PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Inorg Chem. Author manuscript; available in PMC 2010 July 20.
Published in final edited form as:
PMCID: PMC2878369
NIHMSID: NIHMS127127

High Resolution Crystal Structure of the Methylcobalamin Analogs Ethylcobalamin and Butylcobalamin by X-ray Synchrotron Diffraction

Abstract

The X-ray crystal structures of the methylcobalamin (MeCbl) analogs ethylcobalamin (EtCbl) and butylcobalamin (BuCbl) are reported. The X-ray crystal structures of EtCbl and BuCbl were obtained with some of the lowest crystallographic residuals ever achieved for cobalamins (R = 0.0468 and 0.0438, respectively). The Co-C bond distances for EtCbl and BuCbl are 2.023(2) and 2.028(4) Å, whereas the Co-α-5,6-dimethylbenzimidazole (Co-N3B) bond distances were 2.232(1) and 2.244(1) Å, respectively. Although EtCbl and BuCbl displayed a longer Co-N3B bond than that observed in the naturally occurring methylcobalamin, the orientation of the α-5,6-dimethylbenzimidazole moiety with respect to the corrin ring did not vary substantially amongst the structures. The lengthening of both Co-C and Co-N3B bonds in EtCbl and BuCbl can be attributed to the “inverse” trans influence exerted by the σ-donating alkyl groups, typically observed in alkylcobalamins. Analysis of the molecular surface maps showed that the alkyl ligands in EtCbl and BuCbl are directed towards the hydrophobic side of the corrin ring. The corrin fold angles in EtCbl and BuCbl were determined to be 14.7° and 13.1°, respectively. A rough correlation exists between the corrin fold angle and the length of the Co-N3B bond, and both alkylcobalamins follow the same trend.

Keywords: Alkylcobalamin, methylcobalamin, ethylcobalamin, butylcobalamin, X-ray structure, synchrotron diffraction

Introduction

Cobalamins are cobalt complexes coordinated by four equatorial nitrogens from a corrin ring macrocycle. The corrin ring incorporates a total of seven amide side chains - 3 acetamides and 4 propionamides (Figure 1). In addition, the Co center is coordinated to a nitrogen atom from a 5,6-dimethylbenzimidazole (DMB) moiety at the lower (α) axial position. Cleavage of the DMB moiety from Cbl generates the “base-off” conformation of the complex. The upper (β) axial position can be occupied by a range of ligands including alkyl groups such as methyl and 5′-deoxyadenosyl, giving rise to the naturally occurring organometallic corrinoids methylcobalamin (MeCbl) and adenosylcobalamin (AdoCbl), respectively. The first evidence for the existence of a naturally-occurring organometallic compound was presented by Lenhert and Hodgkin in 1961, with the elucidation of the structure of AdoCbl by X-ray crystallography.1 The analysis revealed the presence of a Co-C bond linking the 5′-deoxyadenosyl moiety with the cobalt center of cobalamin. In 1985, the X-ray crystal structure of MeCbl was reported, showing once again the presence of a Co-C bond.2

Figure 1
Chemical structure of cobalamins with atom numbering.

Organometallic corrinoids serve as important cofactors and methyl transfer catalysts in enzymatic reactions, in both prokaryotes and eukaryotes.3, 4 AdoCbl-dependent enzymes catalyze rearrangement reactions in which a key step involves Co-C bond homolysis to produce an adenosyl radical and cob(II)alamin.5, 6 MeCbl-dependent methionine synthases catalyze the remethylation of homocysteine to methionine, generating cob(I)alamin in the process. The latter is remethylated by N5-methyl-tetrahydrofolate to form MeCbl and tetrahydrofolate. Thus, in contrast to AdoCbl-dependent reactions, MeCbl-dependent methionine synthases formally operate via heterolytic cleavage of the Co-C bond.6, 7 Interestingly, MeCbl-dependent methionine synthases are inactivated in vitro upon incubation with alkylating agents. This inactivation is reversed by photolysis, which suggests the formation of a non-functional alkylcobalamin (that is, a MeCbl analog) in the enzyme's active site.8-10 More recently, Yamada and coworkers showed that rat methionine synthase, which requires MeCbl for its catalytic activity, forms inactive complexes with EtCbl and propylcobalamin (PrCbl).11 Again, the enzymatic activity was restored upon exposure to light. Hence, the enzymatic reaction appeared to be very sensitive to nature of the ligand coordinating at the upper axial position.

