507
Fluorescence resonance energy transfer as a structural tool for nucleic acids David MJ Lilley* and Timothy J Wilson Fluorescence resonance energy transfer is a spectroscopic method that provides distance information on macromolecules in solution in the range 20–80 Å. It is particularly suited to the analysis of the global structure of nucleic acids because the long-range distance information provides constraints when modelling these important structures. The application of fluorescence resonance energy transfer to nucleic acid structure has seen a resurgence of interest in the past decade, which continues to increase. An especially exciting development is the recent extension to single-molecule studies. Addresses CRC Nucleic Acid Structure Research Group, Department of Biochemistry, The University of Dundee, Dundee DD1 4HN, UK *e-mail:
[email protected] Current Opinion in Chemical Biology 2000, 4:507–517 1367-5931/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved Abbreviations 3′′-UTR 3′ untranslated region bp base pairs Cy3 cyanine 3 dsDNA double-stranded DNA efficiency of FRET EFRET FRET fluorescence resonance energy transfer HMG high mobility group spFRET single-pair FRET TBP TATA-box-binding protein
Introduction Fluorescence resonance energy transfer (FRET) provides long-range distance information, typically in a range of 20–80 Å, which is not covered by virtually any other solution technique. Although the approach was used in biological systems in the 1970s [1,2], it has seen a resurgence of interest and application in the past decade, particularly in the field of nucleic acid structure. By their very nature, DNA and RNA structures tend to be rather extended, being based on double helical segments. In general, the component polynucleotide chains do not tend to double back on themselves, and thus distances up to, say, 5 Å, will not contain the long-range connectivities of primary sequence that would be present with the path of a polypeptide in a protein. To take a simple example, helices in natural RNA molecules are frequently kinked by an asymmetric loop or bulge. Accurate determination of this angle would best be accomplished by measurement of the end-to-end distance between the distal termini of these helices, which might well be in excess of 60 Å. Because FRET can provide distance information in this range it is potentially very valuable in the study of nucleic acid structure. FRET is complementary to NMR, which can provide information over many short-range (≤ 5 Å) distances, filling in the missing long-range distances and providing a better indication of the global structure.
In this review, we discuss the information available from FRET, as well as the limitations of the technique. This covers the past decade, but we concentrate on work published in the past two years.
Fluorescence resonance energy transfer In a typical FRET experiment, a nucleic acid is labelled with two different fluorophores, a donor and an acceptor, covalently attached at different locations (Figure 1). The absorption of the donor occurs at higher frequency than that of the acceptor. FRET involves a resonance between singlet−singlet electronic transitions of the two fluorophores, generated by a coupling between the emission transition dipole moment of the donor and the corresponding absorption dipole moment of the acceptor [3]. This leads to a transfer of excitation energy from the donor to the acceptor. FRET can be observed in a variety of ways, including a reduction in the fluorescent quantum yield of the donor, a corresponding shortening of the donor excited state lifetime, and an increased fluorescent emission from the acceptor (if fluorescent). Quantum mechanics show that the rate of this process depends on the inverse sixth power of the distance between the two fluorophores. This is the basis of the use of the technique to provide structural information. In the laboratory, the efficiency of FRET (EFRET) is normally determined. This can be defined in a variety of ways, and is the proportion of donor excitation events that lead to excitation of the acceptor by dipolar coupling — a quantum yield of FRET. It can be readily shown that: EFRET =
1 R 6 1 + R0
(1)
where R is the distance between the fluorophores. R0 is the characteristic Förster length for a given donor−acceptor pair, which is given by: R06 = 8.8 × 10–28 . ΦD . κ2 . n−4 . J(λ)
(2)
where ΦD is the fluorescent quantum yield of the donor in the absence of the acceptor, κ is a parameter that depends on the relative orientation of the donor and acceptor transition moments, n is the refractive index of the medium and J(λ) is the spectral overlap between donor emission and acceptor absorption. From Equation 1 it is easily seen that when R = R0, the efficiency of FRET is 50%. FRET experiments can be done in two broad modes. The easiest approach is to use steady-state fluorescence, simply requiring a good fluorimeter equipped with polarizers. The FRET efficiency can be derived by measuring fluorescence
508
Analytical techniques
Figure 1
Donor
Acceptor kFRET = [R0 / R]6 τD
vr
S1D ab
em hνemD
D
hν1
S1A
vr
hνemA
em
dx
S0A
S0D
R µD
µA
Schematic to illustrate the photophysical processes in a FRET experiment. The cylinder in the lower section represents a macromolecule to which donor and acceptor fluorophores have been covalently attached at known positions. The donor and acceptor transition moments are indicated by µD and µA, respectively, and are separated by distance R. A simplified Jablonski diagram is presented in the upper section. Absorption of a photon (of energy hνexD) stimulates a transition (ab) from the ground state (S0D) to a higher vibrational state of the excited singlet state (S1D). This is followed by a rapid loss of energy (vr) to the lowest vibrational state. In the absence of an acceptor, this fluorophore can revert to the ground state by emission of a photon (em) of lower energy (hνemD), or by a variety of non-radiative processes (collected together as dx). In the presence of an acceptor, dipolar energy transfer is possible at a rate kFRET, leading to excitation of the acceptor (state S1A). After loss of vibrational energy (vr), this can return to the ground state by emission of light (em) of energy hνemA. Efficiency of FRET can be measured in a variety of ways, including the enhanced emission of the acceptor, reduced fluorescent quantum yield of the donor and the shortened fluorescent lifetime of the donor (τD).
Current Opinion in Chemical Biology
spectra at two different excitation frequencies, and carrying out a normalisation procedure [4]. Additional information can be obtained by measuring fluorescent lifetimes, which can be performed either in the time or frequency domain. The experimental data can be analysed in terms of multiple decay processes, and thus information on distributions of distances is available from such experiments. There are two main limitations to the exploitation of FRET data in structural studies. The first lies in the parametrisation of Equation 2. Assuming this is solved, the second problem is turning an interfluorophore distance into a distance constraint useful in building a molecular model, which requires information on the location of the fluorophores with respect to the molecule. We shall discuss these problems in greater detail below.
