Tracking The Structural Changes in A Series of Cholesterol Solvates
Tracking The Structural Changes in A Series of Cholesterol Solvates
Tracking The Structural Changes in A Series of Cholesterol Solvates
pubs.acs.org/crystal
■ INTRODUCTION
Cholesterol is one of the major steroids within mammalian cells
Despite the functional use of cholesterol in the body, it has
received notoriety in the press for being a contributing factor in
and forms an integral part of the cell membranes. The causing heart disease if present in high levels in the
molecular structure of cholesterol is ideally suited to these bloodstream. Cholesterol is also a major component of
surroundings, as it consists of a hydroxyl group, that interacts gallstones as the anhydrous form as well as in the form of
with the polar headgroup of the phospholipid, and a tetracyclic the monohydrate, and so there have been a number of studies
steroid body bonded to an eight carbon alkyl chain that relating to the growth of this phase onto calcium carbonate, the
interacts with the hydrophobic fatty acid chains of the other main contributor to gallstones.4−7 Further theoretical
phospholipid bilayer of the cell membrane (Scheme 1). studies of cholesterol have been made investigating formation
of nanoparticles and also the applicability of the GROMOS
Scheme 1. Molecular Structure of Cholesterol force field, designed primarily for biological systems, to model
the crystalline structures of smaller membrane sterols.8,9 At the
heart of all these studies, is the investigation of how cholesterol
molecules interact with one another and surrounding materials.
Further investigation of the solid-state forms of cholesterol and
how the molecules interact with one another in these systems
can contribute to our understanding of cholesterol in these
other environments.
In the solid-state, cholesterol crystallizes in a number of
different forms in addition to the two anhydrous phases.10−14
Once orientated within the membrane, one of the roles of In the Cambridge Structural Database (CSD),15,16 there is a
cholesterol is to alter the physicochemical properties of the cell monohydrate (CSD refcode: CHOLES20),17 a hemimethanol
membrane by subtly altering the interactions of the solvate (CHOLME02),18 two hemiethanol solvates (CHOLEU01,
surrounding fatty acid chains. It has been shown through
CHOLEU10),19 an isobutylphosphocholine/tert-butanol coc-
DSC and X-ray diffraction measurements that the addition of
rystal (MEQKAU),20 and finally a cocrystal with 4-iodophenol
cholesterol to the membrane structure causes changes in the
(WOMHAI).21 The structures of all these compounds can be
physical behavior of the phospholipid layers with respect to
broadly categorized into two distinct molecular architectures
temperature. Pure lipid layers possess a gel/liquid-crystal
transition at elevated temperatures whereby the chains “melt” depending on the size of the additional solvent or coformer. The
and become less ordered. The addition of >40 mol % majority of these compounds crystallize in a bilayer struc-
cholesterol causes the disappearance of this phase transition, ture with alternating hydrophobic and hydrophilic regions.
resulting in an “intermediary” structure that possesses a bilayer The molecules of cholesterol are positioned end-to-end such
thickness between that of the gel and liquid-crystal states.1,2 that the isopropyl groups at the end of the alkyl chains interact
Furthermore, a study by Lund-Katz et al. demonstrated, using with one another. The second architecture is more compact,
surface pressure measurements, that the addition of cholesterol
to a membrane causes it to become more condensed and Received: July 27, 2011
hence harder which has an effect on the functioning of trans- Revised: December 1, 2011
membrane proteins.3 Published: December 5, 2011
© 2011 American Chemical Society 231 dx.doi.org/10.1021/cg200971f | Cryst. Growth Des. 2012, 12, 231−239
Crystal Growth & Design Article
Figure 1. Crystal morphologies of the seven new cocrystals compared to the morphology of the methanol and ethanol solvates. Solvates of
cholesterol with (a) propanol, (b) butanol, (c) pentanol, (d) hexanol, (e) phenol, (f) benzyl alcohol·water, (g) phenylethanol, (h) methanol, or (i)
ethanol.
where the cholesterol molecules from neighboring chains are Phase IdentificationX-ray Powder Diffraction. A small
intercalated with one another in a head-to-tail arrangement. In quantity (1−50 mg) of each recrystallized sample was analyzed
this study we have attempted to systematically investigate the using transmission foil XRPD data collected on a Bruker AXS D8-
Advance transmission diffractometer equipped with a θ/θ geometry,
effect of the size and rigidity of the solvent on the final crystal
primary monochromated radiation (Cu Kα1, λ = 1.54056 Å), a Bruker
structure so that a comparison can be made with those struc- Vantec 1D position sensitive detector (PSD), and an automated
tures already found in the CSD. multiposition x−y sample stage.22 Samples were mounted on a 28
position sample plate supported on a polyimide (Kapton, 7.5 μm
■ EXPERIMENTAL SECTION
Sample Preparation. Cholesterol and solvents were purchased
thickness) film. Data were collected from each sample in the range
4−35° 2θ with a 0.015° 2θ step size and a 1 s·step−1 count time.
