If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
5 D. R. Kolling, J. S. Brunzelle, S. Lhee, A. R. Crofts, and S. K. Nair, submitted for publication. * This work was supported by National Institutes of Health Grants GM 35438 (to A. R. C.) and GM 62954 (to S. A. D.), Fogarty Grant PHS 1 RO3 TW 01495 (to A. R. C. and R. I. S.), and National Institutes of Health/National Center for Research Resources Grant S10-RR15878 for instrumentation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 2 Present address: Dept. of Chemistry, Princeton University, Princeton, NJ 08540.
The interaction of the reduced[2Fe-2S] cluster of isolated Rieske fragment from the bc1 complex of Rhodobacter sphaeroides with nitrogens (14N and 15N) from the local protein environment has been studied by X- and S-band pulsed EPR spectroscopy. The two-dimensional electron spin echo envelope modulation spectra of uniformly 15N-labeled protein show two well resolved cross-peaks with weak couplings of ∼0.3-0.4 and 1.1 MHz in addition to couplings in the range of 6-8 MHz from two coordinating Nδ of histidine ligands. The quadrupole coupling constants for weakly coupled nitrogens determined from S-band electron spin echo envelope modulation spectra identify them as Nϵ of histidine ligands and peptide nitrogen (Np), respectively. Analysis of the line intensities in orientation-selected S-band spectra indicated that Np is the backbone N-atom of Leu-132 residue. The hyperfine couplings from Nϵ and Np demonstrate the predominantly isotropic character resulting from the transfer of unpaired spin density onto the 2s orbitals of the nitrogens. Spectra also show that other peptide nitrogens in the protein environment must carry a 5-10 times smaller amount of spin density than the Np of Leu-132 residue. The appearance of the excess unpaired spin density on the Np of Leu-132 residue indicates its involvement in hydrogen bond formation with the bridging sulfur of the Rieske cluster. The configuration of the hydrogen bond therefore provides a preferred path for spin density transfer. Observation of similar splittings in the 15N spectra of other Rieske-type proteins and [2Fe-2S] ferredoxins suggests that a hydrogen bond between the bridging sulfur and peptide nitrogen is a common structural feature of [2Fe-2S] clusters.
Proteins containing Rieske-type[2Fe-2S] clusters with two histidyl and two cysteinyl ligands play important roles in many biological electron transfer reactions such as aerobic respiration, photosynthesis, and biodegradation of various alkene and aromatic compounds. The distinct biological functions of this protein family are in part associated with the cluster redox potential and the pK of the oxidized form, which are roughly correlated with the number of hydrogen bonds from protein side chains and the peptide backbone to the cluster and its immediate ligands (
The abbreviations used are: ISP, Rieske iron-sulfur protein; ESEEM, electron spin echo envelope modulation; ISF, the water soluble proteolyzed extrinsic domain of the Rieske subunit (the iron-sulfur fragment) of the bc1 complex from R. sphaeroides; HYSCORE, hyperfine sublevel correlation; NQR, nuclear quadrupole resonance; Np, peptide nitrogen; MOPS, 4-morpholinepropanesulfonic acid; mT, millitesla.
4The abbreviations used are: ISP, Rieske iron-sulfur protein; ESEEM, electron spin echo envelope modulation; ISF, the water soluble proteolyzed extrinsic domain of the Rieske subunit (the iron-sulfur fragment) of the bc1 complex from R. sphaeroides; HYSCORE, hyperfine sublevel correlation; NQR, nuclear quadrupole resonance; Np, peptide nitrogen; MOPS, 4-morpholinepropanesulfonic acid; mT, millitesla.
is a constituent of the high potential electron transfer chain that accepts the first electron in the bifurcated reaction at the ubihydroquinone (quinol, QH2) oxidizing Qo site. In the catalytic mechanism at the Qo site, the redox reaction involves extraction of both an electron and a proton from the bound quinol and their transfer to the ISP through a short pathway that includes the ISP His-161 ring and the H-bond between Nϵ and the -OH of the quinol substrate (
). It has been proposed that the electron transfer is gated by the low probability of finding the H-atom of the H-bond in the kinetically favorable position, determined by the pK difference between the quinol and the oxidized ISP (ISPox) (reviewed in Refs.