Relative distance information from FRET Even without addressing the above problems directly, a great deal of useful information can be derived from FRET experiments by carrying them out in a comparative manner. In this approach, no measurements of absolute distance are attempted, but instead comparisons of FRET efficiencies are made between a series of closely similar species designed to reveal various aspects of a given structure. An early example of this approach was the analysis of the global structure of the fourway DNA junction, in which the different end-to-end
distances were compared for a junction with four arms of equal length, leading to the confirmation of the stacked X-structure [5,6]. As in most studies of nucleic acids, the fluorophores were attached to the 5′ termini of different strands, thus placing them at the ends of designated helices. For an n-helix junction there will be n!/2(n−2)! end-to-end distances, giving three vectors for a three-way junction and six for a fourway junction. With this approach it is important to ensure that the environment of the fluorophores is kept as constant as possible, and thus the terminal sequence of each helix would be kept the same over two or more basepairs. In a system such as a helical junction we begin with a lot of information on the nature of the structural components. We know that the helical arms will be essentially B-form helices, and that they radiate from an approximately constant centre. Thus, the determination of the relative end-to-end distances should largely define the global structure, even in the absence of absolute distances. We note that the stacked X-structure of the four-way DNA junction that was deduced by the application of FRET and comparative gel electrophoresis studies has recently (over a decade later) been confirmed in almost every detail by crystallography [7–10]. This success shows that the method can be applied to new systems with confidence. The comparative approach is also valuable for studying a given structure as a function of conditions. In particular, the polyelectrolyte character of nucleic acids frequently
FRET as a structural tool for nucleic acids Lilley and Wilson
results in an extended conformation at low salt concentrations, and metal ion-induced structural transitions can be readily followed by FRET. For example, the four-way DNA junction adopts an extended, unstacked conformation in the absence of ions [11]. By studying the FRET efficiency for a given end-to-end vector as a function of divalent ion concentration, a binding isotherm describing the folding transition can often be obtained, together with an apparent association constant and Hill coefficient (a measure of the cooperativity of binding; in the limit of full cooperativity, this becomes the stoichiometry) [12]. Indeed, considerable thermodynamic information can be extracted from FRET experiments [13]. In principle, the transition may be observed using different end-to-end vectors within the molecule, including some that elongate and others that become shorter in the folding process. Such folding transitions could also be studied as a function of time, perhaps using rapid reaction techniques such as stopped-flow methods. Eis and Millar [14] have used energy transfer in time-resolved fluorescence experiments to analyse distance distributions in DNA junctions. Another example of the study of an ion-induced conformational transition in DNA by FRET was the observation of a B→Z transition (i.e. an inversion of handedness) by Jares-Erijman and Jovin [15]. A further subdivision of the comparative approach arises where a structure varies systematically as a result of small changes to the covalent structure. One example is the base bulges (a series of one or more consecutive, unopposed bases) in DNA and RNA, where an incremental shortening of the end-to-end distance observed as an increase in FRET efficiency reveals an axial kinking that increases with the size of the bulge [16]. The three-way DNA junctions provide another example. Although perfect three-way junctions (3H according to the IUB nomenclature [17]) are conformationally rather rigid, addition of extra unopposed bases at the junction (such as a 3HSn junction) allows pairwise coaxial stacking to occur [18,19].
509
the case of flexible reorientation during the lifetime of the excited state, κ2 averages to 2/3, and as the interfluorophore distance increases this becomes a better approximation [21]. Fluorescein attached via a six-carbon linker to the 5′-terminus of DNA is generally very mobile, with a low fluorescence anisotropy of around 0.1. This indicates that the fluorophore is rapidly reorienting during the lifetime of the excited state. Using fluorescein as the donor, κ2 = 2/3 is generally a good approximation. Fortunately the calculated distances are not as sensitive to the exact value of this parameter as it might appear because the sixth root is taken. The range of distances that can be obtained from FRET is set by the value of R0, and the most significant component is the value of the overlap integral J(λ). This measures the spectral overlap between the fluorescence emission of the donor and the acceptor absorption: ∞
∫ φ λ ⋅ε
λ =
∞
λ ⋅λ λ
∫φ λ λ
(3)
where φD is the spectral shape function for the donor emission, εA is that for acceptor absorption and λ is the wavelength (see Figure 2a). Recently, a number of FRET studies have employed the cyanine dyes, and fluorescein–Cy3 (cyanine 3) is a particularly useful donor−acceptor pair [15,22,23•]. Values of R0 have been calculated using either sulfoindocarbocyanine-3 or indocarbocyanine-3 as the acceptor from fluorescein as 56 Å [15] and 55.7 Å [24••], respectively. This results in the calculated dependence of FRET efficiency on distance shown in Figure 2b. If we argue that measurements of efficiency between 0.1 and 0.9 can be made with sufficient accuracy, an accessible distance range of 35–85 Å is obtained. This is extremely useful on the scale of nucleic acid structure.
Determination of absolute distances In principle, absolute distance information is available from FRET experiments. In practise this requires us to be confident about the parametrisation of Equation 2, and to be able to use inter-fluorophore distances competently in our model building. An early study of a series of DNA duplexes (8–20 bp in length), terminally labelled with fluorescein and tetramethylrhodamine as the donor–acceptor pair, showed that quantitative agreement between FRET efficiency and molecular geometry was possible [20].
Another uncertainty in R0 is the value for the refractive index that should be used. The value for bulk solvent is 1.33. Flow birefringence studies in DNA [25] have indicated that a higher value, in the region of 1.75, is more appropriate for the interior of DNA, and such values have been used to analyse excitation transfer between intercalated dyes [26]. With an extended fluorophore such as fluorescein, however, the environment of the bulk solvent is probably correct, and calculations suggest that the value of 1.33 is appropriate in these circumstances.
One significant impediment to the extraction of absolute distances from FRET concerns the orientation dependence of transfer efficiency, embodied in the term κ in Equation 2. This describes the relative orientation of donor and acceptor transition dipole moments, and can take a value between 0 and 4 for different orientations. In
Any distances that are calculated from FRET refer, of course, to the separation between the fluorophores themselves. In order to interpret this information in a manner that is useful for modelling the structure, we need to know where the dyes are located relative to the nucleic acid. Significant effort has gone into characterising the manner
510
Analytical techniques
Figure 2
(b)
3.5
100,000
3 F
2.5
80,000
Cy3
2
60,000
1.5
40,000
Cy3 molar absorbance
Fluorescein fluorescent emission
(a)
1 20,000
0.5 0 480
0 580
500 520 540 560 Wavelength / nm
1 R0 = 56 Å
0.8 0.6 EFRET
of DNA duplexes of different length, terminally labelled with fluorescein and Cy3, suggested that Cy3 was located close to the helical axis [24••]. This was confirmed in an NMR study of a self-complementary DNA duplex with Cy3 attached at the 5′ termini via a three-carbon linker, where it was found that the Cy3 was stacked onto the end of the helix in the manner of a pseudo-basepair (Figure 3) [24••]. This is important, because it now provides a known point in the structure from which distances are measured.
0.4 0.2
By comparison with tetramethylrhodamine and Cy3, fluorescein attached to the 5′ end of dsDNA or RNA is relatively flexible within a cone. Its charge characteristics will result in repulsion by the phosphodiester groups of the nucleic acid. Clearly, a single position cannot be ascribed to such a dynamic situation. Using the location of Cy3 as established by NMR, together with interfluorophore distances for different-length DNA duplexes from FRET, however, an effective position was determined which indicated that fluorescein is extended laterally from the DNA [24••]. Changing the tether length could give better definition to this position, and provide additional distance constraints.
