Solution 1H NMR Study of the Influence of Distal Hydrogen Bonding and N Terminus Acetylation on the Active Site Electronic and Molecular Structure of Aplysia limacinaCyanomet Myoglobin*

The sea hare Aplysia limacinapossesses a myoglobin in which a distal H-bond is provided by Arg E10 rather than the common His E7. Solution 1H NMR studies of the cyanomet complexes of true wild-type (WT), recombinant wild-type (rWT), and the V(E7)H/R(E10)T and V(E7)H mutants of AplysiaMb designed to mimic the mammalian Mb heme pocket reveal that the distal His in the mutants is rotated out of the heme pocket and is unable to provide a stabilizing H-bond to bound ligand and that WT and rWT differ both in the thermodynamics of heme orientational disorder and in heme contact shift pattern. The mean of the four heme methyl shifts is shown to serve as a sensitive indicator of variations in distal H-bonding among a set of mutant cyanomet globins. The heme pocket perturbations in rWT relative to WT were traced to the absence of the N-terminal acetyl group in rWT that participates in an H-bond to the EF corner in WT. Analysis of dipolar contacts between heme and axial His and between heme and the protein matrix reveal a small ∼2° rotation of the axial His in rWT relative to true WT and a ∼3° rotation of the heme in the double mutant relative to rWT Mb. It is demonstrated that both the direction and magnitude of the rotation of the axial His relative to the heme can be determined from the change in the pattern of the contact-dominated heme methyl shift and from the dipolar-dominated heme meso-H shift. However, only NOE data can determine whether it is the His or heme that actually rotates in the protein matrix.

Myoglobin (Mb) 1 is a member of the globin family of proteins of approximately 140 -150 residues that encapsulate heme and exhibit a remarkably strongly conserved fold of seven to eight helices (A-H) despite a high variability in sequence (1)(2)(3). The heme is wedged between the E and F helices, and only the axial (proximal) His F8 (eighth residue on helix F) and Phe CD1 (first residue on the loop between the C and D helices) are completely conserved. This conserved globin fold, however, results in a very wide range of functionality, which appears to be controlled primarily by the nature of the "distal" residues at position E7, E11, E10, and B10 that line the ligand binding pocket (4). The major interaction that strongly stabilizes O 2 over CO binding has been shown to involve H-bonding to the bound O 2 by a distal residue, although destabilization of the bound CO by distal steric interaction cannot be completely discounted (4 -6). Although the distal H-bond in vertebrate globins is always provided by residue E7 (which is overwhelmingly His, but occasionally Gln (2)), there is much more variation in the nature of the E7 residue and the position of the distal H-bond donor in nonvertebrate globins (3). Such alternate residue H-bond stabilization of the bound O 2 has been established in natural globins from sea hares (Arg E10) (7) and trematodes (Tyr B10) (8 -10) and in synthetic globins at position E11 (Asn and Thr) (11).
An effective strategy for determining the role of individual residues is to perform functional and molecular structural studies of site directed mutant globins (4). X-ray crystallography is generally the most effective tool for describing both the global and heme pocket structures of globins (4,6,12). The description of the globin active site structure, however, is just as effectively pursued by NMR, particularly in the paramagnetic oxidation/spin states, in which structurally exquisitely sensitive hyperfine shifts impart improved active site resolution over a diamagnetic analog (13)(14)(15). Hyperfine shifts reflect electronic and/or magnetic properties of the chromophore, and hence can be extraordinarily sensitive to (and hence render detectable) small structural changes that are unlikely to be detected in either a crystal structure or a solution NMR structure of a diamagnetic analog. Perhaps the best examples of the exquisite sensitivity are the observation of isotope effects on iron-porphyrin covalency due to the distal H-bond to the bound ligand (16 -18) and the characterization of small populations of globins with alternate orientation of the heme about the ␣,␥meso axis (19,20).
