For Winfried Denk, spending time at Janelia Farm is a bit like coming home. For close to nine years, Denk was principal investigator at Bell Laboratories in Murray Hill, New Jersey—one of the research institutes the Janelia campus was modeled after. And one of his former postdocs from that time, Karel Svoboda, is now a group leader at Janelia.
Denk will take time from his current post at the Max Planck Institute for Medical Research in Heidelberg, Germany, to discuss projects with researchers at Janelia, spending about four weeks of the year at the Virginia campus. “At Janelia interactions are freewheeling,” says Denk. “There is more time to do that because groups are smaller so less time is spent doing supervisory things. You have more time to think and talk to other PIs.”
This type of structure, Denk believes, can foster breakthroughs in biomedical technologies. “This is the perfect environment. The resources are here to support research that is risky or, more commonly, referred to as crazy,” he says, laughing. “There is enough of a core of physicists to talk shop with and many biologists who can tell us what they need. If you don’t want to develop ‘bells and whistles’ technology, but rather technology that will help push a field forward, you need frequent feedback from those working in the field.”
Denk has had his share of breakthroughs. As he was completing his Ph.D. at Cornell University, on measuring the motion of inner-ear hair cells, he began to dabble in a different project “almost as a hobby,” he says. The technique he developed, two-photon microscopy, has since revolutionized how cells inside tissues are visualized.
Before this advance, biologists primarily had used the confocal microscope for imaging—an instrument that scans a focused beam of light across a sample containing some molecules labeled with a fluorescent dye. When a dye molecule absorbs a photon of light, it becomes excited and then emits light of a different wavelength that passes through a pinhole and is then measured by a detector. However, small structures can be difficult to visualize because so much of the sample fluoresces at the same time and light from the structure is scattered and lost at the pinhole. Also, the short-wavelength light used by the microscope can damage cells and tissues.
The idea behind two-photon microscopy is that two, instead of one, photons of long wavelength are needed to excite fluorescence. A dye molecule absorbs the two photons at almost the same time, kicking the molecule into an excited state, which then emits light. Because the probability of a near-simultaneous hit by two photons is low, except near the beam focus, there is little or no excitation outside the focus or by scattered light.
To build the microscope, Denk and his colleagues altered a scanning confocal microscope by taking out the confocal pinhole and attaching instead a laser that delivers short pulses of light, built by another group at Cornell. In 1990, an article published in Science magazine by Denk and his colleagues showed that two-photon microscopy could be used to map the distributions of various molecules in cells in three dimensions.
After a stint as a postdoc at IBM Research Lab in Rueschlikon, Switzerland, Denk joined Bell Labs, where he applied two-photon microscopy to various biological problems. One of the first had to do with how networks of neurons relay their signals. In collaboration with a postdoc from David Tank's lab, Rafael Yuste (now an HHMI investigator at Columbia University), Denk imaged dendritic spines—the protrusions of neurons that connect to other neurons—and measured how quantities of calcium, a critical player in signaling, accumulate and then wane in relation to neuronal activity.
This and other studies clearly established that two-photon microscopy is a powerful technique for imaging molecules, typically far outperforming other techniques at depths greater than 100 microns below the surface of a tissue sample. “This is critical when you are working with brain tissue, because if you cut the tissue to make the specimen thinner and increase resolution, you lose the connectivity between nerve cells,” explains Denk. He later showed that two-photon microscopy is also ideal for imaging tissues, such as the retina, that are sensitive to light.
After leaving Bell Labs for the Max Planck Institute in Heidelberg in 1999, Denk started looking for something new on which to work. “We had been mapping the signals through calcium. But to interpret the signal you have to know where the wires come from that deliver the signal to cells,” says Denk.
Because the resolution of traditional confocal microscopy and two-photon microscopy was not sufficient for following the network of connections, or wires, among neurons, Denk turned to the scanning electron microscope—a powerful microscope that scans a beam of electrons on the surface of a specimen and then measures the electrons that are scattered back to produce an image. Denk equipped this microscope with a tool for cutting thin slices of tissue. As each slice is cut away and discarded, the electron beam is scanned over the remaining block face to build three-dimensional images. “We think it is possible to use this technique to track neuronal processes across large distances,” he says.
This latest foray into neuroscience takes Denk back to the days when, as a teenager, he took apart radios and followed the wires to figure out how the instruments worked. “I was always drawn to understanding the function of things. It came naturally to me,” he says. “In a way, the research I am doing now is the most natural extension of doing electronics as a kid.”