Of Terms in Biology: FRET
by Christoph
FRET is the commonly used acronym for "Förster resonance energy transfer", which sounds just as awe-inspiring as it is. One has to dive deep into quantum mechanics and Jablonski diagrams to understand how 'non-radiative intermolecular dipole coupling' works. But fortunately, FRET can be explained in terms that not so physics-savvy microbiologists can grasp (including this author ). Take a chromophore, that is, a molecule that can absorb light. "Absorbing light" means a photon that plunges into the chromophore kicks one of its electrons into a higher-energy level (remember that in the quantum world everything comes in discrete steps, the photon has to have the "right" wavelength, that is, energy to kick the electron into one particular energy level ). When, after a few picoseconds, the electron returns to its ground state it does so by emitting a photon of longer wavelength (this is what fluorescence is about ). If now a second chromophore is close by, the excited electron can sort of "tunnel" its energy to this second chromophore instead of emitting a photon. This energy transfer by a "virtual photon" is called FRET, and it requires 1. that the distance between the donor and the acceptor is in the range of 1 – 10 nm, 2. that the donor emission spectrum and the acceptor absorption (=excitation ) spectrum overlap (Fig. 1A), and 3. that the dipole moments of donor and acceptor are in a favorable relative orientation. It is easily conceivable that if FRET "works" at distances of <10 nm it should make for a perfect ruler to measure, for example, distances between individual domains of protein molecules, or distances between monomers of a protein oligomer, in the test tube or in living cells (ribosomes, for example, have a diameter of 25 – 30 nm, while the oligomeric RNA polymerase of E. coli has a diameter of ~5 nm ). And that's exactly what biochemists and cell biologists have used FRET for during the last decades: measuring molecules. Here are three examples.
FRET-ing proteins
Figure 1. FRET analysis of c‑Raf protein conformation. A Excitation (Abs.) and emission (Em.) spectra of CFP (cyan FP) and YFP (yellow FP). The CFP emission spectrum overlaps the YFP excitation spectrum allowing for FRET (Source). B In the closed conformation of c‑Raf fusion protein, the N‑terminal YFP domain comes into close contact (<10 nm) with the C‑terminal CFP domain. Excitation of CFP by near-UV light at 433 nm results in stimulation of light emisssion at 527 nm by YFP (left). Upon stimulation by membrane-bound Ras protein, the c‑Raf fusion protein switches into the open conformation, which increases the distance between the YFP and CFP domains. Excitation of CFP by near-UV light at 433 nm results in emission at 475 nm (specific for CFP) while no emission at 527 nm (specific for YFP) is observed in this case. Source
One of the most enthralling ways for a biologist to look inside living cells – eukaryotic as well as prokaryotic cells – and localize specific proteins is the use of GFP-fusion proteins (see Elio's post on GFP here ). Although only the successful cases have been published, it is surprising how many different proteins accept ~30 kD GFP tags (that's easily half of the molecular weight of the tagged protein, or more) without a murmur, that is, assuming their normal conformation as monomers or in oligomers, performing their normal function(s), and finding their specific location(s) within the cell (in most cases it is necessary, however, to express the wild-type protein along with the GFP-tagged protein to sustain cell viability ). Meanwhile, several pairs of GFP derivatives are known in which the emission spectrum of one of them overlaps the excitation spectrum of the other; they are thus suitable "partners" for FRET experiments (Fig. 1A). Terai and Matsuda used the CFP/YFP pair (fluorescent protein nomenclature: the first letter stands for the color, that is, C for cyan, Y for yellow, etc. You can guess what FP stands for ) to study a conformational switch in the human c‑Raf protein in living cells by FRET (c‑Raf is a kinase in the MAPK/ERK pathway, a chain of proteins that communicates a signal from a receptor on the surface of the cell to transcription factors in the nucleus that regulate growth and division ). The structure of the 648 amino acids-long c‑Raf is well known, and also its conformational switch between an 'open conformation' and a 'closed conformation'. The known structure of c‑Raf allowed the targeted attachment of CFP and YFP to the protein's N‑terminus and C‑terminus, respectively, without messing up its structure and function (Fig. 1B) (the fusions were constructed for the genes, of course, while the description here is for the expressed proteins ). By observing cells expressing the tagged c‑Raf protein by fluorescence microscopy, Terai and Matsuda could measure, as variation in FRET efficiency, the time course for the conformational switch of c‑Raf upon stimulation by membrane-bound Ras, one step in the MAPK/ERK pathway. Note that they could actually see the colored c‑Raf molecules over the entire time course, either the yellow-greenish 'closed conformation' or the cyan-blue 'open conformation' (you can see it yourself in one of their videos ).