In vivo, cyanocobalamin (CNCbl, vitamin B12) is converted to MeCbl and AdoCbl in both pro- and eukaryotes.7, 12 Randaccio and coworkers have recently re-determined the X-ray crystal structures of AdoCbl, MeCbl and CNCbl using synchrotron diffraction.13, 14 Analysis of the crystal structures of AdoCbl1, MeCbl2 and CNCbl14 showed that the Co-C bond in MeCbl has a strength similar to that of AdoCbl, but it is significantly weaker than that in CNCbl.13 The rate of cleavage of the Co-C bond in several alkylcobalamins including the natural cofactors is strongly influenced by structural, electronic and solvent factors.15 Indeed, density functional theory (DFT) calculations for alkylcorrinoids indicate that weakening of the lower α-Co-N(DMB) bond (trans influence) is important for achieving a low-energy pathway in the electron transfer mechanism of the Co-C bond breaking in alkylcobalamins.13 In addition, DFT studies also suggest that the length of the Co-N3B bond is largely influenced by crystal packing forces.16 This is particularly relevant to alkylcobalamins incorporating bulky upper axial ligands, such as the 5′-deoxyadenosyl ligand in AdoCbl, which is capable of establishing hydrogen bonding and van der Waals interactions with the neighboring atoms, but of less importance for alkylcobalamins possessing small upper axial ligands (like methyl in MeCbl) for which interactions with the rest of the atoms are essentially impeded.17

Another important aspect of alkylcobalamin chemistry has emerged from several studies where cobalamins operate as catalysts in the dechlorination of perchloro- and trichloroethylene,18-20 which are widespread contaminants of soils and aquifers.21-23 The process by which these toxic species are converted to ethylene and acetylene takes place in vivo in certain strains of bacteria24, 25 and involves corrinoid-dependent enzymes. Model studies indicate that the reactions involve breaking and forming of the corrinoid Co-C bond.18-20 The X-ray crystal structures of two model catalysts, vinylcobalamin and cis-chlorovinylcobalamin, were recently reported.19

Herein, we report the high-resolution X-ray crystal structure of ethylcobalamin (EtCbl) and butylcobalamin (BuCbl). Both analogs displayed the so called “inverse” trans influence, typically observed in alkylcobalamins. Analysis of the electronic surface maps indicate that both the ethyl and butyl ligands are directed towards the hydrophobic side of the cobalamin molecule. This particular conformation has been observed in other Co-C bonded cobalamins and appears to be an important feature of cofactor reactivity in vivo.26

Experimental Section

General procedures and chemicals

Hydroxocobalamin hydrochloride (HOCbl•HCl) was purchased from Fluka. Stated purity by manufacturer is ≥96%. However, the aromatic region of the 1H NMR spectrum shows it contains at least ~6% impurities.27 Alkyl halides (bromoethane, 99%, FW 108.97, d 1.46 g/mL, and 1-bromobutane, 99%, FW 137.02, d 1.276 g/mL) were purchased from Sigma. All chemicals were used without further purification. Water was purified using a Barnstead Nanopure Diamond water purification system and/or HPLC grade water was used. All solutions used for the synthesis of the alkylcobalamins were degassed by at least three freeze-pump-thaw cycles under argon using standard Schlenck techniques. Syntheses were carried out in an MBraun Labmaster 130(1250/78) glove box operating under argon atmosphere, or in an anaerobic cuvette with continuous argon bubbling. Both synthesis and handling of the alkylcobalamins were performed in the darkness under red-dim light illumination.

UV-visible spectra were recorded on a Cary 5000 spectrophotometer equipped with a thermostatted cell holder, operating with WinUV Bio software (version 3.00). 1H NMR spectra were recorded on a Varian Inova 500 MHz spectrometer equipped with a 5 mm thermostatted (22.0 ± 0.5 °C) probe. Solutions were prepared in D2O and TSP (3-(trimethylsilyl)propionic-2,2,3,3-d4 acid, Na+ salt) used as an internal reference. pH measurements were made at room temperature with a Corning Model 445 pH meter equipped with a Mettler-Toledo Inlab 423 electrode.

Liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS) was carried out using a Thermo Finnigan Triple Stage Quadrupole (TSQ) Quantum Ultra mass spectrometer (Thermo Electron Corporation, Waltham, MA, USA). Each alkylcobalamin was dissolved in 95%MeOH/5%H2O at a concentration of 10 mg/L and infused at a rate of 10 μL/min.

Synthesis of alkylcobalamins

EtCbl and BuCbl were synthesized by reacting cob(I)alamin with the corresponding alkyl halide for ~10 min in the absence of light under anaerobic conditions according to literature methods.28, 29 In both cases, cob(I)alamin was generated by reduction of aquacobalamin with 10 equivalents of sodium borohydride (NaBH4). Prior to adding the alkyl halide, excess NaBH4 was destroyed by the addition of a small excess of anaerobic acetone. The alkylcobalamin product precipitated upon drop-wise addition into cold acetone (4°C, 20 mL) and was filtered and dried overnight under vacuum at 60°C. Alkyl halides are soluble in acetone; hence it is unlikely that unreacted alkyl halide co-precipitates with the product. The percentage of non-cobalamin impurities was determined by converting the products to dicyanocobalamin, (CN)2Cbl- (0.10 M KCN, pH 10.0, ε368 = 3.04 × 104 M-1cm-1).30

