Synthesis and NMR studies of 13C-labeled vitamin D metabolites

Article abstract of DOI:10.1021/jo011096y

Isotope-labeled drug molecules may be useful for probing by NMR spectroscopy the conformation of ligand associated with biological hosts such as membranes and proteins. Triple-labeled [7,9,19-¹³C₃]-vitamin D₃ (56), its 25-hydroxylated and 1α,25-dihydroxylated metabolites (58 and 68, respectively), and other labeled materials have been synthesized via coupling of [9-¹³C]-Grundmann's ketone 39 or its protected 25-hydroxy derivative 43 with labeled A ring enyne fragments 25 or 26. The labeled CD-ring fragment 39 was prepared by a sequence involving Grignard addition of [¹³C]-methylmagnesium iodide to Grundmann's enone 28, oxidative cleavage, functional group modifications leading to seco-iodide 38, and finally a kinetic enolate S_N2 cycloalkylation. The C-7,19 double labeling of the A-ring enyne was achieved by the Corey-Fuchs/Wittig processes on keto aldehyde 11. By employing these labeled fragments in the Wilson-Mazur route, the C-7,9,19 triple-¹³C-labeled metabolites 56, 58, and 68 as well as other ¹³C-labeled metabolites have been prepared. In an initial NMR investigation of one of the labeled metabolites prepared in this study, namely [7,9,19-¹³C₃]-25-hydroxyvitamin D₃ (58), the three ¹³C-labeled carbons of the otherwise water insoluble steroid could be clearly detected by ¹³C NMR analysis at 0.1 mM in a mixture of CD₃OD/D₂O (60/40) or in aqueous dimethylcyclodextrin solution and at 2 mM in 20 mM sodium dodecyl sulfate (SDS) aqueous micellar solution. In the SDS micellar solution, a double half-filter NOESY experiment revealed that the distance between the H_19Z and H₇ protons is significantly shorter than that of the corresponding distance calculated from the solid state (X-ray) structure of the free ligand. The NMR data in micelles reveals that 58 exists essentially completely in the α-conformer with the 3β-hydroxyl equatorially oriented, just as in the solid state. The shortened distance (H_19Z-H₇) in micellar solutions as compared to that in the solid state is most easily rationalized on the basis that the 5(10)-torsion angle in 58 is decreased in micellar solutions as compared to that in the solid state.

Full text of DOI:10.1021/jo011096y

J . Org. Chem. 2002, 67, 1637-1650
1637
Syn th esis a n d NMR Stu d ies of 13C-La beled Vita m in D Meta bolites1
William H. Okamura,*,1b Gui-Dong Zhu,1b David K. Hill,1b Richard J . Thomas,1b
Kerstin Ringe,1b Daniel B. Borchardt,1b,c Anthony W. Norman,1d and Leonard J . Mueller1b
Department of Chemistry and Department of Biochemistry, University of California,
Riverside, California 92521
william.okamura@ucr.edu
Received November 21, 2001
Isotope-labeled drug molecules may be useful for probing by NMR spectroscopy the conformation
of ligand associated with biological hosts such as membranes and proteins. Triple-labeled [7,9,19-
13C3]-vitamin D3 (56), its 25-hydroxylated and 1R,25-dihydroxylated metabolites (58 and 68,
respectively), and other labeled materials have been synthesized via coupling of [9-13C]-Grundmann’s
ketone 39 or its protected 25-hydroxy derivative 43 with labeled A ring enyne fragments 25 or 26.