The information content is particularly rich in the cyanomet globins, in which the bound cyanide can model both the H-bond acceptor properties of bound O 2 (21)(22)(23) and the potential steric tilt/bend of bound CO (17, 24 -28). The large dipolar shifts, moreover, guarantee that any heme pocket labile proton can be detected (usually resolved), and its role in H-bonding to the ligand elucidated directly by its placement in the distal pocket (17,22,23,26,28), and indirectly by the expected influence of such a distal H-bond on the electronic structure of the heme (16 -18). The pattern of the dominant heme methyl contact shifts reflect the orientation of the axial His relative to the heme (i.e. in Fig. 1) (29 -33), and the pattern of the dominant meso-H dipolar shift reflects the orientation of the rhombic magnetic axes ( in Fig. 1) (34,35). The mean of the heme methyl contact shifts has been shown to be sensitive to distal H-bonding to bound cyanide in models (36), but it has not yet been assessed in globins. Lastly, the dominant dipolar shifts for nonligated residues provide information on the orientation of the paramagnetic susceptibility tensor, which can be related to the tilt/bend of the Fe-CN unit and the orientation of the axial His and thereby facilitate the determination of the orientation of mutated residues in the heme cavity (17, 21, 22, 24 -26, 28).
The Mbs from the sea hare Aplysia limacina (37), like those from Dolabela auricularia (38), possess a Val E7 but still exhibit high O 2 affinity and reasonably slow O 2 off-rates. Both crystallography (12) and solution 1 H NMR on Aplysia Mb (22,39) and solution 1 H NMR of Dolabela Mb (40) have demonstrated that Arg E10 can orient into the heme pocket and provide an H-bond to bound ligand. The sharp increase in k off and decrease in affinity for O 2 in the Aplysia R(E10)T-Mb mutant directly confirms (7) the H-bonding role for this residue. A question that naturally arises is whether the alternate H-bonding residues in sperm whale and Aplysia Mbs can be interchanged solely by interchanging the Val E7/Arg E10 in Aplysia with the His E7/Thr E10 of sperm whale Mb. Similar studies have shown that substitution of the key distal residues can in small or large part transfer an unusual functional property from one globin to another (17,26,28). Thus, although the O 2 off-rate increases sharply and its O 2 affinity decreases sharply upon substituting His E7 by Val in sperm whale Mb, a significant portion of the O 2 affinity can be recovered by inserting Arg E10 in the sperm whale H(E7)V/T(E10)R-Mb mutant (41). Solution 1 H NMR of sperm whale H(E7)V/T(E10)R-met-MbCN found that Arg E10 side chains oriented to provide an H-bond to the bound cyanide (22). Hence, the distal H-bonding scheme of Aplysia Mb can be transferred, albeit less effectively, to sperm whale Mb. The successful cloning and expression of Aplysia Mb has been reported (7), although the N terminus of the recombinant protein, in contrast to true wild-type (12), is not acetylated (7). Substitution of Arg E10 for the Thr found in mammalian Mb led to very large decrease in O 2 affinity and increase in O 2 off-rate and autoxidation and clearly identified Arg E10 as the source of the H-bond stabilization to the O 2 ligand (7). However, engineering the second residue to convert the Aplysia Mb to mimic the sperm whale Mb pocket, V(E7)H/ R(E10)T-Mb, failed to either recover a significant fraction of the O 2 affinity or retard the O 2 off-rates relative to WT Aplysia Mb.
We report herein on the 1 H NMR spectra of the cyanomet complex of the Aplysia mutants V(E7)H-Mb and V(E7)H/ R(E10)T-Mb that show that the individual distal His E7 residues are indeed oriented out of the heme pocket and hence cannot participate in any H-bonding interaction with the bound cyanide. On the other hand, comparison of the 1 H NMR spectra of native wild-type (WT) and recombinant wild-type (rWT) Aplysia metMbCN shows that the N-acetylation missing in rWT Mb (7) leads to changes in both the relative stability of the alternate heme orientations in the heme pocket and a small reorientation of the proximal His imidazole plane relative to the F-helix, when compared with true WT.