DNA structure
0 0
20
40 60 80 Distance (R) / Å
100
Current Opinion in Chemical Biology
Spectral overlap and the distance dependence of energy transfer for a donor−acceptor pair of fluorophores. (a) Spectral overlap between fluorescein and Cy3 terminally attached to duplex DNA. A normalised fluorescence emission spectrum of fluorescein-labelled DNA (F) and an absorption spectrum of Cy3-labelled DNA (Cy3) are plotted together to show the overlap region. The fluorescence emission spectrum of fluorescein-labelled DNA was recorded with excitation at 490 nm. (b) The dependence of FRET efficiency on inter-fluorophore distance. The theoretical profile is calculated for an R0 of 56 Å, corresponding to a fluorescein−Cy3 pair. Vertical lines have been drawn at EFRET of 0.9 and 0.1, showing the range of distances accessible using this donor−acceptor pair.
of association of fluorophores attached at the 5′-termini of double-stranded DNA (dsDNA). Tetramethylrhodamine typically has an anisotropy of around 0.25 when terminally attached to DNA, indicating that it is relatively constrained by interaction with the DNA. Clegg and co-workers [27] made a detailed study of the photophysics of 5-carboxytetramethylrhodamine attached via a six-carbon linker to DNA. They found that three states were required to explain their data (including one non-fluorescent state), the distribution of which was dependent on temperature and ionic strength. However, they suggest that the different states probably do not reflect widely different spatial positioning of the rhodamine group. The cyanine dyes are probably more constrained by terminal attachment to dsDNA, frequently having anisotropy values of 0.3 or higher. Studies of energy transfer in a series
A wide variety of structures can be adopted by dsDNA in the laboratory, although it is not clear that they are all exploited in a biological context. Almost every aspect of DNA can potentially be modified, including the helical repeat and handedness, the trajectory of the axis, the number of strands and the manner of their association, and the continuity of the axis. A full definition of these structures requires the analysis of long-range features, and thus FRET has an important role to play in their study. Of course, the most important DNA structure is the simple duplex. FRET has been used to demonstrate its helical geometry [20], and time-resolved measurements have been employed to analyse the distribution of distances [28,29]. FRET has also been used to follow the hybridisation of the component strands [30,31]. This approach has been exploited to generate technology for the detection of PCR products, hybridisation in cells and karyotyping [32–35]. Vámosi and Clegg [36••] have made a detailed analysis of the thermal dissociation of duplex DNA species using a variety of fluorescence methods. Duplexes of different length terminally labelled with fluorescein and tetramethylrhodamine were analysed by fluorescence intensity, anisotropy and FRET as a function of temperature. It was found that the melting of short duplexes (≤ 20 bp) could be analysed as a two-state process, but that of a 34 bp duplex was more complex and required the use of a zipper model. The methods gave consistent results, and values of ∆H and ∆S per base pair in good agreement with those determined by unrelated methods. This is important, because it demonstrates the reliability of data obtained using fluorescence methods. This study underlines the importance of understanding the photophysics of
FRET as a structural tool for nucleic acids Lilley and Wilson
Figure 3
511
Figure 4
(a) D
A
D
A
S15
(b)
5 µm The structure of Cy3 (shown in magenta) attached to the 5′ terminus of dsDNA [24••]. The minimised average conformation of Cy3 covalently attached to a 10 bp DNA duplex via a three-carbon linker deduced from NMR studies is shown. (a) Side view of the helix, showing stacking of the terminally attached Cy3 onto the end of the helix in the manner of a pseudo basepair. (b) Close-up view of Cy3 stacked onto the terminal G•C basepair of the helix. Note the good overlap between the distal indole ring of the Cy3 and the guanine base.
the fluorophores used, because the analysis depended on the earlier study of tetramethylrhodamine attached to dsDNA [27]. Of course, the fluorophore itself may influence the stability of the duplex (usually a stabilisation would be expected), and this has been shown to be significant in a recent study of peptide nucleic acid (PNA) hybridisation [37]. A competitive FRET titration method has recently been used to extract free energies of base mismatches in a sensitive manner [38]. Axial bending or kinking is a common distortion in nucleic acids, and one that is well suited to analysis by FRET. Kinking due to base bulges has been studied in DNA and RNA [16,39••], showing that substantial axial bending can be achieved. DNA can also adopt a bent or curved trajectory by
Protein-induced folding of an RNA three-way junction observed by FRET in individual molecules. The junction occurs in 16S rRNA, and serves as the binding site for the ribosomal protein S15. (a) Schematic of the system. The RNA junctions are labelled with fluorescein (donor; D) and Cy3 (acceptor; A), and attached to a glass surface via a biotinstreptavidin linkage. In the unfolded conformation, the fluorophores are relatively far apart. Binding of the ribosomal S15 protein induces the formation of a new conformation in which the fluorophores are closer together, such that efficient energy transfer takes place. The slide is examined in a scanning confocal microscope, illuminated by a laser at 488 nm. (b) Image of individual molecules of immobilised RNA junctions after addition of S15 protein. The donor and acceptor fluorescence are imaged separately using false colour, and overlayed. The unfolded molecules contain predominantly fluorescein fluorescence, and are coloured green. On folding, the donor fluorescence is strongly quenched due to energy transfer, leaving predominantly acceptor fluorescence, coloured red. It is clear that this field contains both unfolded and protein-induced folded junctions. Reproduced from [91••] with permission. Copyright (1999) National Academy of Sciences, USA.
virtue of its sequence. Langowski and co-workers [40] have studied the bending of a 31 bp duplex containing three correctly phased oligoadenine tracts. The observed efficiency of energy transfer between terminally attached fluorophores was consistent with sequence-directed bending,
512
Analytical techniques
the magnitude of which was strongly salt-dependent. A right→left handed transition between B-DNA and Z-DNA conformations was detected using FRET by Jares-Erijman and Jovin [15], and this observation has recently been exploited in a molecular device [41]. Multi-stranded structures have been studied by FRET. This includes the formation of triplex structures [42,43], guanine tetraplex [44•] and i-motif structures [45•]. The application of FRET to the analysis of branched DNA junctions has been discussed above. Time-resolved FRET measurements have been used in combination with NMR to demonstrate the existence of different populations of alternative stacking conformers in four-way DNA junctions of different sequence [46].
