A simple protocol to study blue copper proteins by NMR Ioannis Gelis 1 , Nikolaos Katsaros 1 , Claudio Luchinat 2,3 , Mario Piccioli 2,4 and Luisa Poggi 2,4 1 NCSR Demokritos, Institute of Physical Chemistry, Agia Paraskevi Attikis, Greece; 2 Magnetic Resonance Center, 3 Department of Agricultural Biotechnology and 4 Department of Chemistry, University of Florence, Italy In the case of oxidized plastocyanin from Synechocystis sp. PCC6803, an NMR approach based on classical two and three dimensional experiments for sequential assignment leaves unobserved 14 out of 98 amino acids. A protocol which simply makes use of tailored versions of 2D HSQC and 3D CBCA(CO)NH and CBCANH leads to the identi- fication of nine of the above 14 residues. The proposed protocol differs from previous aproaches in that it does not involve the use of unconventional experiments designed specifically for paramagnetic systems, and does not exploit the occurrence of a corresponding diamagnetic species in chemical exchange with the blue copper form. This protocol is expected to extend the popularity of NMR in the struc- tural studies of copper (II) proteins, allowing researchers to increase the amount of information available via NMR on the neighborhood of a paramagnetic center without requi- ring a specific expertise in the field. The resulting 3D spectra are standard spectra that can be handled by any standard software for protein NMR data analysis. Keywords: blue copper proteins; NMR spectroscopy; struc- tural biology; paramagnetic proteins; plastocyanin. There is a strong interest in the structural biology commu- nity for the study of copper trafficking and copper homeostasis [1–11]. This involves the understanding of the role of metal ions in protein folding and misfolding related diseases [12–20], as well as the understanding at the atomic level of protein–protein interactions in electron-transfer processes [21–29]. Within this framework the search for methodological advancements in NMR spectroscopy tail- ored to the structural characterization of copper(II) proteins may play a significant role. In paramagnetic metalloproteins, NMR signals of pro- tons close to the metal ions are broadened, sometimes beyond detection, by the presence of the paramagnetic center [30,31]. The extent of paramagnetic induced line broadening depends on the electronic relaxation times of the metal center [32–35]. Tetragonal Cu(II), found in Type II centers, has long electronic relaxation times [36] which make the NMR lines of residues belonging to the coordination sphere broad beyond detection [37,38]. When Cu(II) adopts a trigonal geometry, such as that provided by two histidines and one cysteine residue in Type I centers, or blue copper centers, the electronic relaxation times are about one order of magnitude shorter. Hence, signals belonging to Cu(II) first coordination sphere, although severely broadened, become observable [39]. Because of the axial symmetry of the g-tensor in Type I copper centers [40], the pseudocontact contribution to the observed shifts is negligible and para- magnetic shifts arise only from through-bond spin density delocalization from the metal to the ligands. Therefore, in Type I copper centers, pseudocontact shifts can not be used for structural purposes, unlike many other classes of metalloproteins [41–43]. As a partial compensation of such a drawback, the shifts can be safely interpreted on the basis of the chemical shift index [44]. It was recently shown that solution structures of copper(II) proteins can be obtained [45]. To this end, the standard protocol for solution structure of biomacromole- cules has been substantially augmented by a number of non conventional strategies for resonances assignments and by the use of paramagnetism-based constraints for structure calculations [46]. This approach often requires specific expertise in the field of electron relaxation and hyperfine interaction [42,46–53] and, in some case, specific hardware [54]. As a consequence, NMR structural characterization of paramagnetic metalloproteins is routinely performed only in a limited number of laboratories [31,55–63]. We would like to present here a different perspective of the NMR study of paramagnetic proteins and to emphasize the fact that paramagnetic proteins should not necessarily be considered as a different field with respect to mainstream biomolecular NMR. We will discuss the information content of basic 2D and 3D experiments when they are collected using a different choice of experimental parameters with respect to the standard ones. The additional experi- ments that we propose are deliberately restricted to simple modifications of the pulse sequences that are routinely used for resonance assignment, like CBCA(CO)NH [64,65] and CBCANH [66,67], in such a way that their implementation does not require any special expertise. This approach should extend significantly the detectability of resonances that sense the hyperfine interaction and therefore should substantially increase the number of assignments in the proximity of the paramagnetic center that can be obtained within a standard protocol [68,69]. The modifications discussed here to CBCA(CO)NH and CBCANH experiments do not substantially alter the coherence transfer pathway with Correspondence to M. Piccioli, Via L. Sacconi 6, 50019 Sesto Fiorentino, Florence, Italy. Fax: + 39 055457 4253, Tel.: + 39 055457 4265, E-mail: piccioli@cerm.unifi.it Abbreviations: INEPT, insensitive nuclei enhanced by polarization transfer; PFG, pulsed field gradients. (Received 12 June 2002, revised 25 October 2002, accepted 27 November 2002) Eur. J. Biochem. 270, 600–609 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03400.x respect to the scheme originally proposed by Bax and coworkers [64], but they make the two sequences much more effective in the presence of contributions to relaxation arising from the hyperfine interaction. The test system chosen is the blue copper protein plastocyanin from the cyanobacterium Synechocystis sp. PCC6803. It contains a typical Type I center extensively spectroscopically characterized [70–72]. Previous NMR studies showed that this is an excellent system to address the efficiency of nonconventional NMR approaches to obtain structural information [45,73]. In the present work we will demonstrate that an approach which does not require any hardware or software dedicated to paramagnetic systems can substantially improve the available assignments close to the copper (II) ion without recourse to metal substitution. Materials and methods Protein expression and purification The expression and purification of Synechocystis sp. PCC6803 plastocyanin in Escherichia coli was performed as previously described [74]. Uniformly 13 C, 15 N-labeled over- expressed plastocyanin was obtained from M9 minimal medium containing ( 15 NH 4 ) 2 SO 4 as the sole nitrogen source and [ 13 C 6 ] D -glucose as the sole carbon source. Samples for NMR spectroscopy (2 m M )werepreparedin50 m M sodium phosphate buffer (either in 90% H 2 O, 10% D 2 O or in 100% D 2 O) at pH 5.2. Complete oxidation of the protein was achieved using a slight excess of ferricyanide, subsequently removed by gel filtration. The samples were kept at 4 °Cin between measurements. NMR Spectroscopy Experiments were