EtCbl

Bromoethane (5 equiv., 1.27 × 10-4 mol, 9.5 μL) was added to HOCbl•HCl (35 mg, 2.53 × 10-5 mol) dissolved in H2O (1.0 mL). Yield: ~ 90%. 1H NMR (D2O, δ, ppm): 6.06, 6.26, 6.27, 7.00 and 7.18,; ~98% purity.27 ES/MS (+ve), m/z: 1358.8 (calcd for [EtCbl + H]+, C64H94N13O14PCo = 1358.6); 1380.7 (calcd for [EtCbl + Na]+, C64H93N13O14PCoNa = 1380.6); 1381.6 (calcd for [EtCbl + H + Na], C64H94N13O14PCoNa = 1381.6); 680.0 (calcd for [EtCbl + 2H]2+, C64H95N13O14PCo = 679.8).

BuCbl

1-Bromobutane (5 equiv., 1.27 × 10-4 moles, 13.6 μL) was added to HOCbl•HCl (35 mg, 2.53 × 10-5 mol) dissolved in H2O (1.0 mL). Yield: ~ 87%. 1H NMR (D2O, δ, ppm): 6.08, 6.26, 6.29, 7.01 and 7.18; ~98% purity.27 ES/MS (+ve), m/z: 1386.7 (calcd for [BuCbl + H]+, C66H98N13O14PCo = 1386.6); 1408.7 (calcd for [BuCbl + Na]+, C66H97N13O14PCoNa = 1408.6); 1409.8 (calcd for [BuCbl + H + Na], C66H98N13O14PCoNa = 1409.6) 693.9 (calcd for [BuCbl + 2H]2+, C66H99N13O14PCo = 693.8).

X-ray diffraction studies

With the aim of obtaining suitable crystals of structural analogs of MeCbl, we synthesized and crystallized EtCbl and BuCbl. The synthesis of propylcobalamin (PrCbl), pentylcobalamin (PnCbl) and hexylcobalamin (HxCbl) was also carried out, but attempts to obtain crystals of these derivatives were unsuccessful.

Crystals of EtCbl and BuCbl were grown from their corresponding synthesis mixtures, at room temperature, under anaerobic conditions. Crystals suitable for X-ray analysis appeared after 1 week. The EtCbl and BuCbl crystals were transferred into paratone oil and any residual synthesis mixture was carefully removed by dragging the crystals through the oil. The crystals were then mounted in thin nylon loops on copper magnetic pins (Hampton Research) and flash-cooled in liquid nitrogen. X-ray diffraction experiments were carried out at beamline BL9-2 at the Stanford Synchrotron Radiation Laboratory (SSRL). Data from both complexes were collected on a MarMosaic 325 CCD detector using X-rays produced by a 16-pole wiggler insertion device through a flat Rh-coated collimating mirror, a liquid nitrogen-cooled double Si(111) crystal monochromator and a toroidal focusing mirror. The X-ray wavelength used for both crystals was 0.78468 Å (15800 eV). For EtCbl, two data sets were collected, both consisting of 360 0.5° images covering the same 180° angular range and with a crystal to detector distance of 95.0 mm. The first data set was recorded with an exposure time of 5 s and no beam attenuation, and the second data set had the same exposure time but the beam was attenuated by 75%. The two data sets were processed with the program XDS and scaled together with the program XSCALE.31 Symmetry-equivalent and Bijvoet pairs were not merged and no absorption correction was applied. A total of 132135 reflections were measured, resulting in 18317 unique reflections to a nominal resolution of 0.71 Å, with a merging R-factor of 0.028 for common reflections on all images. For BuCbl only one data set was recorded comprising 360 0.5 images with an exposure time of 15 s and a crystal to detector distance of 95.0 mm. The data were processed and scaled with XDS and XSCALE,31 with the symmetry-equivalent and Bijvoet pairs unmerged. A total of 66437 reflections were measured, giving a final unique data set of 20451 reflections to a resolution of 0.71 Å with a merging R-factor of 0.026.

Both structures were solved by Patterson methods as implemented in the program SHELXS.32 The cobalt, phosphorus and some of the nitrogen atoms were readily located, the remainder of the lighter atoms later identified by difference Fourier synthesis. The structures were refined by full-matrix least-squares methods using SHELXL.32 A correction for the anomalous scattering from cobalt at 15800 eV was applied during refinement. All non-hydrogen atoms were refined with anisotropic thermal parameters and hydrogen atoms were added in idealized positions and refined in riding positions at the very end of the refinement. Additional difference electron density peaks were modeled as water molecules. Seventeen water molecules were added to the EtCbl structure and refined with full occupancy. Twenty water molecules were added to BuCbl, ten with fractional occupancies. The final crystallographic R factor (R1) for EtCbl was 0.0468 for 18291 reflections with Fo > 4σF, and for BuCbl R1 was 0.0438 for 20171 reflections with Fo > 4σF.