). The pK on ISPox is contributed by one of the cluster ligands, His-161 in bovine numbering, and is associated with H+ dissociation from the Nϵ involved in the H-bond. The gating accounts for the fact that this first electron transfer is slower by more than 3 orders of magnitude than would be expected from the short distance (∼7 Å) involved (
). The first electron transfer is the ratedetermining step in quinol oxidation under conditions of substrate saturation. Rate determination at this reaction is demonstrated by the fact that the overall rate depends on its driving force in a classical Marcus fashion, as shown through use of mutants of the ISP with modified redox potential (
). Briefly, the driving force, ΔGo, for the first electron transfer is determined by the redox potential difference between the donor (SQ/QH2) and acceptor (ISPox/ISPH) couples. Assuming that mutations in the ISP do not change Em(SQ/QH2), a change in Em (ISPox/ISPH) will effect the driving force. Marcus theory relates the rate constant to the driving force, and a plot of log10k v. ΔGo follows the expected form for rates measured in mutants generated in mitochondria or in different species of bacteria (
). Because the oxidation of QH2 is the limiting partial process and the Marcus limitation is seen for the first but not the second electron transfer step, the location of the limiting step is unambiguous (
). It is clear, therefore, that, for this crucial reaction, a deeper analysis of the factors contributing to changes in Em of ISP, and in the pKa values of the Nϵ, will play a key role in our understanding of the overall mechanism (
An informative approach to studying the environment of the reduced Rieske cluster is through application of high resolution EPR techniques. Indeed, the coordination of the Rieske cluster by histidyl ligands was initially established by electron nuclear double resonance spectroscopy (
) have shown that the major contribution in these spectra comes from the coordinating Nδ of histidines. Other more weakly coupled 14N nitrogens present in the cluster environment do not produce recognizable lines in the spectra, because of the influence of the nuclear quadrupole interaction of 14N. Whether or not lines are seen depends on a particular relation between nuclear Zeeman frequency and hyperfine coupling (
), as explained under “Experimental Procedures.” Recently, it was demonstrated that the weakly coupled nitrogens produce readily observed lines in X-band two-dimensional ESEEM (HYSCORE) spectra of 15N-labeled sulredoxin, another Rieske-like protein (
). In particular, two weak couplings of ∼0.7 and 0.25 MHz (where the values reflect a recalculation of frequencies for comparison with 14N) were observed and tentatively assigned to peptide nitrogen of the backbone, and to Nϵ of the histidine ligands, respectively, based on comparison with model complexes and other clusters (
). However, these data did not provide any direct indication for the chemical nature of the nitrogens producing the couplings observed.
This finding is important in the context of functional studies of the bc1/b6f family because it opens the way for the spectroscopic characterization of the Nϵ nitrogen involved in H-bonding with the occupant of the Qo site and the nitrogen(s) Np forming H-bonds with the cluster.
In the present article, we have applied X-band (∼9.7 GHz) and S-band (∼3.1 GHz) spectroscopy to the water soluble proteolyzed extrinsic domain of the Rieske subunit (the iron-sulfur fragment (ISF)) of the bc1 complex from Rhodobacter sphaeroides to further study the weakly coupled nitrogens around the Rieske cluster.
Sample Preparation—Growth of R. sphaeroides, purification of the bc1 complex, and isolation of the ISF were previously described (
), but with the following modifications: 50% less nitrilotriacetic acid was used, l-aspartate and l-glutamate were not used, and 15N ammonium sulfate (Aldrich) replaced 14N ammonium sulfate. The working buffer used for all of the ISF samples was 50 mm KH2PO4, pH 7.0, 400 mm NaCl, and 20% (v/v) glycerol. The samples were reduced with 5 mm buffered (50 mm MOPS, pH 7.0) sodium ascorbate and placed into a quartz cuvette (Wilmad-Labglass, Bueno, NJ) with Teflon tubing to prevent scratching. In all of the spectra reported in this paper, the ISF was used rather than the intact complex. As a consequence, no interactions between the ISP extrinsic domain and occupants of the Qo site could occur, and the spectra therefore lack the features associated with such interactions.
