Slicing Brains DIY

Is Ken Hayworth Building the Instruments for Uploading Brains?

Here at h+ Magazine, we love DIYers and garage entrepreneurs who work in the “NBIC” (nano, bio, info, and cogno) fields. We recently visited Harvard University’s Center for Brain Science where we came across a particularly inspiring example.

In order to understand the brain’s function, neuroscientists must be able to map out its basic neuronal circuits. A neuronal circuit (for example, a thalamocortical circuit) typically spans quite a large volume of brain tissue (tens of cubic millimeters). At the same time, the axonal and dendritic processes comprising such a circuit are so fine and so tortuously interconnected that only electron microscopy of ultrathin (50nm) serial sections can resolve their connectivity. No current techniques can image so large a volume of tissue at such fine resolution. While working as a computer scientist and engineer at the Jet Propulsion Laboratory in Pasadena, Ken Hayworth says that he was inspired by a paper he read by Xerox PARC’s Ralph Merkle on “Large Scale Analysis of Neural Structures.” In 1989, Merkle wrote that we should be able to reverse engineer the human brain in the near future using advanced electron microscopy, and new slicing, staining and computational analysis algorithms.

Hayworth, a cryonicist (he’s signed up with Alcor to be cryopreserved upon his death in the hopes of being repaired and revived by a future technology), immediately saw that understanding the circuitry of the brain was a necessary first step towards mind uploading. His interest in developing new neuroanatomical mapping instruments stems from his frustration with the lack of neuroanatomical knowledge currently available on the circuits and systems of the higher-level visual system in mammals.

So Ken started applying to grad schools and interviewing with neuroscientists about his plans to build a high-throughput brain-slicing machine. Time after time, he was rejected: either because he didn’t have the academic background they were expecting, or because the professors wanted a lab slave and not someone with his own plans. Finally, he was offered admission by a professor at U.S.C. conducting psychophysical research in human vision. Fascinated by the topic, Hayworth accepted, even though it meant putting off his brain-slicing experiments.

After his second year of grad school, Hayworth started working on his brain-slicing idea again. As he began attending neuroscience conferences and telling people about his machine, he faced ridicule (“that can’t be done”) and scorn for focusing on equipment rather than the brain itself. Unperturbed, Ken built a prototype that demonstrated the concept sufficiently well to attract the interest of a Harvard professor interested in high-throughput neural circuit mapping. Together they submitted a proposal to the McKnight Endowment Fund to bring his project to the next level. Much to his surprise, he won a $200,000 grant and with it a chance to work at the Center for Brain Science at Harvard University. There he built the ATLUM (Automatic Tape-collecting Lathe Ultramicrotome), the first machine to demonstrate the fully automatic collection of brain slices thin enough to view in an electron microscope and thus image at nanometer resolution. No one is laughing now. Work on ATLUM2 is reaching nearly $1 million in cost, and Ken is getting requests from neuroscientists worldwide who see the value of obtaining such a machine to build a library of brain slices for examination and scanning using an automated system.

The ATLUM2 Machine

Today only tiny volumes of neuropil (the Free Dictionary: “the complex network of unmyelinated axones, dendrites, and glial branches that form the bulk of the central nervous system’s grey matter and in which nerve cell bodies are embedded”) can be traced at electron microscopy resolution using painstaking manual techniques. Mapping the neuronal network of the nematode worm C. Elegans was a decade-long Herculean task, even though it is less than 0.01mm_³ in volume. Hayworth wants to use his device to cut an entire mouse brain into such thin slices that an electron microscope can scan virtually all of the structures within the tissue. As small as a mouse brain is, at 30nm thickness, that would require approximately 2 million slices! To accomplish this, the mouse brain would be chemically fixed and embedded in plastic. Then, using a special slicing device which Ken is also designing, the plasticized brain would be cut into approximately 400 sub-blocks. Each sub-block would then be loaded into Ken’s ATLUM2 and it would be cut into 5,000 incredibly thin slices. Each slice would be picked up on a long carbon-coated tape for later staining and imaging in a scanning electron microscope (SEM). Because the process is fully automated, volumes as large as tens of cubic millimeters (large enough to span entire multi-region neuronal circuits) can be quickly and reliably reduced to a tape of ultrathin sections.

Ken’s original ATLUM machine has already collected over 1,000 sections of embedded mouse cortex, each 30nm thick and 1mm x 5mm in area. SEM images of these ATLUM-collected sections can attain lateral resolutions of 5nm or better, which is sufficient to image individual synaptic vesicles and to identify and trace all circuit connectivity.

Following collection, the ATLUM tape can be stained with heavy metals (or other markers), and cut into shorter lengths. This allows rows of sections to be attached to large (200mm diameter) imaging plates that can be loaded into a standard SEM for automated random access imaging of any location within any of the hundreds of sections on the plate’s surface. A few dozen of these plates could hold an entire 10mm_³ volume representing an incredible 1x10_¹⁶ voxels of raw image data.

Bulk imaging of such a large volume at the highest resolution would take hundreds of years, but having the ultrathin sections laid bare on a set of tissue plates solves this problem. Luckily, a researcher can quickly produce a lower-resolution image set of the entire volume, setting up a unified coordinate system for the sample volume and plates, and then use robotic loading and positioning of plates to zoom in on any part to obtain the highest resolution SEM images. In the near future, multi-beam scanning electron microscopes will shorten the time it takes to scan such samples in high resolution, and efforts similar to the Human Genome Project, whereby hundreds of sequencing machines were used at once, might allow researchers to use hundreds or thousands of electron microscopes simultaneously to map entire brains within days rather than years.

Once Ken’s ATLUM2 is perfected, it will hopefully go into production and copies of the device will find their way into neuroscience labs around the world. At first, these devices will be used to section areas of the brain that are of particular interest to the individual researchers. The circuitry could be used to emulate those brain functions, run experiments emulating a brain section, and possibly even test pharmaceuticals or therapies. In the future, we might understand brain circuitry so well that such devices could be used to scan and “upload” an individual’s mind to any type of substrate (a new body, robot, or artificial environment). This Matrix-like immortality would be the ultimate backup of ourselves.