The CCDC database (EtCbl: CCDC 721275; BuCbl: CCDC 721276) contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, by emailing ku.ca.mac.cdcc@tseuqer_atad, or by contacting The Cambridge Crystallographic Data Centre, 12, Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

Results and Discussion

Synthesis and characterization of EtCbl and BuCbl

The structural characterization of a number of corrinoid complexes has facilitated our understanding of cobalamin chemistry in vivo.33 For example, comparison of the protein-bound cobalamins to model compounds have shown that the DMB side chain of ring D is displaced from Co in the protein structures of methionine synthase and methylmalonyl-CoA mutase, and some of the propionamide and acetamide side chains have reoriented to hydrogen-bond to protein atoms.34 The plane of the histidine residues that bind Co are oriented similarly to the benzimidazole rings of the model cobalamins.34 Upward folding of the corrin ring and strain induced alterations of the Co-C and Co-N axial bonds have been observed in the protein-bound structures compared to the protein-free MeCbl and AdoCbl X-ray structures.34

We recently reported the X-ray crystal structures of imidazolylcobalamin (ImCbl) and histidinylcobalamin (HisCbl),35 cobalamin models for aquacobalamin bound to the B12 transporter transcobalamin.36, 37 A superposition of the model cobalamins with that of the protein-bound showed that the orientation of the imidazole with respect to the corrin ring is the same in the bound protein and in the isolated cobalamin.35, 37 The Co-N bond distances of the upper and lower axial ligands in the protein-bound cobalamin were considerable longer than those found in the model cobalamins, whereas the corrin fold angle was smaller in the protein-bound cobalamin compared to the corresponding protein-free form.35, 37

The partial characterization of alkylcobalamins other than MeCbl has been previously reported by other laboratories and our UV-visible and 1H NMR spectroscopic data for EtCbl and BuCbl (Table 1) were in excellent agreement with the literature.29, 38-40 UV-vis and 1H NMR spectra were almost identical for EtCbl and BuCbl and differed significantly from that of MeCbl (Table 1). The γ-band of the UV-visible spectrum of corrinoids, which is the most intense band in the near ultraviolet region, reflects the amount of negative charge donated by the upper axial ligand to the cobalt via the σ-bond.29 This was observed for EtCbl and BuCbl when compared to MeCbl (Table 1). On the other hand, the 1H NMR resonance of the C10 hydrogen for EtCbl and BuCbl are considerably downfield compared with MeCbl (6.06 and 6.08 ppm respectively, versus 5.91 ppm for MeCbl). The effect of the upper axial ligands of the Cbl on the hydrogen atom on C10, the bridge carbon between rings B and C (Figure 1) has been previously studied.29 Examination of MeCbl, EtCbl, PrCbl and BuCbl showed that strong electron donating ligands in the β-axial position of hexacoordinated corrinoids induced a downfield shift in the magnetic resonance of the C10 hydrogen.29 Our results follow the same pattern (Table 1). Also, replacement of the C10 hydrogen by another substituent affects the electronic structure of the corrin macrocycle and influences the binding properties of the cobalt center.41-43 These findings indicate that the hydrogen atom in C10 is extremely sensitive to its electronic environment and vice versa. Thus, the chemical shifts of the hydrogen in C10 for EtCbl and BuCbl display the correlation expected for the stronger electron donating ability of the respective alkyl ligands compared to that of the methyl group in MeCbl (Table 1).

Table 1
UV-visible and 1H NMR spectroscopy data.

X-ray structural characterization of EtCbl and BuCbl

Suitable crystals for X-ray analysis were obtained for EtCbl and BuCbl. Both EtCbl and BuCbl crystallized in the orthorhombic space group P212121 with cell parameters 16.00 Å × 21.02 Å × 24.54 Å, and 16.03 Å × 20.73 Å × 24.62 Å, respectively. A summary of the crystallographic data and refinement parameters is presented in Table 2. The crystal structure and the crystal packing of the cobalamin molecule in this space group has been described exhaustively in the literature and the current structures do not deviate markedly from the cobalamin structures known to date. The final R-factors of 0.0468 and 0.0438 for the two structures are amongst the lowest ever achieved for cobalamin structures, and this can be directly correlated with the quality of the diffraction data. The merging R-factors for the synchrotron data were extremely good (0.0281 and 0.0262 for the symmetry-related reflections) and the average signal to noise values (I/σ) were 57.2 and 31.4 for EtCbl and BuCbl, respectively. Comparison of the two data sets indicates that although they both extend to approximately the same resolution (around 0.71 Å), the EtCbl data is almost twice as strong as the BuCbl data (as judged by the I/σ values quoted above). Despite the fact that the data resulted from two data sets merged together, the better signal to noise ratio clearly results in a more accurately-determined crystal structure.