Pulsed EPR Spectroscopy—In pulsed EPR, a magnetization vector, reflecting a population of electron spins aligned along the magnetic field, is rotated to a new alignment, and the kinetics of relaxation back to the equilibrium alignment is followed by detecting the microwave emission (the echo). The rotation of the magnetization is determined by the pulse length, 90° for a π/2 pulse, 180° for a π pulse. The interaction with nuclear spins results in modulation of the decay kinetics, and Fourier transformation of these modulations reveals the frequencies of the interacting species. In practice, the resolution is improved through use of ESEEM (see below). For the Rieske protein in frozen solution, different orientations of the cluster relative to the magnetic field vector can be selected by tuning the magnetic field within the anisotropic EPR line from the rhombic g-tensor. The orientation of the g-tensor axes is strictly connected with the cluster.
Several types of electron spin echo measurements with different pulse sequences were used, with appropriate phase cycling schemes employed to eliminate unwanted features from experimental echo envelopes (
Two-pulse Field Sweep—In experiments using the two-pulse sequence (π/2 - τ - π - τ - echo), the intensity of the echo signal was measured with a fixed interval, τ, between two microwave pulses with spin vector rotation angles π/2 and π. The echo intensity varies with magnetic field strength (units mT) to show the spectrum of the absorbing species. This type of measurement is termed a “field sweep,” and at settings at which modulation from magnetic nuclei are minimized (long pulse lengths, π ≥ 100 ns), the EPR line is comparable with the integral of the derivative spectrum collected by continuous wave-EPR.
One-dimensional Three-pulse ESEEM—In ESEEM spectroscopy, the spin echo envelope, resulting from measurement of changes in amplitude of the echo with variation of pulse timing, provides an averaging of the relaxation kinetics of the spin population, improving resolution. In the one- dimensional three-pulse measurement (π/2 - τ - π/2 - T - π/2 - τ - echo), the intensity of the stimulated echo signal after the third pulse is recorded as a function of time, T, at constant time, τ, to generate an echo envelope. The set of three-pulse envelopes recorded at different τ values forms, after Fourier transformation, a two-dimensional three-pulse data set showing the spectra caused by nuclear spins interacting with the paramagnetic center (
HYSCORE—In the two-dimensional four-pulse experiment (π/2 - τ - π/2 - t1 - π - t2 - π/2 - τ-echo) known as HYSCORE, the intensity of the inverted echo after the fourth pulse was measured with varied t1 and t2 and constant τ (
). Such a two-dimensional set of echo envelopes gives, after Fourier transformation, a four-quadrant spectrum that selects different correlations between nuclear frequencies from two manifolds with opposite electron spin projections, with equal resolution in each frequency coordinate. Because the spectrum is symmetrical with respect to the zero axes, only two quadrants are usually shown (
). Control of the experiment was accomplished through X-epr software using an ELEXSYS console including SpecJet and PatternJet (Bruker BioSpin, Rheinstetten, Germany.) The probe used was an ER 4118CF liquid helium flow cryostat with a Flexline (Bruker) cavity holder and a home-built bridged loop-gap resonator.
Spectral processing of three- and four-pulse ESEEM patterns was performed with Bruker WIN-EPR software. Processing first consisted of subtracting the monotonic component of the decay from time traces (real and imaginary parts) by a cubic or sixth order polynomial to remove the echo decay function. The time trace was then zero-filled to increase the number of experimental data points to a power of one greater than that collected. Following this, a Hamming window function was applied, and the magnitude Fourier spectra were calculated (
The Factors Leading to the “Cancellation Condition” in 14N ESEEM Spectra—Because of the I = 1 nuclear spin, and the quadrupole interactions resulting from this, the 14N nucleus can produce up to six lines in an ESEEM spectrum, three from each of the two electron spin manifolds with mS =+½ or -½. In measurements of amorphous (powder) samples (such as the frozen suspensions of the ISF used here), because of their different orientation dependence, not all transitions contribute equally to the spectra. The type of spectrum expected from 14N with predominantly isotropic hyperfine coupling A is governed by the ratio between the effective nuclear frequency in each manifold, νef±, given by νef± = |νI ± |A|/2|, and the quadrupole coupling constant, K, given by K = e2qQ/4h (
If νef±/K =∼0, i.e. νef± ≅ 0 (the situation called a cancellation condition because νI ≅ A/2, i.e. the external magnetic field matches the local hyperfine field producing a situation of pure nuclear quadrupole resonance), then the three nuclear frequencies from a corresponding manifold will be close to the three pure nuclear quadrupole resonance (NQR) frequencies of 14N. In this case, three narrow peaks at the following frequencies,
would appear in the powder ESEEM spectra, with the property ν+ = ν- + ν0 (the term η is an asymmetry parameter). These frequencies would also be present in orientation-selected spectra. However, their intensities depend on the orientation of the magnetic field relative to the g-tensor (i.e. cluster) and the respective NQR tensor in each particular experiment. The frequencies described by Equations 1-3. can appear in spectra up to a ratio of νef±/K =∼0.75-1 but are broadened as this value departs from 0.
