New Directions in FRET
Recently published studies highlight the improved efficacy of FRET and FRET measurements as well as novel uses of the technique.
Gary Boas, News Editor
Förster resonance energy transfer (FRET) continues to provide researchers with a means to quantify protein-protein interactions and other molecular dynamics by labeling molecules with a donor and an acceptor, which give off a characteristic emission when in proximity to one another.
Recent advances have led to the improved efficacy of FRET measurements as well as to novel applications of the technique — including, on the one hand, development of quantum dots as donors and incorporation of nanopipettes to increase the signal-to-noise ratio in single-molecule FRET measurements and, on the other, use of FRET for new drug delivery platforms and new “caged dyes” with which to track the trajectory, speed and timing of molecules. Thus, investigators have furthered the already tremendous potential of the technique.
Researchers have described a technique with which to address the complicated intermittent emission — known as blinking — exhibited by quantum dots used as donors in FRET measurements by combining time-resolved multispectral microscopy and single-molecule fluorescence intensity change-point analysis. The experimental setup consistedof a picosecond mode-locked laser — with a repetition rate of 20 MHz to allow for the long luminescence lifetimes of quantum dots — and a time- and position-sensitive custom photon-detection system.
Quantum dots recently gained currency as FRET donors because they offer a range of advantages over fluorescent dyes, including increased photostability, broad absorption and narrow, tunable emission. Another significant advantage is that they do not photobleach very easily, whereas fluorescent dyes often do.
Quantum dot-organic dye hybrids (where the quantum dots serve as donors and the dyes as acceptors) can be particularly useful at the single-particle level, where they offer potential for monitoring nanostructures’ local environments and examining conformational fluctuations. However, use of these hybrids for single-particle applications often is confounded by the dots’ complicated intermittent emission — known as blinking — which makes it difficult to extract the changes in FRET efficiency that researchers are trying to measure.
A team from the University of California, Berkeley, and from Sandia National Laboratories in Livermore, Calif., recently reported a method that addresses the blinking problem. As described in the Sept. 12 issue of the Journal of the American Chemical Society, they combined an experimental advance — time-resolved multispectral microscopy — with a methodological advance — single-molecule fluorescence intensity change-point analysis — to distinguish the various signal sources in FRET measurements at the single-particle level.
“The primary challenge is the number of possible interfering signal sources,” said Haw Yang, one of the authors of the study; “namely, quantum dot donor and dye acceptor blinking, direct acceptor excitation and acceptor photobleaching. The time and spectral resolution allow us to detect direct acceptor excitation, while the application of the change-point analysis to the time- and spectrally resolved data identifies the other signal sources.”
The researchers demonstrated the technique by performing multiparameter FRET measurements of quantum dot-Cy5 hybrids consisting of biotinylated quantum dots and Cy5-labeled streptavidin.
The multiparameter fluorescence spectrometer they used, in Carl C. Hayden’s laboratory at Sandia, consists of a conventional confocal microscope with a custom photon-detection system. A picosecond mode-locked laser made by Time-Bandwidth Products of Burgdorf, Switzerland, provided excitation at 480 nm; the scientists reduced the laser repetition rate to 20 MHz to allow for the long luminescence lifetimes of quantum dots. The light was focused at the sample by a 1.45-NA, 60× total internal reflectance microscopy oil-immersion objective from Olympus Corp. of Melville, N.Y.
The custom photon-detection system, which is both time- and position-sensitive, was based on a multianode photomultiplier detector made by Hamamatsu of Bridgewater, N.J., with 32 discrete elements. The signals from a custom readout coupled to the detector were sent to time-to-digital converters operating in the inverted mode, allowing the detection of a photon to signal “start” and the next laser shot to signal “stop.”
The experiments confirmed that, by combining spectral and temporal information, the technique could distinguish the various signals in quantum dot-dye hybrid FRET measurements and could identify confidently the changes in FRET efficiency. Thus, it could prove useful for a variety of single-particle applications.
Yang and Hayden noted that the researchers are working toward applying the technique for biosensor schemes that use FRET to measure discrete binding/unbinding events at the single-particle level, and for probing the physical and chemical nanoscale environment in complex systems such as inside a cell.
Getting in line
Single-molecule FRET has gained popularity in recent years because it helps to address the two main challenges of ensemble measurements: the difficulties in quantifying energy transfer resulting from inhomogeneous samples as well as in acquiring information about molecular dynamics. The single-molecule approach also has drawbacks, however. Namely, the signal from a single molecule is comparatively small, while very low concentrations of molecules in solution must be used, limiting studies to relatively high association constants (a measure of the association between a donor and an acceptor).
Investigators have incorporated nanopipettes into single-molecule FRET measurements to confine molecules to the detection volume, as shown here, thus increasing the concentration of molecules and improving the signal-to-noise ratio. The two platinum electrodes created the flow through the nanopipette.