Table 2
Refinement parameters for EtCbl and BuCbl.

The crystal packing of the cobalamin molecule in the P212121 space group is generally analyzed by comparison of the ratios of the c/a and b/a unit cell dimensions.44 The EtCbl ratios of c/a = 1.534 and b/a = 1.314, and the BuCbl ratios of c/a = 1.536 and b/a = 1.293 indicate that these crystals are typical of cluster I packing.44 Cobalamins that belong to the same crystal packing type display remarkably similar conformations of their side chains as well as a similar pattern of intra and intermolecular hydrogen bonding.14 The cobalamin molecules are oriented in the crystal such that the plane of the corrin ring is almost parallel to the ab plane of the unit cell, tilted by approximately 5-10°, and when viewed down the a-axis, the molecules appear to adopt a slight zig-zag sheet-like pattern due to this tilt, running parallel to the ab plane. Layers of solvent molecules lie between the tilted cobalamin sheets, with the axial base and the upper axial ligand projecting into this layer. Long solvent channels are also evident when viewed down the c axis. In both cases the solvent was modeled very accurately, with 17 placed in EtCbl and 20 in BuCbl, with the residual peaks in the electron density less than 1 e/Å3 in both structures. Thermal ellipsoids plots for EtCbl and BuCbl (drawn at 75% probability) are presented in Figure 2. Of the 21 water molecules in the BuCbl structure, five were in fully occupied positions and the rest were refined with partial occupancy, initially set at 0.5 based upon the observed increase in their anisotropic displacement parameters when refined with full occupancy. Conversely, all 17 water molecules in the EtCbl structure were at fully occupied positions. In the two structures, all the water molecules are in hydrogen bonding contact with either oxygen or nitrogen atoms on the cobalamin moiety, or with other water molecules.

Figure 2
Thermal ellipsoid plots of EtCbl (left) and BuCbl (right). The ethyl β-axial ligand in EtCbl is shown in magenta and the butyl ligand (two conformations) in BuCbl is shown in yellow. The rest of the cobalamin complex is colored green (C), red ...

The conformation of the corrin ring is generally quantified by the fold angle around the Co-C10 axis, between the planes of the conjugated ring systems (plane 1: N21, C4, C5, C6, N22, C9, C10) and plane 2: N24, C16, C15, C14, N23, C11, C10; see Figure 1). In EtCbl and BuCbl this corrin fold angle is 14.8° and 13.2° respectively, the direction of the fold being towards the β face of the corrin (away from the α-5,6-dimethylbenzimidazole group). The corresponding values for a number of other alkylcobalamins are given in Table 3. Recently, van der Donk and coworkers reported the X-ray structure of cis-chlorovinylcobalamin and vinylcobalamin.19 Interestingly, cis-chlorovinylcobalamin exhibited a fold angle of 5.7 (9)° which is the smallest reported value for cobalamins (others fall in the 13.3° (AdoCbl) to 18.7° (H2OCbl+) range).19 The authors attributed this result to interactions between the chlorine and the corrin ring and its β-substituents on the tetrapyrrolic macrocycle.19 It was originally proposed that the extent of folding of the corrin ring was inversely related to the steric bulk of the β–axial ligand but it is now generally believed that the trans influence of the ligand is the critical factor.14, 33 There is a rough correlation between the corrin fold angle and the length of the Co-NB3 bond33 (Figure 3). Although there is considerable scattering, the parameters obtained for EtCbl and BuCbl are similar to those reported for AdoCbl structures (Figure 3). The Co-C bond distances in the two complexes are 2.023(2) Å and 2.028(4) Å for EtCbl and BuCbl, respectively. These distances are comparable to other Co-C bond lengths observed in cobalamin complexes (Table 3). In the EtCbl complex, the second carbon atom (C72) is directed away from the two amide groups which project up from the β face of the corrin at carbons C2 and C7, and points almost directly towards the C46 methyl group (attached to carbon C12 of the corrin, see Figure 1). This results in a Co-C71-C72 angle of 117.5°. By calculating a plane through the cobalt atom, C71 and C46, the degree of rotation of the C72 carbon out of this plane can be estimated at approximately 5°. In the BuCbl complex, two orientations of the butyl group were observed in the electron density, as presented in Figure 4. The first and third carbon atoms of both conformations were in identical positions, with the second and fourth carbon atoms occupying two distinct positions. The net effect is a rotation of the butyl group about an axis running between the first and third carbon atoms. The butyl group in both orientations is in a typical elongated conformation, also directed towards the hydrophobic side of the corrin ring similar to the ethyl group (Figure 4). However, neither of the two conformations of the butyl group have the C72 carbon atom in the same position as in the ethyl group with the C72 atoms rotated by approximately 29° and 62°, respectively, from the Co-C71-C46 plane. The Co-C71-C72 angles in the two conformations are 119.0 and 119.6°.

