Methyl β-allolactoside [methyl β-D-galactopyranosyl-(1→6)-β-D-glucopyran­oside] monohydrate

Methyl β-allolactoside [methyl β-D-galactopyranosyl-(1→6)-β-D-glucopyran­oside], (II), was crystallized from water as a monohydrate, C₁₃H₂₄O₁₁·H₂O. The βGalp and βGlcp residues in (II) assume distorted ⁴C₁ chair conformations, with the former more distorted than the latter. Linkage conformation is characterized by φ′ (C2_Gal—C1_Gal—O1_Gal—C6_Glc), ψ′ (C1_Gal—O1_Gal—C6_Glc—C5_Glc) and ω (C4_Glc—C5_Glc—C6_Glc—O1_Gal) torsion angles of 172.9 (2), −117.9 (3) and −176.2 (2)°, respectively. The ψ′ and ω values differ significantly from those found in the crystal structure of β-gentiobiose, (III) [Rohrer et al. (1980). Acta Cryst. B36, 650–654]. Structural comparisons of (II) with related disaccharides bound to a mutant β-galactosidase reveal significant differences in hydroxy­methyl conformation and in the degree of ring distortion of the βGlcp residue. Structural comparisons of (II) with a DFT-optimized structure, (II_C), suggest a link between hydrogen bonding, pyranosyl ring deformation and linkage conformation.

S1. Comment

The disaccharide β-D-galactopyranosyl-(1→6)-D-glucopyranose, (I) (allolactose), is widely known as an inducer of the lac operon. Allolactose serves as a negative allosteric effector of the lac repressor (LacI) (Jacob & Monod, 1961; Burstein et al., 1965; Jobe & Bourgeois, 1972; Yildirim & Mackey, 2003). In E. coli, lactose [β-D-galactopyranosyl-(1→4)-D-glucopyranose] is converted to (I) via transglycosylation catalyzed by β-galactosidase (Jacob & Monod, 1961; Burstein et al., 1965; Jobe & Bourgeois, 1972). The crystal structure of (I) has not been reported, but it appears in 1.5 Å resolution cocrystals with a mutant β-galactosidase (Juers et al., 2001). In addition to its biological role as a free disaccharide, the β-D-Galp-(1→6)-β-D-Glcp glycosidic linkage is observed in various glycoconjugates, including a ceramide trisaccharide, β-D-Galp(1→6)-β-D-Galp-(1→6)-β-D– Glcp-(1→1)-Cer, isolated from sea urchin eggs (Kubo et al., 1992), and in the soluble protein bovine lactoferrin (Mir et al., 2009).

As structural information on the β-Gal-(1→6)-β-Glc linkage in differing biological contexts grows, the geometry of the isolated disaccharide, to which comparisons with more complex structures can be made, increases in importance. We report here the crystal structure of the methyl glycoside of (I), methyl β-D-galactopyranosyl-(1→6)-β-D-glucopyranoside (methyl β-allolactoside), (II), monohydrate (Fig. 1). Atom numbering is shown in the structure of (II), with primed and unprimed numbers assigned to the βGalp and βGlcp residue atoms, respectively. Structural parameters found for (II) were compared with those observed in β-gentiobiose [β-D-glucopyranosyl-(1→6)-β-D-glucopyranose], (III) (Arène et al., 1979; Rohrer et al., 1980), which is the only structurally related (1→6)-linked disaccharide whose crystal structure is known. Comparisons were also made with structures of (I_β) bound to β-galactosidase (Juers et al., 2001), denoted (I_G1), (I_G2) and (I_G3), and with methyl β-D-galactopyranosyl-(1→4)-β-D-glucopyranoside (methyl β-lactoside), (IV) (Stenutz et al., 1999) (Table 2). The influence of crystal packing on structural parameters was also investigated in an unconstrained in vacuo density functional theory (DFT) geometry optimization using Gaussian03 (Frisch et al., 2004). The crystal structure of (II) was used as the starting geometry and the calculation was performed with the B3LYP functional (Becke, 1993), and the 6–31G* basis set (Hehre et al., 1972). Structural data for the DFT structure, denoted (II_C), are shown in Table 2.

Data in Table 2 for (II)–(IV) yield the following average C—C bond lengths: C1—C2, 1.522 (6) Å; remaining endocyclic C—C, 1.526 (7) Å; C5—C6, 1.511 (3) Å. Exocyclic C5—C6 bonds appear shorter than all endocyclic C—C bonds (Pan et al., 2005), while C1—C2 bond lengths do not differ statistically from the other endocyclic C—C bond lengths.

Average C—O bond lengths in (II)–(IV) are as follows: endocyclic C—O, 1.425 (9) Å; anomeric C—O (exocyclic), 1.392 (6) Å; exocyclic C—O, 1.426 (8) Å; exocyclic C—O involving the anomeric oxygen, 1.434 (8) Å. The equatorial C1—O1 and C1'—O1' bonds are shorter (by ~0.03 Å) than the remaining exocyclic equatorial C—O bonds (Berman et al., 1967), due to optimal anomeric effects (Lemieux, 1971) in these structures; in (II), the C2'—C1'—O1'—C6 and C2—C1—O1—C7 torsion angles are 172.9 (2)° and 177.4 (3)°, respectively.