RNA folding RNA exists largely as a formally single-stranded species, and it can therefore fold into a wide variety of different conformations. RNA structure is integral to its biological function, and there is a huge interest worldwide in RNA folding processes. Long-range distance information is very valuable in the analysis of RNA structure and folding, and therefore FRET has an important role to play in these studies. Branchpoints are very important architectural elements in RNA, but their study is not as advanced as that of the corresponding structures in DNA. Four-way RNA junctions have been analysed using FRET, showing that they undergo pairwise stacking of helical arms to form one of two possible conformers dependent on the choice of stacking partners [47,48]. These stacked species undergo a rotation into an antiparallel form induced by the binding of group II metal cations with a Hill coefficient of 1. One especially exciting feature of RNA function is the generation of catalytic activity in the ribozymes [49]. FRET has made a particularly significant contribution in this area. The hammerhead ribozyme is a kind of elaborated threeway junction [50,51]. Eckstein and co-workers [52] used FRET to analyse the global conformation, generating a structure that was in good agreement with contemporary crystal structures [53,54] and other approaches [55]. FRET was later used to dissect the folding process into two events [12] that could be rationalised in terms of the crystal structures. The first stage of folding was induced by the binding of a metal ion with an apparent association constant in the region of 10,000 M−1. This corresponds to the formation of a domain involved in the coaxial alignment of two helical stems. Binding of a second ion with 10-fold lower affinity induced the second stage of folding, in which the catalytic core of the ribozyme was folded. FRET was used to characterise a number of mutant sequences in which folding was blocked at various stages [56•]. In its natural form, the hairpin ribozyme is based on a fourway helical junction [57]. Two consecutive arms of this junction contain loops that were believed to interact to
generate the active ribozyme, but the first physical confirmation of this came from FRET experiments [23•,58,59]. Ribozyme activity can be obtained in the junction form, or in a minimised hinged form comprising the two loop-carrying duplexes connected by a single strand. Burke and co-workers [58] have used FRET to characterise the equilibrium between the form in which the loops are interacting, and a competing form in which the two duplexes appear to be stacked coaxially. They measured rates for the ‘docking’ process of the order of 0.5 min−1. The junction form of the ribozyme undergoes a folding process in which the stacked arms rotate around the junction to form the structure with the interacting loops. Apparent binding isotherms obtained from the FRET data suggest that this is induced by divalent metal ions binding with high affinity and a Hill coefficient of 2 [59]. The conformation of the junction is critical to the action of the ribozyme, and any changes that alter the conformation of the junction impair cleavage activity of the ribozyme [60]. The importance of the junction was confirmed by timeresolved FRET experiments, in which it was found that the four-way junction significantly stabilised the active conformation of the ribozyme [61•]. The U1A protein binds to an element in the 3′ untranslated region (3′-UTR) of its own pre-mRNA to regulate polyadenylation [62]. The target comprises two asymmetric internal loops (termed 3′-UTR boxes, each composed of seven bases opposing one) on different strands and separated by four basepairs. This structure and its interaction with the U1A protein have been extensively studied by NMR by Varani and co-workers [63,64]. The global structure was not precisely determined, however, and FRET was applied to the problem [39••,65]. The known position of Cy3 and the effective location of the fluorescein were used together with molecular modelling to construct a model of a single 3′-UTR box, and the complete element. The included bend angle of a single box was determined to be 100°, while the outer helical arms of the complete 3′-UTR were calculated to be related by a dihedral angle of 35°. It is certain that analysis of RNA folding will necessitate the investigation of increasingly larger species, and this will require solutions to a new set of problems. The chemical synthesis of RNA on which these studies depend is presently limited to around 60 nucleotides, though this can probably be extended to 90 nucleotides with new forms of oligonucleotide chemistry. But the analysis of larger species will require the development of efficient ligation, and the use of semi-synthetic species. Furthermore, it will probably be necessary to attach fluorophores at positions other than termini, such as internal phosphates and bases.
Protein−nucleic acid interactions FRET has been used to explore a number of enzyme reactions in which nucleic acids are the substrate. This includes the cleavage of DNA by nucleases [66], and unwinding induced by helicases [67,68], from which kinetic
FRET as a structural tool for nucleic acids Lilley and Wilson
properties can be measured. Rates of RecA-mediated strand exchange were measured using FRET between different DNA strands, which were labelled with fluorescein and hexachlorofluorescein [69]. Alteration in protein−protein contacts of the sliding clamp during T4 DNA polymerase passage have been recently observed using FRET [70,71•]. Furey et al. [72] investigated energy transfer between DNA polymerase site-specifically labelled with fluorescein and a DNA template internally labelled on the 5 position of dU with tetramethylrhodamine. They observed a sinusoidal modulation of transfer efficiency as a function of the position of the acceptor along the DNA, which is indicative of a helical conformation of the singlestranded template. Heyduk and co-workers [73•,74] have explored the interactions of the Escherichia coli σ70 subunit of RNA polymerase using FRET. They have measured energy transfer between a europium chelate attached to various positions in the protein and Cy5 attached either to the 5′ or 3′ end of a DNA oligonucleotide. By this means, they identified a close contact between the 5′ terminus and region 2.4 of σ70, and deduced the orientation of the DNA on the protein. FRET is especially suited to the analysis of nucleic acid distortion induced by the binding of proteins. This generally requires that the interaction does not directly involve the fluorophores. For example, fluorescein is frequently used as the donor to tetramethylrhodamine or Cy3, and the assumption that κ2 = 2/3 requires that it remains mobile, as can be readily checked by its anisotropy. Provided that this criterion is met, FRET can be employed reliably to monitor protein-induced structural changes in DNA or RNA. A number of DNA−protein interactions involve substantial distortion of the DNA in the bound form, and there have been several recent FRET studies in different systems. The central protein of eukaryotic transcription, TATA-box-binding protein (TBP), binds to the TATA sequence of pol II promoters and generates both local unwinding and axial bending [75,76]. Parkhurst and coworkers [77,78•] have used FRET between fluorophores attached to the two ends of a 14 bp oligonucleotide containing the sequence TATAAAA. Upon binding TBP from Saccharomyces cerevisiae, a decrease in the end-to-end distance was observed, consistent with the expected bending of the DNA [77]. The interaction was studied in real time using stopped-flow techniques [78•] and detailed kinetics were analysed. Two intermediate states were observed, each of which was bent to a similar degree as the final state. The authors conclude that binding and bending events cannot be separated at this level. The high mobility group (HMG) proteins HMG1 and HMG2 contain an 80-amino-acid domain called the HMG box that has been found in a number of DNA-binding proteins and transcription factors [79]. Proteins such as SRY and LEF-1 bind in the minor groove of DNA, generating a bending of similar magnitude to that induced by
513
TBP [80,81]. However, proteins containing the HMG box also bind with elevated affinity to DNA that is already distorted by the presence of a cis-platinum adduct [82], or to a four-way junction [83]. Similar properties are exhibited by other architectural DNA-binding proteins such as HU [84]. The geometry of DNA bound by HMG box proteins has been analysed by FRET, with somewhat conflicting results [85,86•]. Heyduk et al. [85] estimated a bending angle of 150°, which was induced by the binding of Chironomus HMG1 to a duplex DNA fragment, using energy transfer between terminally-attached Cy5 and europium chelate. One difficulty with studying the HMG1-type proteins is a lack of sequence specificity, and therefore an accurate localisation of the protein on the DNA. Travers, Diekmann and co-workers [86•] used a DNA fragment containing a two-base bulge, which would position the HMG box on the DNA. They bound HMG-D protein to an 18 bp duplex, with a centrally located A2 bulge, terminally labelled with fluorescein and tetramethylrhodamine. As discussed above, such a bulged duplex would be already bent by virtue of the additional bases, but the authors estimated that the bend angle became around 90° in the complex with the protein. This value is closer to other estimates of the bending angle induced by HMG box proteins. The assumption that tetramethylrhodamine is located in a position stacked on the end of the helix is questionable, however, and clearly this will influence any interpretation of inter-fluorophore distances. In a similar study Diekmann and co-workers [87•] studied the effect of binding integration host factor (IHF) to dsDNA. In this case, the binding site is relatively large, and a 55 bp duplex terminally labelled with fluorescein and tetramethylrhodamine was used as before. From the efficiency of energy transfer they modelled the bound complex as a heavily distorted U-shaped structure. Lippard and co-workers [88•] have studied the binding of the HMG box B from HMG1 to a DNA containing an intrastrand cis-platinum adduct. They studied energy transfer between fluorescein and rhodamine attached at the 5′-termini of a 20 bp duplex containing a central GpG sequence modified by cis-diamminedichloroplatinum (II). From measured fluorescein lifetimes they estimated that the binding of the HMG box shortened the interfluorophore distance from 66 Å to 55 Å, and from molecular modelling they calculated a protein-induced bend angle of 80−95°. Stopped-flow experiments were used to measure on- and off-rates for binding; from these studies a rate of binding that was close to the diffusion-limit was estimated. At present there are relatively few examples of RNA−protein interactions that have been studied by FRET. Given the power of the method, however, this must increase in the near future. Grainger et al. [39••] studied the effect of binding the RNA-binding domain of the U1A protein on the global structure of the U1A pre-mRNA 3′-UTR (the motif discussed above) using FRET between terminally-attached fluorescein and Cy3 fluorophores. On binding, an increase
514
Analytical techniques
in efficiency of energy transfer from 0.31 to 0.53 was observed. Because the anisotropy of the fluorescein was essentially unaltered, this increase in efficiency was interpreted in terms of a significant change in the conformation of the RNA. A precise model of the complex could not be constructed because the single distance constraint would leave this under-determined. The change in FRET efficiency could reflect alterations in the bend angles of the individual UTR boxes, the dihedral angle between the outer arms, or both. The problem of insufficiency of constraints becomes especially acute as we move into three dimensions as opposed to a planar bend, and it is instructive to ask how more distance information might be obtained to give a better definition of the structure. One simple method is to exchange the positions of donor and acceptor. Another is to use helices of different length. This alters the rotational placing of the fluorescein in particular, and potentially provides a whole series of distance constraints. Other approaches such as altering tether length are also possible.