performed at 295 K on Bruker Avance spectrometers operating at 700 and 800 MHz. Diamagnetic 1 H- 13 CHSQCand 1 H- 15 N HSQC [75] experiments were performed. The number of real data points acquired were 512 in the t 1 dimension ( 13 Cand 15 N), and 2048 in acquisition (t 2 dimension). Spectral widths of 11 p.p.m. for 1 H dimension, 80 p.p.m. for 13 C dimension and 50 p.p.m. for 15 Ndimen- sion were used. For both experiments, 4 scans and a recycle delay of 800 ms were used. Echo-antiecho acquisition [76] was used to perform quadrature detection in t 1 dimension. Sensitivity improvement [77,78] and crush gradients during the insensitive nuclei enhanced by polarization and transfer (INEPT) and inverse INEPT mixing were also used. Two dimensional tailored 1 H- 13 CHSQCand 1 H- 15 N HSQC experiments were performed to detect fast relaxing signals [79]. All delays (INEPT transfer and recycle) were shortened to 1.6 ms and 100 ms, respectively, in order to detect resonances near the paramagnetic center. Two dimensional nonselective inversion-recovery 1 H- 15 NHSQC experiments ( 15 N IR-HSQC) were performed to measure nonselective longitudinal relaxation rates of protons [79]. In order to measure T 1 values of very fast relaxing protons, INEPT transfer and relaxation delays were shortened to 1.6 ms and 200 ms, respectively. Eight points were collected to fit T 1 values, with the following inversion recovery delays: 2, 4, 8, 16, 32, 64, 128 and 256 ms. A three dimensional HNCO experiment [80] was per- formed to assign backbone resonances. For the above experiment spectral windows of 11 p.p.m. for 1 H, 50 p.p.m. for 15 N, and 30 p.p.m. for 13 C dimensions were typically used. The number of real data points acquired were 128 in the t 1 dimension ( 13 C), 64 in the t 2 dimension ( 15 N), and 1024 in acquisition (t 3 dimension). Three dimensional CBCA(CO)NH [64,65] and CBCANH [66,67] experiments were carried out to sequentially assign 13 C resonances. Spectral widths of 11 p.p.m. for 1 H dimension, 76 p.p.m. for 13 C dimension and 41 p.p.m. for 15 N dimension were used. The number of real data points acquired were 64 points in the 15 N dimension, 256 in the 13 C dimension, and 1024 in acquisition (t 3 dimension) for both experiments. A recycle delay of 800 ms was used and 8 scans per increment were collected. All the data were zero-filled in the indirect dimensions and apodized using cosine squared functions. Linear prediction was always applied in the indirect dimension. All NMR data were processed with the Bruker XWINNMR software packages. The program SPARKY 3 (T. D. Goddard and D. G. Kneller, University of California, San Francisco, USA) was used for the analysis of all NMR spectra. Theory A classical approach toward structure determination in a paramagnetic metalloprotein does not provide information in the proximity of the metal center [81,82], even when careful and extensive studies are performed using double and triple labeled samples [61,83]. CBCA(CO)NH and CBCANH are among the most popular experiments for sequential assignment of macro- molecules in solution [65,66]. CBCA(CO)NH spectra con- nect HN(i) with C b (i-1) and C a (i-1) resonances, while CBCANH spectra connect HN(i) with C b (i), C a (i), C b (i-1) and C a (i-1) resonances, the inter residue peaks being lower in intensity than the intra residue peaks. The standard versions of both experiments make use of several INEPT transfer delays, crush gradients, flip back pulses, sensitivity improvement schemes and echo-antiecho gradient selection. Each of the above building blocks requires coherence transfer delays during which the magnetization of interest is relaxed. In the case of paramagnetic molecules, the presence of the unpaired electron makes large contributions to nuclear relaxation for nuclei nearby. As a consequence, CBCA(CO)NH and CBCANH are expected to be unsuit- able for the study of such systems. However, a series of modifications can be planned that make the two sequences exploitable. The optimization of polarization transfer and recycle delays in heteronuclear experiments has been extensively discussed elsewhere, as well as the choice of the number of scans and data points in t 1 , t 2 and t 3 dimensions [46,84]. On such bases, the NH reverse INEPT and the CH INEPT transfer delays were shortened to 1.6 ms and 1.8 ms, respectively, in the CBCA(CO)NH experiment, while only the NH transfer delay was shortened to 1.6 ms in the CBCANH. The building blocks of the pulse sequences related to the coherence transfers pathway C b -C a -CO-N or C b -C a -N were not modified with respect to the standard version of the sequence. A recycle delay of 300 ms was used Ó FEBS 2003 Blue copper proteins studied by NMR (Eur. J. Biochem. 270) 601 and 64 scans were collected for both experiments. With respect to the diamagnetic version of the experiments, the number of data points in the 15 Nand 13 C dimensions were reduced from 64 to 48 and from 256 to 128, respectively. Besides the choice of INEPT transfer delays, other modifications can be introduced with respect to the diamagnetic version of the sequences. The sensitivity improvement scheme (SI) [78] makes use, during the reverse INEPT, of a double spin-echo which allows the detection of both antiphase components N x H z and N y H z created during 15 N evolution, thus giving a 2 1/2 improvement of the signal to noise ratio [78]. This scheme has twice the duration of a normal reverse INEPT, and different relaxation mechanisms are operative on the various coherence transfer pathways that transform the two above components to observable magnetizations. Even if the transfer delays are shortened, as already extensively discussed [84], the occurrence of a strong contribution to relaxation may be such that, for fast relaxing signals, the elimination of sensitivity improvement scheme gives a better S/N. The use of pulsed field gradients (PFGs) within a pulse sequence to detect paramagnetic signals may be critical. In general, their use to clean observable magnetization from spurious peaks has no drawbacks, provided that PFG do not entail additional delays [85]. However, in the case of echo-antiecho detection schemes [76], the gradient selection requires that two additional gradients are placed in the sequences, together with two additional 180° pulses and refocusing delays. Because hyperfine relaxation depends on c 2 X , where X is the involved nucleus, the loss of signal intensity is critical in those coherence transfer steps in which 1 H R 2 relaxation is involved [86]. This is of course the case of the period immediately preceding t 3 acquisition. Similar considerations hold for the use of crush gradients during the INEPT and reverse INEPT steps. In this case, the loss due to relaxation depends on R 1 . Therefore the use of crush gradients for fast relaxing signals is less destructive that the gradients needed in the echo-antiecho scheme. Of course, a major drawback expected from the elimination of gradient selection and crush gradients is that there is no efficient water suppression scheme left in the sequence. To overcome this problem, a Watergate scheme, with short gradients in order to be compatible with the short delays of the reverse INEPT step [87] can be reintroduced in the final reverse INEPT step. The