In the DFT-calculated structure (II_C), trends in C—C bond lengths mimic the experimental observations, but calculated C—C bond lengths are longer on average than experimental values by ~0.01 Å: C1—C2, 1.531 (3) Å; remaining endocyclic C—C, 1.532 (8) Å; C5—C6, 1.525 (3) Å. In contrast, DFT-calculated C—O bond lengths reproduce the experimental data well, both in terms of trends and absolute values: endocyclic C—O, 1.426 (6) Å; anomeric C—O (exocyclic), 1.393 (6) Å; exocyclic C—O, 1.421 (6) Å; exocyclic C—O involving the anomeric oxygen, 1.430 Å.

C4'—O4' bond lengths in (II) and (III) are identical despite differences in C4 configuration (Table 2). C4'—O4' bond conformation in (II) and (III) cannot be responsible for this finding, since both bonds are in conformations in which the C4' and O4' H atoms are eclipsed. Intermolecular hydrogen bonding in the C4'—O4' fragment of (II) and (III) is also identical. In contrast, r_C4,O4 is considerably shorter in (II_C) than in (II) and (III), presumably due, at least partly, to relief of the eclipsing interaction in the calculated structure [the equivalent C4' and O4' H atoms are approximately gauche in (II_C)]. These findings suggest that the relative lengths of axial and equatorial C—O bonds in saccharides can be influenced significantly by crystal-packing forces, and that eclipsing interactions present in crystal structures caused by the intermolecular hydrogen-bonding lattice may suppress competing intramolecular forces that affect exocyclic C—O bond lengths (e.g. bond orientation). The internal glycosidic C—O—C bond angle in (II) and (III) appears slightly smaller than observed in (IV), possibly due to the reduced steric demands of the 1,6-linkage. This conclusion is supported by glycosidic C—O—C bond angles involving the methyl aglycones in (II) and (III), which compare favorably to the internal glycoside C—O—C angles in the same structures.

The βGlcp and βGalp rings of (II)–(IV) assume slightly distorted ⁴C₁ chair conformations based on Cremer–Pople puckering parameters (Cremer & Pople, 1975) (Table 3) (q₃ >> q₂). In (II), the βGlcp ring is closer to an ideal ⁴C₁ chair form (θ = 2.3°) than the βGalp ring (θ = 8.8°), whereas the opposite is found for (IV). In (III), both rings have more comparable θ values and thus comparable degrees of distortion. The direction of distortion, embodied in ϕ, is context dependent and can be easily visualized using the projection convention of Jeffrey & Yates (1979) (Fig. 2). Overall, global ring shapes for the βGalp ring of (IV) and the βGlcp rings of (III) [ϕ = 28 (6)°] are slightly distorted near ⁰H₁; these conformations differ from the βGalp ring of (II) and the βGlcp rings of (II) and (IV) which are distorted near E₅ and ⁰E/⁰H₅, respectively.

Exocyclic hydroxymethyl (–CH₂OH) conformations in (II)–(IV), denoted by torsion angle ω, differ. In (II), the gt conformation is found in both residues, whereas in (III), gg conformations are observed (Table 2), resulting in significantly different surface topologies for the two disaccharides. In contrast, mixed conformations are found in (IV): gg in βGlcp and gt in βGalp. The gt conformation is highly favored in methyl β-D-galactopyranoside in aqueous solution, based on NMR scalar coupling analysis, whereas a roughly equal mixture of gg and gt forms is observed in methyl β-D-glucopyranoside (Thibaudeau et al., 2004), results consistent with the statistical distribution of rotamers observed in (II)–(IV).

Glycosidic linkage conformation in (II) is determined by torsion angles ϕ' [C2'—C1'—O1'—C6; 172.9 (2)°], ψ' [C1'—O1'—C6—C5; -117.9 (3)°] and ω [O1'—C6—C5—O5; 63.8 (3)°]. These torsions contrast with corresponding values of -176.4, -156.3 and -61.6° observed in (III). The ϕ' values differ by ~ 11°, whereas the ψ' values differ by ~38°. Presumably, different conformations about ω and ω' in (II) and (III) influence ψ', which is controlled mainly by steric factors rather than the stereoelectronic [exoanomeric (Lemieux, 1971)] factors that largely influence ϕ'. It is noteworthy that ϕ' values in (II)–(III) are larger than the corresponding value in (IV) (153.8°), which suggests reduced steric constraints for the linkage in the former and leading to nearly ideal, i.e. perfectly staggered, exoanomeric conformations for both C1'—O1' bonds.

Linkage conformation in the DFT-calculated structure (II_C) is similar to that in (II); the glycosidic torsion angles ϕ' and ψ' differ by ~12° (Table 2). While Cremer–Pople θ values differ by < 2.4°, the types of ring distortions in (II) and (II_C) differ, especially for the βGalp ring, where ϕ values differ by ~90° (Table 3). The discrepancy may be linked to a change in hydrogen bonding involving the C3' and C4' hydroxyls from the intermolecular arrangement in (II) (see below) to an intramolecular hydrogen bond (O4'···O3') in (II_C) [r_O3',O4' in (II_C) = 2.693 Å]. Ring-puckering amplitudes, Q, decrease by 3.2% and 3.5% for the βGalp and βGlcp residues, respectively, in (II_C) relative to (II).