Figure 3
Scatter plot of the corrin fold angle ([var phi]) versus the axial Co-N3B distance for Co-C bonded cobalamins. EtCbl and BuCbl are indicated with circled dots.
Figure 4
Molecular surface representations of BuCbl in two different orientations. The butyl group is directed towards the hydrophobic side of the corrin ring, similar to the ethyl group in EtCbl.
Table 3
Comparison of the Co coordination sphere in EtCbl, BuCbl and other alkylcobalamin complexes. Values are also given for CNCbl.

Analysis of other cobalamin complexes containing a Co-C bond (AdoCbl, vinylCbl, chlorovinylCbl) shows a similar pattern with the upper axial ligand directed towards the hydrophobic side of the β face of the corrin ring, between the C46 methyl and the C54 methyl groups. In vinyl- and chlorovinylcobalamin, the former, possessing a similar steric bulk to that of EtCbl, the second carbon is rotated almost 49° towards the C54 methyl.19 The four independent molecules in the asymmetric unit of the chlorovinylcobalamin complex show a range of values from 29° to 44° from the Co-C71-C46 plane. Adenosylcobalamin shows the largest rotation from this plane of almost 84°, which has the C atoms of the ribose moiety pointed almost directly at the C54 methyl group. Consequently, there appears to be no correlation between the size or the substituents on the second carbon atom and the direction in which the ligand points. Extending the analysis to other cobalamin complexes including those with Co-N, Co-O and Co-S bonds showed that in all cases the substituent is directed toward this side of the corrin ring with a mean rotation from the Co-C71-C46 plane of approximately 44°, which bisects almost exactly the angular spread between the C46 and C54 methyl groups (87°). In fact, the 15° variation observed in the four chlorovinylcobalamin molecules19 would indicate a large energy minimum delineated by the C46 and C54 methyl groups and dominated primarily by the hydrophobicity of this side of the corrin ring and the lack of any ordered hydrogen-bonded water molecules. Importantly, it has been proposed that the methyl groups on C46 and C54, and the methylene groups in side chains a and c protect the reactive methyl or adenosyl groups in the coenzymes during catalytic turnover26 (see also references therein).

A superposition of the structure of EtCbl with that of the naturally occurring MeCbl is shown in Figure 5. There is high structural similarity amongst the structures as expected. The positions of the carbon atoms are visually similar in both complexes. However, a clear distinction of the two alkylcobalamins was possible by direct superposition of the two structures. Although the Co-N3B bond is significantly longer in EtCbl (2.232(1) Å) compared to MeCbl (2.162(4) Å), the orientation of the α-5,6-dimethylbenzimidazole moiety with respect to the corrin ring is the same in both complexes.

Figure 5
Superposition of the structures of MeCbl and EtCbl. Two approximately perpendicular views are shown. The methyl group is shown in orange and the ethyl group in magenta.

There are seven amide side chains incorporated in the corrin ring. Three short acetamide side chains (a, c, and g), which project above the plane of the corrin and four propionamide (b, d, e and f) side chains which extend below the plane of the corrin ring. An analysis of the amide side chains of thirteen cobalamin structures performed by Perry et al. suggested that there is remarkable consistency in the conformation of the side chains amongst cobalamin structures, although side chains c, d, and e display greater flexibility with respect to a, b, g and f side chains.26 However, a more extensive examination of all the cobalamin structures elucidated in the last 20 years reported by Randaccio et al.,14 showed that chains a and d assume very similar conformations in all structures, chain g appears in only two distinct conformations, whereas the other side chains display a large conformational freedom.14 The analysis confirmed that in general, cobalamins belonging to the same cluster type display a similar conformation of their side chains.14, 33 EtCbl and BuCbl belong to cluster type I, which includes CNCbl•KCl, (NH2)2CSCbl, HOCbl, (R)-2,3-dihydroxopropylCbl and (S)-2,3-dihydroxopropylCbl and CNCbl (dry). In agreement with the literature,14 the overall conformation of the side chains in EtCbl and BuCbl does not differ from that observed in other cobalamins in the same cluster, except for (NH2)2CSCbl. Figure 6 shows a superposition of the structures of EtCbl and (NH2)2CSCbl. Side chains b, e, g and f showed similar conformations, whereas side chains a, c and to a lesser extent d, differed in both structures.

Figure 6
Superposition of the structures of EtCbl (white) and (NH2)2CSCbl (cyan). For clarity, only the seven amide side chains (labeled a-g) are shown for the (NH2)2CSCbl structure. The location of the bound ethyl (magenta) and (NH2)2CS (yellow) groups are also ...

The superposition of the structures of MeCbl and EtCbl showed that there is considerable variation in the general orientation of their respective side chains. This could be explained by the presence of van der Waals forces between the longer alkyl β-axial ligands (such as ethyl and butyl groups in EtCbl and BuCbl) with the side chains in the corrin ring, compared to the inability of the methyl group in MeCbl to establish interactions with rather distant side chains projecting above and below the plane of the corrin ring.