FRET-ing nucleic acids
Figure 2. A Folding of the χ‑form Holliday junction into one of the two possible "stacked X" forms in the presence of cations. B Schematic of the junction folding that is induced by ion binding. The R and X strands have a linker at the 5' end for attachment of the fluorescein (FAM) + tetramethylrhodamine (TAMRA) dyes, which are in close proximity in the "iso II" conformation (high FRET). (adapted from Source)
Several techniques have been developed to couple fluorophores to nucleic acids as, for example, to oligodeoxyribonucleotides ('oligos' in lab slang ). Fluorophores – a veritable zoo of different compounds is known and widely applied – are usually coupled to the α‑phosphate of nucleotides via a short linker and do thus not interfere sterically with inter-strand hybridizations (= the formation of double-stranded DNA or RNA ). 'Fluorescence in situ hybridization' (FISH) is the most widely known application for fluorophore-labelled oligos (see here in STC ). However, by using two compatible fluorophores to 'tag' individual DNA strands differently, FRET has become a reliable technique to probe conformational changes in nucleic acids. Litke et al. used FRET in a recent study to monitor a conformational switch in Holliday junctions. Huntington Potter had explained the origin of the χ‑form Holliday junction for STC a while ago (see Pictures Considered #28 ). As beautiful and revealing these electron micrographs were, a spread-out "χ" (Greek Chi ) is not the only structure such Holliday junctions assume (Fig. 2A, left). It is known that Holliday junctions fold into "stacked X" forms in the presence of cations (Fig. 2A, right). Litke et al. measured the increase in FRET efficiency during the conformational switch from the open form to the "stacked X" (iso II) conformation with increasing cation concentrations (Fig. 2B). They showed that the folding reaction requires site-specific binding of two cations, for example Mg2+, which are located within 7 Å (0.7 nm) of each other (Fig. 2A, right), and that this interaction involves at least one site of direct coordination to the phosphate backbone of the Holliday junction substrate.
FRET-ing forces
Figure 3. Molecular tension sensor (MTS) overview. a Sensors are site-specifically labeled with biotin (Avitag), the FRET donor Alexa 546 (KC), and FRET acceptor CoA 647 (ACP tag) and present the RGD sequence from fibronectin (TVYAVTGRGDSPASSAA). The (GPGGA)8 sequence acts as an entropic spring that is stretched upon application of force. b Sensor molecules are attached to a coverslip via biotin and Neutravidin; the biotinylated PEG brush prevents nonspecific cell and sensor attachment. Integrin heterodimers attach to the RGD domain and apply load generated by the cell cytoskeleton. c Single fibroblast (HFF) in the process of focal adhesion. yellow: high FRET/low force signal from MTS not interacting with integrins. green dots: low FRET/high force signals for MTS interacting with integrin; blue dots: eGFP‑paxillin expressed by the HFF cells as indicator for focal adhesion. Note the co‑occurrence of green coloring and blue dots at the cell periphery. Source
If FRET is such a versatile "ruler" for measuring intra- and inter-molecular distances in the nanometer range, as described above, would it also be possible to "tweak" the distances between two chromophores and thus obtain a measure for the forces that pull chromophores apart? Headline: Cell biology goes nanomechanics! Neither 'fake news' nor hype, because this is exactly what the Dunn lab at Stanford University, CA, did in their recent study. Briefly, they took an expandable peptide and attached to either end one of a pair of FRET-proficient chromophores. Next they added to one end of this 'molecular expander' a biotin tag to allow its fixation on a streptavidin-coated surface, and to the other end a recognition sequence for integrin, a membrane protein that is involved in focal adhesion of mammalian fibroblasts (Fig. 3a). When they observed fibroblasts crawling over a glass plate covered with 'expander' molecules (Fig. 3b) in the fluorescence microscope they found the moving fibroblasts literally pulling on the expanders by a decrease in FRET efficiency in cellular regions active in focal adhesion (Fig. 3c).
The Grashoff lab at the MPI for Biochemistry in Munich, Germany, took this 'expander' approach one step further: why not measure forces that interacting proteins exert on each other within cells? They describe in their newest study a set of FRET-linked 'expanders' inserted at different positions into the talin‑1 protein, which respond to a range of forces in the lower piconewton range. They measured with this 'expander kit' the forces excerted on talin‑1 during focal adhesion by its interaction with membrane-bound integrin and the intracellular actin cytoskeleton (see frontpage picture ). Their aim is to study the contribution of individual talin‑1 molecules to the efforts of the entire talin‑1 population, as Carsten Grashoff explained in an interview: "In a tug-of-war, many people pull on a rope with different strengths. Some may take it easy and let the person in front do the work. Proteins work in a very similar manner. We can now determine which proteins contribute to cellular force development and what percentage of those molecules are actually involved."
Frontispiece: Stable coexpression of two orthogonal FL-based talin sensors (talin-447-LmO-FL-mK2 and talin-1973-TFP-FL-ShG); the constructs display indistinguishable subcellular localization (overlay of the individual photographs taken at different wavelengths). © MPI für Biochemistry, München, Germany (adapted from Source)