FRET and the single molecule The study of single molecules can provide information about heterogeneous populations and dynamic processes that is unavailable from ensemble measurements. FRET between a single donor–acceptor pair (single-pair FRET; spFRET) was first demonstrated by Weiss and co-workers [89] using near-field scanning optical microscopy. This study used complementary DNA oligonucleotides 5′ labelled with tetramethylrhodamine and Texas red fluorophores, dried onto glass slides. The authors observed a highly heterogeneous population of molecules, but uncertainties about the relatively fixed orientations of the fluorophores (κ2 ≠ 2/3) and the effects of drying on DNA structure precluded quantitative distance measurements. Subsequent spFRET studies have investigated molecules in the aqueous phase but tethered to a surface, which permits the observation of an individual molecule for an extended time. Through numerous observations, data are obtained that are consistent with ensemble measurements but which also display the range of dynamic fluctuations expected in a population. Weiss and co-workers [90•] investigated the conformational dynamics and mechanism of Staphylococcal nuclease, using spFRET between labelled enzyme and DNA substrate to demonstrate a 3′ to 5′ processive cleavage reaction. Williamson, Chu and co-workers [91••] have studied the protein-induced and magnesiumion-induced folding of an RNA three-way junction (Figure 4). They obtained thermodynamic and kinetic data consistent with ensemble observations and found evidence for inhomogeneity in folding, possibly due to multiple intermediate states. By oscillating the magnesium ion concentration they were able to observe reversible folding of a single molecule. The folding and catalysis of the group I ribozyme of Tetrahymena has been studied by Herschlag, Chu and co-workers [92••]. Multiple folding pathways and intermediate folding states were observed, including a novel state and a fast folding pathway not observed by ensemble
measurements. This work also illustrates a limitation of all FRET studies, namely the model-dependent interpretation of a small number of distance constraints. Two intermediate states were observed with similar FRET efficiencies, distinguishable by different rate constants. Further studies will be required to differentiate these two states in order to describe fully the folding pathways. Freely diffusing single molecules can also be observed as they traverse the excitation volume of a confocal microscope. Although the short time of observation limits the study of individual molecules, populations of molecules can be observed free of potentially perturbing surface effects. Deniz et al. [93••] have used this technique to measure the mean distance and distribution width of FRET efficiencies for donor−acceptor labelled DNA duplexes and to distinguish subpopulations within a mixture. Such information can potentially be obtained from time-resolved FRET studies; however, this requires assumptions to be made about the underlying population distribution. In contrast, spFRET permits the direct observation of the population distribution.
Conclusions and perspectives FRET is particularly useful in structural studies of nucleic acids, the extended nature of which makes long-range distance information of some kind essential. Methods that provide distances of limited range will leave the global structure underdetermined. Thus FRET is potentially complementary to NMR, and comparisons with structural information coming from dipolar couplings will be of great interest in the near future. At present, there remain some difficulties in extracting absolute distance information with confidence, but this situation is improving rapidly. FRET has enjoyed some spectacular successes, such as with the four-way DNA junction. It is proving to be increasingly useful in folding studies of RNA, where the method has a great deal to offer in the future. The most exciting new area for FRET is the extension to single-molecule studies, which exploits the great sensitivity of fluorescence. This frees us from the requirement to deduce the properties of individual molecules from those in bulk, and is a tremendous leap forward for the biological chemist. We are likely to see fantastic developments in this area over the coming few years.
Acknowledgements We thank T Ha, J Williamson and S Chu for provision of Figure 4, Dan Herschlag for provision of data prior to publication, and DG Norman for discussion. Work in Dundee is supported by the Cancer Research Campaign and the Biotechnology and Biological Sciences Research Council.
References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as:
• of special interest •• of outstanding interest 1.
Stryer L, Haugland RP: Energy transfer: a spectroscopic ruler. Proc Natl Acad Sci USA 1967, 58:719-726.
FRET as a structural tool for nucleic acids Lilley and Wilson
2.
Fairclough RH, Cantor CR: The use of singlet-singlet energy transfer to study macromolecular assemblies. Methods Enzymol 1978, 48:347-379.
3.
Förster T: Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann Phys 1948, 2:55-75. [Title translation: Intermolecular energy transfer and fluorescence.]
4.
Clegg RM: Fluorescence resonance energy transfer and nucleic acids. Methods Enzymol 1992, 211:353-388.
5.
Murchie AIH, Clegg RM, von Kitzing E, Duckett DR, Diekmann S, Lilley DMJ: Fluorescence energy transfer shows that the four-way DNA junction is a right-handed cross of antiparallel molecules. Nature 1989, 341:763-766.
6.
Clegg RM, Murchie AIH, Zechel A, Carlberg C, Diekmann S, Lilley DMJ: Fluorescence resonance energy transfer analysis of the structure of the four-way DNA junction. Biochemistry 1992, 31:4846-4856.
7.
Nowakowski J, Shim PJ, Prasad GS, Stout CD, Joyce GF: Crystal structure of an 82 nucleotide RNA-DNA complex formed by the 10-23 DNA enzyme. Nat Struct Biol 1999, 6:151-156.
8.
Ortiz-Lombardía M, González A, Erijta R, Aymamí J, Azorín F, Coll M: Crystal structure of a DNA Holliday junction. Nat Struct Biol 1999, 6:913-917.