If νef±/K > 1, only a single line is expected from each corresponding manifold without any pronounced orientation dependence. This line is produced by a transition at the maximum frequency, which is actually a double-quantum transition between the two outer states with mI =-1 and 1. The frequency of this transition is well described by the following formula,
where κ = K2(3 + η2). Such a line might show a change in its frequency of the order of 2Δ(νI) that is dependent on orientation selection because of variation of the Zeeman frequency. However, it has an order of ∼0.12 MHz in S-band experiments, where the width of the EPR spectrum is ∼20 mT.
Two other single-quantum transitions, involving the central level with mI = 0, have a significant orientation dependence from quadrupole interaction and could produce lines at varying frequencies in the orientation-selected spectra.
Two-pulse field sweep X- and S-band electron spin echo spectra of dithionite-reduced Rieske [2Fe-2S] center in ISF show a rhombic EPR line shape consistent with a g-tensor having principal values (gz,y,x = 2.03, 1.90, 1.76) (Fig. 1). The width of the EPR line in S-band is smaller than that in X-band and is proportional to the ratio of microwave frequencies used (3.1 GHz for S-band and 9.7 GHz for X-band) (
). In addition, the shape of the field sweep spectrum is influenced by the ESEEM. This influence is different at different regions of the line shape and is stronger in S-band than in X-band. The effect smoothes the gx feature in the S-band spectrum, which is, however, clearly seen in the X-band spectrum of the same sample.
X-band 15N HYSCORE—X-band 15N HYSCORE spectra measured at different magnetic fields along the EPR line contain the cross-features produced by different types of nitrogens (Fig. 2). In the (+-) quadrant, two pairs of cross-peaks with a contour parallel to the diagonal are detected that have been attributed to the two coordinated 15Nδ1,2 with the hyperfine splittings of the order of 6 and 8 MHz (
) gave values for the hyperfine tensors in the axial approximation of a = 6.6 and T = 1.6 MHz for Nδ1 and of a = 7.6 MHz and T = 1.5 MHz for Nδ2in 15N-ISF. These tensors are very similar to those reported for other Rieske-type proteins obtained by orientation-selected 15N Q-band electron nuclear double resonance (
Of particular interest in the HYSCORE spectra of 15N-ISF is the (++) quadrant, in which two well resolved pairs of the cross-features are clearly detected at (2.05, 0.95) MHz (Np) and (1.68, 1.28) MHz (Nϵ) near gz. These features are centered symmetrically around the diagonal point with 15N Zeeman frequency and are attributed to weakly coupled 15N in the immediate cluster environment. They were also observed in the HYSCORE spectra recorded near gx and gy and also at some intermediate field positions. The splittings are practically the same within the range 1.1-1.2 and 0.3-0.4 MHz in all of the spectra recorded, indicating their predominantly isotropic character. This is a result of the transfer of unpaired spin density onto the corresponding nuclei. Similar couplings were previously observed in the 15N HYSCORE spectra of the archaeal Rieske protein sulredoxin and assigned to the peptide nitrogen of the backbone (larger coupling) and to the remote Nϵ of coordinating histidine ligands (smaller coupling) (
). From these results, we can tentatively suggest the same assignment for ISF. However, neither ISF nor sulredoxin show spectra that resolve two different couplings from Nϵ-atoms, probably because of the small differences between them.