EXPERIMENTAL PROCEDURES
Protein Preparation-A. limacina V(E7)H and V(E7)H/R(E10)T mutant myoglobins were expressed and purified as described previously (7). 1 H NMR Spectra-1 H NMR spectra were collected on a GE Omega 500 MHz spectrometer. The strongly relaxed signals were optimally detected in WEFT spectra (42). Nonselective T 1 values for the resolved strongly relaxed protons were measured via inversion-recovery experiments. Steady-state NOEs were recorded as described in detail previously (43). The phase-sensitive TOCSY (44), NOESY (45), and conventional magnitude COSY (46) employed the method described by States et al. (47) to provide quadrature detection in the t1 dimension. Solvent suppression, when required, was achieved by direct saturation in the relaxation delay period. 512 blocks were collected with 25.0 kHz spectral width to include all resonances and 10 kHz to improve resolution for the diamagnetic envelope. 128 -256 scans were accumulated with a repetition rate of 0.7 or 1.2 s Ϫ1 for each block with free induction decays of 2048 complex points. The data were processed as described previously (22); details are given in the figure legends. All two-dimensional data were processed on Silicon Graphics workstation using the software package Felix from Biosym/MSI (San Diego, CA).
Determination of Magnetic Axes-The magnetic axes were determined as described in detail previously (21, 22, 24 -26, 28, 34, 35). Experimental dipolar shifts (n values) for the structurally conserved proximal side of the heme were used as input to search for the Euler rotation angles, ⌫(␣, ␤, ␥), which transform the molecular pseudosymmetry coordinates (xЈ, yЈ, zЈ or R, Ј, ⍀Ј (Fig. 1)), readily obtained from crystal coordinates (12), into magnetic axes x*, y*, and z*, (where is diagonal) by minimizing the global error function, F/n, where ␦ dip (calc) and ␦ dip (obs) are given by respectively. ⌬ ax and ⌬ rh are axial and rhombic anisotropies, and ␦ DSS (obs) is the observed chemical shift referenced to DSS. ␦ dia is the shift in the isostructural diamagnetic complex that is calculated via ␦ dia ϭ ␦ tetr ϩ ␦ sec ϩ ␦ rc , where ␦ tetr is the shift in an unfolded tetra peptide (48); ␦ sec is the shift of an amino acid proton typical for ␣-helices, ␤-strand, coils, etc. (49); and ␦ rc is the ring current shift (50). Minimizing the error function F/n in Equation 1 was performed over three parameters, ␣, ␤, and ␥, using the ⌬ ax and ⌬ rh from WT metMbCN (22) or extended to all five parameters to yield both the Euler angles and anisotropies, using the A. limacina MbF crystal coordinates (12), as described in detail previously (22). Dipolar Shift Simulations-The position of a substituted or perturbed residue can be determined by minimizing a local error function (22-24, 26, 28). This local error function, designated F*(residue)/nЈ (where nЈ is the number of protons) to distinguish it from that global error function in Equation 1 is given by the following equation, where ␦ dip ( 1 , 2 . . . ) represents the ␦ dip (calc) as a function of a bond rotations 1 , 2 . . . using the magnetic axes derived from conserved structural elements. The bond angle that minimizes the residual error function F*(residue)/nЈ defines structural changes, as described in detail previously (17, 22-24, 26, 28). The molecular modeling was carried out on a Silicon Graphics Indigo work station from available crystal coordinates using the Insight II program (Biosym/MSI).