calculated effects of the stepwise removal of the crush gradients, echo-antiecho and sensitivity improvement schemes are shown in Fig. 1. The calculations are per- formed for the transfer function from a 15 N nucleus to a bound proton in either the CBCANH or CBCA(CO)NH pulse schemes. The proton is considered to be at 6 A ˚ from the copper(II) center, assuming a s s ¼ 0.5 ns [73] and a s r ¼ 5.9 ns [74]. Under these conditions R 2 c. ¼ 600 s )1 , while R 1 is about 5 times smaller. If we use the standard values for duration and recovery of gradients of 1 ms and 0.5 ms, respectively, the transfer function has a maximum at about 1 ms (Fig. 1A). Its intensity is about 2% of the intensity expected for the corresponding peak in a normal reverse INEPT when relaxation is neglected. Elimination of the crush gradients, during which 1 H R 1 relaxation occurs, leads to a gain in intensity of about 15% (Fig. 1B). The most important effect arises from the elimination of the echo-antiecho scheme. The effect of removing the echo- antiecho building block is observed in the calculated transfer functions shown in Fig. 1C. Considering as a test case the signal discussed above, the replacement of the echo-antiecho block with any other quadrature detection scheme that does not rely on gradient selection of coherences, increases signal intensity by about a factor of five. Of course the relative gain in intensity is reduced when, in the diamagnetic version of the sequence, shorter gradients and recovery delays are used. When gradient and recovery delays in the diamagnetic experiment are shortened down to 150 lsand100ls, respectively, the gain of signal intensity under the above conditions is still of about a factor of two. This shows that even if very short values of gradient and recovery delays are used within the diamagnetic version of CBCA(CO)NH and CBCANH (and this would not be the ÔdefaultÕ choice in the absence of fast relaxation), the use of echo-antiecho quadrature detection is not recommended with respect to States-TPPI [88] or any other quadrature detection scheme methods that does not rely on gradient selection of coherences. Finally the effects of the replacement of the sensitivity improvement step with the usual reverse INEPT step is illustrated in the transfer function shown in Fig. 1D. It can be seen that the single reverse INEPT step, not only gives about a 10% increase in the maximum of the transfer function with respect to the sensitivity improvement scheme but also it gives a transfer function which is much less sensitive to optimization of the transfer delay, as observed in Fig. 1 when transfer delays longer than 1.8–2 ms are considered. Fig. 1. Calculated transfer functions for the NH reverse INEPT transfer step of CBCA(CO)NH or CBCANH experiments with: (A) diamagnetic pulse sequence, using sensitivity improvement detection scheme and echo- antiecho quadrature detection (all applied gradients were 1 ms with a recovery delay of 0.5 ms); (B) same as (A) without the use of crush gradient occurring in between the 90° pulses; (C) same as (B) without the echo-antiecho detection, i.e. with the elimination of the additional delays needed for the gradients of the echo-antiecho; (D) same as (C) with the removal of the SI scheme. All transfer functions are normalized with respect to a normal reverse INEPT under optimized condition for the transfer delay and neglecting losses due to 1 H- 15 N relaxation. Transfer functions have been calculated for a 1 H signal of a proton at about 6 A ˚ from the metal center (R 2 ¼ 600 s )1 , R 1 ¼ 120 s )1 assuming a s s ¼ 0.5 ns and a s r ¼ 5.9 ns). 602 I. Gelis et al. (Eur. J. Biochem. 270) Ó FEBS 2003 Results and discussion Spectral assignment of oxidized plastocyanin: the standard approach Synechocystis sp. PCC6803 plastocyanin was overexpressed in E. coli to obtain large amounts of 13 C, 15 N-enriched protein. The already available assignment of 1 Hand 15 N resonances [45] was extended to 13 C resonances of backbone and side chains by a combination of classic triple resonance experiments. 3D HNCO [89], CBCA(CO)NH [64,65] and CBCANH [66,67] were collected at 700 and 800 MHz spectrometers. The analysis of these spectra has lead to the assignment of 73% of C¢, 81% of C a and 79% of C b . Because of broadening effects induced by the paramagnetic center, no sequential backbone assignment is available [45] in the loop regions encompassing residues 7–8, 38–42, 61–62, and 82–88. Detection of fast relaxing signals: 15 N- and 13 C-HSQC experiments Tailored versions of 1 H- 15 N HSQC [90], 1 H- 13 CHSQC[79], CBCA(CO)NH and CBCANH were used to detect reso- nances in the proximity of Cu(II). The comparison of two 1 H- 15 N HSQC spectra recorded with different recycle and polarization transfer delays allows to identify 14 resonances that clearly experience a substan- tial gain in signal intensity when comparing a diamagnetic HSQC experiment with a tailored experiment. The overlay of the two spectra is shown in Fig. 2, and the 14 resonances are highlighted. Of these, 7 are observed with much lower intensity in the diamagnetic experiment while 7 were completely missing in the diamagnetic experiment. The former 7 signals were already assigned in a previous study [45], and correspond to residues Leu14, Phe16, Asn34, Lys35, Ser37, Ile41 and Ala89. The seven new signals are listed in Table 1. In order to measure the proton T 1 values of the previously unobserved fast relaxing signals detected in the tailored 1 H- 15 N HSQC, a series of two dimensional nonselective inversion-recovery 1 H- 15 N HSQC experiments was performed [79]. As our present interest is focused on relatively fast relaxing signals, we used for the inversion recovery experiment a recycle delay of 200 ms. Therefore the inversion recovery experiment gave fully reliable results only for those resonances having a T 1 values < 60 ms. The T 1 values obtained for the above signals, together with the 1 Hand 15 Nshifts,arealsoreportedinTable1. Similarly to the 1 H- 15 N HSQC experiment, the compari- son of two 1 H- 13 C HSQC spectra recorded with different recycle and polarization transfer delays allows to identify 11 resonances that clearly experience a substantial gain in signal intensity when comparing a diamagnetic HSQC experiment with a tailored experiment. Of these, four belong to C a s peaks 1–4 in Table 2 and 7 to C b s peaks 5–11 in Table 2. They are also highlighted in Fig. 3A and 3B. Detection of fast relaxing signals: tailored CBCA(CO)NH and CBCANH A 3D version of these experiments tailored as discussed above to optimize the detection of fast relaxing signals has been performed. The new peaks identified through 1 H- 15 N HSQC were monitored in CBCA(CO)NH. While in the diamagnetic version of CBCA(CO)NH experiment only Fig. 2. Overlay of diamagnetic and tailored 1 H- 15 NHSQCspectra.Fourteen resonances are highlighted. The seven signals completely missing in the diamagnetic experiment are labelled A-G, while the seven observed with much lower intensity in the diamagnetic experiment are labelled with their corres- ponding assignment. Table 1. Previously unobserved signals found in the tailored 1 H- 15 N HSQC. dN (p.p.m) dHN (p.p.m) T 1 (ms) A 108.29 8.80 34.0 B 118.52 7.24 17.7 C 121.62 8.1 > 60 D 126.57 8.6 25.3 E 128.23 8.28 40.9 F 107.45 8.1 27.8 G 125.02 9.23 55.1 Ó FEBS 2003 Blue copper proteins studied by NMR (Eur. J. Biochem. 