Machines are quite comfortable with human cells. Embeddable sensors record how and when neurons fire; neuron-like devices could even boost faster regrowth after being implanted in the brain; and electrodes spark heart cells to beat or brain cells to fire.
In the near future, so-called brain-machine interfaces could achieve even more: deliver a blueprint to design artificial intelligence, monitor and treat symptoms of neurological disorders like Parkinson’s disease, or even facilitate brain-to-brain communication.
To realize the reachable and the quixotic, devices require a way to actually dive deeper into human cells to carry out reconnaissance. Understanding how neurons work in a better manner, will help scientists emulate, reproduce, and treat them with machines.
Currently, in a paper published in Nature Nanotechnology, Charles M. Lieber, the Joshua and Beth Friedman University Professor, provides an update to his original nanoscale devices for intracellular recording, the first nanotechnology developed to record electrical chatter within a live cell. Nine years later, Lieber and his team have formulated a way to create several of these devices at once, producing a nanoscale army that could accelerate efforts to discover what is taking place inside human cells.
Before Lieber’s work, analogous devices faced a Goldilocks conundrum: Very big, and they would record internal signals but destroy the cell. Very small, and they failed to pierce the cell’s membrane—recordings ended up noisy and vague.
Lieber’s new nanowires were spot on. Engineered and reported in 2010, the originals had a nanoscale “V” shaped tip with a transistor at the bottom of the “V.” This design could penetrate cell membranes and transmit accurate data back to the team without damaging the cell.
However, there was an issue. The silicon nanowires are much longer than they are wide, making them unsteady and hard to wrangle. “They’re as flexible as cooked noodles,” says Anqi Zhang, a graduate student in the Lieber Lab and one of the authors on the team’s recent work.
To develop the original devices, lab members had to trap nanowire noodles one by one, find each arm of the “V,” and then knit the wires into the recording device. A few devices took two to three weeks to create. “It was very tedious work,” says Zhang.
But nanowires are not created one at a time; they are made collectively like the very things they look like: cooked noodles. Using the nanocluster catalyzed vapor-liquid-solid technique, which Lieber used to make the first nanowires, the team shapes an environment where the wires can grow on their own. They can pre-define each wire’s length and diameter but not how the wires are placed once ready. Even though they grow thousands or even millions of nanowires at a time, the final result was a mess of invisible noodles.
To unravel the mess, Lieber and his team engineered a trap for their loose cooked noodles: They form U-shaped trenches on a silicon wafer and then comb the nanowires across the surface. This “combing” process unravels the mess and positions each nanowire into a neat U-shaped hole. Then, each “U” curve gets a miniature transistor, akin to the bottom of their “V” shaped devices.
With the “combing” technique, Lieber and his team create hundreds of nanowire devices in the same amount of time taken to make just a few. “Because they’re very well-aligned, they’re very easy to control,” Zhang says.
Thus far, Zhang and her colleagues have used the “U” shaped nanoscale devices to record intracellular signals in both cardiac and neural cells in cultures. Coated with a substance that copies the feel of a cell membrane, the nanowires can cross this barrier with the least effort or harm to the cell. Moreover, they can record intracellular chatter with the same level of accuracy as their biggest contender: patch clamp electrodes.
Patch clamp electrodes are around 100 times bigger than nanowires. As the name indicates, the tool clamps down on a cell’s membrane, causing permanent damage. The patch clamp electrode can trap stable recording of the electrical signals within the cells. But, Zhang says, “after recording, the cell dies.”
The “U” shaped nanoscale devices created by Lieber team are friendlier to their cell hosts. “They can be inserted into multiple cells in parallel without causing damage,” Zhang says.
At present, the devices are so flexible that the cell membrane prods them out after approximately 10 minutes of recording. To stretch this window with their next design, the team may incorporate a little of biochemical glue to the tip or roughen the edges so the wire catches against the membrane.
The nanoscale devices possess another benefit over the patch clamp: They can record additional cells in parallel. With the clamps, scientists can gather just a few cell recordings per session. For this research, Zhang recorded up to ten cells at once. “Potentially, that can be much greater,” she says. The more cells they can record at one time, the more they can observe how networks of cells interrelate with each other as they do in living creatures.
In the process of expanding their nanowire design, the team also managed to confirm a long-established theory, known as the curvature hypothesis. After Lieber invented the first nanowires, scientists wondered if the width of a nanowire’s tip (the bottom of the “V” or “U”) can influence a cell’s response to the wire. For this research, the team tested with numerous “U” curves and transistor sizes. The results established the original hypothesis: Cells prefer a narrow tip and a small transistor.
The beauty of science to many, ourselves included, is having such challenges to drive hypotheses and future work.
Charles M. Lieber, the Joshua and Beth Friedman University Professor, Harvard University
With the scalability challenge overcome, the team hopes to trap even more precise recordings, maybe within subcellular structures, and record cells in living creatures. But for Lieber, one brain-machine challenge is more alluring than all others: “bringing cyborgs to reality.”