Determination of the distance between two spin labels attached to a macromolecule.

Abstract

An EPR "spectroscopic ruler" was developed using a series of a-helical polypeptides, each modified with two nitroxide spin labels. The EPR line broadening due to electron-electron dipolar interactions in the frozen state was determined using the Fourier deconvolution method. These dipolar spectra were then used to estimate the distances between the two nitroxides separated by 8-25 A. Results agreed well with a simple a-helical model. The standard deviation from the model system was 0.9 A in the range of [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] A. This technique is applicable to complex systems such as membrane receptors and channels, which are difficult to access with high-resolution NMR or x-ray crystallography, and is expected to be particularly useful for systems for which optical methods are hampered by the presence of lightinterfering membranes or chromophores.Many studies in structural biology are dependent on the physical techniques to measure distances in proteins and nucleic acids. X-ray crystallography and high-resolution NMR have been useful in determining the three-dimensional structures of relatively simple biological macromolecules. For complex systems such as membrane proteins, fluorescence energy transfer (FET) has been the main alternative for measuring distances up to 80 A. FET has been successful for studies of intermolecular organization in biological systems (1, 2), ligandreceptor interactions (3), and structures of nucleic acids (4).Recently, site-directed spin labeling EPR has become useful for studying proteins (5, 6). One or two native residues are mutated to cysteines, which are then labeled with thiol-specific nitroxide spin labels. This technique can also be used to study local secondary structure (7,8). Nucleic acids also appear to be amenable to spin labeling (9). Although spin labeling has been used to estimate distances in the past (10-12), no EPR "spectroscopic ruler" similar to that developed by Stryer and Haugland (13) for FET has been constructed or tested on model systems.In this work a convenient and accurate EPR method to determine distances between two site-specifically placed nitroxides in the range of 8-25 A in biomacromolecules is presented. In this method the pure dipolar spectrum for two interacting spins in the frozen state is directly Fourier deconvoluted from the dipolar broadened continuous-wave EPR spectrum. The average interspin distance and the variance of its distribution are obtained from this dipolar spectrum.The method was tested using a-helical peptides as a model system. The peptides were alanine-based helices with spinlabeled cysteines substituted for alanines at two locations from 1 to 13 residues apart. There is excellent agreement between the spin-spin distances from a simple model and experimental results in the range of 8-25 A. It is also shown that this method is useful for systems that have impurities of singly labeled species. Although this methodology is complementary to FET, EPR has the advantages of easier...

A novel 5′ displacement spin-labeling technique for electron paramagnetic resonance spectroscopy of RNA

“…The gel shift assays indicated that spin-labeled RRE and Rev form a specific complex, so we expected to see a broadening of the spectra due to the motional constraints placed on the spin label when Rev was added to the RRE constructs+ Indeed, broadening upon complex formation is clearly visible in Figure 3 as an increased broadening for RRE plus unlabeled Rev (dotted line)+ Additional broadening is seen in RRE plus labeled Rev (solid line) due to dipolar interactions, which we examined in detail at low temperatures (see below)+ The broadening due to complex formation was also observed for RRE2 and RRE3 (data not shown); however, the additional broadening due to dipolar interactions was not seen in RRE3, consistent with the separation (.25 Å) of the two probes predicted by the NMR structure (Rabenstein & Shin, 1995)+ The relative spectral broadening diminished when Rev was added in more than, or less than, the optimal ratio of 1:1 (data not shown)+ The peaks became less broad due to the presence of free Rev or RRE+ Ultimately, we are interested in using this new technique to find distances within complexed RNA structures and obtain information about RNA function+ To FIGURE 3. Room-temperature EPR absorption spectra of bound and unbound Rev and RRE1+ All spectra were taken at 22 8C in a Bruker ESP300 EPR spectrometer with a modulation amplitude of 0+5 G+ The broadening in the bound and singly labeled systems is due to decreased nitroxide mobility+ Additional broadening in the bound and doubly labeled system is due to spin-spin dipolar interactions+ measure distances between the spin labels in the RRERev complexes, EPR spectra were collected at Ϫ140 8C+ These absorption spectra are shown overlapped in Figure 4+ The spectrum for RRE2-Rev is the broadest, followed closely by RRE1-Rev and RRE3-Rev+ At low temperatures at which the molecular motion is completely frozen, the broadening of EPR spectra is exclusively due to dipole-dipole interactions between the two spin labels (Rabenstein & Shin, 1995)+ For labels that are proximal and more interacting, spectra are broader+ Conversely, distant spin labels yield narrower peaks+ Qualitatively, then, it is clear that the distance between labels in the RRE3-Rev complex is greater than that in either the RRE1 or RRE2 complexes+ Quantitatively, the overall spectral broadening, due to the splitting of EPR lines resulting from spin dipolar interactions, depends on the inverse cube of the distance (r ) between the spins+ This is given by the equation 2B ϭ 1+5g e b(3 cos 3 u Ϫ 1)/r 3 where 2B is the splitting due to the spin-spin interactions, g e is the isotropic g value of an electron, b is the electron Bohr magneton, and u is the angle between the interspin vector and the external magnetic field, which, for a frozen molecule, is fixed in time and is isotropically distributed+ The magnitude of B is obtained from a Fourier analysis of a frozen EPR spectrum, a method developed by Rabenstein and Shin (1995)+…”