Crystal structures of mutant β-galactosidase complexed with (I_β) (Juers et al., 2001) reveal the degree to which the βGal(1→6)βGlc structure is deformed upon protein binding (Table 4). The glycosidic torsions ϕ' are very similar for (II) and (I_G1—G3) (C2'—C1'—O1'—C6 torsions of 170.0–174.0°), but ψ' values differ considerably (C1'—O1'—C6—C5 torsions of -117.9–168.3°). Cremer–Pople θ values are similar for the βGalp residue in the free and protein-bound states (Table 3). However, considerable βGlcp ring distortion is observed in the bound state, as indicated by the significantly enhanced θ values (Table 3), and the direction of distortion varies widely in the bound geometries. Equally important, exocyclic hydroxymethyl conformation is affected by binding; bound βGalp residues assume the tg conformation, whereas gg or gt is adopted in the βGlcp residues (Table 4). In solution, βGalp residues prefer gt and tg conformations, with gg virtually absent (Thibaudeau et al., 2004). The statistical findings for ω' in bound conformations of (I_β) are thus consistent with solution behavior. Similar arguments apply to the βGlcp residue.

Ring pucker can exert an important effect on glycosidic linkage conformation; deformation of one or both pyranose rings in a β-(1→6) linkage may allow conformations that are not normally accessible. For example, inspections of (I_G1—G3) and (III) reveal a qualitative correlation between departure from an ideal ⁴C₁conformation and linkage conformation (Fig. 2). This connectivity suggests that, in solution, a correlation between pyranose ring-puckering dynamics and linkage reorientation may exist. Related behavior has been observed previously in nucleosides (Foloppe & Nilsson, 2005). While furanosyl rings are commonly viewed as more conformationally flexible than pyranosyl rings (Altona & Sundaralingam, 1972, 1973; French & Finch, 1999), pyranose pseudorotation may become more energetically favorable by extrinsic macromolecular forces. This expectation appears to be borne out in the structures of (I_β) bound to mutant β-galactosidase.

The crystal structure of (II) contains no intramolecular hydrogen bonds. Six of the seven hydroxyls (O3, O4, O2', O3', O4' and O6') participate as donors in strong intermolecular hydrogen bonds, while the remaining hydroxyl, O2, has longer contacts and is bifurcated between O2' and O3' (Table 1, Fig. 3). Of the seven exocyclic hydroxyl O atoms present in (II), five (O2, O3, O3', O4', O6') participate in intermolecular hydrogen bonding as monoacceptors. Hydroxyl O2' accepts hydrogen bonds from two donors; one is a weaker bifurcated hydrogen bond described above and the other a more regular hydrogen bond from O6'. The remaining exocyclic hydroxyl, O4, does not receive hydrogen bonds. The two endocyclic O atoms (O5, O5') and the internal glycosidic O atom (O1') are not hydrogen bonded. The remaining terminal glycosidic O atom (O1) serves as a monoacceptor. The water of crystallization (O1W) participates as a node, accepting and donating two hydrgoen bonds.

An intermolecular hydrophobic contact in the crystal is observed between the methyl aglycone and the βGlcp H6R and H6S H atoms adjacent to the glycoside from a neighboring unit cell, with methyl hydrogen to hydroxymethyl hydrogen internuclear distances of 2.602 and 2.522 Å, respectively. The third methyl proton has a weaker intramolecular contact with H1' (3.014 Å).

Related literature

For related literature, see: Altona & Sundaralingam (1972, 1973); Angyal et al. (1979); Austin et al. (1963); Becke (1993); Berman et al. (1967); Burstein et al. (1965); Conchie & Levvy (1963); Cremer & Pople (1975); Foloppe & Nilsson (2005); French & Finch (1999); Frisch et al. (2004); Gravatt & Gross (1969); Hehre et al. (1972); Hodge & Hofreiter (1962); Hope (1971); Jacob & Monod (1961); Jeffrey & Yates (1979); Jobe & Bourgeois (1972); Juers et al. (2001); Kubo et al. (1992); Lemieux (1971); Nilsson (1987, 1988); Pan et al. (2005); Rohrer et al. (1980); Serianni et al. (1979, 1990); Stenutz et al. (1999); Thibaudeau et al. (2004); Yildirim & Mackey (2003).

S2. Experimental

The reagents o-nitrophenol, Sepharose G10, Dowex 1 × 2 (200–400 mesh) (Cl^-) ion-exchange resin, β-galactosidase (E. C. 3.2.1.23) (E. coli) and methyl β-D-glucopyranoside were purchased from Sigma.