RESULTS
Heme Orientational Heterogeneity-1 H NMR spectra for both metMb (51) and metMbCN (19,20) had shown that in solution, ϳ20 -25% of the globin possesses a heme rotated 180°about the ␣,␥-meso axis with respect to the unique orientation reported in the crystal structure (12) (shown in Fig. 1). The metMb spectra at pH 6.0 exhibit low field heme methyl peaks very similar to those of WT (51). However, as shown in the 1 H NMR spectra for the lowest field pair of heme methyls of WT, rWT, Fig. 2, the relative intensities of the major component (M) to minor compo-nent (m) methyl peaks is similar at a 5.0 Ϯ 0.5:1.0 ratio in WT and rWT, but somewhat lower, with a 3.5 Ϯ 0.3:1.0 ratio, for the two mutants. The upfield portion of equilibrated cyanomet complexes of Aplysia WT, rWT, V(E7)H/R(E10)T-Mb and V(E7)H-Mb 1 H NMR spectra, where the vinyl H ␤ peaks (Fig. 3, H 2␤ for major components and h 2␤ for minor components) for both isomers resonate (19,20), are illustrated in Fig. 3, A-D, respectively, and reveal equilibrium major to minor isomer ratios of 3.  Fig. 1 or the position on His 95 (F8)). The spectra are all very similar, as noted previously for mutants and WT sperm whale metMbCN complexes (17, 23-26, 28, 34). However, even cursory examination of Fig. 4, A and B, reveals that the spectra of Aplysia WT and rWT metMbCN differ not only in the ratio of the two isomers noted above, but also in the chemical shifts of heme and axial His resonances, with the heme 1-CH 3 and 5-CH 3 resonating further to low field, and the 3-CH 3 and 8-CH 3 appearing further to high field in rWT than WT metMbCN, clearly showing that the heme electronic structure differs for WT and rWT. This difference in shifts is maintained throughout the alkaline pH range (19). We consider here the 1 H NMR spectral properties and heme pocket structure only for the major isomers in solution for each metMbCN complex; extensive and detailed assignments for the major isomer and some for the minor isomer of WT metMbCN have been reported previously (20,22).
metMbCN Assignments-The procedure for obtaining complete assignment of the heme and the heme cavity residues have been presented in detail for Aplysia WT metMbCN (20) and applied to both WT and numerous mutants of sperm whale Mb (17, 23-26, 34, 35). Hence, NMR data are shown only to substantiate changes in molecular structure from that of WT metMbCN. The complete heme resonances were assigned as reported previously (20) by detecting the characteristic NOESY connections among four low field methyls, two TOCSY-detected three-spin fragments (vinyls), two TOCSY-detected four-spin systems (propionates), and the four strongly relaxed meso-H signals, each with large, temperature-dependent hyperfine shifts. The heme chemical shifts for the four complexes of interest are listed in Table I.
Standard sequential (52) assignments for Aplysia V(E7)H/ R(E10)T-metMbCN locate three segments with the inter-residue NOESY contacts characteristic of three ␣-helical fragments, labeled I, and III, AMX k -Z kϩ1 -AMX kϩ2 -Val kϩ3 (Z residue with more than three nonlabile proton signals). Segment I must arise from Phe 91 (F4)-Val 96 (F9), with AMX iϩ4 exhibiting the large, low field dipolar shifts characteristic of His 95 (F8), as confirmed by the NOE to the peptide NH from a strongly hyperfine-shifted and relaxed labile proton at 14 ppm readily identified as the His 95 (F8) N ␦ H (20). The residues exhibit all the NOESY contacts predicted from the WT crystal structure (12), with the exception of small differences in intensity for His F8 protons as FIG. 1. Schematic representation of the crystal structure-based reference coordinate system, x, y, z; the magnetic coordinate system, x*, y*, z*, in which the paramagnetic susceptibility tensor is diagonal; and the electronic coordinate system, x, y, z, in which d xz , d yz are eigenfunctions, as determined by the axial His orientation relative to the x axis with angle . The reference and magnetic coordinate systems are related by the Euler rotation ⌫(␣, ␤, ␥) according to (x*, y*, z*) ϭ (xЈ, yЈ, zЈ)⌫(␣, ␤, ␥), where ␤ is the tilt of the major axis from the heme normal (not shown), ␣ is the angle between the projection of the tilt on the heme plane and the xЈ axis (not shown), and ϳ ␣ ϩ ␥ corresponds to the location of the rhombic magnetic axes relative to the xЈ and yЈ axes (shown). Theoretical considerations demand that ϭ Ϫ. Panel A shows the effect of a counterclockwise rotation of the axial His by ⌬ relative to heme and protein, with the result that the new † ϭ Ϫ ⌬, and the new † ϭ Ϫ ⌬. Panel B shows the effect of a counterclockwise rotation of the heme by ⌬ relative to a stationary axial His and protein matrix. The new † ϭ ϩ ⌬, but the new †, still referenced to the original axial xЈ, yЈ, must be represented as † ϭ ϩ 2⌬. considered in detail below.
