A new microscope invented by scientists at Howard Hughes Medical Institute’s Janelia Farm Research Campus will let researchers use an exquisitely thin sheet of light -- similar to that used in supermarket bar-code scanners -- to peer inside single living cells, revealing the three-dimensional shapes of cellular landmarks in unprecedented detail. The microscopy technique images at high speed, so researchers can create dazzling movies that make biological processes, such as cell division, come alive.
The technique, called Bessel beam plane illumination microscopy, is described in a research article published online on March 4, 2011, in the journal Nature Methods.
"The question was, is there a way of minimizing the amount of damage you're doing so that you can then study cells in a physiological manner while also studying them at high spatial and temporal resolution for a long time?" Betzig says.
Long before arriving at Janelia Farm in 2006, Betzig began thinking about ways to improve live-cell microscopy. He put those thoughts on hold while he focused on designing new microscopy techniques that would ultimately shatter the limits of spatial resolution (imposed by the laws of diffraction). Until recently, microscopes could see objects no smaller than 200 nanometers in size. Several years ago, Betzig and his Janelia Farm colleague Harald Hess invented photoactivated localization microscopy, PALM, which can produce images of objects only 10-20 nanometers in size.
PALM and most other microscopes — even the ones college students use in their biology classes —work by exposing the sample through one objective lens and then collecting the light that comes back through that same lens. That approach causes light to damage the sample and induces blur, making it difficult to observe live cells.
In 2008, Betzig began working on ways to overcome these challenges. One idea he had was to use plane illumination microscopy. First proposed about 100 years ago, plane illumination involves shining a sheet of light through the side of the sample rather than the top. To do that, microscopists use two different objective lenses that are perpendicular to one another. "Because you come from the side, plane illumination confines the excitation much closer to the part that's in focus," Betzig says.
Although other researchers, including Janelia Farm Fellow Philipp Keller, have used plane illumination to great effect to study multicellular organisms hundreds of microns in size, the light sheets were still too thick to work effectively for imaging within single cells only tens of microns in size. The main problem is that the wide swath of light used in plane illumination exposed more of the cell than Betzig's group wanted. This caused excessive blur and light toxicity. To circumvent this problem, his group used a Bessel beam, a special type of non-diffracting light beam studied by physicists in the late 1980s, and used today in applications including bar-code scanners in supermarkets. Sweeping the beam across the sample creates a thinner light sheet, his group found.
Bessel beams behave a bit strangely, though, and this is what has kept Betzig's postdoctoral researchers — Thomas Planchon and Liang Gao — busy over the past few years. Although they produce a very narrow light beam, Bessel beams also create somewhat weaker light that flanks the focal point, making the pattern of illumination look like a bull's eye. The extra light lobes are a hindrance because they excite too much of the sample. To compensate for this problem, Betzig's group used two tricks. The first is a concept called structured illumination, where instead of sweeping the beam continuously, they turned it on and off rapidly, like firing a machine gun. This creates a periodic grating of excitation that can be used to eliminate any out-of-focus blur. (Structured illumination, used by Janelia Farm Group Leader Mats Gustafsson, is also one way of achieving super-resolution.)
Another strategy Betzig's group used is two-photon microscopy, a method commonly used in neuroscience to visualize thick pieces of brain tissue. One of the advantages of two-photon microscopes is that very little fluorescence signal is generated from weakly exposed regions. Thus, when they applied two-photon methods, the background from the Bessel side lobes was eliminated, and all that remained was the light from the narrow central part of the Bessel beam.
They then set out to image as fast as possible. The Bessel beam sweeps quickly through the sample, allowing the group to take nearly 200 images/second and build three-dimensional stacks from hundreds of two-dimensional images in one to 10 seconds. As they had hoped, they found that they could take hundreds of such three-dimensional image sets without harming the cell, generating amazing movies of cellular processes such as mitosis, where chromosomes divide as one cell becomes two. "There's no other technique that comes close to imaging as long with such high spatial and temporal detail," Betzig says.
Last summer, as soon as they got their first live cell images, Betzig, Planchon and Gao packed up the new instrument in a rented sport utility vehicle and took it to the Woods Hole Marine Biological Laboratory in Massachusetts for the neurobiology course, where they worked with co-authors Jim and Cathy Galbraith from the National Institutes of Health. "We learned a lot about what works and what doesn't and ways to treat the cells in a way that maintains their physiological state while we're doing the imaging," he says. "Like every microscope, the instrumentation is only part of the puzzle. A lot of it is finding the right samples, and right preparation methods to make it work."
The new microscope is also exciting because it may be used in the future to improve super-resolution microscopy. PALM and other super-resolution techniques are limited to looking at thin, dead samples, and can be very damaging when looking at live ones. "That's what's really great about the Bessel — we can confine that excitation and really start to think about applying super-resolution microscopy to study structure or dynamics in thicker cells," says Betzig. Even without super-resolution, Bessel beam plane illumination microscopy will be a powerful tool for cell biologists, Betzig says, since it noninvasively images the rapidly evolving three-dimensional complexity of cells.
Light from laser (L) is reflected from x-axis galvanometer (XG) and transmitted in turn by relay lenses (RL) to z-axis galvanometer (ZG) and annular apodization mask (AM). XG, ZG, and AM are all at conjugate planes, so that the Gaussian beam falling on AM does not oscillate as XG and ZG are scanned. Similarly, AM is conjugate to the rear pupil plane of excitation objective (XO) so that the thin annular illumination transmitted through AM produces a Bessel beam within specimen (S) that translates along x and z without tilting. The light sheet created by scanning XG creates fluorescence at the focal plane of detection objective (DO), which is imaged at camera (C) by tube lens (TL). Different planes within S are imaged by translating DO with z-axis piezoelectric collar (ZP) in synchronization with the z axis motion of the Bessel beam provided by ZG. XO, DO and S reside in medium-filled specimen chamber (SC), and epiobjective (EO) provides a conventional view of specimen (S), for view finding purposes.