For the synthesis of o-nitrophenyl β-D-[1–¹³C]galactopyranoside, D-[1-¹³C]galactose was prepared from D-lyxose and K¹³CN (Cambridge Isotope Laboratories; 99 atom-% ¹³C) by cyanohydrin reduction, yielding D-[1-¹³C]galactose and D-[1-¹³C]talose (Serianni et al., 1979, 1990). The galacto and talo epimers were separated by chromatography on Dowex 50 × 8 (200–400 mesh) (Ca²⁺) (Angyal et al., 1979), with the galacto isomer eluting first. o-Nitrophenyl β-D-[1-¹³C]galactopyranoside was prepared from D-[1-¹³C]galactose in an overall yield of 30% (Conchie & Levvy, 1963).

For the synthesis of (II) by enzyme-catalyzed transglycosylation, the conditions for the transglycosylation reaction were identical to those reported by Nilsson (Nilsson, 1987, 1988); the reaction was conducted with 1.35 g (4.48 mmol) of the acceptor (methyl β-D-glucopyranoside) and 2.50 g (12.9 mmol) of the donor (o-nitrophenyl β-D-[1-¹³C]galactopyranoside). After the reaction was quenched, the mixture was concentrated in vacuo at 313 K to ~45 ml and applied to a column (2.5 × 55 cm) of Sepharose G10. Elution with distilled water gave a phenol–sulfuric acid (Hodge & Hofreiter, 1962) positive peak near the column void volume. This disaccharide-containing fraction was collected, concentrated to ~35 ml, and the solution was applied to a column (2.5 × 55 cm) of Dowex 1 × 2 (200–400 mesh) (OH^-) ion-exchange resin (Austin et al., 1963). Elution with distilled water (3.8 ml fraction^-1, 0.8 ml min^-1) gave three phenol–sulfuric acid positive peaks. NMR analysis of each peak indicated the following: peak 1, residual acceptor; peak 2, (II); peak 3, methyl β-D-[1-¹³C]galactopyranosyl-(1→3)-β-D-glucopyranoside (byproduct). Fractions containing (II) were pooled and the volume reduced to give a clear colorless viscous syrup for characterization by ¹H and ¹³C NMR.

Crystals of (II) suitable for X-ray diffraction were grown using a λ tube (Gravatt & Gross, 1969) and a modification of published protocols (Hope, 1971). Approximately 20 mg of (II) were dissolved in 300 ml of distilled water and the solution was placed in the sample reservoir of a λ tube. A Nichrome wire was wrapped around the arm closer to the sample reservoir for heating purposes, and a length of copper wire was closely wound around the cross arm to act as cooling ribs. The apparatus was gently filled with a 1:3:6 (v/v/v) mixture of n-butanol:methanol:ethanol to a level above the cross arm, the open ends were stoppered, and a low electric current was applied across the Nichrome wire to heat the ascending arm to ~323 K. After 10 d, a tiny needle-shaped crystal formed in the cooler descending arm of the λ tube. This crystal was carefully removed, placed in a capped vial with mother liquor that had been significantly reduced in volume, and the vessel was set at room temperature. The vial was not tightly capped to allow slow evaporation of solvent. After ~18 months, colorless needle-shaped crystals formed, which were harvested for structure determination.

Crystallographic data were collected through the SCrALS (Service Crystallography at Advanced Light Source) program at the Small-Crystal Crystallography Beamline 11.3.1 (developed by the Experimental Systems Group) at the Advanced Light Source (ALS).

S3. Refinement

Hydroxyl and water H atoms were located from a difference Fourier map and included in their observed positions. Their positions were subsequently tied to the O atom to which they are bonded. The H atoms on the water of crystallization were restrained to have O—H distances of 0.84 Å. Hydrogen thermal parameters were tied to that of the atom to which they are bonded.

The assignment of the correct absolute configuration of the molecule and enantiomorph of the space group were determined by comparison with the same compound that had been previously characterized with Cu Kα radiation. That previous analysis had determined the correct configuration by comparison of intensities of Friedel pairs of reflections and by the known configuration of the allolactoside.

Computing details

Data collection: APEX2 (Bruker, 2007); cell refinement: APEX2 and SAINT (Bruker, 2007); data reduction: SAINT (Bruker, 2007) and XPREP (Sheldrick, 2008); program(s) used to solve structure: XS (Sheldrick, 2008); program(s) used to refine structure: XL (Sheldrick, 2008); molecular graphics: XP (Sheldrick, 2008), POV-RAY (Cason, 2003) and DIAMOND (Brandenburg, 2009); software used to prepare material for publication: XCIF (Sheldrick, 2008), enCIFer (CCDC, 2005), publCIF (Westrip, 2009).

Figures

	Fig. 1. Crystal structure and labeling scheme of (II). Thermal displacement ellipsoids are depicted at 50% probability.
	Fig. 2. Stereographic projection of the pyranose pseudorotational itinerary (northern hemisphere only) with ring-puckering coordinates for (I)–(V). Ring-puckering parameters ϕ and θ are represented by angular and radial displacements about the ^4C1 origin (center point), respectively, and displacements are based on individual total puckering amplitudes, Q. In the inset, displacements are based on an average amplitude for the depicted data (Qavg = 0.5734). Inner and outer black rings represent the minimal and maximal Q values for the total data set; colored rings correspond to Q values for the Gal and Glc rings of (II) and (IIC) as per the color code shown below the graphic. The southern hemisphere (not shown) relates via symmetry to the northern with ^1C4 at the origin. ^aRing at the non-reducing end of (III) is Glc not Gal.
	Fig. 3. Hydrogen-bonding scheme viewed along the a axis.

