Parallel analysis method cracks scalable single-molecule imaging

When researchers study galaxies and stars, specialized satellites, such as the famed Hubble Space Telescope, are necessary to capture the best views of space without interference from the Earth’s atmosphere. However, the need to build and launch satellites into orbit limits how large such satellites can be and how far they can see into the universe.

On the other end of the spectrum, biology operates on the scale of single molecules. In order to visualize life processes at this scale, scientists strive to develop single-molecule fluorescence (smF) and single-molecule Förster resonance energy transfer (smFRET) methods to enable investigations of these tiny specimens in real-time.

“You can think of the need for single-molecule imaging as analogous to trying eavesdrop on a cocktail party,” said co-author Daniel Terry, PhD, Single-Molecule Imaging Center director, Department of Structural Biology. “Traditional, ensemble methods listen to everything everyone is saying all at once; there’s a lot of noise, and you aren’t going to learn much about any individual. Single-molecule imaging enables us to engage with each ongoing conversation separately.”

Directly visualizing single molecules has led to many important discoveries; however, the technical issues that arise from working at such a small scale have become major barriers to advancing the field. A newly published technique from St. Jude called Parallel Rapid Exchange (PRE)-smF and PRE-smFRET may help scientists overcome those limits.

“We found ways to improve the efficiency, reproducibility, sensitivity and reliability of single-molecule microscopes,” explained corresponding author Scott Blanchard, PhD, Department of Structural Biology, whose lab members innovated the PRE-smFRET technique, which was published recently in Nature Methods. “This method forms the basis of a generalizable and scalable approach for parallelized single-molecule imaging that we hope will reveal subtle, yet fundamentally important functional distinctions in the molecules that support life at unprecedented resolution.”

Overcoming single-molecule imaging limits

As with most experimental techniques, single-molecule imaging is typically done on one sample at a time by measuring a solution full of identical molecules. In certain circumstances, a mixed population of molecules can be studied, but it requires a follow-up with sophisticated techniques that determine the identity of each molecule and demultiplex the samples. When experiments are performed in series, instrument noise and other variables sometimes mask minor differences that are biologically important. To increase the sensitivity and reproducibility of comparative measurements, St. Jude researchers developed a technique that interrogates distinct populations of individual molecules simultaneously.

“Parallelization is possible because we engineered the means to have prior knowledge of each molecule’s identity in a given experiment,” said co-first author Ryan Brady. “To do so, we print short DNA barcode strands onto the quartz slide, placing multiple spots in the microscope field of view. We then label our molecules of interest with the DNA strands complementary to the printed ones, letting them stick only to the corresponding spot in the field of view. This way, we can look at multiple different samples at once, multiplying our experimental throughput.”

To create DNA barcode microarrays, the group used a specialized instrument that dispenses tiny volumes of liquid in precise patterns that can then be encapsulated within specialized microfluidic devices. Using computer-controlled fluidics handling systems, these microarrays can be flooded with reagents, so that all samples are stimulated at the same time. The simultaneity of stimulus addition or removal gives researchers the ability to make quantitative comparisons between highly similar systems, such as wild-type and mutant samples, to uncover subtle, yet biologically important insights. These nonequilibrium experiments can be automatically repeated to collect more data and to ensure reproducibility. This new approach has the potential to enable drug screening on the single-molecule level.

These parallelized measurements required new developments in microscope design. “We had to increase the field of view to see multiple samples side-by-side, while capturing full-frame images at 500 frames per second,” said co-first author Roman Kiselev, PhD, Department of Structural Biology. This was done using high-speed cameras with large sensors. “We also had to ensure a uniform illumination across the field of view, which required prism-based excitation combined with special flat-field beam shaping optics. The new technique gives us more robust statistics and greater sensitivity, as we no longer have variance introduced from cross-experiment comparisons, letting us uncover new biology.”

Uncovering subtleties at single-molecule scale

Armed with their new technique, Blanchard’s group chose to tackle several biological problems that variance, inherent to traditional methods, had previously obscured. The first was β-arrestin1, a protein that regulates many G-protein-coupled receptors (GPCRs), the most common targets for therapeutic drugs. With PRE-smFRET, they could study four different β-arrestin1 samples tagged with fluorophores on different domains, so that they could examine the relative timing of known shape (conformation) changes.

“We knew there were three big conformational changes,” Blanchard said. “With parallelization, we showed they don’t happen simultaneously; instead, two of them occur at about the same time, and only then, the third, slower one, is allowed to happen. This was not possible to decisively determine before.”

To show that the technique could be used on more than proteins, the Blanchard team set their sights on ribosomes, the protein-making machines that themselves comprise RNA and protein. They first uncovered how transient hydrogen bonds contribute to a bacterial ribosome’s rate of making proteins, and how an antibiotic alters those interactions, to reveal that prior ensemble measurements of the same process had been misinterpreted.

Finally, they looked at the importance of small differences in the makeup of two bacterial ribosomes, both of which occur naturally. The researchers aimed to study how a difference of 10 RNA nucleotides, of the roughly 4500 in the whole ribosome, led to different sensitivity to closely related antibiotics. With PRE-smFRET, they found that ribosomes bearing these 10 nucleotide variants were resistant to tetracycline but sensitive to oxytetracycline, despite the drugs’ relative similarity.

“We’ve broken through the classic threshold for detecting functional distinctions,” Brady said. “We are able to distinguish functions of molecules that are more than 99% identical. The statistical power to compare small functional differences using this approach opens up new areas of biology that have yet to be examined.”

Moving forward with PRE-smF and PRE-smFRET

As the method shows promise and has already uncovered new biology, the researchers are also focusing on helping others adapt the new technique.

“We have optimized every step of this technique,” Kiselev said. “Now, single-molecule experiments that used to take multiple days or weeks can be done within hours, providing higher data content, while also being accessible to scientists.”

Together, the optimized experimental method and its early results provide strong evidence that PRE-smF and PRE-smFRET can uncover important functions at the single-molecule scale for a wide variety of biological systems that were previously beyond the limits of these techniques.

“We’ve demonstrated that PRE-smF and PRE-smFRET can reveal some of the most subtle structural and kinetic single-molecule insights,” Blanchard said. “We hope that its adoption opens an entirely new landscape, increasing resolution and throughput enough to push the whole single-molecule imaging field forward.”