9.
Eichman BF, Vargason JM, Mooers BHM, Ho PS: The Holliday junction in an inverted repeat DNA sequence: sequence effects on the structure of four-way junctions. Proc Natl Acad Sci USA 2000, 97:3971-3976.
515
23. Murchie AIH, Thomson JB, Walter F, Lilley DMJ: Folding of the • hairpin ribozyme in its natural conformation achieves close physical proximity of the loops. Mol Cell 1998, 1:873-881. FRET provided physical evidence for the close approach of the loops in the hairpin ribozyme. 24. Norman DG, Grainger RJ, Uhrin D, Lilley DMJ: The location of •• cyanine-3 on double-stranded DNA; importance for fluorescence resonance energy transfer studies. Biochemistry 2000, 39:6317-6324. Fluorescence and NMR data showing that Cy3 is stacked onto the end of double-stranded DNA helices, thus providing a known measuring point in FRET experiments. 25. Harrington RE: The flow birefringence of persistence length deoxyribonucleic acid. Hydrodynamic properties, optical anisotropy, and hydration shell anisotropy. J Am Chem Soc 1970, 92:6957-6964. 26. Wu PG, Fujimoto BS, Song L, Schurr JM: Effect of ethidium on the torsion constants of linear and supercoiled DNA. Biophys Chem 1991, 41:217-236. 27.
Vámosi G, Gohlke C, Clegg RM: Fluorescence characteristics of 5-carboxytetramethylrhodamine linked covalently to the 5′′ end of oligonucleotides: multiple conformers of single-stranded and double-stranded dye-DNA complexes. Biophys J 1996, 71:972-994.
28. Hochstrasser RA, Chen SM, Millar DP: Distance distribution in a dye-linked oligonucleotide determined by time-resolved fluorescence energy transfer. Biophys Chem 1992, 45:133-141.
10. Lilley DMJ, Norman DG: The Holliday junction is finally seen with crystal clarity. Nat Struct Biol 1999, 6:897-899.
29. Parkhurst KM, Parkhurst LJ: Donor-acceptor distance distributions in a double-labeled fluorescent oligonucleotide both as a single strand and in duplexes. Biochemistry 1995, 34:293-300.
11. Clegg RM, Murchie AIH, Zechel A, Lilley DMJ: The solution structure of the four-way DNA junction at low salt concentration; a fluorescence resonance energy transfer analysis. Biophys J 1994, 66:99-109.
30. Cardullo RA, Agrawal S, Flores C, Zamecnik PC, Wolf DE: Detection of nucleic acid hybridisation by nonradiative fluorescence resonance energy transfer. Proc Natl Acad Sci USA 1988, 85:8790-8794.
12. Bassi GS, Murchie AIH, Walter F, Clegg RM, Lilley DMJ: Ion-induced folding of the hammerhead ribozyme: a fluorescence resonance energy transfer study. EMBO J 1997, 16:7481-7489.
31. Parkhurst KM, Parkhurst LJ: Kinetic studies by fluorescence resonance energy transfer employing a double-labeled oligonucleotide: hybridization to the oligonucleotide complement and to single-stranded DNA. Biochemistry 1995, 34:285-292.
13. Stühmeier F, Lilley DMJ, Clegg RM: The effect of bulges on the stability of three-way DNA junctions studied by fluorescence techniques. Biochemistry 1997, 36:13539-13551.
32. Tyagi S, Bratu DP, Kramer FR: Multicolor molecular beacons for allele discrimination. Nat Biotechnol 1998, 16:49-53.
14. Eis PS, Millar DP: Conformational distributions of a 4-way DNA junction revealed by time-resolved fluorescence resonance energy transfer. Biochemistry 1993, 32:13852-13860. 15. Jares-Erijman EA, Jovin TM: Determination of DNA helical handedness by fluorescence resonance energy transfer. J Mol Biol 1996, 257:597-617. 16. Gohlke C, Murchie AIH, Lilley DMJ, Clegg RM: The kinking of DNA and RNA helices by bulged nucleotides observed by fluorescence resonance energy transfer. Proc Natl Acad Sci USA 1994, 91:11660-11664. 17.
Lilley DMJ, Clegg RM, Diekmann S, Seeman NC, von Kitzing E, Hagerman P: Nomenclature Committee of the International Union of Biochemistry: a nomenclature of junctions and branchpoints in nucleic acids. Recommendations 1994. Eur J Biochem 1995, 230:1-2.
18. Stühmeier F, Welch JB, Murchie AIH, Lilley DMJ, Clegg RM: The global structure of three-way DNA junctions with and without bulges: fluorescence studies. Biochemistry 1997, 36:13530-13538. 19. Yang MS, Millar DP: Conformational flexibility of three-way DNA junctions containing unpaired nucleotides. Biochemistry 1996, 35:7959-7967. 20. Clegg RM, Murchie AIH, Zechel A, Lilley DMJ: Observing the helical geometry of double-stranded DNA in solution by fluorescence resonance energy transfer. Proc Natl Acad Sci USA 1993, 90:2994-2998. 21. Wu P, Brand L: Orientation factor in steady-state and time resolved resonance energy transfer measurements. Biochemistry 1992, 31:7939-7947. 22. Hung SC, Mathies RA, Glazer AN: Cyanine dyes with high absorption cross section as donor chromophores in energy transfer primers. Anal Biochem 1996, 243:15-27.
33. Hung SC, Mathies RA, Glazer AN: Comparison of fluorescence energy transfer primers with different donor-acceptor dye combinations. Anal Biochem 1998, 255:32-38. 34. Ota N, Hirano K, Warashina M, Andrus A, Mullah B, Hatanaka K, Taira K: Determination of interactions between structured nucleic acids by fluorescence resonance energy transfer (FRET): selection of target sites for functional nucleic acids. Nucleic Acids Res 1998, 26:735-743. 35. Sixou S, Szoka FC Jr, Green GA, Giusti B, Zon G, Chin DJ: Intracellular oligonucleotide hybridization detected by fluorescence resonance energy transfer (FRET). Nucleic Acids Res 1994, 22:662-668. 36. Vámosi G, Clegg RM: The helix-coil transition of DNA duplexes •• and hairpins observed by multiple fluorescence parameters. Biochemistry 1998, 37:14300-14316. Detailed analysis of the thermal dissociation of duplex DNA species using fluorescence intensity, anisotropy and FRET. Enthalpy and entropy changes per basepair were found to be in good agreement with those determined by unrelated methods. This shows the reliability of data obtained using fluorescence methods, and underlines the importance of understanding the photophysics of the fluorophores used. 37.
Ratilainen T, Holmen A, Tuite E, Haaima G, Christensen L, Nielsen PE, Norden B: Hybridization of peptide nucleic acid. Biochemistry 1998, 37:12331-12342.