Methyl β-allolactoside monohydrate

Crystal data

C13H24O11·H2O	F(000) = 400
Mr = 374.34	Dx = 1.511 Mg m^−3
Monoclinic, P21	Synchrotron radiation, λ = 0.77490 Å
Hall symbol: P 2yb	Cell parameters from 2490 reflections
a = 7.528 (2) Å	θ = 3.0–29.5°
b = 8.744 (4) Å	µ = 0.14 mm^−1
c = 12.695 (11) Å	T = 150 K
β = 100.15 (4)°	Blade, clear colorless
V = 822.6 (8) Å^3	0.12 × 0.06 × 0.01 mm
Z = 2

Data collection

Bruker d8 APEXII CCD diffractometer	2192 independent reflections
Radiation source: synchrotron, Advanced Light Source beamline 11.3.1 (SCrALS)	1920 reflections with I > 2σ(I)
Si<111> channel cut crystal monochromator	Rint = 0.087
Detector resolution: 83.33 pixels mm^-1	θmax = 31.4°, θmin = 3.0°
combination of ω and ϕ–scans	h = −10→10
Absorption correction: multi-scan (SADABS; Sheldrick, 2007)	k = −11→11
Tmin = 0.984, Tmax = 0.999	l = −16→16
10449 measured reflections

Refinement

Refinement on F^2	Primary atom site location: structure-invariant direct methods
Least-squares matrix: full	Secondary atom site location: difference Fourier map
R[F^2 > 2σ(F^2)] = 0.049	Hydrogen site location: inferred from neighbouring sites
wR(F^2) = 0.121	H atoms treated by a mixture of independent and constrained refinement
S = 1.05	w = 1/[σ^2(Fo^2) + (0.0592P)^2] where P = (Fo^2 + 2Fc^2)/3
2192 reflections	(Δ/σ)max = 0.021
239 parameters	Δρmax = 0.41 e Å^−3
3 restraints	Δρmin = −0.39 e Å^−3

Special details

Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes.

Refinement. Refinement of F² against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F², conventional R-factors R are based on F, with F set to zero for negative F². The threshold expression of F² > σ(F²) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F² are statistically about twice as large as those based on F, and R- factors based on ALL data will be even larger.

Hydroxyl and water H atoms were located from a difference Fourier map and included in their observed positions. Their positions were subsequently tied to the oxygen to which they are bonded. The H atoms on the water of crystallization were restrained to have O—H distances of 0.84 Å. Hydrogen thermal parameters were tied to that of the atom to which they are bonded.

