Researchers from the Georgia Institute of Technology and University of California,
San Francisco
have advanced scientists’ ability to view a clear picture of a single cellular
structure in motion. By identifying molecules using compressed sensing, this
new method provides needed spatial resolution plus a faster temporal resolution
than previously possible.
Despite many achievements in the field of super-resolution microscopy in the
past few years with spatial resolution advances, live-cell imaging has remained
a challenge because of the need for high temporal resolution.
Now, Lei Zhu, assistant professor in Georgia Tech’s George W. Woodruff
School of Mechanical Engineering, and Bo Huang, assistant professor in UCSF’s
Department of Pharmaceutical Chemistry and Department of Biochemistry and
Biophysics, have developed an advanced approach using super-resolution
microscopy to resolve cellular features an order of magnitude smaller than what
could be seen before. This allows the researchers to tap previously
inaccessible information and answer new biological questions.
The research was published in Nature Methods. The research is funded
by the National Institutes of Health, UCSF Program for Breakthrough Biomedical
Research, Searle Scholarship and Packard Fellowship for Science and
Engineering.
The previous technology using the single-molecule-switching approach for
super-resolution microscopy depends on spreading single molecule images
sparsely into many, often thousands of, camera frames. It is extremely limited
in its temporal resolution and does not provide the ability to follow dynamic
processes in live cells.
“We can now use our discovery using super-resolution microscopy with seconds
or even sub-second temporal resolution for a large field of view to follow many
more dynamic cellular processes,” said Zhu. “Much of our knowledge of the life
of a cell comes from our ability to see the small structures within it.”
Huang noted, “One application, for example, is to investigate how
mitochondria, the power house of the cell, interact with other organelles and
the cytoskeleton to reshape the structure during the life cycle of the cell.”
Currently, light microscopy, especially in the modern form of fluorescence
microscopy, is still used frequently by many biologists. However, the authors
say, conventional light microscopy has one major limitation: the inability to
resolve two objects closer than half the wavelength of the light because of the
phenomenon called diffraction. With diffraction, the images look blurry and
overlapped no matter how high the magnification that is used.
“The diffraction limit has long been regarded as one of the fundamental
constraints for light microscopy until the recent inventions of
super-resolution fluorescence microscopy techniques,” said Zhu.
Super-resolution microscopy methods, such as stochastic optical reconstruction
microscopy (STORM) or photoactivated localization microscopy (PALM), rely on
the ability to record light emission from a single molecule in the sample.
Using probe molecules that can be switched between a visible and an
invisible state, STORM/PALM determines the position of each molecule of
interest. These positions ultimately define a structure.
The new finding is significant, said Zhu and Huang, because they have shown
that the technology allows for following the dynamics of a microtubule
cytoskeleton with a three-second time resolution, which would allow researchers
to study the active transports of vesicles and other cargos inside the cell.
Using the same optical system and detector as in conventional light
microscopy, super-resolution microscopy naturally requires longer acquisition
time to obtain more spatial information, leading to a trade-off between its
spatial and temporal resolution. In super-resolution microscopy methods based
on STORM/PALM, each camera image samples a very sparse subset of probe
molecules in the sample.
An alternative approach is to increase the density of activated fluorophores
so that each camera frame samples more molecules. However, this high density of
fluorescent spots causes them to overlap, invalidating the widely used
single-molecule localization method.
The authors said that a number of methods have been reported recently that
can efficiently retrieve single-molecule positions even when the single
fluorophore signals overlap. These methods are based on fitting clusters of
overlapped spots with a variable number of point-spread functions (PSFs) with
either maximum likelihood estimation or Bayesian statistics. The Bayesian
method has also been applied to the whole image set.
As a result of new research, Zhu and Huang present a new approach based on
global optimization using compressed sensing, which does not involve estimating
or assuming the number of molecules in the image. They show that compressed
sensing can work with much higher molecule densities compared to other
technologies and demonstrate live cell imaging of fluorescent protein-labeled
microtubules with three-second temporal resolution.
The STORM experiment used by the authors, with immunostained microtubules in
Drosophila melanogaster S2 cells, demonstrated that nearby
microtubules can be resolved by compressed sensing using as few as 100 camera
frames, whereas they were not discernible by the single-molecule fitting
method. They have also performed live STORM on S2 cells stably expressing
tubulin fused to mEos2.
At the commonly used camera frame rate of 56.4 Hertz, a super-resolution
movie was constructed with a time resolution of three seconds (169 frames) and
a Nyquist resolution of 60 nanometers, much faster than previously reported,
said Zhu and Huang. These results have proven that compressed sensing can
enable STORM to monitor live cellular processes with second-scale time
resolution, or even sub-second-scale resolution if a faster camera can be used.