The labeled CD-ring fragment 39 was prepared by a sequence involving Grignard addition of [13C]-
methylmagnesium iodide to Grundmann’s enone 28, oxidative cleavage, functional group modifica-
tions leading to seco-iodide 38, and finally a kinetic enolate SN2 cycloalkylation. The C-7,19 double
labeling of the A-ring enyne was achieved by the Corey-Fuchs/Wittig processes on keto aldehyde
11. By employing these labeled fragments in the Wilson-Mazur route, the C-7,9,19 triple-13C-
labeled metabolites 56, 58, and 68 as well as other 13C-labeled metabolites have been prepared. In
an initial NMR investigation of one of the labeled metabolites prepared in this study, namely [7,9,19-
13C3]-25-hydroxyvitamin D3 (58), the three 13C-labeled carbons of the otherwise water insoluble
steroid could be clearly detected by 13C NMR analysis at 0.1 mM in a mixture of CD3OD/D2O
(60/40) or in aqueous dimethylcyclodextrin solution and at 2 mM in 20 mM sodium dodecyl sulfate
(SDS) aqueous micellar solution. In the SDS micellar solution, a double half-filter NOESY
experiment revealed that the distance between the H19Z and H7 protons is significantly shorter
than that of the corresponding distance calculated from the solid state (X-ray) structure of the free
ligand. The NMR data in micelles reveals that 58 exists essentially completely in the R-conformer
with the 3â-hydroxyl equatorially oriented, just as in the solid state. The shortened distance (H19Z
-
H7) in micellar solutions as compared to that in the solid state is most easily rationalized on the
basis that the 5(10)-torsion angle in 58 is decreased in micellar solutions as compared to that in
the solid state.
In tr od u ction
mechanism of vitamin D action at the molecular level
remains incomplete, but recent progress in the X-ray
crystallographic studies of the ligand bound vitamin D
nuclear receptor (n-VDR)3a-e and plasma vitamin D
binding protein (DBP)3f,g have been major steps forward
in this regard.
It has been found in recent years that the vitamin D
metabolic system exhibits a broad spectrum of biological
activities, well beyond its classical functions in regulating
calcium and phosphorus metabolism.2 This steroid hor-
mone system and its analogues have been used or have
high potential in treating a diverse range of human
diseases such as cancer, skin diseases, and diseases
associated with aberrant immunological responses or
calcium metabolism.2a,d An understanding of the detailed
The seco-steroid vitamin D in its principal metabolic
forms [Scheme 1: vitamin D3 (1, D3), 25-hydroxyvitamin
D3 (2, 25-D3) and the hormonally active form 1R,25-
dihydroxyvitamin D3 (3, 1,25-D3)] involve association
with a variety of host systems including membranes and
proteins during the course of mediating their biological
actions. To gain insight into ligand design, extensive
effort has gone into solid-state, solution, and computa-
tional studies of vitamin D ligands (1-3).2b,4-7 X-ray
crystallographic results for a number of vitamin D
structures including 1-3 reveal that the ∆5,7-diene is 6-s-
trans and that this diene subunit deviates from planarity
within a range of only (8.1°.2b,8 For crystalline 3, the
deviation from planarity is less than 3°.8c The ∆10(19)
double bond deviates from the ∆5,7-diene plane, however,
wherein the 5,10-s-cis bond deviates from planarity
* To whom correspondence should be addressed. Tel: (909) 787-
3509. Fax: (909) 787-4713.
(1) (a) This study was supported by grants from the National
Institutes of Health (NIH Grant No. DK-16595), the National Science
Foundation facilities programs (NSF-MRI-CHE-9724392, NSF-ARI-
9602708, NSF-ARI-9601831, and NSF-ARI-ST-193-13519), the Cam-
bridge Isotopes Laboratories (CIL Research Grant Program), and the
Committee on Research of UC Riverside. Selected chemicals were also
provided by Solvay, Hoffmann-LaRoche, and Leo Pharmaceuticals. (b)
Department of Chemstry. (c) Also, the Analytical Chemistry Instru-
mentation Facility. (d) Department of Biochemistry.
(2) (a) Bouillon, R.; Okamura, W. H.; Norman, A. W. Structure-
Function Relationships in the Vitamin D Endocrine System. Endocr.