S-band ESEEM and HYSCORE—Although the 15N HYSCORE spectra at X-band provide evidence for the presence of weakly coupled nitrogens, we could only speculate about their chemical nature. To obtain additional information about these nitrogens, we have employed S-band ESEEM spectroscopy.
Fig. 3 shows typical three-pulse S-band ESEEM spectra of ISF with natural isotope abundance in the region appropriate for the 14N nuclei, recorded at three different magnetic fields that select the principal values of the g-tensor (Fig. 1). Comparison of the spectra shows that a complete set of frequencies appeared at all three fields, including peaks at 0.6, 1.5, 1.8, 2.3, 2.9, and 3.4 MHz. These frequencies were also seen along the diagonal of the HYSCORE spectra. One can note with high accuracy, that the frequencies from the set that are observed in at least two spectra (0.6, 1.5, 2.3, and 2.9 MHz) are independent of the applied magnetic field and the excitation point within the EPR spectrum. Additional information about relative relations between these frequencies was obtained from the S-band HYSCORE spectra. The spectrum recorded at 129 mT (near gy) showed cross-peaks at (1.5, 1.8) MHz, indicating that these frequencies belong to two opposite electron spin manifolds of the same nucleus.
We have taken advantage of two features of nitrogen nuclear spin systems to examine the properties of atoms interacting weakly with the electron spin of the [2Fe-2S] cluster. The first of these relates to the different energy levels (frequencies) from quadrupolar spin interactions of the 14N nucleus and the different magnetic fields needed to achieve the optimal conditions for its observation with reference to the Zeeman frequency on switching between X-band and S-band (see discussion of the cancellation condition under “Experimental Procedures”). The second arises from the simplification in spectra when dealing with the I = ½ 15N nucleus compared with the I = 1 14N nucleus, because the former lack the quadrupolar features of the latter.
X-band Versus S-band—The lines observed in S-band 14N ESEEM spectra of the Rieske cluster appear at different fields from those seen in 14N X-band spectra, because the resonance condition depends on field strength. Previous studies of the Rieske cluster in different proteins by X-band one- and two-dimensional ESEEM (
) have shown that the major contributions in these spectra come from the coordinating Nδ of histidines with hyperfine couplings of the order of 4-5 MHz. The dominating features of X-band ESEEM spectra of ISF are similar to the spectra of other Rieske clusters and result from two Nδ atoms. The weakly coupled nitrogens do not produce readily recognizable lines in 14N X-band spectra; this simplifying feature aids unambiguous assignment but at the expense of further information. However, their contribution is clearly seen in the 15N spectra, together with the peaks from Nδ as shown above (Fig. 2).
The basic parameter distinguishing between X-band and S-band ESEEM spectra is the difference in field strength needed to obtain resonance, which leads to a different value for the Zeeman frequency. It is about three times smaller in S-band (∼0.37-0.42 MHz) than in X-band (∼1.05-1.1 MHz). The hyperfine couplings of the coordinated Nδ exceed the Zeeman frequency by a factor of 4-5 in X-band. Usually the 14N nuclei produce well resolved and intense ESEEM spectra under such conditions. However, any further increase of the couplings relative to the Zeeman frequency leads to the suppression of the spectral intensity (
). The ratio of hyperfine couplings and the Zeeman frequency is increased from 4-5 to 12-15 in S-band, and the spectral lines from Nδ have a negligible intensity, i.e. they are not observed in the spectra (Table 1). On the other hand, the 14N Zeeman frequency in X-band exceeds the couplings (0.7 and 0.25 MHz recalculated from 15N couplings) from weakly coupled Np and Nϵ, whereas in S-band the Zeeman frequency is smaller or comparable with these couplings. At this scale, the powder spectra observed for 14N ESEEM also depend on the value of the quadrupole coupling constant.
TABLE 1The contribution of strongly and weakly coupled nitrogens to 14N ESEEM spectra
Assignment of the Frequencies in S-band Spectra—Lines with stable frequencies observed in at least two S-band spectra recorded at different positions along the EPR line are likely to be close to the pure NQR frequencies of nitrogens or their double-quantum transitions.
