very distributed circuit. As a result, we really need to be able to understand how different parts of the brain are interacting — how they’re talking with one another," said Lo4en Frank, UCSF professor of physiology, who headed the implantation procedures and experiments. "Our goal was to build a system that we could basically assemble out of parts: where individual recording components would let us target different areas and then be able to reassign those components based on our experimental needs."
The multi-electrode arrays were fabricated at LLNL and combined with electrical hardware built by private neuroscience tool manufacturer SpikeGadgets. The flexible polymer probes used in the study were designed and developed in collaboration with UCSF’s Frank Lab by a group led by LLNL’s Tooker and former LLNL engineer Vanessa Tolosa, who was formerly the project’s principal investigator and currently works at the private company Neuralink.
Because they are both biocompatible and flexible, the arrays can be safely implanted and move with the brain, reducing damage to neural tissue. Each of the 16 probes has 64 electrodes, for a total of 1,024 channels. Theoretically, each channel could record one neuron apiece, researchers said, depending on where they’re positioned in the brain.
"If you get them in the hands of grad students around the country, some really exciting things could happen," said LLNL engineer Jeanie Pebbles, a co-author on the paper who performed post-processing on the devices. "They’re simple to use once you get past the learning curve."
With a simple algorithm called Mountain Sort, the UCSF team was able to track more than 2,000 neurons for at least an entire day, and about 21 percent could be tracked for more than a week. This capability would enable scientists to research changes over longer timescales, potentially making it possible to understand how the brain processes information that can take days, weeks or even months, such as forming, consolidating and retaining memories.
The National Institute of Neurological Disorders and Stroke and the National Institutes of Health (NIH) funded the study, with the stated purpose of gaining a better understanding of the mechanisms of learning.
Contributors included Supin Chen and Kye Lee, formerly of LLNL. Other contributors were Hannah Joo, Jiang Lan Fan, Daniel Liu, Charlotte Geaghan-Breiner and Hexin Liang of UCSF; Alex Barnett, Jeremy Magland and Leslie Greengard of the Flatiron Institute; and Mattias Karlsson and Magnus Karlsson of SpikeGadgets.
The Lab’s Implantable Technologies Group (part of the Center for Bioengineering), led by LLNL engineer Razi Haque, is continuing to help distribute and share the proven LLNL-developed technology with more research groups. This is being accomplished by expanding neural probe arrays for additional animal models, such as primates, as well as collaborating with more neuroscientists at UCSF and elsewhere. This work, funded by a new NIH BRAIN (Brain Research through Advancing Innovative Neurotechnologies) grant, also is led by UCSF’s Frank.
Further enhancements to the technology are aimed at providing neuroscientists with powerful new tools to perform new studies while achieving high-quality, repeatable data. One approach the group is taking is developing multi-modal arrays, adding electrical stimulation capability, optical stimulation and chemical sensing. Another area the group is working toward is enhancing the electrode density to enable much higher channel counts than are available today. Finally, the group will continue to take the experience and knowledge acquired from these programs and apply them to the ultimate goal of developing technologies directed toward enabling long-term human-use implants.
For more, visit the web.
thomas244 [at] llnl.gov