Rev. 1995, 16, 200-257. (b) Okamura, W. H.; Zhu, G.-D. Chemistry
and Design: Structural Biology of Vitamin D Action. In Vitamin D;
Feldman, D., Glorieux, F. H., Pike, J . W., Eds.; Academic Press: San
Diego, 1997; pp 939-971. (c) Zhu, G.; Okamura, W. H. Synthesis of
Vitamin D (Calciferol). Chem. Rev. 1995, 95, 1877-1952. (d) Feldman,
D., Glorieux, F. H., Pike, J . W., Eds. Vitamin D; Academic Press: San
Diego, 1997.
within the range (52.3°.2b The near planarity of the ∆5,7
-
diene as well as the 6-s-trans orientation in 1-3 is also
apparent from solution NMR Karplus relationship stud-
ies.9 The only completed structures of protein-bound
10.1021/jo011096y CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/02/2002

1638 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
Sch em e 1
vitamin D compounds are the X-ray studies of 3 and
several analogues bound to a doubly truncated n-VDR-
LBD (n-VDR-ligand binding domain).3a,c,d In these re-
cently reported studies by the Moras Laboratory, amino
acids 1-117 and 165-215 were deleted from the full
length, wild-type human n-VDR (427 amino acids) in
order to achieve good crystallinity. Interestingly, there
is in the protein bound ligand 3 an out-of-plane 31.4°
twist of the s-trans 6,7-single bond (as compared to only
2.5° for the free ligand8c). The cisoid 5(10)-single bond of
the same bound ligand is twisted by 52.8° out of plane
(as compared to 48.3° in the free ligand8c).
Thus far, there is no direct structural evidence for 6-s-
cis forms 1′-3′ of metabolites 1-3, respectively.2b,4 It is
known, however, that these metabolites interequilibrate
with previtamin structures 4-6, respectively, via
antarafacial [1,7]-sigmatropic hydrogen shifts between
the C9 and C19 positions.10 To undergo this intramolecular
valence tautomerism, the 6-s-cis forms 1′-3′ must be
available at kinetically competent concentrations. While
the equilibrium concentrations of 6-s-trans vitamins 1-3
(∼90%) relative to the previtamins 4-6 (∼10%) are well
documented, information concerning the concentrations
of 6-s-cis vitamins 1′-3′, respectively, at equilibrium
remains elusive.
In view of the conformational flexibility of vitamin D
and the ability of vitamin D to equilibrate tautomerically
with the previtamins (Scheme 1), a pertinent question
is whether these higher energy forms are in fact involved
in native protein binding. That this may indeed be the
case first emerged from our finding that 1,25-pre-D3 (6)
is as active as the parent hormone 1,25-D3 (3) in
stimulating at least one of its biological responses,
namely the rapidly acting calcium transport response
known as transcaltachia.11-13 This response is thought
to be initiated through binding to a membrane receptor.
The 6-s-cis locked previtamin 6 was 5% as effective as
the parent hormone 3, however, in binding to the n-VDR.
(3) (a) Rochel, N.; Wurtz, J . M.; Mitschler, A.; Klahholz, B.; Moras,
D. The Crystal Structure of the Nuclear Receptor for Vitamin D Bound
to its Natural Ligand. Molecular Cell 2000 5, 173-179. (b) J untunen,
K.; Rochel, N.; Moras, D.; Vihko, P. Large-Scale Expression and
Purification of the Human Vitamin D Receptor and Its Ligand-Binding
Domain for Structural Studies. Biochem. J . 1999, 344. 297-303. (c)
Tocchini-Valentini, G. D.; Rochel, N.; Wurtz, J . M.; Mitschler, A.;
Moras, D. Ligand Binding to the Nuclear Receptor for Vitamin D. In
Vitamin D Endocrine System: Structural, Biological, Genetic and
Clinical Aspects; Norman, A. W., Bouillon, R., Thomasset, M., Eds.;
Printing and Reprographics, University of California, Riverside: Riv-
erside, 2000; pp 205-212. (d) Tocchini-Valentini, G.; Rochel, N.; Wurtz,
J . M.; Mitschler, A.; Moras, D. Crystal Structures of the Vitamin D
Receptor Complexed to Superagonist 20-epi Ligands. Proc. Natl. Acad.
Sci. U.S.A. 2001, 98, 5491-5496. (e) Rochel, N.; Tocchini-Valentini,
G.; Egea, P. F.; J untunen, K.; Garnier, J .-M.; Vihko, P.; Moras, D.
Functional and Structural Characterization of the Insertion Region
in the Ligand Binding Domain of the Vitamin D Nuclear Receptor.