Previous data have indicated that the protonated Nϵ-H of the imidazole ring possesses a stable (quadrupole) coupling constant of K = 0.35 MHz and a high asymmetry parameter η = 0.915-0.995 in noncoordinated imidazole and histidine (
). One might therefore expect a set of NQR lines from Nϵ in the range 1.4-1.6 MHz for ν+ and 0.7-0.8 MHz for ν- and ν0.
The hyperfine coupling 0.3-0.4 MHz tentatively assigned above to the Nϵ from the 15N HYSCORE is equivalent to 0.25 ± 0.03 MHz when recalculated to 14N. This gives νef- = 0.28 MHz for νI = 0.4 MHz, and defines the ratio νef-/K = 0.65-0.8 for K = 0.35-0.43 MHz. This ratio is quite high, at the borderline of the cancellation condition. We do not therefore expect all peaks to show the narrow lines close to pure quadrupole frequencies of 14N as anticipated under cancellation conditions (
). The line at 1.5 MHz observed in the S-band spectra (Fig. 3) can therefore be assigned to this highest frequency, which comes from the manifold with νef- = |νI - |A|/2|of the protonated Nϵ-atoms, with hyperfine coupling at ∼0.25 MHz.
From the above, one can calculate the frequency of the formal double-quantum transition (Equation 4) from the opposite manifold with νef+ = |νI + |A|/2| = 0.65 MHz and νef+/K = 1.5-1.9. This gives an estimate of 1.76-1.9 MHz for K =∼0.4 MHz and η varying between 0 and 1. This value is consistent with the frequency 1.8 MHz observed in the three-pulse spectra of Fig. 3, which is involved in cross-correlation with the frequency 1.5 MHz in the HYSCORE spectrum.
The exact cancellation condition, νef- = |νI - |A/2||∼0, is well approximated in the S-band experiment for the 14N (Fig. 3), with the second hyperfine coupling A =∼0.7 MHz determined from the X-band HYSCORE experiment of Fig. 2. Three frequencies, at 0.6, 2.3, and 2.9 MHz, which are also observed in the S-band spectra, satisfy the condition for the NQR frequencies, i.e. the sum of the two lower frequencies is equal to the maximum frequency. These three frequencies allow one to determine K = 0.87 MHz and η = 0.34. The value of the quadrupole coupling constant K is close to the typical values ∼0.75-0.85 MHz previously reported for peptide nitrogens (
). Calculation of the double-quantum transition for this nitrogen, with A = 0.7 MHz and the nuclear quadrupole interaction parameters found, gives a value 3.4 MHz, consistent with the frequency observed in three-pulse S-band spectrum at gz (Fig. 3).
Thus, the S-band 14N ESEEM spectra of the Rieske cluster in ISF could best be explained as a superposition of the lines from two types of nitrogens, with quadrupole couplings typical of an imidazole amine nitrogen, Nϵ, and of amide peptide nitrogen, Np, with the hyperfine couplings determined from 15N HYSCORE spectra. This result directly defines the weakly coupled nitrogens as Nϵ and Np, in agreement with our previous tentative assignments (
Spin Density on the Nϵ—The intense pair of cross-features with smaller splittings of 0.3-0.4 MHz seen in spectra of 15N-ISF belong to Nϵ-atoms of the histidyl ligands. The literature values for isotropic hyperfine couplings for coordinated Nδ and remote Nϵ of the imidazole rings ligated to Cu(II) and VO(II) in various model complexes and in proteins of well defined structure have an almost constant ratio of about 20 between the Nδ and Nϵ couplings (although both couplings in the Cu(II) complexes are much larger than those in the VO(II) complexes) (
). Thus, the ratio ∼20 between hyperfine couplings of the Nδ and Nϵ applies to the coordinated imidazoles in the Rieske cluster. This stability probably reflects the analogous mechanism of spin density transfer from the metal over the imidazole ring to the Nϵ, which is also sensitive to protonation state (
). In view of this result, it would be interesting to study the Nϵ hyperfine couplings for the Rieske cluster in the bc1 complex to explore the interaction of ISP with occupants of the Qo site via the H-bond with the Nϵ-H of one of the histidine ligands.