38. Gelfand CA, Plum GE, Mielewczyk S, Remeta DP, Breslauer KJ: A quantitative method for evaluating the stabilities of nucleic acids. Proc Natl Acad Sci USA 1999, 96:6113-6118. 39. Grainger RJ, Norman DG, Lilley DMJ: Binding of U1A protein to the •• 3′′ untranslated region of its pre-mRNA. J Mol Biol 1999, 288:585-594. The global structure of the U1A-binding element of the U1A pre-mRNA 3′UTR was deduced using FRET analysis. FRET was also used to demonstrate a change in the structure on binding the U1A protein, and was one of the first examples of FRET used to study RNA−protein interactions.
516
Analytical techniques
40. Toth K, Sauermann V, Langowski J: DNA curvature in solution measured by fluorescence resonance energy transfer. Biochemistry 1998, 37:8173-8179.
These are analysed in terms of two forms — the folded and open conformations. It is found that the junction conformation (see [23•]) is the most stable folded form.
41. Mao C, Sun W, Shen Z, Seeman NC: A nanomechanical device based on the B-Z transition of DNA. Nature 1999, 397:144-146.
62. van Gelder CW, Gunderson SI, Jansen EJ, Boelens WC, Polycarpou-Schwarz M, Mattaj IW, van Venrooij WJ: A complex secondary structure in U1A pre-mRNA that binds two molecules of U1A protein is required for regulation of polyadenylation. EMBO J 1993, 12:5191-5200.
42. Mergny JL, Garestier T, Rougee M, Lebedev AV, Chassignol M, Thuong NT, Helene C: Fluorescence energy transfer between two triple helix-forming oligonucleotides bound to duplex DNA. Biochemistry 1994, 33:15321-15328. 43. Yang M, Ghosh SS, Millar DP: Direct measurement of thermodynamic and kinetic parameters of DNA triple helix formation by fluorescence spectroscopy. Biochemistry 1994, 33:15329-15337. 44. Simonsson T, Sjöback R: DNA tetraplex formation studied with • fluorescence resonance energy transfer. J Biol Chem 1999, 274:17379-17383. FRET used to demonstrate antiparallel guanine tetraplex formation. 45. Mergny JL: Fluorescence energy transfer as a probe for tetraplex • formation: the i-motif. Biochemistry 1999, 38:1573-1581. FRET used to demonstrate i-motif tetraplex formation in oligocytosinecontaining sequences. 46. Miick SM, Fee RS, Millar DP, Chazin WJ: Crossover isomer bias is the primary sequence-dependent property of immobilized Holliday junctions. Proc Natl Acad Sci USA 1997, 94:9080-8084. 47.
64. Howe PW, Allain FH, Varani G, Neuhaus D: Determination of the NMR structure of the complex between U1A protein and its RNA polyadenylation inhibition element. J Biomol NMR 1998, 11:59-84. 65. Grainger RJ, Murchie AIH, Norman DG, Lilley DMJ: Severe axial bending of RNA induced by the U1A binding element present in the 3′′ untranslated region of the U1A mRNA. J Mol Biol 1997, 273:84-92. 66. Ghosh SS, Eis PS, Blumeyer K, Fearon K, Millar DP: Real time kinetics of restriction endonuclease cleavage monitored by fluorescence resonance energy transfer. Nucleic Acids Res 1994, 22:3155-3159. 67.
Walter F, Murchie AIH, Duckett DR, Lilley DMJ: Global structure of four-way RNA junctions studied using fluorescence resonance energy transfer. RNA 1998, 4:719-728.
48. Walter F, Murchie AIH, Lilley DMJ: The folding of the four-way RNA junction of the hairpin ribozyme. Biochemistry 1998, 37:17629-17636. 49. Lilley DMJ: Structure, folding and catalysis in the small nucleolytic ribozymes. Curr Opin Struct Biol 1999, 9:330-338. 50. Forster AC, Symons RH: Self-cleavage of plus and minus RNAs of a virusoid and a structural model for the active sites. Cell 1987, 49:211-220. 51. Hazeloff JP, Gerlach WL: Simple RNA enzymes with new and highly specific endoribonuclease activities. Nature 1988, 334:585-591. 52. Tuschl T, Gohlke C, Jovin TM, Westhof E, Eckstein F: A threedimensional model for the hammerhead ribozyme based on fluorescence measurements. Science 1994, 266:785-789. 53. Pley HW, Flaherty KM, McKay DB: Three-dimensional structure of a hammerhead ribozyme. Nature 1994, 372:68-74. 54. Scott WG, Finch JT, Klug A: The crystal structure of an all-RNA hammerhead ribozyme: a proposed mechanism for RNA catalytic cleavage. Cell 1995, 81:991-1002. 55. Bassi G, Møllegaard NE, Murchie AIH, von Kitzing E, Lilley DMJ: Ionic interactions and the global conformations of the hammerhead ribozyme. Nat Struct Biol 1995, 2:45-55. 56 •
Bassi GS, Møllegaard NE, Murchie AIH, Lilley DMJ: RNA folding and misfolding of the hammerhead ribozyme. Biochemistry 1999, 38:3345-3354. FRET has been used to show that the folding of the hammerhead ribozyme occurs in two single ion-induced stages [12], corresponding to the formation of domains 2 and 1. This work uses FRET to analyse the folding properties of various point mutations in the hammerhead core. A series of mutants that are blocked at different stages of the folding process are presented. 57.
63. Gubser CC, Varani G: Structure of the polyadenylation regulatory element of the human U1A pre-mRNA 3′′-untranslated region and interaction with the U1A protein. Biochemistry 1996, 35:2253-2267.
Hampel A, Tritz R: RNA catalytic properties of the minimum (-) sTRSV sequence. Biochemistry 1989, 28:4929-4933.
58. Walter NG, Hampel KJ, Brown KM, Burke JM: Tertiary structure formation in the hairpin ribozyme monitored by fluorescence resonance energy transfer. EMBO J 1998, 17:2378-2391. 59. Walter F, Murchie AIH, Thomson JB, Lilley DMJ: Structure and activity of the hairpin ribozyme in its natural junction conformation; effect of metal ions. Biochemistry 1998, 37:14195-14203. 60. Thomson JB, Lilley DMJ: The influence of junction conformation on RNA cleavage by the hairpin ribozyme in its natural junction form. RNA 1999, 5:180-187. 61. Walter NG, Burke JM, Millar DP: Stability of hairpin ribozyme • tertiary structure is governed by the interdomain junction. Nat Struct Biol 1999, 6:544-549. The hairpin ribozyme in its hinged form and variants have been analysed using time-resolved FRET studies to calculate conformational populations.
Bjornson KP, Amaratunga M, Moore KJ, Lohman TM: Single-turnover kinetics of helicase-catalyzed DNA unwinding monitored continuously by fluorescence energy transfer. Biochemistry 1994, 33:14306-14316.