Eur. J . Biochem. 2001, 268, 971-979. (f) Verboven, C. Three-
Dimensional Crystal Structure Determination of the Human Vitamin
D Binding Protein. Ph.D. Thesis. Katholieke Universiteit Leuven,
Leuven, Belgium, 1998. (g) Bogaerts, I.; Verboven, C.; Rabijns, A.; Van
Baelen, H.; Bouillon, R.; De Ranter, C. In Crystallization and Prelimi-
nary X-ray Investigation of the Human Vitamin D-Binding Protein in
Complex with 25-Hydroxyvitamin D3; Norman, A. W., Bouillon, R.,
Thomasset, M., Eds.; Printing and Reprographics, University of
California, Riverside: Riverside, 2000; pp 117-120.
(4) Okamura, W. H.; Midland, M. M.; Hammond, M. W.; Abd.
Rahman, N.; Dormanen, M. C.; Nemere, I.; Norman, A. W. Chemistry
and Conformation of Vitamin D Molecules. J . Steroid Biochem. Mol.
Biol. 1995, 53, 603-613.
(5) Okamura, W. H.; Midland, M. M.; Norman, A. W., Hammond,
M. W.; Dormanen, M. C.; Nemere, I. Biochemical Significance of the
6-s-cis Conformation of the Steroid Hormone 1R,25-Dihydroxyvitamin
D3 Based on the Provitamin D Skeleton Ann. New York Acad. Sci.
1995, 761, 344-3486.
(6) Okamura, William H.; Palenzuela, J . A.; Plumet, J .; Midland,
M. M. Vitamin D: Structure-Function Analyses and the Design of
Analogs. J . Cell. Biochem. 1992, 49, 10-18.
(7) Okamura, W. H.; Do, S.; Kim, H.; J eganathan, S.; Vu, T.; Zhu,
G.-D.; Norman, A. W. Conformationally Restricted Mimics of Vitamin
D Rotamers. Steroids 2001, 66, 239-247.

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1639
Ch a r t 1
might be involved in nuclear receptor binding. However,
on the basis of the recent protein X-ray result, the Moras
group has logically suggested3a that the inability of 8 to
bind to the n-VDR is that 8 is unable to twist (by 30°)
about its 6,7-single bond because of the presence of the
double bond. It is clear that there is a 360° continuum of
topologies available to the hormone 1,25-D3 by rotation
about its 6,7-single bond.5 Determination of the topology
of 1,25-D3 as well as the other metabolites shown in
Scheme 1 by 6,7-single bond rotation in the active site of
protein would provide insight into drug design. To
evaluate this concept for this particular series of flexible
ligands, we describe herein a strategy for determining
the conformation of the seco-B region of vitamin D
metabolites and analogues in ligand-protein complexes.
The essential feature of the strategy is the synthesis of
C-9,19 and C-7,19 double 13C and C-7,9,19 triple 13C
labeled vitamin D compounds since the NMR techniques
for measuring the distance between protons attached to
pairs of 13C atoms, or directly between pairs of 13C atoms
themselves, are available. Thus, a combination of C-9,19
and C-7,19 double 13C labeled vitamin D metabolites
would nicely provide a means of distinguishing 6-s-trans
vitamin D from 6-s-cis vitamin D in the protein-ligand
complex. For example, in solution, the NOESY spectrum
containing exclusively NOEs of 13C-bound protons could
be achieved with use of a heteronuclear-resolved double
half-filter experiment.15,16 Utilizing these solution tech-
The recent X-ray structure noted above of the hormone
1,25-D3 bound to the n-VDR-LBD3a-c suggests that a 30°
twisted s-trans form of 3 is required for binding, but it
is difficult to explain why 6 shows even a 5% binding
affinity. Moreover, additional evidence has accrued14
suggesting that this same hormone binds to the mem-
