This article, and the one accompanying it, were written by Seattle Post-Intelligencer business reporter Warren Wilson. Originally published Nov. 20, 1995, they appear here courtesy of the Seattle Post-Intelligencer.
Chromosomes, genes, DNA strands, enzymes. Over the past few decades scientists have begun to catalog the ingredients in the nucleus, a cell's vital core.
But how those ingredients are organized is still mostly a mystery.
"Is the nucleus like a shopping bag, where all the chromosomes and genes kind of rattle around?" asks biochemist Dr. John Sedat. "Or do you park this gene in this place, that gene in that place?"
Sedat and protein crystallographer Dr. David Agard, colleagues at the University of California -San Francisco, have devoted a good share of their careers to answering such questions.
"We're trying to understand the dynamics of biochemistry on a kind of cell-by-cell basis,"
Sedat says. The goal is to know "not only where your molecules are (but also) how much of this is
here, how much is moving over there, at what rate, and the consequences."
To reach that goal, they developed a new microscope, now called the DeltaVision, that is not so much an invention as a collection of inventions each a stunning advance from where things stood when they began.
One of the innovations is Agard's deconvolution process, which addresses a key challenge with any light-based microscope: how to handle blurring caused by out-of-focus light from above and below the target section. The problem is similar to one photographers face in controlling depth of field the larger the aperture, the shallower the zone in focus.
The approach is straightforward. Put a simple object such as a tiny spherical bead under the microscope, record the blurred image, and develop a formula that describes how the light has been scattered. Mathematically invert the formula and apply it to the bead image to refocus or "deconvolute" it.
Here's the payoff: Because the distortion follows a consistent pattern, the deconvolution for mula will sharpen a complicated shape just as it does the image of the bead.
In practice, of course, that process is complex enough to require substantial computer power power that was prohibitively expensive 15 years ago.
Just two years ago, a state-of-the-art computer required three hours to fully process an image, says Ron Seubert, one of the founders of Applied Precision, the Mercer Island company that helped develop and commercialize Sedat and Agard's system.
"Today, a computer that sells for about one-half the price does the same computation in about one hour," Seubert said.
Deconvolution is only half the microscope's computer-intensive process, however. The other half is the task of assembling the three- to five-dozen two-dimensional images into a final, three -dimensional version.
To gather the individual sections, the microscope must move the object through a series of tiny steps each exceedingly precise, if the result is to be accurate.
The solution came from Applied Precision, which had dealt with similar motion-control issues in developing test equipment for the semiconductor industry.
Seubert said that when he and partners Don Snow and John Stewart couldn't find a commercial device exacting enough for their needs, they simply invented their own. Named the "Nanomover," it can move an object in precise steps as small as 10 nanometers. For comparison, a typical human hair is 100,000 nanometers thick.
When Sedat and Agard heard about the device, they purchased one and found it so well-suited to their system that they eventually approached Applied Precision about engineering the microscope for commercial sale.
Just as important as the hardware improvements, Sedat said, have been dramatic improvements in fluorescent "probes," chemicals that let the researcher zero in on specific components inside the cell.
As Sedat and Agard developed their system, they eventually found it in such great demand that they were being distracted from their biology research.
"There were a lot of people on our backs to acquire one of these things," Sedat says. Each microscope had to be assembled individually, but the process is complicated and requires skills that most biologists don't have.
"It would always work out that I would have to hand-make them," Sedat says.
The two were also under some pressure from federal agencies such as the National Institutes of Health to make the technology more widely available to researchers, Sedat says.
"The goal was not a monetary one," he says. Applied Precision paid a small licensing fee to the University of California and pays a small royalty on each system it sells, but Sedat and Agard don't see any of it.
"The main goal was to get it out and into an arena where people could use the technology," he says. "It was recognized early on that just the hardware was reasonably expensive, and to charge a prohibitive fee would make it that much more inaccessible to people."
In 1991 they approached Applied Precision about building a commercial version. It took a couple of years to evaluate the idea, and in 1993 the company decided to go ahead.
The unit at the Fred Hutchinson Cancer Research Center was installed in 1994 and is the only one in the Northwest. Prices range from $130,000 to $200,000 depending mainly on how powerful a work-station computer is included.
Of course, with so many advances occurring in so many fields, no single system the DeltaVision included is the last word in microscopy.
Baylor College of Medicine researcher Dr. Michael Mancini faults its lengthy processing time, and says its computer interface is unnecessarily difficult to use. But further gains in computer power will bring faster output, and Seubert says Applied Precision is eager to listen to people such as Mancini to help refine the interface.
Though the time lag can be a serious drawback, the images themselves are clearly superior to those of the closest competitor, called a confocal microscope, says Steven Smith, a professor of molecular and cellular physiology at Stanford Medical School.
"For the highest resolution, most beautiful and most quantitative images . . . you get the best results with digital image restoration," he says. "There's pretty much no question about it."
UW researcher Dr. David Battaglia said the DeltaVision represents an advance of the same order of importance as electron microscopy and fluorescence microscopy.
"It's not just an increment, it's a big change," he says. "It's a new way of thinking about how to capture and work with images."
| Section Head | Front Page | Contents | Search |