The assignment of the correct absolute configuration of the molecule and enantiomorph of the space group were determined by comparison with the same compound that had been previously characterized with Cu—Kα radiation. That previous analysis had determined the correct configuration by comparison of intensities of Friedel pairs of reflections and by the known configuration of the allolactoside.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
O1	1.0049 (3)	0.6971 (3)	0.60348 (18)	0.0264 (5)
O2	0.9786 (3)	0.6609 (3)	0.82792 (17)	0.0223 (5)
H2	1.0217	0.7497	0.8304	0.027*
O3	0.6245 (3)	0.5785 (2)	0.86435 (17)	0.0217 (5)
H3	0.6470	0.6713	0.8776	0.026*
O4	0.4249 (3)	0.3872 (3)	0.69935 (17)	0.0217 (5)
H4	0.3687	0.4147	0.7477	0.026*
O5	0.7392 (3)	0.5748 (2)	0.55624 (16)	0.0181 (4)
C1	0.8897 (4)	0.5910 (3)	0.6388 (2)	0.0195 (6)
H1A	0.9526	0.4905	0.6534	0.023*
C2	0.8297 (4)	0.6500 (4)	0.7400 (2)	0.0192 (6)
H2A	0.7707	0.7523	0.7260	0.023*
C3	0.6944 (4)	0.5347 (3)	0.7709 (2)	0.0178 (6)
H3A	0.7590	0.4352	0.7868	0.021*
C4	0.5382 (4)	0.5081 (3)	0.6767 (2)	0.0167 (5)
H4A	0.4663	0.6040	0.6604	0.020*
C5	0.6180 (4)	0.4594 (3)	0.5798 (2)	0.0163 (6)
H5A	0.6855	0.3613	0.5963	0.020*
C6	0.4791 (4)	0.4400 (4)	0.4788 (2)	0.0196 (6)
H6A	0.3850	0.3668	0.4919	0.024*
H6B	0.4202	0.5394	0.4580	0.024*
C7	1.0724 (6)	0.6444 (5)	0.5105 (3)	0.0427 (11)
H7A	1.1532	0.7217	0.4890	0.064*
H7B	0.9711	0.6269	0.4518	0.064*
H7C	1.1389	0.5486	0.5276	0.064*
O1'	0.5658 (3)	0.3842 (2)	0.39385 (15)	0.0179 (4)
O2'	0.8582 (3)	0.4358 (3)	0.27454 (18)	0.0225 (5)
H2'	0.8873	0.3750	0.3259	0.027*
O3'	0.7358 (3)	0.4179 (3)	0.05500 (17)	0.0263 (5)
H3'	0.7020	0.4468	−0.0085	0.032*
O4'	0.3622 (3)	0.3618 (2)	0.05100 (17)	0.0250 (5)
H4'	0.3081	0.3613	−0.0128	0.030*
O5'	0.3694 (3)	0.4879 (2)	0.25623 (16)	0.0174 (4)
O6'	0.0581 (3)	0.6722 (3)	0.20671 (19)	0.0253 (5)
H6'	0.0010	0.6053	0.2345	0.030*
C1'	0.5544 (4)	0.4835 (3)	0.3064 (2)	0.0154 (5)
H1'A	0.5971	0.5880	0.3309	0.018*
C2'	0.6721 (4)	0.4174 (3)	0.2311 (2)	0.0172 (6)
H2'A	0.6449	0.3059	0.2202	0.021*
C3'	0.6379 (4)	0.4980 (3)	0.1228 (2)	0.0187 (6)
H3'A	0.6884	0.6038	0.1330	0.022*
C4'	0.4353 (4)	0.5110 (3)	0.0778 (2)	0.0186 (6)
H4'A	0.4170	0.5766	0.0123	0.022*
C5'	0.3421 (4)	0.5839 (3)	0.1624 (2)	0.0173 (6)
H5'A	0.3973	0.6864	0.1818	0.021*
C6'	0.1416 (4)	0.6035 (4)	0.1250 (2)	0.0230 (6)
H6'A	0.1205	0.6687	0.0602	0.028*
H6'B	0.0860	0.5024	0.1057	0.028*
O1W	0.8132 (3)	0.9006 (3)	0.15167 (18)	0.0223 (5)
H1WA	0.861 (5)	0.986 (3)	0.154 (3)	0.027*
H1WB	0.897 (4)	0.836 (4)	0.165 (3)	0.027*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
O1	0.0262 (12)	0.0294 (13)	0.0257 (11)	−0.0113 (10)	0.0108 (9)	−0.0085 (10)
O2	0.0263 (11)	0.0174 (11)	0.0220 (11)	−0.0041 (9)	0.0009 (9)	−0.0036 (9)
O3	0.0323 (12)	0.0165 (11)	0.0177 (10)	−0.0014 (9)	0.0084 (9)	−0.0010 (8)
O4	0.0268 (11)	0.0189 (10)	0.0215 (11)	−0.0069 (9)	0.0101 (8)	−0.0014 (9)
O5	0.0188 (10)	0.0200 (10)	0.0155 (10)	−0.0055 (8)	0.0030 (8)	0.0017 (8)
C1	0.0196 (14)	0.0192 (15)	0.0205 (14)	−0.0015 (12)	0.0058 (11)	−0.0031 (11)
C2	0.0195 (14)	0.0193 (14)	0.0189 (14)	−0.0013 (11)	0.0035 (11)	−0.0029 (11)
C3	0.0242 (14)	0.0159 (14)	0.0147 (13)	−0.0002 (10)	0.0070 (11)	0.0007 (10)
C4	0.0183 (13)	0.0152 (13)	0.0169 (13)	−0.0003 (10)	0.0036 (10)	0.0010 (10)
C5	0.0177 (13)	0.0129 (13)	0.0187 (13)	0.0004 (10)	0.0047 (11)	−0.0001 (10)
C6	0.0218 (15)	0.0225 (15)	0.0153 (14)	0.0012 (11)	0.0056 (11)	−0.0005 (11)
C7	0.044 (2)	0.053 (3)	0.038 (2)	−0.023 (2)	0.0266 (18)	−0.0217 (19)
O1'	0.0222 (10)	0.0173 (10)	0.0147 (9)	0.0016 (8)	0.0044 (8)	0.0009 (8)
O2'	0.0172 (10)	0.0245 (12)	0.0258 (11)	0.0035 (8)	0.0035 (8)	0.0108 (9)
O3'	0.0317 (12)	0.0304 (13)	0.0188 (10)	0.0105 (10)	0.0100 (9)	0.0036 (9)
O4'	0.0367 (13)	0.0149 (11)	0.0211 (11)	−0.0050 (9)	−0.0015 (9)	−0.0013 (8)
O5'	0.0150 (9)	0.0183 (10)	0.0183 (10)	−0.0014 (8)	0.0012 (8)	0.0021 (8)
O6'	0.0210 (11)	0.0230 (12)	0.0330 (12)	0.0015 (9)	0.0074 (9)	0.0045 (10)
C1'	0.0172 (13)	0.0134 (12)	0.0149 (12)	−0.0002 (10)	0.0011 (10)	0.0018 (10)
C2'	0.0179 (13)	0.0152 (14)	0.0188 (13)	0.0018 (10)	0.0041 (10)	0.0014 (10)
C3'	0.0219 (14)	0.0167 (14)	0.0189 (14)	0.0024 (11)	0.0076 (11)	0.0037 (11)
C4'	0.0262 (15)	0.0132 (13)	0.0159 (13)	−0.0036 (11)	0.0023 (11)	0.0008 (10)
C5'	0.0196 (13)	0.0149 (13)	0.0166 (13)	−0.0010 (11)	0.0008 (10)	0.0033 (10)
C6'	0.0197 (14)	0.0248 (16)	0.0237 (16)	0.0006 (12)	0.0012 (12)	0.0030 (12)
O1W	0.0206 (11)	0.0181 (11)	0.0270 (11)	0.0013 (9)	0.0009 (8)	0.0017 (9)