brane receptor in its spectroscopically invisible s-cis form
3′, the rationale being that the latter is mimicked
effectively by the s-cis-locked 1,25-pre-D3 (6). A study of
1R,25-dihydroxylumisterol (7, “lumi”) shown in Chart 1
supported this s-cis interpretation in that the “lumi”
analogue bound equally as well as 6 and 3/3′ to the
membrane receptor. Moreover, the lumi analogue 7, like
6, bound weakly to the nuclear receptor. It was also
shown14 that the 6,7-trans-locked analogues 1R,25-dihy-
droxytachysterol (8) and 1R,25-dihydroxyisotachysterol
(9) failed to bind effectively to either the nuclear or
membrane receptors.14 It was suggested therefore that
an out-of-plane 6,7-twisted s-cis conformer of 3 or 3′
(10) (a) Curtin, M. L.; Okamura, W. H. 1R,25-Dihydroxyprevitamin
D3: Synthesis of the 9,14,19,19,19-Pentadeuterio Derivative and a
Kinetic Study of Its [1,7]-Sigmatropic Shift to 1R,25-Dihydroxyvitamin
D3. J . Am. Chem. Soc. 1991, 113, 6958-6966. (b) Okamura, W. H.;
Elnagar, H. Y.; Ruther, M.; Dobreff, S. Thermal [1,7]-Sigmatropic Shift
of Previtamin D3 to Vitamin D3: Synthesis and Study of Pentadeuterio
Derivatives. J Org. Chem. 1993, 58, 600-610. (c) Hoeger, C. A.;
Okamura, W. H. On the Antarafacial Stereochemistry of the Thermal
[1,7]-Sigmatropic Hydrogen Shift. J . Am. Chem. Soc. 1985, 107, 268-
270.
(11) Norman, A. W.; Okamura, W. H.; Farach-Carson, M. C.;
Allewaert, K.; Branisteanu, D.; Nemere, I.; Muralidharan, K. R.;
Bouillon, R. Structure-Function Studies of 1,25-Dihydroxyvitamin D3
and the Vitamin D Endocrine System: 1,25-Dihydroxy-Pentadeuterio-
Previtamin D3 (As a 6-s-Cis Analog) Stimulates Nongenomic But Not
Genomic Biological Responses. J . Biol. Chem. 1993, 268, 13811-13819.
(12) Dormanen, M. C.; Bishop, J . E.; Hammond, M. W.; Okamura,
W. H.; Nemere, I.; Norman, A. W. Non-Nuclear Effects of the Steroid
Hormone 1R,25-(OH)2-Vitamin D3: Analogs Are Able to Functionally
Differentiate Between Nuclear and Membrane Receptors. Biochem.
Biophys. Res. Commun. 1994, 201, 394-401.
(8) For X-ray crystallographic results for the principal metabolites
1-3, see: (a) Trinh-Toan; DeLuca, H. F.; Dahl, L. F Solid-State
Conformations of Vitamin D3. J . Org. Chem. 1976, 41, 3476-3478. (b)
Trinh-Toan; Ryan, R. C.; Simon, G. L.; Calabrese, J . C.; Dahl, L. F.
Crystal Structure of 25-Hydroxy-vitamin D3 Monohydrate: A Stereo-
chemical Analysis of Vitamin D Molecules. J . Chem. Soc., Perkin Trans.
2 1977, 393-401. (c) Suwinska, K.; Kutner, A. Crystal and Molecular
Structure of 1,25-Dihydroxycholecalciferol. Acta Crystallogr. 1996, B52,
550-554. See also: (d) Tan, E. S.; Tham, F. S.; Okamura, W. H.
Vitamin D1. Chem. Commun. 2000, 2345-2346 and ref 2b for leading
(13) (a) Nemere, I.; Dormanen, M. C.; Hammond, M. W.; Okamura,
W. H.; Norman, A. W. Identification of a Specific Binding Protein for
1R,25-Dihydroxyvitamin D3 in Basal Lateral Membranes of Chick
Intestinal Epithelium and Relationship to Transcaltachia. J . Biol.
Chem. 1994, 269, 23750-23756. (b) Nemere, I.; Schwartz, Z.; Pedrozo,
H.; Sylvia, V. L.; Dean, D. D.; Boyan, B. D. Identification of a
Membrane Receptor for 1R,25-Dihydroxyvitamin D3 Which Mediates
Rapid Activation of Protein Kinase C. J . Bone Min. Res. 1988, 13,
1353-1359. (c) Baran, D. T.; Quail, J . M.; Ray, R.; Leszyk, J .;
Honeyman, T. Annexin II Is the Membrane Receptor that Mediates
the Rapid Actions of 1R,25-Dihydroxyvitamin D3. J . Cellular Biochem.