Identification of the Peptide Nitrogen—Comparison of the intensities of the lines in the S-band ESEEM spectra assigned to the NQR frequencies of the peptide nitrogen recorded at different points along the EPR line makes it possible to identify a particular peptide nitrogen that carries the maximum unpaired spin density from among all of the peptide nitrogens around the Rieske cluster (
), has a narrow range of quadrupole coupling constants, K = 0.75-0.85 MHz, determined by the electronic structure and the geometry of the planar peptide group. This coupling constant is only slightly perturbed by hydrogen bonding, which has been confirmed by calculations of the quadrupole coupling tensor (
). Calculations based on the isolated molecules yield quadrupole coupling tensors for the peptide nitrogen very close to experimental values, confirming the negligible influence of hydrogen bonding. From these calculations, the maximum principal value is normal to the local peptide plane; the intermediate element almost coincides with the C(O)-N(H) bond, and the minimal element points about 30° from the N-H-bond.
A significant increase in the intensity of the lowest NQR line, ν0, at 0.6 MHz is clearly seen from the three-pulse ESEEM spectra (Fig. 3) when the magnetic field is applied close to the gx principal value of the Rieske cluster. Theoretical considerations (
) indicate that the increase of ν0 intensity occurs when both vectors Bo and gx closely coincide with the axis of maximal principal value of the NQR tensor.
To interpret the magnetic resonance data in conjunction with crystal structures, one needs to know the orientation of the g-tensor principal axes. Recently, a single-crystal EPR study of the reduced Rieske cluster in cytochrome bc1 complex with stigmatellin, showed that the gz, gx, and gy axes are not oriented exactly along the sulfur-sulfur and iron-iron directions and nearly normal to the cluster plane of the cluster, respectively, as theoretically predicted (
). The g-tensor principal axes are skewed with respect to the iron-iron and sulfur-sulfur atom direction in the [2Fe-2S] cluster. For instance, the gx ∼ 1.79 axis makes an average angle of 30° with respect to the Fe-Fe direction and the gz ∼ 2.024 axis an average angle of 26° with respect to the sulfur-sulfur direction.
We have used the available x-ray structure of the ISF from R. sphaeroides (Fig. 4) to characterize the orientation of the peptide plane and the normal to this plane for the peptide nitrogens located within 5 Å around Rieske cluster.
). This information has allowed us to calculate the angles between the axis of the maximum principal value of the NQR tensor (i.e. normal to the peptide plane determined by the location of N- and C=O-atoms) for different peptide nitrogens, and gx principal direction of the g-tensor. These calculations (Table 2) have found that the minimum angle between these two axes belongs to the peptide nitrogen from Leu-132. This angle in the intact bc1 complex is equal to 6° for monomer A (i.e. deviation from coincidence of two directions is very small) and increased up to 24° for monomer B, with an average value of 15°. All of the other angles significantly exceed this angle and thus would not satisfy the observed variations caused by orientation in ESEEM spectra.
TABLE 2Angles between the gx axis of the g-tensor and the axis of maximum principal value of the NQR tensor of peptide nitrogens
One can conclude that the peptide nitrogen possessing the highest unpaired spin density is that of Leu-132. Although the -N-H-S2 distance of 3.36 Å is relatively long, the spin density observed shows that this nitrogen is involved in hydrogen bond formation with the cluster, with a favorable path for the spin density transfer. Peptide nitrogens from the other 5 residues located within H-bond distance of the ISP cluster carry at least 5-10 times less spin density and give neither resolved splittings in the spectra of 15N-labeled protein nor intensive narrow lines from the 14N nuclei in the native protein. One can note however, that the line shape of the central doublet caused by splitting by the 15Nϵ (Fig. 2) is not identical in different Rieske proteins. We therefore cannot exclude some small additional nonequivalent unresolved contribution from other weakly coupled nitrogens to this line. The differences observed could be attributed to other variations in the H-bond network in different proteins.