270) 603 one of the peaks (signal C) listed in Table 1 was observed, the tailored CBCA(CO)NH has allowed us to observe connectivities with previous amino acid for 5 out of 7 residues, as reported in Table 3. As far as signal C is concerned, the very weak connectivities observed in the diamagnetic version of CBCA(CO)NH are observed with much larger intensity (a factor of 2) in the tailored experiment. Similar considerations hold for CBCANH. The sensiti- vity of CBCANH is expected to be smaller than CBCA (CO)NH, as already proven extensively in diamagnetic systems. None of the peaks listed in Table 1 was observed in the diamagnetic experiment while four out of seven gave connectivities in the tailored CBCANH, as reported in Table 3, which summarizes the information obtained using modified CBCA(CO)NH and CBCANH. Assignment of fast relaxing signals The assignment of the new signals found in the tailored 1 H- 15 N HSQC can be performed considering the following: (a) limited number of missing assignments in the 1 H- 15 N HSQC spectrum (14, listed in Table 4); (b) C b and C a chemical shifts provide substantial information on the nature of the amino acid under investigation [44,91,92]. Of course, such assignment is feasible only under the assump- tion that contributions arising form pseudocontact shifts are negligible with respect to the chemical shift index tolerance [93]. As outlined above, this is a very reasonable assumption as shown by the available literature on Cu(II) proteins [39,94]. Let us consider signal A [Table 3]: the intra residue C a peak at 43.8 p.p.m. shows unambiguously that signal A belongs to a Gly residue, while inter residue C a and C b peaks at 58.3 and 27.9 p.p.m. are primarily consistent with Met, Arg or His residues. Therefore the only possible assignment is Gly8, preceded by Met7. In previous works only some sparse 1 H assignments were available for residues 7, 8, 61 and 62 [45]. No assignment can be performed for signal B, for which no connectivities are found in both CBCA(CO)NH and CBCANH. Signal C shows no connectivities in the CBCANH spectrum, but the inter residue C a peak found in the CBCA(CO)NH at 43.1 p.p.m. is only consistent with a Gly as preceding residue. Given the limited number of missing assignments, this is in agreement only with the assignment of signal C as the HN of Leu61, preceded by Gly60. Signal D shows in the CBCA(CO)NH spectrum inter residue peaks at 56.7 and 36.3 p.p.m., while among the intra residue peaks only the C a is found in the CBCANH at 59.2 p.p.m. These connectivities perfectly fit the assignment of signal D as NH of Val42, preceded by Ile41. The identification of Val42 is also confirmed by the pattern observed in the CBCA(CO)NH for Phe43, which presents inter residue connectivities at 59.2 and 30.8 p.p.m. As far as signal E is concerned, the four peaks corres- ponding to intra and inter residue C a and C b do not permit a fully consistent assignment. Inter and intra residue C a ’s are observed at 52.2 and 55 p.p.m., respectively, and they match with His86-Arg87 residues. This assignment is supported by Table 2. Signals that experience a substantial gain in signal intensity in the tailored 1 H- 13 C HSQC compared with the diamagnetic experiment. d 1 H (p.p.m) d 13 C (p.p.m) Assignment 1 5.44 56.2 Val15 (C a -H a ) 2 5.14 58.0 Met7 (C a -H a ) 3 4.92 42.3 Gly88 (C a -H a ) 4 4.67 43.0 Gly8 (C a -H a ) 5 1.27 37.3 Tyr 81 (C b -H b ) 6 0.93 37.9 Arg87 (C b -H b ) a 7 0.72 37.3 Tyr 81 (C b -H b ) 8 2.66 33.8 Val15 (C b -H b ) 9 1.00 31.2 Val42 (C b -H b ) 10 2.97 30.1 – 11 ) 0.30 11.6 Ile41 (C d -H d ) a Tentative assignment. Fig. 3. Overlay of diamagnetic and tailored 1 H- 13 CHSQCspectra.(A) C a region. (B) C b region. The 11 resonances that substantially increase their intensity in the tailored experi- ment are highlighted and labelled 1–11. 604 I. Gelis et al. (Eur. J. Biochem. 270) Ó FEBS 2003 the inter residue C b , which is observed at 35.6 p.p.m. (a typical His region), but does not fit with the intra residue C b that is observed at 38 p.p.m., i.e. out of the region where C b of Arg residues are expected to fall. Therefore, we assign signal E as the HN of Arg 87 only tentatively. The 15 N shift of signal F is only consistent with a Gly residue. As Gly8 has been already identified as signal A, signal F can be safely assigned as the only other glycine residue missing, i.e. Gly88, even if no connectivities are found in both CBCA(CO)NH and CBCANH. For signal G, in the CBCANH spectrum only the intra residue connectivities are observed while those with the previous residue are observed only in CBCA(CO)NH. The intra residue peaks at 48.9 p.p.m. for the C a and 17.3 p.p.m. for C b are only consistent with an Ala residue, while side chain carbons observed from signal G in CBCA(CO)NH (53.2 and 40.3 p.p.m) are only consistent with an Asn or a Leu residues. The only possible assignment for signal G is thus Ala62 NH, preceded by Leu61. In summary, the tailored experiments described above allowed us to detect and assign six new HN signals that were previously completely unobserved. Another seven signals showed a sizable increase in their S/N ratio. With the only exception of signal B, all these newly identified signals in the tailored 1 H- 15 N HSQC could be assigned. Figure 4 shows, as an example, comparison of diamag- netic and tailored CBCA(CO)NH as far as signal D is concerned. As observed, the two spectra are processed and displayed with the same resolution. While the two peaks arising from signal D are unambiguously detected in the paramagnetic spectrum, there is no evidence of them in the diamagnetic experiment. Some of the 13 C resonances that were identified as arising from the proximity of the paramagnetic center can be also identified in the tailored 1 H- 13 C HSQC. This is the case of the H a -C a peaks 1–4 shown in Fig. 3A, whose shifts match with the C a resonances of Val15, Met7, Gly88 and Gly8. Analogous considerations hold for the 7 H b -C b resonances identified (Fig. 3B), which are assigned on the basis of the already available 1 H assignment [45]. The only exception to this criterion is peak 6 which has a C b shift that corresponds to the identified Arg87 C b and for which no. 1 H assignment is available. Therefore, we tentatively assign peak 6 as Table 3. Connectivities found for signals A-G in tailored CBCA(CO)NH and CBCANH spectra. HN(i) dC a (i-1) (p.p.m) dC b (i-1) (p.p.m) dC a (i) (p.p.m) dC b (i) (p.p.m) A 58.3 27.9 43.8 B C 43.1 D 56.7 36.3 59.2 E 52.2 35.6 55 38 F G 53.2 40.3 48.9 17.3 Table 4. New assignments obtained for oxidized plastocyanin from Synechocystis sp. PCC6803. Copper(II) ligands are highlighted in bold. In the right column N-Cu and H N -Cu distances are reported for each amino acid. dN (p.p.m) dNH (p.p.m) dH a (p.p.m) dC a (p.p.m) dH b (p.p.m) dC b (p.p.m) N, HN distances (A ˚ ) Met7 5.14 58.3 27.9 7.9–8.4 Gly8 108.29 8.80 4.67 43.8 6.7–7.0 Val15 5.44 56.3 2.68 33.1 7.9–8.5 His39 4.9–5.8 Asn40 4.6–3.7 Val42 126.57 8.6 59.2 30.8 8.1–7.3 Leu61 121.62 8.1 53.2 40.3 10.3–11.3 Ala62 125.02 9.23 48.9 17.3 8.6–7.7 Tyr82 8.2–8.4 Cys83 5.6–5.8 Glu84 6.6–7.0 His86 52.2 35.6 5.7–5.4 Arg87 128.23 8.28 55 38 7.2–7.5 Gly88 107.45 8.1 4.92 42.3 8.9–8.9 Fig. 4. Strip plot of tailored (left) and diamagnetic (right) CBCA (CO)NH spectra in the region corresponding to signal D. While inter- residue C a and C b peaks are present in the tailored spectrum, no correlation is found in the diamagnetic one. Ó FEBS 2003 Blue copper proteins studied by NMR (Eur. J. Biochem. 