2000, 78, 34-36. (d) Baran, D. T.; Quail, J . M.; Ray, R.; Honeyman, T.
Binding of 1R,25-Dihydroxyvitamin D3 to Annexin II: Effect of Vitamin
D Metabolites and Calcium. J . Cell. Biochem. 2000, 80, 259-265.
(14) Norman, A. W.; Okamura, W. H.; Hammond, M. W.; Bishop, J .
E.; Dormanen M. C.; Bouillon, R.; van Baelen, H.; Ridall, A. L.; Daane,
E.; Khoury, R.; Farach-Carson, M. C. Comparison of 6-s-cis and 6-s-
trans Locked Analogs of 1R,25-Dihydroxyvitamin D3 Indicates that the
6-s-cis Conformation is Preferred for Rapid Nongenomic Biological
Responses and that Neither 6-s-cis nor 6-s-trans Locked Analogues
are Preferred for Genomic Biological Responses. Mol. Endocrinol. 1997,
11, 1518-1531.
references to X-ray crystallographic studies of other vitamin
compounds.
D
(9) (a) Delaroff, V.; Rathle, P.; Legrand, M. No. 280. EÄ tude de La
Re´sonance Magne´tique Nucle´aire du Pre´calcife´rol, du Tachyste´rol et
du Calcife´rol (NMR Study of Precalciferol, Tachysterol and Calciferol).
Bull. Soc. Chim. Fr. 1963, 1739-1741. (b) Wing, R. M.; Okamura, W.
H.; Pirio, M. R.; Sine, S. M.; Norman, A. W. Vitamin D in Solution:
Conformations of Vitamin D3, 1R,25-Dihydroxyvitamin D3, and Dihy-
drotachysterol3. Science, 1974, 186, 939-941. (c) La Mar, G. N.; Budd,
D. L. Elucidation of the Solution Conformation of the A Ring in Vitamin
D Using Proton Coupling Constants and a Shift Reagent. J . Am. Chem.
Soc. 1974, 96, 7317-7324. (d) Wing, R. M.; Okamura, W. H.; Rego, A.;
Pirio, M. R.; Norman, A. W. Studies on Vitamin D and Its Analogs.
VII. Solution Conformations of Vitamin D3 and 1R,25-Dihydroxyvita-
min D3 by High-Resolution Proton Magnetic Resonance Spectroscopy.
J . Am. Chem. Soc. 1975, 97, 4980-4985. (e) Helmer, B.; Schnoes, H.
K.; DeLuca, H. F. 1H Nuclear Magnetic Resonance Studies of the
Conformations of Vitamin D Compounds in Various Solvents. Arch.
Biochem. Biophys. 1985, 241, 608-615.
(15) (a) Wider, G.; Weber, C.; Traber, R.; Widmer, H.; Wuthrich, K.
Use of a Double half-filter in two-dimensional proton NMR studies of
receptor-bound cyclosporin. J . Am. Chem. Soc. 1990, 112, 9015-9016.
(b) Wider, G.; Weber, C.; Wuthrich, K. Proton-Proton Overhauser
Effects of Receptor-Bound Cyclosporin A Observed with the Use of a
Heteronuclear-Resolved Half-Filter Experiment. J . Am. Chem. Soc.
1991, 113, 4676-4678.

1640 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
Sch em e 2
niques, it is anticipated that a significant NOE between
H9 and H19 would implicate the steroidal s-cis conforma-
tion (1′, 2′, and 3′) while an enhanced NOE between H7
and H19 would suggest an s-trans conformation (1, 2, and
3). The use of triple-labeled (C-7,9,19) material rather
than pairs of double-labeled materials (C-7,19 and C-9,19)
could be more economical in probing for both possibilities
in a single experiment. Alternatively, using solid-state
NMR spectroscopy, the distance between pairs of 13C-
atoms can be obtained through rotationally resonant
magnetization exchange with high accuracy ((0.1 Å up
to 6.5 Å).17,18 To demonstrate the feasibility of this
approach, it is first necessary to synthesize doubly and
triply 13C-labeled vitamin D compounds19 and to demon-
strate that 13C NMR studies of these steroids in water
are feasible. We report herein the synthesis of multiply
labeled D3, 25-D3, and 1,25-D3 and an initial NMR
experiment on a model micellar system.