It has been established that all alkylcobalamins display the so called “inverse” trans influence,33 where the distances of both axial ligands lengthen or shorten when the electron donating ability of the upper axial ligand (R) increases or decreases, respectively. EtCbl and BuCbl are no exception, as evidenced by comparison with other available structures (Table 3). A linear correlation (r2= 0.919) was obtained by plotting Co-C versus Co-N3B bond distances for a number of alkylcobalamins, as shown in Figure 7. Lengthening of the Co-C bond in alkylcobalamins is thought to be a consequence of the steric repulsion caused by bulky β-axial ligands (such as ethyl and butyl groups), whereas lengthening of the Co-N3B bond appears to be electronic in nature, as a consequence of the strong σ-donation from the alkyl ligands.33

Figure 7
Plot of the axial Co-C against Co-N3B for Co-C bonded cobalamins. EtCbl and BuCbl are indicated with circled dots. The equation of linear regression is: 0.552 + 0.649X, r2 = 0.919, (N= 14).

Summary

The X-ray crystal structures of EtCbl and BuCbl were obtained with some of the lowest crystallographic residuals (R = 0.0468 and 0.0438, respectively), making full use of the high intensity X-rays from a synchrotron source. Both EtCbl and BuCbl exhibited an “inverse” trans influence, as depicted by the lengthening of their Co-C and Co-N3B bonds. Analysis of the molecular surface maps showed that both alkyl ligands in EtCbl and BuCbl project towards the hydrophobic side of the corrin ring. Comparison with other cobalamins showed that in general the β-axial substituent is directed towards the hydrophobic side of the β face of the corrin somewhere between the C46 methyl and the C54 methyl groups. To that end, it has been proposed that the hydrophobic axial groups (methyl groups on C46 and C54 and the methylene groups of the a and c side chains) protect the reactive methyl or adenosyl groups in the coenzymes during catalytic turnover.26 These findings together with the results presented herein strengthen the notion that very fine structural and conformational relationships must be involved in the biological roles of Nature's most beautiful cofactor.45

Supplementary Material

1_si_001

2_si_002

3_si_003

4_si_004

5_si_005

Acknowledgments

This research was funded by the National Heart, Lung and Blood Institute of the National Institutes of Health (HL71907 to DWJ) and Kent State University (NEB). SSRL is a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is supported by the Department of Energy (BES, BER) and by the National Institutes of Health (NCRR, BTP, NIGMS).