“…The resulting distances are given in Figure 5 in black with the modeled distances from two lowest-energy NMR shown in blue and green+ The difference between the numbers in blue and green (Fig+ 5) gives an estimate of the range of distances as determined by NMR+ For RRE1 the nitroxide tip-to-tip distance from the 59 end to Rev position 51 is 14 Å, whereas the modeled NMR distance, C a -P, is 15 Å+ The RRE3-Rev distance, on the other hand, is modeled to be over 40 Å based on NMR data, which is well out of the 25-Å range of EPR (Rabenstein & Shin, 1995;Battiste et al+, 1996)+ Thus, the EPR lineshape of RRE3-Rev at low temperatures (Fig+ 4) is indistinguishable from a singly labeled com-FIGURE 4. Low-temperature EPR absorption spectra of bound and unbound Rev and RRE constructs+ The spectra were collected in a quartz vacuum dewer held at Ϫ140 8C with the modulation amplitude set to 3+2 G+ The broadening of the RRE1 and RRE2 spectra indicates spin-spin interactions+ plexed spectrum+ In the case of RRE2-Rev, the measured tip-to-tip distance is 12 Å and the modeled C a -P based on NOE constraints is between 6 and 10 Å+…”

“…Room-temperature EPR absorption spectra of bound and unbound Rev and RRE1+ All spectra were taken at 22 8C in a Bruker ESP300 EPR spectrometer with a modulation amplitude of 0+5 G+ The broadening in the bound and singly labeled systems is due to decreased nitroxide mobility+ Additional broadening in the bound and doubly labeled system is due to spin-spin dipolar interactions+ measure distances between the spin labels in the RRERev complexes, EPR spectra were collected at Ϫ140 8C+ These absorption spectra are shown overlapped in Figure 4+ The spectrum for RRE2-Rev is the broadest, followed closely by RRE1-Rev and RRE3-Rev+ At low temperatures at which the molecular motion is completely frozen, the broadening of EPR spectra is exclusively due to dipole-dipole interactions between the two spin labels (Rabenstein & Shin, 1995)+ For labels that are proximal and more interacting, spectra are broader+ Conversely, distant spin labels yield narrower peaks+ Qualitatively, then, it is clear that the distance between labels in the RRE3-Rev complex is greater than that in either the RRE1 or RRE2 complexes+ Quantitatively, the overall spectral broadening, due to the splitting of EPR lines resulting from spin dipolar interactions, depends on the inverse cube of the distance (r ) between the spins+ This is given by the equation 2B ϭ 1+5g e b(3 cos 3 u Ϫ 1)/r 3 where 2B is the splitting due to the spin-spin interactions, g e is the isotropic g value of an electron, b is the electron Bohr magneton, and u is the angle between the interspin vector and the external magnetic field, which, for a frozen molecule, is fixed in time and is isotropically distributed+ The magnitude of B is obtained from a Fourier analysis of a frozen EPR spectrum, a method developed by Rabenstein and Shin (1995)+…”