Resu lts a n d Discu ssion
A number of general approaches toward construction
of the vitamin D skeleton have been developed,2c but
because of the central seco-B location of the desired labels
in the targeted steroids, the Wilson-Mazur coupling
protocol20-25 entailing sovolysis of intermediates of type
10 (Chart 1) seemed best to suit our NMR requirements.
This route employs keto aldehyde 11 to which Wittig
methodology was anticipated to allow introduction of one
or two 13C labels into the triene/A-ring at a reasonably
late stage. Several useful syntheses of derivatives of the
keto aldehyde 11 have been reported20a,22-23 and the
synthesis reported herein extends these results. A num-
ber of total synthesis routes to derivatives of an appropri-
ate CD fragment of the type 12 have been reported.2c
Because the reported total syntheses including label
incorporation appeared to require far too many steps, it
was anticipated that proper transformations of Grund-
(16) (a) Norris, A. W.; Rong, D.; d’Avignon, D. A.; Rosenberger, M.;
Tasaki, K.; Li, E. Nuclear Magnetic Resonance Studies Demonstrate
Differences in the Interaction of Retinoic Acid with Two Highly
Homologous Cellular Retinoic Acid Binding Proteins. Biochemistry
1995, 34, 15564-15573. (b) Lu, J .; Lin, C.-L.; Tang, C.; Ponder, J . W.;
Kao, J . L. F.; Cistola, D. P.; Li, E. Binding of Retinol Induces Changes
in Rat Cellular Retinol-Binding Protein II Conformation and Backbone
Dynamics. J . Mol. Biol. 2000, 300, 619-632. (c) Curley, R. W., J r.;
Sundaram, A. K.; Fowble, J . W.; Abildgaard, F.; Westler, W. M.;
Markley, J . L. NMR studies of retinoid-protein interactions: the
conformation of [13C]-â-ionones bound to â-lactoglobulin B. Pharm.
Res. 1999, 16, 651-659.
(17) Creuzet, F.; McDermott, A.; Gebhard, R.; van der Hoef, K.;
Spijker-Assink, M. B.; Herzfeid, J .; Lugtenburg, J .; Levitt, M. H.;
Griffin, R. G. Determination of Membrane Protein Structure by
Rotational Resonance NMR: Bacteriorhodopsin. Science 1991, 251,
783-786.
mann’s ketone (12) would better allow introduction of 13
C
label R to the carbonyl group at position-9. The easily
available, parent vitamin D3 would nicely provide by
oxidative degradation the key keto aldehyde 11 as well
as Grundmann’s ketone (12).
Solvolysis of cholecalciferyl tosylate (aqueous acetone,
NaHCO3), which was synthesized from vitamin D3 and
p-toluenesufonyl chloride, furnished the known, starting
cyclovitamin 13 (Scheme 2).24 Hydroxylation of the cy-
clovitamin 13 with stoichiometric osmium tetraoxide in
pyridine afforded pentaol 14 in poor yield (20-30%),
whereas the catalytic version (NMO/OsO4/acetone/H2O/
pyridine) gave this material in very good yield (80%). This
interesting difference in reactivity can be rationalized on
the basis that initial osmylation occurs at the ∆10,19
double bond, presumably to afford an osmate ester (with
the stereochemistry corresponding to that in 14).25 Under
stoichiometric conditions, it is suggested that the steri-
cally bulky osmate ester hinders further osmylation
across the ∆7,8 double bond. Under catalytic conditions,
the less hindered 10,19-diol is produced in situ, permit-
ting further oxidation of the ∆7,8-double bond. It is known
that the corresponding methyl ether of 13 is selectively
osmylated across the ∆10,19 double bond under stoichio-
metric conditions.19,25 Under even our catalytic conditions
(NMO, OsO4/acetone/H2O/pyridine) using excess reagent
(18) Griffiths, J . M.; Lakshmi, K.V.; Benneth, A. E.; Raap, J .; van
der Wielen, C. M.; Lugtenburg, J .; Herzbeld, J .; Griffin, R. G. Dipolar
Correlation NMR Spectroscopy of a Membrane Protein. J . Am. Chem.