Fragment II, the TOCSY/NOESY data of which are shown in Fig. 5, is identified by the sequence as due to His 63 (E7)-Arg 70 (E14), with His 63 (E7) and Ile 67 (E11) exhibiting significant dipolar shifts, and with residues E10, E11, and E14 exhibiting the NOESY cross peak pattern for rWT connectivities that is identical to that of WT (20) and very similar but not identical (see below) for the mutants relative to rWT. The largest dipolar shifts are displayed by the His 63 (E7) C ␤ Hs. Although the NHC ␣ HC ␤ H 2 fragment of His 63 (E7) could be unambiguously identified in the two mutants, it was not possible to locate the cross-peak to the ring C ␦ H. Lastly, helical segment III arises from Phe 105 (G5)-Val 108 (G8), with the residues exhibiting NOESY contacts to the heme in all three complexes as observed for WT (20). The complete substituted Thr E10 in the double mutant was readily located by TOCSY; spectral congestion prevented assignment of more than the NHC ␣ HC ␤ H 2 fragment of Arg 66 (E10) in the single mutant. The dipolar-shifted residue at two interhelical corners, Phe 43 (CD1), Phe 98 (FG2), and Val 100 (FG4), were identified by TOCSY spectra, characteristic paramagnetic relaxation, and the NOESY contacts to the heme predicted by the crystal structure and previously reported for Aplysia WT metMbCN (20). The chemical shifts for nonligated residues are listed in Table II.
Structural Differences between WT and rWT-The difference in heme methyl contact shifts between WT and rWT dictates (see under "Discussion") a difference in the orientation of the axial His 95 (F8) imidazole plane relative to the heme (29,30,32,33), with in Fig. 1 predicted to decrease (rotate counterclockwise in Fig. 1A) in rWT relative to WT. However, the altered heme methyl contact shift pattern does not distinguish between rotation of the heme relative to a stationary His 95 (F8) (Fig. 1B) and rotation of the His 95 (F8) imidazole ring relative to a stationary heme and protein (Fig. 1A). These alternate movements, however, can be distinguished by changes either in distance between heme methyls (i.e. 1-CH 3 ) and the E helix backbone (i.e. Ile 67 (E11) C ␣ H) to detect heme rotation (24,34) or between the His 95 (F8) N ␦ H and the F helix to detect rotation of His 95 (F8), as described in detail previously (34).
Saturation of the heme 1-CH 3 peak in WT and rWT metM-bCN leads to identical NOEs to Ile 67 (E11) C ␣ H, as shown in Fig. 6, A and B, dictating that the heme orientation relative to the protein matrix is the same in WT and rWT. On the other hand, saturating the His 95 (F8) N ␦ H signal in WT and rWT reveals numerous differences in the pattern of NOEs, as shown in Fig. 7. Quantitative comparison of the NOEs is complicated by overlap of the His 95 (F8) N ␦ H for both the major and minor isomers in each globin and the fact that the His 95 (F8) C ␤ H shifts differ, but also overlap for the two isomers, rendering the interpretation of the NOEs to the C ␤ H peaks ambiguous. Analysis of the WT crystal structure reveals that a small 2-3°r otation of His 95 (F8) ring does not significantly alter the dis- C ␣ H signals are degenerate (not shown). However, saturation of the heme 1-CH 3 signal in the double mutant leads to a ϳ30% larger NOE to Ile 67 (E11) C ␣ H than in rWT (Fig. 6C), and the implied decrease in the distance of 5% requires a 3°counterclockwise rotation of the heme in the mutant relative to conserved His 95 (F8) and protein matrix, as illustrated in Fig. 1B. The heme rotational position of the single mutant was found unchanged from that of rWT on the basis of the magnitude of the 1-CH 3 NOE to Ile 67 (E11) C ␣ H (not shown).
