Summary: Harvard researchers have created a silicon chip that records intracellular signals from thousands of neurons in parallel, enabling the mapping and characterization of more than 70,000 synaptic connections from roughly 2,000 rat neurons. Using an array of 4,096 microhole electrodes, the team achieved high-sensitivity recordings that reveal not only which neurons connect, but also the strength and type of those connections—an advance beyond visual-only methods such as electron microscopy.
This microhole electrode array functions like millions of tiny patch-clamp tools integrated on a single chip, producing intracellular access and capturing faint synaptic events across a broad network. Compared with the group’s earlier nanoneedle array, the microhole design delivered dramatically higher coupling rates and data quality, enabling the extraction of hundreds of times more synaptic links and supporting classification of connections by electrical versus chemical, and by inhibitory versus excitatory strength.
The technology opens new possibilities for large-scale functional mapping of neuronal circuits and could become a powerful platform for studying brain function, development, and disease. The researchers are now focused on adapting the approach for use in living brains to observe synaptic communication in real time.
Key Facts
- Massive synaptic mapping: The device recorded over 70,000 putative synaptic connections from approximately 2,000 neurons, far exceeding previous experimental scales.
- High intracellular coupling: The microhole electrodes achieved an average intracellular coupling rate near 90% (about 3,600 of 4,096 electrodes), producing high-fidelity signals suitable for detailed analysis.
- Classification of connections: Recorded synapses were catalogued into electrical synapses and chemical synapses (inhibitory, weak/uneventful excitatory, and strong/eventful excitatory), with an estimated overall error rate of roughly 5%.
- Path toward in vivo mapping: Ongoing work aims to realize versions of the system suitable for deployment in live animal brains to track real-time network dynamics.
Source: Harvard
Harvard engineers and neuroscientists have combined microfabrication and integrated electronics to scale intracellular neuronal recording to an unprecedented level.
Higher-order brain functions emerge from patterns of connectivity among neurons. To understand how networks compute and behave, scientists need not only anatomical maps that show which neurons touch each other, but also functional maps that measure the strength, polarity, and dynamics of those synaptic connections.
Traditional electron microscopy produces highly detailed structural maps of synapses but does not report synaptic strength or functional influence. The patch-clamp technique, by contrast, provides sensitive, intracellular recordings that reveal faint synaptic currents and allow direct measurement of connection strength—but it has historically been slow and limited to one or a few neurons at a time.
To bridge that gap, the team led by Donhee Ham at the Harvard John A. Paulson School of Engineering and Applied Sciences developed a complementary metal-oxide-semiconductor (CMOS) chip patterned with 4,096 platinum/platinum-black microhole electrodes. Rat neurons cultured on the chip formed interfaces with the microholes; controlled, gentle current injections through integrated electronics provided intracellular access across the array, while the same circuitry recorded the resulting signals.
Co-lead authors Jun Wang and Woo-Bin Jung oversaw the microhole electrode design, chip fabrication, electrophysiological recordings, and data analysis. The microhole geometry mimics the aperture of a patch-clamp pipette and couples more effectively to neuron interiors than the team’s prior vertical nanoneedle electrodes. In addition to improved coupling, the microhole array is simpler to fabricate, increasing its practical utility.
Performance exceeded expectations: on average, roughly 90% of the 4,096 electrodes achieved intracellular coupling. From the network-scale intracellular recordings, the researchers identified more than 70,000 plausible synaptic connections—about 200 times the number they previously extracted from the nanoneedle device. The improved signal quality allowed the group to classify each connection by type and strength.
The integrated electronics were essential: they delivered carefully controlled currents to open membrane interfaces without excessive damage while simultaneously amplifying and digitizing intracellular voltages. That combination of gentle interfacing and high-throughput acquisition yielded dense, analyzable data across a cultured neuronal network.
Analyzing such large datasets posed a major challenge. The team developed pipelines to detect synaptic events, infer connectivity, and estimate classification confidence. Their analysis produced a catalog of electrical and chemical synaptic links and estimated the mapping error rate at about 5%—a level that supports meaningful biological interpretation and further methodological refinement.
Looking ahead, the researchers are designing next-generation devices that can operate in vivo to capture functional connectivity in live brains, tracking how synaptic networks change over time and in response to behavior or disease. Such capabilities could accelerate research into learning, memory, neurodevelopmental disorders, and neurodegeneration.
Paper co-authors include Rona S. Gertner (Department of Chemistry and Chemical Biology) and Hongkun Park (Mark Hyman, Jr. Professor of Chemistry and Professor of Physics). Funding for the research was provided by the Samsung Advanced Institute of Technology of Samsung Electronics.
About this research coverage
Author: Anne Manning
Source: Harvard
Contact: Anne Manning – Harvard
Image credit: Neuroscience News
Original research: Closed access. Title: “Synaptic connectivity mapping among thousands of neurons via parallelized intracellular recording with a microhole electrode array” by Donhee Ham et al., published in Nature Biomedical Engineering.
Abstract
Synaptic connectivity mapping among thousands of neurons via parallelized intracellular recording with a microhole electrode array
Massively parallel intracellular recording that captures synaptic signals across a neuronal network enables mapping and characterization of synaptic connections at scale. Historically, mapping efforts have been limited to roughly 300 synaptic connections. Here, a 4,096 platinum/platinum-black microhole electrode array fabricated on a CMOS chip enabled parallel intracellular recording in rat neuronal cultures, reaching a 90% average intracellular coupling rate. Network-wide recordings produced abundant synaptic signals, from which the team extracted more than 70,000 plausible synaptic connections among over 2,000 neurons. Connections were catalogued as electrical synapses and as inhibitory, weak/uneventful excitatory, and strong/eventful excitatory chemical synapses, with an estimated overall error rate near 5%. This scale of functional mapping and the ability to characterize connection types and strengths represent a major step toward comprehensive functional connectivity mapping of large-scale neuronal networks.