Figure ES2 of the Supporting Information contains the X-ray powder
diffraction patterns of all the bulk samples and confirms the identity of
from Sigma Aldrich and used as received.
the cocrystals. Slight differences in the diffraction patterns can be
Formation of Compounds 1−7. These compounds were
attributed to the data collection temperature (293 K cf. 123 K for the
obtained from a 1:1 mixture of diethylether and the target solvent
single crystal experiments) and preferred orientation.
(propanol, butanol, etc.). Cholesterol (103 mg, 0.266 mmol for 1; 100
Crystal Structure Determination. X-ray diffraction intensities
mg, 0.259 mmol for 2; 98 mg for 3; 106 mg, 0.275 mmol for 4; 101
were collected with Mo Kα radiation on a Bruker KAPPA Apex II
mg, 0.261 mmol for 5; 151 mg, 0.391 mmol for 6; and 256 mg, 0.663
CCD diffractometer equipped with an Oxford Cryosystems Cryo-
mmol for 7) was dissolved in 3 cm3 of the 1:1 mixture of diethyl ether
stream-Plus variable-temperature device operating at 123 K.23
and target solvent. The solvent was left to evaporate, allowing crystals
Absorption corrections were carried out using the multiscan procedure
to form. In the case of propanol, after a few days, some small crystals
SADABS (Sheldrick, 2004, based on the procedure described by
appeared on the side of the vial which were pushed into the mother
Blessing, 1995).24,25 The structures were solved by direct methods and
liquor in order to grow a suitable crystal for single crystal X-ray refined by full-matrix least-squares against F2 using all data
diffraction. For compound 5, an excess of phenol was added (44 mg, (SHELX).26 Due to the wavelength of the X-ray source (0.71073 Å),
0.468 mmol) to 3 cm3 of dietheyl ether. The cocrystals from pentanol the absolute configuration of the cholesterol molecule was not deter-
and hexanol needed further addition of cholesterol in order to form mined but the absolute structure was chosen to reflect the chirality
the precipitate on evaporation of the diethyl ether. observed for the vast majority of structures in the literature in order to
Crystal Morphology. Figure 1 shows the morphology of the make a direct comparison of the crystal structures. All hydrogen atoms
crystals from each of the crystallizations. The crystals from propanol, attached to carbon atoms were geometrically placed, and those partici-
butanol, and benzyl alcohol show a needle-like morphology whereas pating in hydrogen bonding, i.e. hydroxyl hydrogens, were found in the
the crystals from the other solvents possessed a lathe or plate difference map. All non-H atoms were modeled with anisotropic dis-
morphologythese are compared with the crystals of the methanol placement parameters. Where disorder was present, the bond lengths
and ethanol solvates. and angles were restrained to values found in the CSD.
Differential Scanning Calorimetry. DSC plots were obtained Additional programs used included Materials Mercury 2.4,27
using dynamic DSC (DSC 822e, Mettler Toledo, U.K.). Samples were PLATON as incorporated in WINGX.28,29 Mercury, ChemBioDraw
prepared by carefully weighing between 2.31 and 6.03 mg of each 12.0, and GIMP 2.630 were used in the production of the figures.
■
sample into a 40 μL aluminum pan, which was then hermetically
sealed with a pinhole in the lid. An empty pin-holed 40 μL aluminum
pan was used as a reference. Both pans were subjected to a nitrogen
RESULTS AND DISCUSSION
atmosphere. The pans were then heated at a rate of 10 °C/min from Cholesterol was crystallized with seven alcoholic solvents of increas-
293 to 463 K (well above the melting point of cholesterol). The ing size: propanol, butanol, pentanol, hexanol, phenol, benzyl
temperature and heat flow of the DSC instrument were calibrated with alcohol, and phenylethanol. The descriptions of the crystal
indium and zinc. The results were analyzed using Mettler STAR structures follow, and the crystallographic parameters for the
software. Figure ES1 of the Supporting Information shows the DSC compounds under study and also those of previous studies
and TGA traces for each of the samples.
can be found in Tables 1 and 2.