270) 605 Arg87 C b -H b and we identify an H b 1 H signal of Arg87 at 0.93 p.p.m. No assignment is proposed for peak 10. All the new assignments are summarized in Table 4. A simple NMR protocol is an important tool to study paramagnetic proteins Plastocyanin is a good model system to address features of paramagnetic copper proteins in terms of assignment strategy. Indeed, the previously available assignment on oxidized plastocyanin from Synechocystis sp. PCC6803 [45] was obtained through a combination of methods which basically rely on saturation transfer techniques [73]. Within such a frame, dedicated experiments overcome the difficul- ties arising from the presence of the paramagnetic center and, eventually, permit the assignment for most of the amino acids, including those directly bound to the copper ion. Such nonconventional experiments include saturation transfer [95,96] from signals broadened beyond detection [97], mono dimensional NOEs over 1 H signals very broad and shifted in the region 100/)50 p.p.m. [98,99], NOESY and TOCSY experiments that allowed several 1 H assign- ment only on the basis of relative line broadening (i.e. based on a metal-to proton distance predictable by means of relaxation rates) [79], NOESY cross peaks between protons that were not identified in a classical sequential assignment work [100], the occurrence of signals with unusual chemical shift behaviour [39]. The above approach, which had lead to extensive assignment of paramagnetic copper proteins even in the first coordination sphere, required the occurrence of favourable exchange rates between the oxidized form and the reduced diamagnetic form. Of course, such requirements limit the application of the approach. Therefore we have designed experiments to extend the assignment of plasto- cyanin without relying on its reduced state and without any specific aprioriknowledge. A standard approach to resonance assignment, i.e. CBCA(CO)NH, and CBCANH, applied on plastocyanin permitted the identification of 80 out of 94 non proline residues [85%] with 14 amino acid for which no information were available. All missing residues belong to the northern loops of the protein surrounding the copper ion and fall within a 11 A ˚ sphere from the metal center. The protocol proposed in the present work allowed assignment of 9 out of the above 14 residues. Indeed, no information was obtained only for two of the three strong ligands of copper(II) (His39 and Cys83), for Asn40, whose HN group is directly involved in a hydrogen bond with the copper-bound Cys83 S c atom [45,101–103], and for Tyr82 and Glu84. It is noteworthy that both C a and C b resonances of the binding residue His86 can be assigned. This permits the identification of reso- nances as close as 3.6 A ˚ from the copper center without relying on any knowledge on the electron-nucleus coupling. Missing residues also provide a picture of the electron spin density delocalization on the ligands. Experimental evidence and theoretical calculations show that a larger amount of spin density is expected on Cys83 [97,104–106]. Consistently, not only Cys83 but also the surrounding residues (Tyr82, Glu84) are missing in the present assign- ment. Electron spin density is delocalized also through the H-bond between Cys83 Sc and Asn40. This makes Asn40 unobservable. The missing assignment of Asn40 prevents, in turn, the identification of the preceding residue His39. Indeed both 14 N ENDOR [107] and 1 H NMR data [45] on plastocyanin indicate that metal bound imidazoles from His86 and His39 experience a similar spin density delocali- zation, thus supporting the hypothesis that the H-bond between Cys83–Asn 40 is indeed responsible for the non identification of His39 with this approach. In summary, such an approach allows identification, in a sequence specific fashion, 89 out of 94 non proline residues (95%) providing 89%, 87% and 92% of the assignment of C a ,C b and N–H, respectively. With the above approach we can reach metal-to-nucleus distances of 7.2, 3.6, and 7.5 A ˚ , for H, C a and N, respectively. Conclusions In the case of the oxidized plastocyanin from Synechocystis sp. PCC6803, an NMR approach based on classical two and three dimensional experiments for sequential assign- ment leaves unobserved 14 residues out of 98 amino acids. A protocol that simply makes use of tailored version of 2D HSQC and 3D CBCA(CO)NH and CBCANH leads to the identification of 9 of the above 14 residues. Although it is clear that such improvement does not circumvent all the limitations arising from the presence of an oxidized copper center and actually still prevents the complete characteriza- tion of the first coordination sphere, we should stress that the approach proposed allows those structural biologists that are not experts nor familiar with paramagnetic proteins to substantially increase their knowledge. Acknowledgements We are grateful to Prof. Ivano Bertini for his advice and support. The expression system of Synechocystis sp. PCC6803 plastocyanin was a generous gift of Prof S. Ciurli. This work was supported by the European Union Research and Training Network (RTN) Project ÔCross correlation between the fluctuations of different interactions: a new avenue for biomolecular NMRÕ (Contract no. HPRN-CT-2000– 00092). I.G. is a Fellow of the Marie Curie Training Site ÔNMR in Inorganic Structural BiologyÕ, contract no. HPMT-2000–000137. Support from PARABIO (HPRT-CT-00009) is acknowledged. References 1. Brown,D.R.,Qin,K.,Herms,J.W.,Madlung,A.,Manson,J., Strome, R., Fraser, P.E., Kruck, T.A., Von Bohlen, A., Schulz- Schaeffer, W., Giese, A., Westaway, D.&Kretzschmar, H.A. (1997) The cellular prion protein binds copper in vivo. Nature 684–687. 2. Poulos, T.L. (1999) Helping copper find a home. Nat Struct. Biol. 6, 709–711. 3. Viles, J.H., Cohen, F.E., Prusiner, S.B., Goodin, D.B., Wright, P.E. & Dyson, H.J. (1999) Copper binding to the prion protein: structural implications of four identical cooperative binding sites. Proc.NatlAcad.Sci.USA96, 2042–2047. 4. O’Halloran, T.V. & Culotta, V.C. (2000) Metallochaperones: An Intracellular Shuttle Service for Metal Ions. J. Biol. Chem. 275, 25057–25060. 5. Goto, J.J., Zhu, H., Sanchez, R.J., Gralla, E.B. & Valentine, J.S. (2000) Loss of in vitro metal ion binding specificity in mutant copper-zinc superoxide dismutase associated with familial amyo- trophic lateral sclerosis. J. Biol. Chem. 14, 1007–1014. 606 I. Gelis et al. (Eur. J. Biochem. 270) Ó FEBS 2003 6. Harrison, M.D., Jones, C.E., Solioz, M. & Dameron, C.T. (2000) Intracellular copper routing: the role of copper chaperones. Trends Biochem. Sci. 25, 29–32. 7. Heaton, D., Nittis, T., Srinivasan, C. & Winge, D.R. (2000) Mutational analysis of the mitochondrial copper metallochaper- one Cox17. J. Biol. Chem. 275, 37582–37587. 8. McMahon, H.E., Mange, A., Nishida, N., Creminon, C., Casa- nova, D. & Lehmann, S. (2001) Cleavage of the amino terminus of the prion protein by reactive oxygen species. J. Biol. Chem. 276, 2286–2291. 