68. Bjornson KP, Hsieh J, Amaratunga M, Lohman TM: Kinetic mechanism for the sequential binding of two single-stranded oligodeoxynucleotides to the Escherichia coli Rep helicase dimer. Biochemistry 1998, 37:891-899. 69. Gumbs OH, Shaner SL: Three mechanistic steps detected by FRET after presynaptic filament formation in homologous recombination. ATP hydrolysis required for release of oligonucleotide heteroduplex product from RecA. Biochemistry 1998, 37:11692-11706. 70. Soumillion P, Sexton DJ, Benkovic SJ: Clamp subunit dissociation dictates bacteriophage T4 DNA polymerase holoenzyme disassembly. Biochemistry 1998, 37:1819-1827. 71. Alley SC, Shier VK, Abel-Santos E, Sexton DJ, Soumillion P, • Benkovic SJ: Sliding clamp of the bacteriophage T4 polymerase has open and closed subunit interfaces in solution. Biochemistry 1999, 38:7696-7709. FRET is used to measure distances between amino acids in bacteriophage T4 gp45 in open and closed states. 72. Furey WS, Joyce CM, Osborne MA, Klenerman D, Peliska JA, Balasubramanian S: Use of fluorescence resonance energy transfer to investigate the conformation of DNA substrates bound to the Klenow fragment. Biochemistry 1998, 37:2979-2990. 73. Heyduk E, Heyduk T: Architecture of a complex between the σ70 • subunit of Escherichia coli RNA polymerase and the nontemplate strand oligonucleotide. Luminescence resonance energy transfer study. J Biol Chem 1999, 274:3315-3322. Energy transfer between europium attached to various sites on the sigma subunit of RNA polymerase, and Cy5 attached to DNA is described. 74. Callaci S, Heyduk E, Heyduk T: Core RNA polymerase from E. coli induces a major change in the domain arrangement of the σ70 subunit. Mol Cell 1999, 3:229-238. 75. Kim JL, Nikolov DB, Burley SK: Co-crystal structure of TBP recognizing the minor groove of a TATA element. Nature 1993, 365:520-527. 76. Kim YC, Geiger JH, Hahn S, Sigler PB: Crystal structure of a yeast TBP/TATA-box complex. Nature 1993, 365:512-520. 77.
Parkhurst KM, Brenowitz M, Parkhurst LJ: Simultaneous binding and bending of promoter DNA by the TATA binding protein: real time kinetic measurements. Biochemistry 1996, 35:7459-7465.
78. Parkhurst KM, Richards RM, Brenowitz M, Parkhurst LJ: Intermediate • species possessing bent DNA are present along the pathway to formation of a final TBP-TATA complex. J Mol Biol 1999, 289:1327-1341. Detailed kinetic study of the interaction between yeast TATA-box-binding protein and dsDNA, using FRET between fluorophores terminally-attached to the DNA. End-to-end energy transfer was observed consistent with the expected
FRET as a structural tool for nucleic acids Lilley and Wilson
bending of the DNA. Stopped-flow analysis revealed bent intermediates in the binding process. 79. Bianchi ME: The HMG-box domain. In DNA−Protein: Structural Interactions. Edited by Lilley DMJ. Oxford: IRL Press at Oxford University Press; 1995:177-200. 80. Werner MH, Bianchi ME, Gronenborn AM, Clore GM: NMR spectroscopic analysis of the DNA conformation induced by the human testis determining factor SRY. Biochemistry 1995, 34:11998-12004. 81. Love JJ, Li X, Case DA, Giese K, Grosschedl R, Wright PE: Structural basis for DNA bending by the architectural transcription factor LEF-1. Nature 1995, 376:791-795. 82. Chow CS, Barnes CM, Lippard SJ: A single HMG domain in highmobility group 1 protein binds to DNAs as small as 20 base pairs containing the major cisplatin adduct. Biochemistry 1995, 34:2956-2964. 83. Bianchi ME, Beltrame M, Paonessa G: Specific recognition of cruciform DNA by nuclear protein HMG1. Science 1989, 243:1056-1059. 84. Pontiggia A, Negri A, Beltrame M, Bianchi ME: Protein HU binds specifically to kinked DNA. Mol Microbiol 1993, 7:343-350. 85. Heyduk E, Heyduk T, Claus P, Wisniewski JR: Conformational changes of DNA induced by binding of Chironomus high mobility group protein 1a (cHMG1a). Regions flanking an HMG1 box domain do not influence the bend angle of the DNA. J Biol Chem 1997, 272:19763-19770. 86. Lorenz M, Hillisch A, Payet D, Buttinelli M, Travers A, Diekmann S: • DNA bending induced by high mobility group proteins studied by fluorescence resonance energy transfer. Biochemistry 1999, 39:12150-12158. Bending of bulge-containing DNA induced by the binding of Drosophila HMG box domains. A bending angle of 90−100° was estimated from FRET between fluorophores terminally attached to the DNA. 87. •
Lorenz M, Hillisch A, Goodman SD, Diekmann S: Global structure similarities of intact and nicked DNA complexed with IHF measured in solution by fluorescence resonance energy transfer. Nucleic Acids Res 1999, 27:4619-4625. Bending of DNA induced by the binding of E. coli integration host factor measured using FRET.
517
88. Jamieson ER, Jacobson MP, Barnes CM, Chow CS, Lippard SJ: • Structural and kinetic studies of a cisplatin-modified DNA icosamer binding to HMG1 domain B. J Biol Chem 1999, 274:12346-12354. Bending of DNA containing a cisplatin adduct induced by the binding of an HMG box. A bending angle of 80−95° was deduced from FRET between fluorophores terminally attached to the DNA. 89. Ha T, Enderle T, Ogletree DF, Chemla DS, Selvin PR, Weiss S: Probing the interaction between two single molecules: fluorescence resonance energy transfer between a single donor and a single acceptor. Proc Natl Acad Sci USA 1996, 93:6264-6268. 90. Ha T, Ting AY, Liang J, WB C, Deniz AA, Chemla DS, Schultz PG, • Weiss S: Single-molecule fluorescence spectroscopy of enzyme conformational dynamics and cleavage mechanism. Proc Natl Acad Sci USA 1999, 96:893-898. The first single-molecule protein–nucleic acid study. The authors use FRET to measure the lifetimes of nuclease−DNA complexes. 91. Ha T, Zhuang X, Kim HD, Orr JW, Williamson JR, Chu S: Ligand•• induced conformational changes observed in single RNA molecules. Proc Natl Acad Sci USA 1999, 96:9077-9082. FRET in single RNA molecules attached to a surface. This study demonstrates the folding of a three-way RNA junction found in 16S rRNA on addition of magnesium ions or the binding of S15 protein. 92. Zhuang X, Bartley LE, Babcock HP, Russel R, Ha T, Herschlag D, •• Chu S: A single molecule study of RNA catalysis and folding. Science 2000, 288:2048-2051. Another important study showing the potential of spFRET, in this case for investigating the thermodynamics and kinetics of nucleic acid folding in the group I ribozyme. 93. Deniz AA, Dahan M, Grunwell JR, Ha T, Faulhaber AE, Chemla DS, •• Weiss S, Schultz PG: Single-pair fluorescence resonance energy transfer on freely diffusing molecules: observation of Forster distance dependence and subpopulations. Proc Natl Acad Sci USA 1999, 96:3670-3675. FRET in single RNA molecules in free solution. Analysis of the mean distance and distribution width of FRET efficiencies for donor−acceptor labelled DNA duplexes as they traverse the excitation volume of a confocal microscope.