Thermal Gravimetric Analysis. TGA measurements were
performed on a Mettler Toledo TGA 751e. Each sample (7.86− Thermal Analysis. Figure ES1 of the Supporting Informa-
22.48 mg) was placed in a ceramic pan. An empty ceramic pan was tion shows the thermal analysis plots for each of the cocrystals.
used as a reference, and both pans were subjected to a nitrogen atmo- It can be observed that each of the cocrystals 1−4 shows a
sphere. The pans were then heated from 303 to 463 K at 10°/min. desolvation event followed by the melting of the cholesterol at
The results were analyzed using Mettler STAR software. ∼423 K. The desolvations of the propanol and butanol solvates
232 dx.doi.org/10.1021/cg200971f | Cryst. Growth Des. 2012, 12, 231−239
Crystal Growth & Design Article
Table 2. Unit Cell Parameters for the Crystal Structures of Cholesterol in the Cambridge Structural Database15,16
cholesterol cholesterol cholesterol cholesterol cholesterol isobutyl- cholesterol
cholesterol cholesterol monohydrate hemimethanol hemiethanol hemiethanol phosphocholine 4-iodophenol
form I form II form III solvate solvate form I solvate form II isobutanol solvate cocrystal
CSD CHOEST20 CHOEST21 CHOLES20 CHOLME02 CHOLEU01 CHOLEU10 MEQKAU WOMHAI
Refcode
crystal triclinic, P1 triclinic, P1 triclinic, P1 triclinic, P1 triclinic, P1 monoclinic, P21 monoclinic, C2 monoclinic,
system, P21
space
group
a, b, c (Å) 14.172(7) 27.565(10) 12.390(30) 12.2735(17) 12.787(2) 12.775(2) 16.994(10) 6.302(<1)
34.209(18) 38.624(16) 12.410(30) 34.237(7) 35.310(11) 68.668(15) 11.314(7) 10.295(<1)
10.481(5) 10.748(4) 34.360(60) 6.2739(8) 12.225(1) 12.213(2) 28.164(15) 41.964(3)
α, β, γ (deg) 94.64(4) 93.49(3) 91.90(10) 90.224(14) 97.80(2)
90.67(4) 90.90(3) 98.10(10) 93.705(10) 100.40(2) 100.43(1) 104.07(3) 91.03(<1)
96.32(4) 117.15(3) 100.80(10) 91.576(14) 99.06(2)
V (Å3) 5033(4) 10151(7) 5128 2629.8(7) 5284(2) 105367(3) 5252 2722.3(3)
Z 8 16 8 4 8 16 2 2
are quite clean events; the desolvations of the pentanol and during the prepartion of the sample and subsequent equili-
hexanol show other events surrounding the desolvation which bration of the pan inside the TGA instrument, some solvent
may be due to an increase in disorder of the solvent molecule in was lost from the crytal structure; for crystal structure analysis,
the crystal lattice before it leaves the lattice. The TGA plots the crystals were placed under oil before putting them onto the
show that compounds 1, 2, and 4 (Figures ES1a, ES1b, and instrument under a nitrogen cold stream which would preserve
ES1d) all lose a mass that is equivalent to the stoichiometry the solvent in the crystal.
that is found in the crystal structure analysis (7, 12, and 9%). Compounds 5 and 7 (Figures ES1e and ES1g) show an
The hexanol solvate shows a change in the gradient of the TGA endothermic event at ∼373 K with a constant loss of mass
which signifies a change from the loss of residual solvent to the throughout the experiment. There was no separate endother-
loss of solvent from the crystalline lattice. The second loss is mic event that could be assigned to the melting of cholesterol.