In the Oct. 1 issue of Analytical Chemistry, scientists with the University of Bielefeld and with Ludwig Maximilians Universität of Munich, both in Germany, reported a relatively simple yet effective means by which to overcome both of these obstacles. As described in the paper, the researchers incorporated fairly common nanopipettes — which are used for scanning ion-conductance microscopy — to localize the molecules before performing FRET.
“The nanopipettes confine the molecules to a volume that is much smaller than the excitation focus defined by diffraction-limited laser illumination,” explained Philip Tinnefeld, the principal investigator of the study. “This allows us to increase the concentration approximately fiftyfold and to significantly improve the signal-to-noise ratio, since all molecules have to cross the focus through its very center.”
The researchers fabricated nanopipettes with an inner diameter of approximately 50 to 100 nm using a laser-based pipette puller and tested them by performing single-molecule FRET with donor/acceptor-labeled double-stranded DNA conjugates. They used a custom-built confocal microscope modified to allow alternating laser excitation. Alternating light from a 531-nm Ar+/Kr+ mixed-gas ion laser from Spectra-Physics of Mountain View and from a 638-nm diode laser from Coherent Inc. of Santa Clara, both in California, was transmitted into the microscope and coupled into a 60×, 1.2-NA water-immersion objective from Olympus. The same objective collected fluorescence, which was sent to two avalanche photodiodes that were made by PerkinElmer of Wiesbaden, Germany, using the appropriate filters for the detection of the donor and acceptor fluorescence.
The researchers tested the technique by performing single-molecule FRET using nanopipettes with an inner diameter of ~50 to 100 nm and a custom-built confocal microscope. The experiments demonstrated the improved signal-to-noise ratio and confirmed that considerably lower sample amounts are required when using the technique.
Potential of nanopipettes
The experiments confirmed the potential of the nanopipettes, showing an improved signal-to-noise ratio and demonstrating the considerably lower sample amounts required. “The nanopipette approach utilizes minute amounts of sample (<5 μl),” Tinnefeld said, “and, as every molecule passing the tip is detected, the number of molecules applied can be equal to the number of molecules measured.”
A possible disadvantage, he noted, is the proximity of the glass walls in the confined volume of the nanopipettes. “This means that one always has to be aware of possible surface interactions. This is less of a problem when working with DNA but can become important with sticky protein samples.” Such surface influences can be detected using autocorrelation, however, and then addressed with static or dynamic coatings of the pipettes.
The researchers continue to develop nanopipettes for use with single-molecule FRET. Tinnefeld added that reducing the pipette size to 10 to 20 nm at the tip — possibly by using quartz pipettes — might allow them to extend the dynamic concentration range of single-molecule measurements still further, toward the micro- or even millimolar regime. In addition, they are applying the technique to biologically relevant issues such as determining the structure of DNA-protein complexes that exhibit relatively low association constants.
Researchers also are looking to improve existing applications with FRET. About a year ago, a team with Brigham and Women’s Hospital in Boston and with MIT in Cambridge reported a targeted nanoparticle platform that could deliver drugs to treat cancer. The study generated a significant amount of interest, but the investigators were not satisfied.
“Those particles were capable of doing one thing and doing it well,” said principal investigator Omid C. Farokhzad, “killing prostate cancer cells. In the next phase, we decided to develop multifunctional nanoparticles to do more than treat cancer, to build on our own work and what others had done to make them even smarter.”
In the October issue of Nano Letters, therefore, in collaboration with researchers from the Gwangju Institute of Science and Technology in South Korea, and from Dana Farber Cancer Institute in Boston, they described a nanoparticle that can image and deliver drugs and furthermore can “sense” the delivery of the drugs to the targeted tumor cells in a relatively simple manner, based on the mechanism of FRET.
Researchers have described a FRET-based nanoparticle that can image and deliver drugs to targeted tumor cells as well as “sense” the drugs’ delivery. The three components of the nanoparticle — quantum dots (QD), the source of fluorescence; RNA aptamers (Apt), the targeting molecules, which also deliver the drug; and the drug itself (doxorubicin, or Dox), with known fluorescence properties — constitute a bi-FRET complex: between the quantum dot and the drug and between the drug and the aptamers. The fluorescence of both the quantum dot and the drug is “off” when the nanoparticle is loaded with the drug but is switched to “on” when the drug is released and delivered to the target cells. Reprinted with permission of Nano Letters.
The nanoparticle consists of three components: quantum dots, which provide the fluorescence (CdSe/ZnS core-shell QD490, in this study); RNA aptamers attached to the dots’ surface, which both deliver the drug and serve as targeting molecules; and doxorubicin, the drug itself, which has known fluorescent properties. The system constitutes a bi-FRET complex: a donor-acceptor model FRET that exists between the quantum dot and the drug, where the dot’s fluorescence is quenched when the drug is absorbed; and a donor-quencher model FRET that exists between the drug and the aptamers, where the aptamers quench the fluorescent drug.