“…Since the publication of the "EPR spectroscopic ruler" for the determination of distances in protein by Rabenstein and Shin (1995), this method has been applied successfully to many problems involving proteins (Rabenstein & Shin, 1996;Hall et al+, 1997;Thorgeirsson et al+, 1997;Ottemann et al+, 1998)+ Now, we extend the same spin-labeling EPR strategy to RNA or RNAprotein systems+ For RRE-Rev model systems, we applied the ruler to three categories of 59-displacement spin-labeling constructs+ These constructs (RRE1, RRE2, and RRE3), representing native, permutated, and modulated 59 ends, form a reasonable test of the method and yield three unique distance measurements+ The distances measured for the RRE constructs correlate well with the NMR-based structure+ Specifically, the EPR-determined distance for RRE1-Rev is slightly shorter relative to the average NMR-based structure, whereas the RRE3-Rev distance is relatively longer+ This could be explained by the methanethiosulfonate spin label (MTSSL) moiety being positioned in such a way that the nitroxide group is pointing towards the C-terminus of Rev+ If the position of the nitroxide was shifted along the 5-Å length of MTSSL, it would account for the observed EPR distances in RRE1 and RRE2+ Also, the 19 low-energy structures calculated from NMR data show high flexibility in the C-terminal region of Rev+ This higher flexibility could account for FIGURE 5. The average NMR structure of RRE (violet, Protein Data Bank (PDB) file 1ETF) complexed with two out of the 19 lowestenergy Rev structures (blue and green, PDB file 1ETG), which show the range of Rev-RRE distances+ EPR-measured nitroxide tip-to-tip distances are shown in black, and the corresponding C a -P distances of the two different NMR structures are shown in green and blue+ The green helix is a structure chosen to represent the closest NMR modeled distances from the 19 lowest-energy conformations, whereas the blue helix represents the longest modeled Rev-RRE distance+ Together, these helices give an estimate of the upper and lower limit in the range of possible C a -P distances based on NMR data+ For EPR nitroxide tip-to-tip distances, the main source of error is due to the flexibility of the nitroxide-bearing side chains (Rabenstein & Shin, 1995)+ A simple molecular-dynamics calculation indicated that this error is in the range of 2 to 2+5 Å (Fiori & Millhauser, 1995)+ Errors resulting from EPR spectral measurement are negligible+ the larger distance observed by EPR spectroscopy+ However, some differences between the interpretation of the data in this model RNA study and in previous protein EPR studies must be examined+ From extensive studies in proteins, it has been established that a large number of EPR distances compensate for the spatial uncertainty of the nitroxide groups in structural determination+ Distance determination is the primary prerequisite for structural studies+ It is encouraging that the EPR method outlined in this study compares favorably with the previously determined NMR structure+ More importantly, this distance technique is poised to study larger, more complicated RNA systems, whose size would prohibit NMR strategies from being effectively employed+ The application of this technique in the future promises to be a fruitful area of RNA research+…”

“…Although progress has been made in the past decade in understanding the various functions of RNA, biologically important nucleic acid molecules are still more difficult to characterize than proteins+ Several wellestablished biophysical techniques have been recruited in an effort to better understand the biological role of RNA (Holbrook & Kim, 1998;Chang & Varani, 1997)+ A newly established method to quantitate distances by electron paramagnetic resonance (EPR) of spin-labeled molecules has been useful in probing protein structure and function, but has yet to be applied to RNA molecules+ Given the utility of EPR in resolving problems in the structural biology of proteins (Hubbell et al+, 1998;Shin, 1998), it would appear highly desirable to extend EPR as a tool for studying RNA+ Various aspects of spin-labeling EPR make it ideal for both protein and nucleic acid research+ Spin labeling involves the site-specific attachment of a nitroxide group containing an unpaired electron to a macromolecule+ Consequently, the rotational correlation time of the nitroxide probe, which can be obtained from the EPR signal, yields information on macromolecular dynamics proximal to the site of attachment+ The EPR method can be applied to very large molecules, including RNA-protein complexes such as the ribosome+ Most importantly, EPR provides long-range distance measurements between two labeling sites up to 25 Å apart (Rabenstein & Shin, 1995)+ Furthermore, distance changes can be followed by EPR on the submillisecond time scale, which can provide information about the time course of conformational changes (Thorgeirsson et al+, 1997)+…”