Determination of Magnetic Axes-The magnetic axes for rWT and WT Aplysia metMbCN are essentially identical, as expected from the fact that the dipolar shifts for nonligated residues are largely indistinguishable (Table II), with ␣ ϭ 65°, ␤ ϭ 7.0°, and ϭ 25°(and with uncertainties of Ϯ 10°, Ϯ1°, and Ϯ 10°, respectively) with ⌬ ax ϭ 2.38 ϫ 10 Ϫ8 m 3 /mol and ⌬ rh ϭ Ϫ0.55 ϫ 10 Ϫ8 m 3 /mol, as reported previously 2 (22). Both fiveparameter and three-parameter searches based on the WT (35) ⌬ ax and ⌬ rh using a variety of input data sets yielded highly clustered angles for each mutant with that using the same 21 2 It is noted that the convention for xЈ, yЈ, zЈ differs from that used previously for A. limacina metMbCN (22) Fig. 1). The resolved signal for the axial His 95 (F8) for the major isomer is also labeled. proximal side protons yields ␣ ϭ 75°, ␤ ϭ 11°, and ϭ 35°for V(E7)H/R(E10)T-metMbCN and ␣ ϭ 95°, ␤ ϭ 8.5°, and ϭ 25°f or V(E7)H-metMbCN, with the optimized anisotropies in the five-parameter search inconsequentially altered from those of WT. In each case, the correlation between ␦ dip (obs) and ␦ dip -(calc) was excellent not only for the input data protons but also for the majority of the distal side, excluding the mutated residues (not shown).
Orientation of Distal His 63 (E7) and Thr 66 (E10) in Mutants-Introduction of Thr 66 (E10) into Aplysia metMbF crystal structure (12) with the orientation as found in sperm whale carbonmonoxymyoglobin (6) resulted in an excellent correlation between ␦ dip (obs) and ␦ dip (calc) (not shown). In contrast, the ␦ dip (obs) for one His 63 (E7) C ␤ H in both mutants is much larger than predicted by the His 63 (E7) orientation of sperm whale carbonmonoxymyoglobin. Plots of the residue error function (Equation 4) for the C ␤ Hs of His 63 (E7) as a function of 1 are shown in Fig. 8

for V(E7)H/R(E10)T-metMbCN (line A) and V(E7)H-metMbCN (line B).
A His 63 (E7) orientation in the heme pocket such as in sperm whale Mb corresponds to 1 ϭ Ϫ163°, as shown by the vertical arrow on the left, which predicts shifts clearly inconsistent with those observed in either mutant. However, the plots in Fig. 8 1 is rotated by ϳ120°from that found in sperm whale Mb (6). The magnitude and direction of the change in 1 orients the imidazole ring out of the heme pocket and in the direction of the protein surface. The temperature dependence of the ␦ dip (obs) for the His 63 (E7) C ␤ Hs correlates well with that for other dipolar shifted protons (21), indicating that the His 63 (E7) orientation is relatively well defined by 1 ϳ Ϫ40°and does not represent any equilibrium between an "in" and "out" orientation, as observed in crystals of Chironomus HbIII (53). The His 63 (E7) orientation deduced herein for the mutant metMbCN complexes is completely consistent with the observation (7) of low O 2 affinity and rapid O 2 off-rate indicative of the absence of significant H-bond stabilization of the bound O 2 . Superposition of the crystallographically defined heme cavities of sperm whale (6) and Aplysia (12) Mb indicates that the His E7 C ␣ H is ϳ1.1 Å closer to the iron in Aplysia than sperm whale Mb, such that the His E7 "in" orientation in Aplysia Mb, like that of sperm whale, could not be accommodated in the ligated state.