A similar coupling of ∼1.1-1.5 MHz (for 15N) has been observed in the spectra of other Rieske proteins (high potential sulredoxin (
T. Iwasaki and S. A. Dikanov, unpublished observations.
which may indicate the presence of a similar hydrogen bond transferring an excess of unpaired spin density onto the peptide nitrogen. If so, this configuration might be a common motif, not linked specifically to the differences in cluster potential, but a structural component of all [2Fe-2S] clusters. Additional support for this view comes from NMR experiments with the Rieskeferredoxin component of toluene-4-monooxygenase (
), selectively labeled with 15N on specific residues. The NMR spectra show that the peptide nitrogen of Gln-48 (the analogue of Leu-132 in ISF) in the reduced protein possesses a hyperfine-shifted resonance with maximal chemical shift of ∼426 ppm. This nitrogen also undergoes the largest change of chemical shift (∼300 ppm) on reduction of the Rieske cluster. This large change is probably due to a contribution of contact shift, resulting from the appearance of a large unpaired spin density on its nucleus.
Mechanistic Implications—Detailed density function theory/electrostatic calculations have been able to account for the redox and pK value differences between strains in terms of H-bonding and local negative charge from side chains substitutions (
). Because the H-bond donor in Leu-132 is a backbone N-H, mutagenesis cannot be used to test its role. The backbone configuration of the loop following the first pair of ligands (the conserved sequence CTHLGCVP, where the ligands are in bold italic) is determined by residues Leu-132, Gly-133, and Cys-134 (residues 142-144 in the beef or chicken complexes), which are conserved in the Rieske proteins from all α-proteobacterial bc1 complexes. Each of these residues has an interesting structural role. In the mechanism of QH2 oxidation in the intact complex, Leu-132 provides a contact with the b-interface at which the ISP mobile domain is docked on cytochrome b. The interface is tightly packed, and this strongly constrains the configuration. It also constrains the possibilities for mutational analysis. Residues tolerated at this position in R. sphaeroides include alanine and tyrosine,
S. Lhee and A. R. Crofts, unpublished observations.
but function is impaired, indicating that the constrained configuration in the wild type is important for optimal turnover. Other mutations have been explored through recombinant expression of the bovine protein in Escherichia coli and show relatively weak variation in thermodynamic characteristics but substantial changes in EPR spectra (
). However, structures are not yet available for any of these proteins. A wider range of substitutions is tolerated at Gly-133, but any substitution gives rise to some impairment of function. The explanation is immediately apparent on examination of the φ and ψ torsional angles for this (or the equivalent) span in solved structures (Table 3). Those for this glycine occupy a region of the Ramachandra space accessible only to glycine, so that any substitution will force a reorientation of the backbone. Although this site was among those generated as spontaneous mutations in earlier studies, detailed mutagenic analysis has been limited (
). In work currently in progress, we have constructed a range of mutants and are currently characterizing these. For one strain, G133S, the crystallographic structure has been solved at ∼1.5 Å resolution, and the φ and ψ torsional angles for this strain are included in Table 3. The reconfiguration of the backbone orientation results in a substantial change in local dipole orientations, so that the -C=Oδ- points toward the cluster, and the -N-Hδ+ points away, which is the reverse of the wild type. This dipolar reorientation would likely contribute to a change in thermodynamic properties, but the extent of this is currently under investigation using the wider set of mutant strains. The reconfiguration of the backbone does not propagate far; on the N-side, the -N-H-S2 distance for Leu-132 is changed by only 0.05 Å (3.41 instead of 3.36 Å), and the backbone configuration is not changed beyond the next residue, Cys-144. This is involved in a conserved disulfide linkage that stabilizes the loop, and the side chain and sulfur-sulfur bond have the same position in wild type and G133S mutant. The disulfide linkage appears to have little effect in redox properties (
). From this discussion, it will be apparent that the structural information currently available and results from mutagenesis favor a nonspecific role for the H-bond from Leu-132 -NH to S2 and that further exploration of this span through mutagenesis will likely be most informative in the context of the Leu-132 and Gly-133 mutants currently under study.
TABLE 3The φ and ψ torsional angles for the protein backbone in the span LGCV obtained from high resolution structures (all angles are in degrees)