Cliff Slaughterbeck, Ph.D., is showing off some of his babies.

“Nobody sells this whole thing off the shelf,” he said, gesturing at a microscope housed in a large black box, one of eight identical rigs set up in the Allen Institute’s electrophysiology lab, where neuroscientists use impossibly thin electrodes to listen in on neurons’ electrical chatter. The rigs combine commercial microscopes with several in-house-made pieces and software programs that adapt the scope to the specialized experiments run in this lab.

In the 1990s, Slaughterbeck, a former physicist, built a special kind of microscope to study the surface of ice. Now, he’s a principal engineer on the Allen Institute’s engineering team, working on equipment and software that allow the Institute’s scientists to carry out their work.

On a recent Monday, Slaughterbeck took me and Allen Institute Digital Media Specialist Madeline Burchard on a whirlwind tour of some of the machines that neuroscientists are using to solve the mysteries of the brain.

From a special microscope that lets scientists record electrical activity from up to eight neurons at once to a new high-tech brain slicer under development, these contraptions allow their users to get more data, faster. For 2023’s Brain Awareness Week, we learned the ins and outs of four of these mechanical Allen Institute team members.

Meet the machines.

Octo-patch

At the back of the electrophysiology lab, behind the rows of eight identical rigs, sit two huge black boxes. Each holds a machine known as a multipatch rig, also dubbed octo-patch by the scientists who use them. They look like beefed up versions of the single electrophysiology rigs; each microscope is surrounded by eight chunky mechanical limbs.

The “patch” in octo-patch refers to patch-clamp, a neuroscience technique that uses a tiny glass needle with an even tinier electrode inside to “patch” onto the edge of a neuron, using that electrode to both stimulate and record electrical activity from a single brain cell. Octo-patches, as their name implies, can record from up to eight neighboring neurons at once.

These giant rigs were recently used to catalog more than 1,700 connections between neurons in the mouse and human brain, using one of the eight limbs to supercharge a neuron with an electrical current and see which of seven other nearby neurons respond. Manipulating the eight probes to “patch” onto eight different neurons is tedious and technically challenging – even a specialist in the technique can only conduct about three full experiments a day.

Now, the machines are being used in a slightly different context. On the day we drop in, neuroscientists Travis Hage, Ph.D., and Jessica Trinh are using two to four of the machines’ eight limbs in their work. They’re looking at the decay time of the electrical signals in certain neurons after they’re stimulated, with the goal of understanding how a single gene affects the speed of neurons’ recovery after it fires an electrical signal.

LASSO

The first thing you need to know about LASSO, an in-development automated brain tissue slicer, is that it’s run with an Xbox controller. This isn’t an Xbox-brain science crossover – but how cool would that be? The gaming controller is a stop-gap on the way to true automation LASSO’s developer, Dan Bumbarger, Ph.D., is quick to point out. Eventually the machine will be able to move, slice and grab on its own.

LASSO, which stands for Loop-based Automated Serial Sectioning Operation, lives in the electron microscopy lab at the Allen Institute. Using a knife made out of literal diamonds, LASSO’s job is to slice a piece of brain tissue into inconceivably thin pieces — LASSO’s older counterpart cut a grain-of-sand sized piece of mouse brain into 25,000 slices, each of which needed to be imaged consecutively. These thin slices are imaged using electron microscopes, which capture images of their subjects at incredibly high resolution, allowing the scientists to image detailed shapes of brain cells and even smaller structures inside the cells, as well as all their connections in the piece of tissue under study.

And yes, there’s a picture of beloved TV character Ted Lasso taped to LASSO.

If that all sounds thrilling — diamond knives! Thousands of brain slices! — know that there’s also a fair amount of tedium baked into the work. The previous 25,000-slice project required two trained scientists at a time to babysit the slicing machine around the clock, making sure the slices came off the machine and onto the imaging tape in consecutive order. Mapping the cells and structures in that tiny piece of brain requires putting all the slice images back together, in order, like a weird science puzzle.