Effect of N-Acetylation on WT Mb Structure-A logical link between the orientation of the His 95 (F8) ring and N-acetylation in Aplysia Mb is found in the unique, direct interaction between the N terminus and the beginning of the F helix in the form of an H-bond between Leu 2 NH to the Ala 78 (EF) carbonyl at the end of the E-helix, as depicted (12) in Fig. 9. Deletion of the acetyl group (7) can be expected to seriously impact this interaction. The absence of N-acetyl group in rWT metMbCN leads to ϳ0.3 Kcal/mol decrease in the relative stabilization of the major versus minor heme orientation and a ϳ2°counterclockwise rotation for the axial His plane relative to WT met-MbCN. The change in the ratio of the heme orientations demands a small difference in the contacts between the protein and pyrroles A and B of the heme, and the rotation of the His suggests a small translation of the F-helix relative to the heme (rather than heme relative to the F-helix) (see below).
Rotation of the His 95 (F8) about 2 in the Aplysia Mb crystal structure shows that the ring N ⑀ ligated to the iron translates laterally by ϳ0.03 Å per 1°counterclockwise rotation of the axial His, in the direction indicated by the arrow in Fig. 9. The conserved contact shifts for the axial His argue against such a distortion for the iron-His bond. The ϳ2°axial His imidazole rotation, however, would leave the axial His bond intact if the F-helix translated relative to the heme by ϳ0.07 Å in the direction of the ␥-meso position as shown by the arrow in Fig.  9. Such a movement of the central portion of the F-helix is consistent with a ϳ0.15 Å movement of the N terminus of the F-helix in the same direction. The direction of the movement of the F-helix terminus would suggest that the deletion of the N-acetyl group either strongly destabilizes or abolishes the H-bond between the N terminus and the EF corner in rWT. It is noted that such a small translation of the F-helix could account for the altered stabilities for the two heme orientations, inasmuch as Phe 91 (F4) represents an important contact with pyrrole A. Heme Methyl Shifts as Indicator of Distal H-Bonding-It has been observed that the heme mean methyl hyperfine shift in low spin hemin models with ligated cyanide are systematically dependent on the H-bond strength of the solvent, with decreasing H-bond donation leading to increasingly upfield shifted heme methyl shifts (36). Comprehensive published assignments of sperm whale WT (43) and mutants (17,23,24,26), in fact, support a correlation between heme mean methyl shift and presence of labile protons in contact with the bound ligand, as shown in Table III The labile proton for Arg 66 (E10), observed clearly in both WT (22) and rWT Aplysia metMbCN, was not detected in the V(E7)H-metMbCN NMR spectrum. However, even a weak H-bond or only a fractional populated, dynamic H-bond would lead to rapid exchange with solvent. Indeed the O 2 off-rates were found to be very similar for both the single and double mutants (7).
The present data confirm a valuable role of the variable heme mean methyl shift as an indicator of differentiation distal H-bonding among a series of mutant metMbCN complexes. However, because it has been shown that the heme mean methyl shift varies with the axial His orientation (33), as defined by in Fig. 1, it may serve as an indicator of distal H-bonding among a series of point mutants, but not among a The NOESY data were processed by applying 30°-shift sine-bell-squared window over 1024 t 1 ϫ 256 t 2 points prior to zero-filling to 2048 ϫ 1048 data points and Fourier transformation. The TOCSY data were processed by applying an 30°-shifted sine-bell-squared window over 512 t 1 ϫ 256 t 2 points prior to zero-filling to 2048 ϫ 1048 data points and Fourier transformation. series of natural genetic variants that exhibit significantly different , such as Aplysia ( ϭ Ϫ22°) and sperm whale ( ϭ Ϫ5°) metMbCN.