Geometric parameters (Å, º)

O1—C1	1.397 (4)	O1'—C1'	1.400 (3)
O1—C7	1.440 (4)	O2'—C2'	1.422 (4)
O2—C2	1.439 (4)	O2'—H2'	0.8400
O2—H2	0.8400	O3'—C3'	1.413 (3)
O3—C3	1.432 (3)	O3'—H3'	0.8400
O3—H3	0.8400	O4'—C4'	1.434 (4)
O4—C4	1.419 (3)	O4'—H4'	0.8400
O4—H4	0.8400	O5'—C1'	1.426 (3)
O5—C1	1.408 (3)	O5'—C5'	1.442 (3)
O5—C5	1.427 (3)	O6'—C6'	1.437 (4)
C1—C2	1.525 (4)	O6'—H6'	0.8400
C1—H1A	1.0000	C1'—C2'	1.528 (4)
C2—C3	1.532 (4)	C1'—H1'A	1.0000
C2—H2A	1.0000	C2'—C3'	1.526 (4)
C3—C4	1.540 (4)	C2'—H2'A	1.0000
C3—H3A	1.0000	C3'—C4'	1.536 (4)
C4—C5	1.521 (4)	C3'—H3'A	1.0000
C4—H4A	1.0000	C4'—C5'	1.523 (4)
C5—C6	1.515 (4)	C4'—H4'A	1.0000
C5—H5A	1.0000	C5'—C6'	1.510 (4)
C6—O1'	1.441 (3)	C5'—H5'A	1.0000
C6—H6A	0.9900	C6'—H6'A	0.9900
C6—H6B	0.9900	C6'—H6'B	0.9900
C7—H7A	0.9800	O1W—H1WA	0.829 (19)
C7—H7B	0.9800	O1W—H1WB	0.840 (19)
C7—H7C	0.9800
C1—O1—C7	112.3 (3)	H7A—C7—H7C	109.5
C2—O2—H2	109.5	H7B—C7—H7C	109.5
C3—O3—H3	109.5	C1'—O1'—C6	114.2 (2)
C4—O4—H4	109.5	C2'—O2'—H2'	109.5
C1—O5—C5	112.0 (2)	C3'—O3'—H3'	109.5
O1—C1—O5	106.9 (2)	C4'—O4'—H4'	109.5
O1—C1—C2	109.8 (2)	C1'—O5'—C5'	111.8 (2)
O5—C1—C2	110.0 (2)	C6'—O6'—H6'	109.5
O1—C1—H1A	110.0	O1'—C1'—O5'	106.9 (2)
O5—C1—H1A	110.0	O1'—C1'—C2'	107.5 (2)
C2—C1—H1A	110.0	O5'—C1'—C2'	111.0 (2)
O2—C2—C1	111.7 (2)	O1'—C1'—H1'A	110.5
O2—C2—C3	107.9 (2)	O5'—C1'—H1'A	110.5
C1—C2—C3	107.5 (2)	C2'—C1'—H1'A	110.5
O2—C2—H2A	109.9	O2'—C2'—C3'	107.2 (2)
C1—C2—H2A	109.9	O2'—C2'—C1'	110.8 (2)
C3—C2—H2A	109.9	C3'—C2'—C1'	111.4 (2)
O3—C3—C2	113.2 (2)	O2'—C2'—H2'A	109.1
O3—C3—C4	110.0 (2)	C3'—C2'—H2'A	109.1
C2—C3—C4	110.6 (2)	C1'—C2'—H2'A	109.1
O3—C3—H3A	107.6	O3'—C3'—C2'	107.2 (2)
C2—C3—H3A	107.6	O3'—C3'—C4'	113.6 (2)
C4—C3—H3A	107.6	C2'—C3'—C4'	111.6 (2)
O4—C4—C5	107.0 (2)	O3'—C3'—H3'A	108.1
O4—C4—C3	111.0 (2)	C2'—C3'—H3'A	108.1
C5—C4—C3	108.3 (2)	C4'—C3'—H3'A	108.1
O4—C4—H4A	110.1	O4'—C4'—C5'	110.1 (2)
C5—C4—H4A	110.1	O4'—C4'—C3'	109.6 (2)
C3—C4—H4A	110.1	C5'—C4'—C3'	108.6 (2)
O5—C5—C6	105.8 (2)	O4'—C4'—H4'A	109.5
O5—C5—C4	109.3 (2)	C5'—C4'—H4'A	109.5
C6—C5—C4	113.8 (2)	C3'—C4'—H4'A	109.5
O5—C5—H5A	109.3	O5'—C5'—C6'	108.4 (2)
C6—C5—H5A	109.3	O5'—C5'—C4'	108.8 (2)
C4—C5—H5A	109.3	C6'—C5'—C4'	112.6 (2)
O1'—C6—C5	109.6 (2)	O5'—C5'—H5'A	109.0