9. Quaglio, E., Chiesa, R. & Harris, D.A. (2001) Copper converts the cellular prion protein into a protease-resistant species that is dis- tinct from the scrapie isoform. J. Biol. Chem. 276, 11432–11438. 10. Rosenzweig, A.C. (2001) Copper delivery by metallochaperone proteins. Acc Chem Res. 34, 119–128. 11. Hayward, L.J., Rodriguez, J.A., Kim, J.W., Tiwari, A., Goto, J.J., Cabelli, D.E., Valentine, J.S. & Brown, R.H.J. (2002) Decreased metallation and activity in subsets of mutant superoxide dis- mutases associated with familial ALS. J. Biol. Chem. 277, 15923– 15931. 12. Hughson, F.M., Wright, P.E. & Baldwin, R.L. (1987) Structural characterization of a partly folded apomyoglobin intermediate. Science 249, 1544–1548. 13. Arcus, V.L., Vuilleumier, S., Freund, S.M.V., Bycroft, M. & Fersht, A.R. (1995) A comparison of the pH, urea, and tem- perature-denaturated states of barnase by heteronuclear NMR: implications for the initiation of protein folding. J. Mol. Biol. 254, 305–321. 14. Frank, M.K., Clore, G.M. & Gronenborn, A.M. (1995) Structural and dynamic characterization of the urea denatured state of the immunoglobin binding domain of streptococcal protein G by multidimensional heteronuclear NMR spectroscopy. Protein Sci. 4, 2605–2615. 15. Serrano, L. (1995) Comparison between the phi distribution of amino acids in the protein data base and NMR data indicates that amino acids have various phi propensities in the random coil conformation. J. Mol. Biol. 254, 322–333. 16. Bertini, I., Cowan, J.A., Luchinat, C., Natarajan, K. & Piccioli, M. (1997) Characterization of a partially unfolded high potential iron protein relevant to the folding pathway and cluster stability. Biochemistry 36, 9332–9339. 17. Zhang, O., Kay, L.E., Shortle, D. & Forman-Kay, J.D. (1997) Comprehensive NOE characterization of a partially folded large fragment of staphylococcal nuclease Delta131Delta, using NMR methods with improved resolution. J. Mol. Biol. 272, 9–20. 18. Bertini, I., Luchinat, C., Piccioli, M. & Soriano, A. (1998) Folding properties of iron sulfur proteins. (Dedicated to Prof. O. Yamauchi). Inorg. Chim. Acta 283, 12–16. 19. Dyson, H.J. & Wright, P.E. (1998) Equilibrium NMR studies of unfolded and partially folded proteins. Nat. Struct. Biol. 5, 499–503. 20. Kuhn, T. & Schwalbe, H. (2000) Monitoring the Kinetics of Ion- Dependent Protein Folding by Time-Resolved NMR Spectro- scopy at Atomic Resolution. J. Am. Chem. Soc. 122, 6169–6174. 21. Farrar, J.A., Neese, F., Lappalainen, P., Kroneck, P.M.H., Saraste, M., Zumft, W.G. & Thomson, A.J. (1996) The Electronic Structure of Cu A : a novel mixed-valence dinuclear copper elec- tron-transfer center. J. Am. Chem. Soc. 118, 11501–11514. 22. Klomp, L.W., Lin, S.J., Yuan, D., Klausner, R.D., Culotta, V.C. & Gitlin, J.D. (1997) Identification and functional expression of HAH1, a novel human gene involved in copper homeostasis. J. Biol. Chem. 272, 9221–9226. 23. Van Pouderoyen, G., Cigna, G., Rolli, G., Cutruzzola, F., Malatesta, F., Silvestrini, M.C., Brunori, M. & Canters, G.W. (1997) Electron-transfer properties of Pseudomonas aeruginosa [Lys44,Glu64]azurin. Eur. J. Biochem. 247, 322–331. 24. Casareno, R.L., Waggoner, D.J. & Gitlin, J.D. (1998) The copper chaperone CCS directly interacts with copper/zinc superoxide dismutase. J. Biol. Chem. 273, 23625–23628. 25. Wittung-Stafshede, P., Malmstro ¨ m, B.G., Sanders, D., Fee, J.A., Winkler, J.R. & Gray, H.B. (1998) Effect of redox state on the folding free energy of a thermostable electron-transfer metallo- protein: the CuA domain of cytochrome oxidase from Thermus thermophilus. Biochemistry 37, 3172–3177. 26. Hamza, I., Schafer, M., Klomp, L.W. & Gitlin, J.D. (1999) Interaction of the copper chaperone HAH1 with the Wilson dis- ease protein is essential for copper homeostasis. Proc. Natl Acad. Sci. USA 96, 13363–13368. 27. Huffman, D.L. & O’Halloran, T.V. (2000) Energetics of Copper Trafficking Between the Atx1 Metallochaperone and the Intra- cellularCopper-transporter, Ccc2. J. Biol. Chem.275, 18611–18614. 28. Crowley, P.B., Otting, G., Schlarb-Ridley, B., Canters, G.W. & Ubbink, M. (2001) Hydrophobic interactions in a cyanobacterial plastocyanin-cytochrome f complex. J. Am. Chem. Soc. 123, 10444–10453. 29. Banci, L., Bertini, I., Del Conte, R., Markey, J. & Ruiz-Duen ˜ as, F.J. (2001) Copper trafficking: the solution structure of Bacillus subtilis CopZ. Biochemistry 40, 15660–15668. 30. Bertini, I. & Luchinat, C. (1996) NMR of paramagnetic sub- stances. Coord. Chem. Rev. 150, 1–225. 31. Bertini, I., Luchinat, C. & Parigi, G. (2001) Solution NMR of Paramagnetic Molecules. Elsevier, Amsterdam. 32. Banci,L.,Bertini,I.&Luchinat,C.(1991)Nuclear and Electron Relaxation. The Magnetic Nucleus-Unpaired Electron Coupling in Solution. VCH, Weinheim. 33. Xia, B., Westler, W.M., Cheng, H., Meyer, J., Moulis, J M. & Markley, J.L. (1995) Detection and Classification of Hyperfine- Shifted 1 H, 2 H, and 15 N Resonances from the Four Cysteines That Ligate Iron in Oxidized and Reduced Clostridium pasteurianum Rubredoxin. J. Am. Chem. Soc. 117, 5347–5350. 34. Wilkens, S.J., Xia, B., Weinhold, F., Markley, J.L. & Westler, W.M. (1998) NMR investigations of Clostridium pasteurianum rubredoxin. Origin of hyperfine 1 H, 2 H, 13 Cand 15 NNMRche- mical shifts in iron-sulfur proteins as determined by comparison of experimental data with hybrid density functional calculations. J. Am. Chem. Soc. 120, 4806–4814. 35. Hurley, J.K., Weber-Main, A.M., Hodges, A.E., Stankovic, M.T., Benning, M.M., Holden, H.M., Cheng, H., Xia, B., Markley, J.L., Genzor, C., Gomez-Moreno, C., Hafezi, R. & Tollin, G. (1997) Iron-sulfur cluster cysteine-to-serine mutants of Anabaena 2Fe-2S- ferredoxin exhibit unexpected redox properties and are competent in electron transfer to ferredoxin: NADP+ reductase. Biochem- istry 36, 15109–15117. 36. Fee, J.A. & Briggs, R.G. (1975) Studies on the reconstitution of bovine erythrocyte superoxide dismutase. V. Preparation and properties of derivatives in which both zinc and copper sites contain copper. Biochim. Biophys. Acta 400, 439–439. 37. Lippard, S.J., Burger, A.R., Ugurbil, K., Pantoliano, M.W. & Valentine, J.S. (1977) Nuclear magnetic resonance and chemical modification studies of bovine erythrocyte superoxide dismutase: evidence for zinc-promoted organization of the active site struc- ture. Biochemistry 16, 1136–1141. 38. Bertini, I., Lanini, G., Luchinat, C., Messori, L., Monnanni, R. & Scozzafava, A. (1985) Investigation of Cu 2 Co 2 SOD and its anion derivatives. 1 H NMR and electronic spectra. J. Am. Chem. Soc. 107, 4391–4396. 39. Kalverda, A.P., Salgado, J., Dennison, C. & Canters, G.W. (1996) Analysis of the paramagnetic copper (II) site of amicyanin by 1 H NMR spectroscopy. Biochemistry 35, 3085–3092. 40. Penfield, K.W., Gewirth, A.A. & Solomon, E.I. (1985) Electronic structure and bonding of the blue copper site in plastocyanin. J. Am. Chem. Soc. 107, 4519–4529. Ó FEBS 2003 Blue copper proteins studied by NMR (Eur. J. Biochem. 270) 607 41. Banci, L., Bertini, I., Cremonini, M.A., Gori Savellini, G., Luch- inat, C., Wu ¨ thrich,K.&Gu ¨ ntert, P. (1998) PSEUDODYANA for NMR structure calculation of paramagnetic metalloproteins using torsion angle molecular dynamics. J. Biomol. NMR 12, 553–557. 