equivalent to the quantity of solvent in the crystal structure. Intriguingly, hot-stage microscopy showed that the samples
The proportion of solvent that is lost from compound 3 remained beyond 373 K and melted at the melting point of
(Figures ES1c) is a little lower than would be expected (7% cf. cholesterol with indications that solvent was being lost on
10%), but refining the crystal structure with a lower occupancy heating. Figures ES1e and ES1g show that, at the melting point
of solvent resulted in extra electron density surrounding the of cholesterol, compound 5 has lost 18% of its mass and
solvent molecules; that is, the stoichiometry is correct at 2:1 compound 7 has lost 14% of its mass, which are consistent with
cholesterol/pentanol. A possible explanation could be that the weights of solvent in the crystal structures identified by
233 dx.doi.org/10.1021/cg200971f | Cryst. Growth Des. 2012, 12, 231−239
Crystal Growth & Design Article
single crystal diffraction (20% and 14%, respectively). Beyond The three molecules interact through hydrogen bonding
the melting point, the mass reduced in each of these solids between the alcohol moieties on each of the molecules to form
indicated the sublimation of cholesterol. chains along the b-direction (O1···O1S, 2.650(3) Å; O2···O1,
Compound 6 (Figure ES1f) showed two endothermic peaks 2.676(2) Å; O1S···O2, 2.683(3) Å). Neighboring chains are
at ∼330 K which could be attributed to the loss of benzyl packed through the application of the 21-screw axis so that the
alcohol and water. Increasing the end point of the experiment cholesterol molecules interact in a head-to-tail manner. Due to
to 230 K showed no further events that could be attributed to the paucity of hydrogen-bonding groups in the alkyl chain, the
the melting of another compound. The mass lost at the melting cholesterol molecules only interact through van der Waals
point of cholesterol was 21%, which is greater than the mass contacts, and so one observes different molecular confomations
of benzyl alcohol and water, which indicates that during heat- of the two independent molecules.
ing it is likely that some of the cholesterol was lost through This structure is not extensively layered into hydrogen
sublimation. bonded and hydrophobic regions, in contrast to the cases of the
Crystal Structure Analysis. 2:1 Cholesterol/Propanol methanol and ethanol solvates. Instead, the propanol molecules
Solvate (1). In 2008, Uskoković investigated the effects of are in discrete locations in the crystal structure (Figure 3).
varying the crystallization conditions on the morphology of
crystals of cholesterol.12 As part of this study, Uskoković
crystallized cholesterol from a mixture of propanol and water
and analyzed these crystals using both DSC and X-ray powder
diffraction. Despite the use of propanol in the reaction mixture,
the author did not observe any changes in the diffraction
pattern or the thermal analysis trace that would suggest the
presence of another phase. Nevertheless, the formation of a
potential propanol solvate was investigated due to the previous
successful isolation of methanol and ethanol solvates by various
groups.18,19
A new compound was observed and found to crystallize in
space group P21 with two molecules of cholesterol and one
propanol molecule in the asymmetric unit (Figure 2). All three
Figure 7. Comparison of the packing diagrams of compounds 1 (lower) and 6 (upper) viewed along the b-direction. The cholesterol molecules in
compound 6 interact via their α-sides, which has previously been calculated as being a favorable motif.8
Figure 12. Crystal structures of compound 6 (left) and compound 5 (right) with an overlay depicting the similarity of the dimer chain of compound
6 and the single hydrogen bonded chain in compound 5.
■ CONCLUSIONS
In this paper we have described the formation of seven new
solvates of cholesterol. All of these novel forms adopt crystal
structures that are removed from the bilayer structure observed
for the simple solvates presently in the CSD. In each of these
structures, the cholesterol molecules interact in a head-to-tail
manner as opposed to tail-to-tail structure of the bilayer. The
solvates of the simplest two alcohols and benzyl alcohol form a
new molecular architecture where the solvent molecules are
found in discrete regions in the crystal structure. Phenol forms
an intermediary structure with more extended hydrophobic and
Figure 13. Crystal structure of compound 7. Note that the hydrophilic regions than the previous solvates. The solvates
phenylethanol participating in the hydrogen bonded chain lies only with pentanol, hexanol, and phenylethanol are all isostructural
on one side of the cholesterol chain (cf. compound 5). with each other and are essentially isostructural with an
238 dx.doi.org/10.1021/cg200971f | Cryst. Growth Des. 2012, 12, 231−239
Crystal Growth & Design Article
iodophenol cocrystal in the CSD but for the change in space (26) Sheldrick, G. M. Acta Crystallogr., Sect. A: Fundam. Crystallogr.
group symmetry. 2008, 64, 112.
■
(27) Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.;
ASSOCIATED CONTENT McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de
Streek, J.; Wood, P. A. J. Appl. Crystallogr. 2008, 41, 466.