O1'—C6—H6A	109.8	C6'—C5'—H5'A	109.0
C5—C6—H6A	109.8	C4'—C5'—H5'A	109.0
O1'—C6—H6B	109.8	O6'—C6'—C5'	111.1 (2)
C5—C6—H6B	109.8	O6'—C6'—H6'A	109.4
H6A—C6—H6B	108.2	C5'—C6'—H6'A	109.4
O1—C7—H7A	109.5	O6'—C6'—H6'B	109.4
O1—C7—H7B	109.5	C5'—C6'—H6'B	109.4
H7A—C7—H7B	109.5	H6'A—C6'—H6'B	108.0
O1—C7—H7C	109.5	H1WA—O1W—H1WB	107 (4)
C7—O1—C1—O5	−63.2 (3)	C6—O1'—C1'—O5'	−67.8 (3)
C7—O1—C1—C2	177.4 (3)	C6—O1'—C1'—C2'	172.9 (2)
C5—O5—C1—O1	175.3 (2)	C5'—O5'—C1'—O1'	−178.3 (2)
C5—O5—C1—C2	−65.5 (3)	C5'—O5'—C1'—C2'	−61.4 (3)
O1—C1—C2—O2	−65.5 (3)	O1'—C1'—C2'—O2'	−72.9 (3)
O5—C1—C2—O2	177.1 (2)	O5'—C1'—C2'—O2'	170.6 (2)
O1—C1—C2—C3	176.3 (2)	O1'—C1'—C2'—C3'	167.9 (2)
O5—C1—C2—C3	58.9 (3)	O5'—C1'—C2'—C3'	51.4 (3)
O2—C2—C3—O3	60.5 (3)	O2'—C2'—C3'—O3'	65.4 (3)
C1—C2—C3—O3	−178.9 (2)	C1'—C2'—C3'—O3'	−173.2 (2)
O2—C2—C3—C4	−175.6 (2)	O2'—C2'—C3'—C4'	−169.5 (2)
C1—C2—C3—C4	−54.9 (3)	C1'—C2'—C3'—C4'	−48.2 (3)
O3—C3—C4—O4	−61.9 (3)	O3'—C3'—C4'—O4'	53.8 (3)
C2—C3—C4—O4	172.4 (2)	C2'—C3'—C4'—O4'	−67.6 (3)
O3—C3—C4—C5	−179.1 (2)	O3'—C3'—C4'—C5'	174.1 (2)
C2—C3—C4—C5	55.2 (3)	C2'—C3'—C4'—C5'	52.7 (3)
C1—O5—C5—C6	−172.3 (2)	C1'—O5'—C5'—C6'	−170.7 (2)
C1—O5—C5—C4	64.8 (3)	C1'—O5'—C5'—C4'	66.5 (3)
O4—C4—C5—O5	−177.7 (2)	O4'—C4'—C5'—O5'	59.5 (3)
C3—C4—C5—O5	−57.9 (3)	C3'—C4'—C5'—O5'	−60.5 (3)
O4—C4—C5—C6	64.3 (3)	O4'—C4'—C5'—C6'	−60.7 (3)
C3—C4—C5—C6	−175.9 (2)	C3'—C4'—C5'—C6'	179.3 (2)
O5—C5—C6—O1'	63.8 (3)	O5'—C5'—C6'—O6'	59.5 (3)
C4—C5—C6—O1'	−176.2 (2)	C4'—C5'—C6'—O6'	179.9 (2)
C5—C6—O1'—C1'	−117.9 (3)

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
O2—H2···O2′^i	0.84	2.38	3.089 (3)	142
O2—H2···O3′^i	0.84	2.58	3.280 (3)	141
O3—H3···O4′^ii	0.84	1.90	2.695 (3)	157
O4—H4···O1W^iii	0.84	2.03	2.828 (3)	157
O2′—H2′···O1^iv	0.84	1.90	2.693 (3)	156
O3′—H3′···O3^v	0.84	1.99	2.796 (4)	162
O4′—H4′···O1W^vi	0.84	1.87	2.698 (4)	169
O6′—H6′···O2′^vii	0.84	1.95	2.781 (3)	170
O1W—H1WA···O2^i	0.83 (2)	1.93 (2)	2.750 (3)	168 (4)
O1W—H1WB···O6′^viii	0.84 (2)	1.89 (2)	2.724 (3)	170 (4)

Symmetry codes: (i) −x+2, y+1/2, −z+1; (ii) −x+1, y+1/2, −z+1; (iii) −x+1, y−1/2, −z+1; (iv) −x+2, y−1/2, −z+1; (v) x, y, z−1; (vi) −x+1, y−1/2, −z; (vii) x−1, y, z; (viii) x+1, y, z.