Full size image available through contact |
An interdisciplinary team of scientists, headed by physicists at the University of California, San Diego, has developed a new technique for slicing and imaging brain tissue that makes use of ultra-fast lasers.
This technique, detailed in the July 3 issue of the journal Neuron, provides an important new tool scientists can use to automate and modernize histology, the study of tissues at the microscopic level, as well as to map the production of neurotransmitters and other proteins involved in communication between cells and normal cell function, which are produced in different regions of the brain.
"We can now look at a large portion of the brain and do statistics on the anatomy, because this new technique is able to more precisely measure structures in the brain and do it without the tedious and laborious methods of thin slicing sections of the brain," says David Kleinfeld, a professor of physics at UCSD who headed the study. "It solves a big problem and it potentially changes the way a lot of science is done today."
Until now, to get microscopic images of the brain, scientists have relied on a procedure developed in the early twentieth century that involves manually cutting thin slices of frozen brain, and viewing them through a light microscope. Not only is this painstaking, but freezing the tissue can damage it.
"This project began as a response to a lament from Beth Friedman, a researcher in the neurosciences department at UCSD and the neuroanatomist on the project," recalls Kleinfeld. "Beth felt that current methods to reconstruct the architecture of regions of the brain were simply tedious and inaccurate in comparison with the rapid assays for genes and proteins brought about by the revolution in molecular biology. The technique we have developed makes it possible to take a chunk of tissue and come up with hard numbers on sizes of structures or expression of proteins."
In the new technique developed by the Kleinfeld team, the tissue is imaged and ablated, or vaporized, in successive iterations. The ablation requires a femtosecond laser, which produces light pulses lasting one quadrillionth of a second. This is as short in comparison with a second as a second is in comparison with all of human history.
"The pulse is so short that it doesn't generate heat, which would damage the tissue," according to Philbert Tsai, a graduate student in Kleinfeld's lab and the first author of the study. "Electrons are ripped away from atoms to form a plasma on a time scale too short for the atom cores to vibrate and produce heat."
The layer of tissue removed in each successive laser ablation is less in thickness than that of a human hair. After each ablation, the newly uncovered tissue is labeled with a fluorescent dye, and lower intensity laser light is used to take an image of the surface. Successive snapshots of each layer can be recombined to create a 3-D image of the tissue. While it takes about a day to image the entire brain of a mouse, this procedure can be fully automated.
One particularly valuable application of this technique is to image brains of animals that have been genetically engineered to produce a fluorescent protein in certain cells. By studying 3-D images of the brains of mice that produced fluorescent protein in the blood vessels of the brain, Kleinfeld's team was able to determine the volume of the blood vessels relative to the rest of the brain. Images of brains that express a particular fluorescent protein could also be used to quantify how much of that protein there is in different regions of the brain. This technique is not limited to the brain, either. Other tissues in the body could be imaged with the femtosecond laser. The only disadvantage is that the sample is destroyed in the process, so it would not be suitable if, for example, a clinician wanted to preserve a sample for later reference.
Kleinfeld credits the open atmosphere among UCSD's different departments and divisions for the success of the project. "Things are free here, but at a lot of schools people can be kind of territorial," he adds.
Indeed, the project drew together researchers with a wide range of expertise. Jeffrey Squier, a physicist and laser expert in the chemistry department at UCSD and now at the Colorado School of Mines, played an essential role from the beginning. Other researchers involved in the study include Varda Lev-Ram of UCSD's department of pharmacology; Chris Schaffer of UCSD's department of physics; Roger Tsien of UCSD's departments of chemisty and pharmacology and the Howard Hughes Medical Institute; Qing Xiong of HHMI; and Agustin Ifarraguerri and Beverly Thompson of Science Applications International Corporation.
The project was funded by the National Institute of Health's National Institute of Neurological Disorders and Stroke, the National Science Foundation and the David and Lucile Packard Foundation. The university has applied for a patent on the new technique.
Journal
Neuron