42. Turner, D.L., Brennan, L., Chamberlin, S.G., Louro, R.O. & Xavier, A.V. (1998) Determination of solution structures of paramagnetic proteins by NMR. Eur. Biophys. J. 27, 367–375. 43. Gochin, M. & Roder, H. (1995) Protein Structure Refinement based on Paramagnetic NMR shifts. Applications to Wild-Type and mutants forms of cytochrome c. Protein Sci. 4, 296–305. 44. Wishart, D.S., Sykes, B.D. & Richards, F.M. (1992) The chemical shift index: a fast and simple method for the assignment of protein secondary structure through NMR spectroscopy. Biochemistry 31, 1647–1651. 45. Bertini, I., Ciurli, S., Dikiy, A., Ferna ´ ndez, C.O., Luchinat, C., Safarov, N., Shumilin, S. & Vila, A.J. (2001) The first solution structure of an oxidized paramagnetic copper (II) protein: the case of plastocyanin from the cyanobacterium Synechocystis PCC6803. J. Am. Chem. Soc. 123, 2405–2413. 46. Bertini, I., Luchinat, C. & Piccioli, M. (2001) Paramagnetic probes in metalloproteins. turning limitations into advantages. Meth Enzymol. 339, 314–340. 47. Boisbouvier, J., Gans, P., Blackledge, M., Brutscher, B. & Marion, D. (1999) Long-range structral information in NMR studies of paramagnetic molecules from electron spin-nuclear spin cross- correlated relaxation. J. Am. Chem. Soc. 121, 7700–7701. 48. Nguyen, B.D., Xia, Z., Yeh, D.C., Vyas, K., Deaguero, H. & La Mar., G.N. (1999) Solution NMR determination of the ani- sotropy and orientation of the paramagnetic susceptibility tensor as a function of temperature for metmyoglobin cyanide: implica- tions for the population of excited electron states. J. Am. Chem. Soc. 121, 208–217. 49. Xia, Z., Nguyen, B.D. & La Mar., G.N. (2000) The use of che- mical shift temperature gradients to establish the paramagnetic susceptibility tensor orientation: implication for structure determination/refinement in paramagnetic metalloproteins. J. Biomol. NMR 17, 167–174. 50. Hus, J.C., Marion, D. & Blackledge, M. (2000) De novo determination of protein structure by nmr using orientational and long-range order restraints. J. Mol. Biol. 298, 927–936. 51. Tsan, P., Caffrey, M., Daku, M.L., Cusanovich, M., Marion, D. & Gans, P. (2001) Magnetic susceptibility tensor and heme contact shifts determinations in the Rhodobacter capsulatus ferricyto- chrome c’: NMR and magnetic susceptibility studies. J. Am. Chem Soc. 123, 2231–2242. 52. Epperson, J., Ming, L J., Baker, G. & Newkome, G. (2001) Paramagnetic cobalt (II) as an NMR probe of dendrimer struc- ture: mobility and cooperativity of dendritic arms. J. Am. Chem Soc. 123, 8583–8592. 53. Piccioli, M. (1996) The application of selective-excitation pulse sequences in NMR spectroscopy of paramagnetic proteins. J. Magn. Reson. B110, 202–204. 54. Luchinat, C., Piccioli, M., Pierattelli, R., Engelke, F., Marquardsen, T. & Ruin, R. (2001) Development of NMR intrumentation to achieve excitation of large bandwidths in high resolution spectra at high-fields. J. Magn. Reson. 150, 161–166. 55. Ghose, R. & Prestegard, J.H. (1997) Electron spin-nuclear spin cross-correlation effects on multiplet splittings in paramagnetic proteins. J. Magn. Reson. 128, 138–143. 56. Bougault, C.M., Dou, Y., Ikeda-Saito, M., Langry, K.C., Smith, K.M. & La Mar., G.N. (1998) Solution 1 H NMR study of the electronic structure and magnetic properties of high-spin ferrous or deoxy myoglobins. J. Am. Chem. Soc. 120, 2113–2123. 57. Hus, J.C., Marion, D. & Blackledge, M. (2000) De novo determination of protein structure by NMR using orientational and long-range order restraints. J. Mol Biol. 298, 927–936. 58. Brennan, L., Turner, D.L., Messias, A.C., Teodoro, M.L., LeGall, J., Santos, H. & Xavier, A.V. (2000) Structural basis for the net- work of functional cooperativities in cytochrome c3 from Desulfovibrio gigas: solution structures of the oxidised and reduced states. J. Mol. Biol. 298, 61–82. 59. Ubbink, M., Worrall, J.A.R., Canters, G.W., Groenen, E.J.J. & Huber, M. (2002) Paramagnetic resonance of biological metal centers. Annu. Rev. Biophys. Biomol. Struct. 31, 393–422. 60. Walker, F.A. (2000) Proton NMR and EPR spectroscopy of Paramagnetic Metalloporphyrins. In The Porphyrin Handbook (Kadish, K.M., Smith, K.M. & Guilard, R., eds), pp. 81–183. Academic Press, San Diego, CA, USA. 61. Vathyam, S., Byrd, R.A. & Miller, A.F. (2000) Mapping the effects of metal ion reduction and substrate analog binding to Fe-superoxide dismutase by NMR spectroscopy. Magn. Reson. Chem. 38, 536–542. 62. Machonkin, T.E., Westler, W.M. & Markley, J.L. (2002) 13 C [ 13 C]2D NMR: a novel strategy for the study of paramagnetic proteins with slow electronic relaxation times. J. Am. Chem. Soc. 124, 3204–3205. 63. Fernandez, C.O., Cricco, J.A., Slutter, C.E., Richards, J.H., Gray, H.B. & Vila, A.J. (2002) Axial ligand modulation of the electronic structures of binuclear cooper sites: analysis of paramagnetic 1 H NMR spectra of Met160Gln Cu (A). J. Am. Chem. Soc. 123, 11678–11685. 64. Grzesiek, S. & Bax, A. (1992) Correlating backbone amide and side chain resonances in larger proteins by multiple relayed triple resonance NMR. J. Am. Chem. Soc. 114, 6291– 6293. 65. Muhandiram, D.R. & Kay, L.E. (1994) Gradient-enhanced triple resonance three-dimensional NMR experiments with improved sensitivity. J. Magn. Reson. B103, 203–216. 66. Grzesiek, S. & Bax, A. (1992) An efficient experiment for sequential backbone assignment of medium-sized isotopically enriched proteins. J. Magn. Reson. 99, 201–207. 67. Meissner,A.&Sorensen,O.W.(2001)SequentialHNCACBand CBCANH protein NMR pulse sequences. J. Magn. Reson. 151, 328–331. 68. Wu ¨ thrich, K. (1996) Biological macromolecules: structure deter- mination in solution. In Encyclopedia of Nuclear Magnetic Res- onance (Grant, D.M. & Harris, R.K., eds), pp. 932–939. John Wiley & Sons, Chichester, UK. 69. Riek, R., Wider, G., Billeter, M., Hornemann, S., Glockshuber, R. &Wu ¨ thrich, K. (1998) Prion protein NMR structure and familial human spongiform encephalopathies. Proc. Natl Acad. Sci. USA 95, 11667–11672. 70. Adman, E.T. (1985) Structure and functin of blue copper proteins. In Metalloproteins (Harrison, P., ed.), pp. 1–42. Macmillan, New York. 71. Adman, E.T. (1991) Copper protein structures. Adv. Prot. Chem. 42, 144–197. 72. Solomon, E.I., Baldwin, M.J. & Lowery, M.D. (1992) Electronic structures of active sites in copper proteins: contributions to reactivity. Chem. Rev. 92, 521. 73. Bertini, I., Ciurli, S., Dikiy, A., Gasanov, R., Luchinat, C., Martini, G. & Safarov, N. (1999) High-field NMR studies of oxidized blue copper proteins: the case of spinach plastocyanin. J. Am. Chem. Soc. 121, 2037–2046. 74. Bertini, I., Bryant, D.A., Ciurli, S., Dikiy, A., Ferna ´ ndez, C.O., Luchinat, C., Safarov, N., Vila, A.J. & Zhao, J. (2001) Backbone dynamics of plastocyanin in both oxidation states. Solution structure of the reduced form and comparison with the oxidized state. J. Biol. Chem. 276, 47217–47226. 75. Bodenhausen, G. & Ruben, D.J. (1980) Natural abundance nitrogen-15 NMR by enhanced heteronuclear spectroscopy. Chem. Phys. Lett. 69, 185–188. 608 I. Gelis et al. (Eur. J. Biochem. 270) Ó FEBS 2003 76. Kay, L.E., Keifer, P. & Saarinen, T. (1992) Pure absorption gradient enhanced heteronucler single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114, 10663–10665. 