Effect of His/Heme Rotation on 1 H NMR Spectral
Parameters-The perturbed NOE pattern between heme and the protein matrix or axial His and protein matrix indicate that differs slightly between WT and rWT metMbCN and between rWT and V(E7)H/R(E10)T-metMbCN. Such changes in are expected (32,54) to lead to changes in (Fig. 1). The values of obtained from the magnetic axes, however, have Ϯ10°uncertainties (35) and indicate that for the four complexes of interest is the same within the uncertainties. Nevertheless, changes in and manifest themselves in extraordinarily sensitive manners in the heme hyperfine shift patterns (30 -34). It is recognized that the asymmetry in the heme methyl hyperfine shift pattern is dominated by the contact interaction that imposes strong heme methyl shifts dependence of the on the His/heme orientation defined by in Fig. 1. In contrast, it has been shown (34,55) that the asymmetry of the heme meso-H shifts is dominated by the rhombic term of the dipolar shift, which reflects in Fig. 1. Theoretical considerations (54) lead to the expectation that ϭ Ϫ. The changes in heme methyl shifts going from WT 3 rWT metMbCN, based on published modeling of the contact shift (33), estimate a ϳ2°c ounterclockwise rotation of .
For the meso-H, the hyperfine shift asymmetry can be cast in the form,  For WT metMbCN, ⌬␦(meso) calc ϭ Ϫ5.26 ppm, which is in excellent agreement with the observed value of Ϫ5.21 ppm in Table I. It is noted, however, that in rWT metMbCN, ⌬␦(meso) obs becomes more positive to Ϫ4.94, which corresponds to a calculated ⌬␦(meso-H) with decreased by ϳ2°(⌬ in Fig. 1A). Hence, the change in pattern of methyl contact shifts and meso-H dipolar shifts independently confirm both the 2°decrease in detected by altered NOEs of His 95 (F8) N ␦ H to the F helix and the counter-rotation rule (54) that demands that ϭ Ϫ. With the exception of the heme and axial His, only very minor shift differences are observed between WT and rWT metMbCN (see Table II). The observed shift changes, moreover, correlate well in the direction and, at least qualitatively, in magnitude with changes in predicted dipolar shifts due to a change in by 2°, as shown in Fig. 10.
Evidence for a change in in comparing rWT and V(E7)H/ R(E10)T-metMbCN is difficult to detect in the heme methyl contact shift pattern, because the mean shift decreases by ϳ0.8 ppm on abolishing the distal H-bond (see above). The ⌬␦(meso-H) obs , nonetheless, is much more negative in the double mutant (Ϫ6.16 ppm) than in rWT (Ϫ4.94 ppm), and this difference is accounted for in Equation 6 by a ϳ4 -5°increase in . Again, the changes in are in opposite direction to the changes in , and both confirm the direction of the rotation of the heme. It is noted that ⌬ ϳ 4 -5°is larger than the ⌬ ϳ 3°deduced from the NOE data. However, as detailed consideration shows in Fig. 1B, if the heme rather than the axial His rotates, the   experimental , defined with respect to the original reference coordinates xЈ,yЈ, must change by 2⌬ for a net heme rotation by .
Conclusions-Solution 1 H NMR spectra of Aplysia metM-bCN show that the insertion by mutagenesis of His E7 in both V(E7)H-Mb and V(E7)H/R(E10)T-Mb mutants is oriented out of the heme pocket and not able to provide a stabilizing H-bond to bound ligands. Comparison of WT and rWT metMbCN, moreover, shows that the abolished N-acetylation in rWT leads to both a change in the relative stabilities of the alternate heme orientations and a small change in the orientation of the axial His ring in each isomer. It is shown that the rotation of axial His in rWT and of the heme in the double mutant relative to WT lead to a series of changes in the asymmetry of the heme methyl dominant contact shifts and heme meso-H dominant dipolar shifts that are completely consistent with the currently accepted relationship between the active site molecular and electronic structures. Lastly, analysis of the heme mean methyl shift among sperm whale and the present Aplysia metMbCN mutants reveals that it serves as a valuable empirical indicator of distal H-bonding to cyanide. The facility with which the electronic structure changes can be related to small, but possibly functionally relevant, changes in molecular structure, is a testament to the exquisite sensitivity of hyperfine shifts to molecular structure.