References

1. Lenhert PG, Hodgkin DC. Nature. 1961;192:937–8. [PubMed]
2. Rossi M, Glusker JP, Randaccio L, Summers MF, Toscano PJ, Marzilli LG. J Am Chem Soc. 1985;107:1729–1738.
3. Banerjee R, Chowdhury S. Chemistry and Biochemistry of B12. Wiley-IEEE; New York: 1999. Methylmalonyl-CoA mutase; p. 921.
4. Matthews RG. Chemistry and Biochemistry of B12. Wiley-IEEE; New York: 1999. Cobalamin-dependent methionine synthase; p. 921.
5. Banerjee RV, Matthews RG. Faseb J. 1990;4:1450–9. [PubMed]
6. Ludwig ML, Matthews RG. Annu Rev Biochem. 1997;66:269–313. [PubMed]
7. Banerjee R, Ragsdale SW. Annu Rev Biochem. 2003;72:209–47. [PubMed]
8. Burke GT, Mangum JH, Brodie JD. Biochemistry. 1971;10:3079–85. [PubMed]
9. Taylor RT, Weissbach H. J Biol Chem. 1967;242:1509–16. [PubMed]
10. Taylor RT, Whitfield C, Weissbach H. Arch Biochem Biophys. 1968;125:240–52. [PubMed]
11. Yamada K, Yamada S, Tobimatsu T, Toraya T. J Biol Chem. 1999;274:35571–6. [PubMed]
12. Banerjee R. Chem Biol. 1997;4:175–86. [PubMed]
13. Ouyang L, Rulis P, Ching WY, Nardin G, Randaccio L. Inorg Chem. 2004;43:1235–41. [PubMed]
14. Randaccio L, Furlan M, Geremia S, Slouf M, Srnova I, Toffoli D. Inorg Chem. 2000;39:3403–13. [PubMed]
15. Birke RL, Huang Q, Spataru T, Gosser DK., Jr J Am Chem Soc. 2006;128:1922–36. [PubMed]
16. Rovira C, Kozlowski PM. J Phys Chem B. 2007;111:3251–7. [PubMed]
17. Rovira C, Biarnes X, Kunc K. Inorg Chem. 2004;43:6628–32. [PubMed]
18. Hatfield K, Deng B, Burris DR, Buck LE, Delcomyn CA. Environ Toxicol Chem. 1998;17:1681.
19. McCauley KM, Pratt DA, Wilson SR, Shey J, Burkey TJ, van der Donk WA. J Am Chem Soc. 2005;127:1126–36. [PubMed]
20. Pratt DA, van der Donk WA. J Am Chem Soc. 2005;127:384–96. [PubMed]
21. Magara Y, Furuichi T. Dev Toxicol Environ Sci. 1986;12:231–43. [PubMed]
22. McLaughlin JK, Blot WJ. Int Arch Occup Environ Health. 1997;70:222–31. [PubMed]
23. Squillace PJ, Moran MJ, Lapham WW, Price CV, Clawges RM, Zogorski JS. Environ Sci Technol. 1999;33:4176–4187.
24. Holliger C, Hahn D, Harmsen H, Ludwig W, Schumacher W, Tindall B, Vazquez F, Weiss N, Zehnder AJ. Arch Microbiol. 1998;169:313–21. [PubMed]
25. Holliger C, Wohlfarth G, Diekert G. FEMS Microbiol Rev. 1998;22:383–397.
26. Perry CB, Marques HM. South Afr J Sci. 2004;100:368–380.
27. Brasch NE, Finke RG. J Inorg Biochem. 1999;73:215–9. [PubMed]
28. Kim SH, Chen HL, Feilchenfeld N, Halpern J. J Am Chem Soc. 1988;110:3120–3126.
29. Pratt JM. Inorganic chemistry of vitamin B12. Academic Press Inc; New York: 1972. p. 347.
30. Barker HA, Smyth RD, Weissbach H, Toohey JI, Ladd JN, Volcani BE. J Biol Chem. 1960;235:480–8. [PubMed]
31. Kabsch W. J Appl Cryst. 1993;2:795–800.
32. Sheldrick GM, Schneider TR. SHELXL: High Resolution Refinement. In: Sweet RM, Carter CWJ, editors. Method Enzymol. Vol. 277. Academic Press; Orlando, Florida: 1997. pp. 319–343. [PubMed]
33. Randaccio L, Geremia S, Nardin G, Wuerges J. Coord Chem Rev. 2006;250:1332–1350.
34. Ludwig ML, Evans PR. Chemistry and Biochemistry of B12. Wiley-IEEE; New York: 1999. X-ray crystallography of B12 enzymes: methylmalonyl-CoA mutase and methionine synthase; p. 921.
35. Hannibal L, Bunge SD, van Eldik R, Jacobsen DW, Kratky C, Gruber K, Brasch NE. Inorg Chem. 2007;46:3613–8. [PMC free article] [PubMed]
36. Garau G, Fedosov SN, Petersen TE, Geremia S, Randaccio L. Acta Crystallogr D Biol Crystallogr. 2001;57:1890–2. [PubMed]
37. Wuerges J, Garau G, Geremia S, Fedosov SN, Petersen TE, Randaccio L. Proc Natl Acad Sci U S A. 2006;103:4386–91. [PubMed]
38. Cole AG, Yoder LM, Shiang JJ, Anderson NA, Walker LA, 2nd, Banaszak Holl MM, Sension RJ. J Am Chem Soc. 2002;124:434–41. [PubMed]
39. Taylor RT, Smucker L, Hanna ML, Gill J. Arch Biochem Biophys. 1973;156:521–33. [PubMed]
40. Smith EL, Mervyn L, Muggleton PW, Johnson AW, Shaw N. Ann N Y Acad Sci. 1964;112:565–74. [PubMed]
41. Brown KL, Cheng S, Zou X, Zubkowski JD, Valente EJ, Knapton L, Marques HM. Inorg Chem. 1997;36:3666–3675. [PubMed]
42. Knapton L, Marques HM. Dalton Trans. 2005:889–95. [PubMed]
43. Marques HM, Knapton L, Zou X, Brown KL. J Chem Soc Dalton Trans. 2002:3195–3200.
44. Garau G, Geremia S, Marzilli LG, Nardin G, Randaccio L, Tauzher G. Acta Crystallogr B. 2003;59:51–9. [PubMed]
45. Stubbe J. Science. 1994;266:1663–4. [PubMed]
46. Lenhert PG. Proc R Soc London, Ser A. 1968;A303:45–84.
47. Pagano TG, Marzilli LG, Flocco MM, Tsai C, Carrell HL, Glusker JP. J Am Chem Soc. 1991;113:531–542.
48. Wagner T, Afshar CE, Carrell HL, Glusker JP, Englert U, Hogenkamp HP. Inorg Chem. 1999;38:1785–1794. [PubMed]
49. Zou X, Brown KL. Inorg Chim Acta. 1998;267:305–308.
50. Perry CB, Fernandes MA, Marques HM. Acta Crystallogr C. 2004;60:m165–7. [PubMed]
51. Alcock NW, Dixon RM, Golding BT. J Chem Soc Chem Commun. 1985:603–605.