When we drop in to see LASSO in action, scientists Steven Cook, Ph.D., and Jenna Schardt are literally watching water dry. The machine drops its tiny brain slices into a dish of water, and then a thin metal loop (manipulated with the Xbox controller) grabs a small drop of water containing the newly cut slice and deposits it onto the imaging surface, which is sitting on top of a heated metal block. The water droplet dries down in the heat, leaving the slice in its correct spot.

Once the automated part of LASSO is in place, the machine will use computer vision to guide the metal loop to the correct spot to pick up and deposit the slices of brain. It’s not just about saving the scientists from needing to work through the night, it’s also about democratizing the technology, Cook said.

“These kinds of machines are out of reach for a lot of labs,” he said. “We want to build a less user-intensive experience and a less financially intensive experience.”

VERSA

The Allen Institute’s imaging lab has had a few updates since I last visited. The researchers in this lab make beautiful high-resolution photos of neurons and brain slices, images that are used by other teams at the institute to recreate and study the details of neurons’ 3D shapes. New in the lab is a VERSA slide loader and scanner, a machine that helps automate some of the team’s work in placing slides and capturing images.

Neuroscientist Zoe Juneau (she/they) oversees VERSA’s work. This machine is the only completely off-the-shelf equipment we’ve visited on our tour, although new software was written in-house to integrate the files from this new microscope into the Institute’s image processing and storage system. VERSA can grab, sort and scan dozens of microscope slides at a time, allowing the scientists to set up stacks of slides to be imaged and let the machine do its job overnight or on the weekend. The Scanning Lab’s small team produced more than 5,000 scans last year, so automation is a great help.

On the day we drop in, Juneau is using VERSA to take photos of full brain slices where many different cells are labeled in glowing, fluorescent colors. These images will go to experiments where scientists need detailed information about which kinds of cells get those fluorescent tags and which don’t. VERSA takes an initial scan of the slice and then Juneau selects different points on that low-res scan that are important focus areas.

“It’s creating a topographical map of this brain slice,” she said. “Like the autofocus on your phone, you have to tell the camera to focus on the right thing.”

Engineering: Behavior upgrade

The Allen Institute’s engineering lab is not exactly easy to find. Tucked away down a hallway behind the elevators sits a maker’s dream of a space, every wall lined with dozens and dozens of neatly labeled boxes of hardware and tools.

“It’s like the Mythbusters lab!” Burchard exclaimed when we peeked inside.

“Actually, that show’s a huge part of the reason I’m doing this job,” said Heston Smith, system design engineer at the Allen Institute.

On the day we pop in, Smith is working on upgrades to what looks like a stack of small black lockers — four boxes with doors that close to block out distracting light and noise. These boxes are built to house mice being trained for experiments in the MindScope Program, where scientists aim to understand the links between electrical activity in the brain and animal behavior. (Mice are nocturnal and more active in the dark — hence the light-blocking.)

Previous experiments relied on visual cues — scientists would show the animals different pictures or movies and train them to lick a waterspout when a certain image popped into view. Now, Smith is adding speakers and extra cameras to the boxes that will let scientists train the animals to respond to different sounds, part of a project to understand the brain’s response to auditory cues.

What kinds of sounds do the mice hear? we asked. The system isn’t hooked up to play them for us right now, Smith said, but the experiments will play tones — just your basic beeps and boops.

Smith joined the institute two years ago from a previous job designing fuel lines and filters for long-haul trucking. Now, he spends his days designing parts of training suites for mice and other experimental equipment.

In an era where we’re constantly being told that machines are coming for our jobs, it’s somewhat comforting to know there’s still a human force behind those machines.

On that note, I decided to ask the infamous artificial intelligence bot ChatGPT to finish this article for me.

Me: ChatGPT, please write an ending to this article

ChatGPT: Overall, the machines being used by neuroscientists at the Allen Institute are allowing researchers to collect more data, faster, and to better understand the complexities of the brain. From the octo-patch, which can record from up to eight neighboring neurons at once, to LASSO, the in-development automated brain tissue slicer, these cutting-edge technologies are making significant contributions to the field of neuroscience.

Me: write a shorter ending than that

ChatGPT: Got it! How about: “That’s it, done!”

There you have it.

‍