77. Cavanagh, J., Palmer, A.G., IIIWright, P.E. & Rance, M. (1991) Sensitivity improvement in proton-detected two-dimen- sional heteronuclear relay spectroscopy. J. Magn. Reson. 91,429– 436. 78. Palmer,A.G., IIICavanagh, J., Wright, P.E. & Rance, M. (1991) Sensitivity improvement in proton-detected two dimension al heteronuclear correlation NMR spectroscopy. J. Magn. Reson. 93, 151–170. 79. Bertini, I., Couture, M.M.J., Donaire, A., Eltis, L.D., Felli, I.C., Luchinat,C.,Piccioli,M.&Rosato,A.(1996)Thesolution structure refinement of the paramagnetic reduced HiPIP I from Ectothiorhodospira halophila by using stable isotope labeling and nuclear relaxation. Eur. J. Biochem. 241, 440–452. 80. Kay, L.E., Ikura, M., Tschudin, R. & Bax, A. (1990) Three- dimensional triple-resonance NMR spectroscopy of isotopically enriched proteins. J. Magn. Reson. 89, 496–514. 81. Pochapsky, S.S., Jain, N.U., Kuti, M., Lyons, T.A. & Heymont, J. (1999) A refined model for the solution structure of oxidized putidaredoxin. Biochemistry 38, 4681–4690. 82. Baumann, B., Sticht, H., Scha ¨ rpf, M., Sutter, M., Haehnel, W. & Roesch, P. (1996) Structure of synechococcus-elongatus [Fe 2 S 2 ] ferredoxininsolution.Biochemistry 35, 12831–12841. 83. Vathyam, S., Byrd, R.A. & Miller, A.F. (1999) Assignment of the backbone resonances of oxidized Fe-superoxide dismutase, a 42 kDa paramagnet-containing enzyme. J. Biomol. NMR 14,293– 294. 84. Piccioli, M. & Poggi, L. (2002) Tailored HCCH-TOCSY experi- ment for resonance assignment in the proximity of a paramagnetic center. J. Magn. Reson. 155, 236–243. 85. Skidmore, K. & Simonis, U. (1996) Novel strategy for assigning hyperfine shifts using pulsed-field gradients heteronuclear multi- ple-bond correlation spectroscopy. Inorg. Chem. 35, 7470–7471. 86. Bertini, I., Lee, Y M., Luchinat, C., Piccioli, M. & Poggi, L. (2001) Locating the metal ion in calcium-binding proteins by using cerium (III) as a probe. Chembiochem. 2, 550–558. 87. Bertini, I., Dalvit, C., Huber, J.G., Luchinat, C. & Piccioli, M. (1997) ePHOGSY experiment on a paramagnetic protein: location of the catalytic water molecule in the heme crevice of the oxidized form of horse heart cytochrome c. FEBS Lett. 415, 45–48. 88. States, D.J., Habenkorn, R.A. & Ruben, D.J. (1982) A two- dimensional nuclear overhauser experiment with pure absorption phase in four quadrants. J. Magn. Reson. 48, 286–292. 89. Grzesiek, S. & Bax, A. (1992) Improved 3D triple-resonance NMR techniques applied to a 31 KDa protein. J. Magn. Reson. 96, 432–440. 90. Bertini, I., Luchinat, C., Macinai, R., Piccioli, M., Scozzafava, A. & Viezzoli, M.S. (1994) Paramagnetic metal centers in proteins can be investigated through heterocorrelated NMR spectroscopy. J. Magn. Reson. B104, 95–98. 91. Wishart, D.S., Sykes, B.D. & Richards, F.M. (1991) Relationship between nuclear magnetic resonance chemical shift and protein secondary structure. J. Mol. Biol. 222, 311–333. 92. Wishart, D.S. & Sykes, B.D. (1994) The 13 Cchemicalshift index: a simple method for the identification of protein secondary structure using 13 C chemical shift data. J. Biomol. NMR 4,171– 180. 93. Spera, S. & Bax, A. (1991) Empirical correlation between protein backbone conformation and 13 Ca and C b 13 C nuclear magnetic resonance chemical shifts. J. Am. Chem. Soc. 113, 5490–5492. 94. Bertini, I., Bren, K.L., Clemente, A., Fee, J.A., Gray, H.B., Luchinat, C., Malmstro ¨ m, B.G., Richards, J.H., Sanders, D. & Slutter, C.E. (1996) The Cu a center of a soluble domain from thermus cytochrome ba 3 :anNMRinvestigationofthepara- magnetic protein. J. Am. Chem. Soc. 46, 11658–11659. 95. Jensen, M., Hansen, D.F. & Led, J.J. (2002) A general method for determining the electron self-exchange rates of blue copper pro- teins by longitudinal NMR relaxation. J. Am. Chem Soc. 124, 4093–4096. 96. Ma, L., Philipp, E. & Led, J.J. (2001) Determination of the elec- tron self-exchange rates of blue copper proteins by super-WEFT NMR spectroscopy. J. Biomol. NMR 19, 199–208. 97. Bertini, I., Ferna ´ ndez,C.O.,Karlsson,B.G.,Leckner,J.,Luch- inat, C., Malmstro ¨ m, B.G., Nersissian, A.M., Pierattelli, R., Shipp, E., Valentine, J.S. & Vila, A.J. (2000) Structural informa- tion through NMR hyperfine shifts in blue copper proteins. J. Am. Chem. Soc. 122, 3701–3707. 98. Banci, L., Bertini, I., Luchinat, C., Piccioli, M., Scozzafava, A. & Turano, P. (1989) 1 H NOE studies on dicopper (II) dicobalt (II) superoxide dismutase. Inorg. Chem. 28, 4650–4656. 99. Goasdoue, N., Riviere, D.G., Correia, I., Convert, O. & Piccioli, M. (2000) Multiple selective excitation as a tool for NMR studies of paramagnetic proteins. Magn. Reson. Chem. 38, 827–832. 100. Banci, L., Bertini, I., Eltis, L.D., Felli, I.C., Kastrau, D.H.W., Luchinat, C., Piccioli, M., Pierattelli, R. & Smith, M. (1994) The three dimensional structure in solution of the paramagnetic pro- tein high-potential iron-sulfur protein I from Ectothiorhodospira halophila through nuclear magnetic resonance. Eur. J. Biochem. 225, 715–725. 101. Guss, J.M. & Freeman, H.C. (1983) Structure of Oxidized Poplar Plastocyanin at 1.6 A Resolution. J. Mol. Biol. 169, 521–563. 102. Donaire, A., Salgado, J. & Moratal, J.M. (1998) Determination of the magnetic axes of cobalt (II) and nickel (II) azurins from 1H NMR data: influence of the metal and axial ligands on the origin of magnetic anisotropy in blue copper proteins. Biochemistry 37, 8659–8673. 103. Donaire, A., Jimenez, B., Moratal Mascarell, J.M., Hall, J.F. & Hasnain, S.S. (2001) Electronic characterization of the oxidized state of the blue copper protein rusticyanin by 1H NMR: is the axial methionine the dominant influence for the high redox potential. Biochemistry 40, 837–846. 104. Randall, D.W., Gamelin, D.R., LaCroix, L.B. & Solomon, E.I. (2000) Electronic structure contributions to electron transfer in blue Cu and Cu A . JBIC 5, 16–29. 105. LaCroix, L.B., Randall, D.W., Nersissian, A.M., Hoitink, C.W.G., Canters, G.W., Valentine, J.S. & Solomon, E.I. (1998) Spectroscopi and geometric variation in perturbed blue copper centers: electronic structures of stellacyanin and cucumber pasic protein. J. Am. Chem. Soc. 120, 9621–9631. 106. Solomon, E.I., Penfield, K.W., Gewirth, A.A., Lowery, M.D., Shadle, S.E., Guckert, J.A. & LaCroix, L.B. (1996) Electronic structure of the oxidized and reduced blue copper sites: con- tributions to the electron transfer pathway, reduction potential, and geometry. Inorg. Chim. Acta 243, 67–78. 107. Werst, M.M., Davoust, E.E. & Hoffman, B.M. (1991) Ligand spin densities in blue copper proteins by Q-band 1 Hand 14 N ENDOR spectroscopy. J. Am. Chem. Soc. 113, 1533–1538. Supplementary material The following material is available from http://www. blackwellpublishing.com/products/journals/suppmat/EJB/ EJB3400/EJB3400sm.htm Table S1. 13 C assignment obtained for oxidized plastocya- nin from Synechocystis sp. PCC6803 (BMRB accession number: 5584). Ó FEBS 2003 Blue copper proteins studied by NMR (Eur. J. Biochem. 270) 609 . peak 10. All the new assignments are summarized in Table 4. A simple NMR protocol is an important tool to study paramagnetic proteins Plastocyanin is a. resulting 3D spectra are standard spectra that can be handled by any standard software for protein NMR data analysis. Keywords: blue copper proteins; NMR spectroscopy;