The fluorescence of both the quantum dot and the drug is switched to the “off” state when the initial conjugate is loaded with the drug. When the drug is released from the conjugate during delivery to the target cells, the fluorescence of both the quantum dot and the drug is switched to the “on” state. Thus, the system can sense delivery of the drug by activating the fluorescence of the quantum dot, which then images the cancer cells.
This is the first example of a nanoparticle system that can perform three tasks — targeted therapy, cell imaging and sensing of drug delivery. A novel feature of the nanoparticle is that each of the three components performs two functions: The drug kills cancer cells but also quenches the fluorescence of the quantum dot, and the quantum dot enables imaging but also responds to the quenching to add the sensing capability. This is very important from a translational perspective, Farokhzad said. The more complex the nanoparticles, the less likely that they will find use in clinical environments because they would be difficult to manufacture and difficult to scale up. “To make them useful, you have to make them very smart but almost very dumb at the same time with regards to their design simplicity.”
The researchers expect to begin animal studies with the nanoparticle sometime in 2008. A series of optimization studies will follow, and then “we can reassess where we are with regards to the clinic,” Farokhzad said.
Into the cage
Researchers have developed photoactivatable fluorophores — also known as caged dyes — to track the trajectory, speed and timing of molecules and cells in biological systems. Advances in imaging technology have spurred further progress, including the recent introduction of dyes with improved water solubility and biocompatibility, increased photostability and high photolytic efficiency.
Recently reported FRET-based photoactivatable fluorophores known as caged dyes offer much improved contrast enhancement (>14 times) over previously described caged dyes. Shown here are images of the dyes before (left), shortly after (center) and approximately 2 min after (right) uncaging, for 460-nm and 530-nm emissions (top and bottom, respectively).
In 2004, investigators with the University of Texas Southwestern Medical Center in Dallas described a new class of caged coumarins with high uncaging efficiency. Now they have introduced a dye based on the principles of FRET that links a caged coumarin and a green dye, reporting it in the Sept. 5 issue of the Journal of the American Chemical Society. The dye offers high uncaging efficiency using both ultraviolet and two-photon excitation — roughly two orders of magnitude higher than that achieved with previously reported caged green dyes.
The caged dye includes a caged coumarin 3-carboxamide and a calcein 6-carboxamide connected by a cyclohexyl linker (hence, CCC-1; Caged Coumarin-Calcein 1). Of the green-emitting dyes, the investigators chose calcein because it offers good water solubility as well as long cytoplasmic retention time, both of which are useful when labeling and tracking biological specimens, especially over the long term. Also, the extensive spectral overlap between coumarin emission and fluorescein excitation is established, and researchers already have exploited it to design a variety of FRET sensors of reporter enzymes.
The primary challenge of the study, said principal investigator Wen-hong Li, was synthesizing the FRET probe with high purity. Because the commercially available fluorescein derivative they started with was a mixture of two isomers (5- and 6-isomer), they had to separate the isomers efficiently along the synthesis. They eventually obtained highly pure 5- and 6-isomers. “It was well worth the effort,” Li said, “because these two isomers show quite different FRET efficiency, with the 6-isomer showing a remarkable 95 percent FRET efficiency.”
They demonstrated the dye by uncaging and imaging CCC-1/dextran in living cells, using an inverted fluorescence microscope made by Carl Zeiss Group of Oberkochen, Germany, with a 40× objective. A 75-W xenon lamp excited the cells. A cooled CCD camera made by Hamamatsu of Bridgewater, N.J., detected the epifluorescence. Imaging acquisition and analysis were performed using integrated imaging software from Improvision Ltd. of Coventry, UK.
The dye did not disappoint. As noted, it exhibited high FRET efficiency (95%) and high contrast enhancement (>14 times), as well as high uncaging efficiency.
Another unique feature is that it enables researchers to identify cells or molecules labeled with it prior to uncaging, by exciting the dye at 490 nm. “This feature would be especially desirable,” Li said, “for experiments demanding localized photoactivation in three dimensions (for example, by two-photon uncaging), in which knowing the spatial distribution of the label can guide the researcher to determine local uncaging areas.”
He added that caged FRET dyes such as CCC-1 are attractive for a number of other biological imaging applications, including examining the molecular selectivity of gap junction channels, tracking neuronal processes in three dimensions in tissues and cell fate mapping during development.
- quantum dots
- Also known as QDs. Nanocrystals of semiconductor materials that fluoresce when excited by external light sources, primarily in narrow visible and near-infrared regions; they are commonly used as alternatives to organic dyes.
MORE FROM PHOTONICS MEDIA