*
S Supporting Information (28) Spek, A. L. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2009, 65,
DSC and TGA analysis of the solvates along with the powder 148.
X-ray diffraction results of the solvates formed. This material is (29) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
available free of charge via the Internet at http://pubs.acs.org. (30) Mattis, P.; Kimball, S. GNU Image Manipulation Program (www.
■
gimp.org); Copyright 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009,
AUTHOR INFORMATION 2010.
Corresponding Author
*E-mail: [email protected]. Fax: +44 141 552 2562.
Telephone: +44 141 548 2157.
■ ACKNOWLEDGMENTS
The authors would like to thank Alan R. Kennedy and Alastair
J. Florence for the use of the X-ray diffractometers and Naomi
Watts for help with PXRD. The authors would like to thank the
reviewers for their useful comments. The authors would also
like to thank the Nuffield Trust for a summer scholarship for
R.J.G. and the EPSRC for provision of X-ray equipment under
the Glasgow Centre for Physical Organic Chemistry.
■ REFERENCES
(1) Maulik, P. R.; Shipley, G. G. Biophys. J. 1996, 70, 2256.
(2) Calhoun, W. I.; Shipley, G. G. Biochemistry 1979, 18, 1717.
(3) Lund-Katz, S.; Laboda, H. M.; McLean, L. R.; Phillips, M. C.
Biochemistry 1988, 27 (9), 3416.
(4) D’Ascenzo, G.; Cardarelli, E.; Domenico Magri, A.; Bica, T.;
Sabbatini, M. Clin. Chem. 1978, 24, 119.
(5) Kumar, S.; Burns, S. J.; Blanton, T. N. J. Mater. Res. 1995, 10, 216.
(6) Frincu, M. C.; Sharpe, R. E.; Swift, J. A. Cryst. Growth Des. 2004,
4, 223.
(7) Frincu, M. C.; Fleming, S. D.; Rohl, A. L.; Swift, J. A. J. Am. Chem.
Soc. 2004, 126, 7915.
(8) Cromie, S. R. T.; Ballone, P. J. Chem. Phys. 2009, 131 (3),
034906.
(9) Baran, M.; Mazerski, J. Chem. Phys. Lipids 2002, 120, 21.
(10) Sheih, H. S.; Hoard, L. G.; Nordman, C. E. Nature 1977, 267,
287.
(11) Sheih, H. S.; Hoard, L. G.; Nordman, C. E. Acta Crystallogr.,
Sect. B: Struct. Sci. 1981, 37, 1538.
(12) Uskoković, V. Steroids 2008, 73, 356.
(13) Hsu, L. Y.; Nordman, C. E. Science 1983, 220, 604.
(14) Hsu, L. Y.; Kampf, J. W.; Nordman, C. E. Acta Crystallogr., Sect.
B: Struct. Sci. 2002, 58, 260.
(15) Allen, F. H. Acta Crystallogr., Sect. B: Struct. Sci. 2002, 58, 380.
(16) Allen, F. H.; Motherwell, W. D. S. Acta Crystallogr., Sect. B:
Struct. Sci. 2002, 58, 407.
(17) Craven, B. M. Acta Crystallogr., Sect. B: Struct. Sci. 1979, 35,
1123.
(18) Sheih, H. S.; Nordman, C. E. Acta Crystallogr., Sect. E: Struct.
Rep. Online 2002, 58, o79.
(19) Sheih, H. S.; Hoard, L. G.; Nordman, C. E. Acta Crystallogr.,
Sect. B: Struct. Sci. 1982, 38, 2411.
(20) Karasev, V. A.; Fundamensky, V. S.; Bannova, I. I.; Franke, V.
D.; Stefanov, V. E. Biochim. Biophys. Acta 2000, 1466, 26.
(21) Bhatt, P. M.; Desiraju, G. R. CrystEngComm 2008, 10, 1747.
(22) Florence, A. J.; Baumgartner, B.; Weston, C.; Shankland, N.;
Kennedy, A. R.; Shankland, K.; David, W. I. F. J. Pharm. Sci. 2003,
92 (9), 1930.
(23) Cosier, J.; Glazer, A. M. J. Appl. Crystallogr. 1986, 19, 105.
(24) Sheldrick, G. M. SADABS, Version 2004-1; Bruker-AXS:
Madison, WI, 2004.
(25) Blessing, R. H. Acta Crystallogr. Sect. A: Fundam. Crystallogr.
1995, 51, 33.