Soc. 1994, 116, 10178-10181.
(19) Eguchi, T.; Kakinuma, K.; Ikekawa, N. Synthesis of 1R-[19-
13C]Hydroxyvitamin D3 and 13C NMR Analysis of the Conformational
Equilibrium of the A-Ring. Bioorg. Chem. 1991, 19, 327-332.
(20) (a) Wilson, S. R.; Venkatesan, A. M.; Augelli-Szafran, C. E.;
Yasmin, A. A New Synthetic Route to 1,25-Dihydroxy-vitamin D3.
Tetrahedron Lett. 1991, 32, 2339-2342 (b) Wilson, S. R.; Venkatesan,
A. M.; Augelli-Szafran, C. E. A Practical Synthesis of 1-Alpha-OH
Vitamin D3 Analogs. In Vitamin D: Molecular, Cellular and Clinical
Endocrinology; Norman, A. W., Schaefer, K., Grigoleit, H.-G., v.
Herrath, D., Eds.; Walter de Gruyter: Berlin, New York, 1988; pp 43-
50.
(21) Sheves, M.; Mazur, Y. The Vitamin D-3,5-Cyclovitamin
D
Rearrangement. J . Am. Chem. Soc. 1975, 97, 6249-6250.
(22) (a) Kabat, M. M.; Kiegiel, J .; Cohen, N.; Toth, K.; Wovkulich,
P. M.; Uskokovic, M. R. Convergent Synthesis of Vitamin D3 Metabo-
lites. Control of the Stereoselectivity in Samarium-Induced Cyclopro-
panations of Cyclopentenes. J . Org. Chem. 1996, 61, 118-124. (b)
Kabat, M.; Kiegiel, J .; Cohen, N.; Toth, K.; Wovkulich, P. M.; Usko-
kovic, M. R. Control of Stereoselectivity in Samarium Metal Induced
Cyclopropanations. Synthesis of 1,25-Dihydroxycholecalciferol. Tetra-
hedron Lett. 1991, 32, 2343-2346.
(23) (a) Moriarty, R. M.; Kim, J .; Guo, L. Intramolecular Cyclopro-
panation Using Iodonium Ylides. The 3,5-Cyclovitamin D Ring A
Synthon. Tetrahedron Lett. 1993, 34, 4129-4132 (b) Moriarty, R. M.;
Prakash, O.; Vaid, R. K.; Zhao, L. A Novel Intramolecular Cyclopro-
panation Using Iodonium Ylides. J . Am. Chem. Soc. 1989, 111, 6443-
6444. (c) Dauben, W. G.; Hendricks, R. T.; Pandy, B.; Wu, S. C.; Zhang,
X.; Luzzio, M. J . Stereoselective Intramolecular Cyclopropanations:
Enantioselective Syntheses of 1R,25-Dihydroxyvitamin D3 A-Ring
Precursors. Tetrahedron Lett. 1995, 36, 2385-2388.
(24) Sheves, M.; Mazur, Y. Cyclovitamins D3: The Preparation of
C6-Ketocyclovitamin and Its Conversion to A Ring B/C Spirosteroid.
Tetrahedron Lett. 1976, 2987-2990.
(25) Paaren, H. E.; Schnoes, H. K.; DeLuca, H. F. Electrophilic
Addition of OsO4 to 25-Hydroxycholecaliferol and Its 3,5-Cyclo Deriva-
tive. J . Org. Chem. 1983, 48, 3819-3820.

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1641
Sch em e 3
mann’s enone 28 was synthesized from Grundmann’s
ketone (12) in 50-60% yield as previously described.27
Addition of [13C]-methylmagnesium iodide to the enone
in the presence of a 2-fold excess of cerium(III) chloride
gave 29 in high yield. The allylic alcohol 29 was not stable
to column chromatography (silica gel) but could be
purified on a small scale by HPLC. The crude product
was directly ozonized in methanol followed by NaBH4
work up to afford triol 31. Reduction of the methoxy-
hydroperoxide intermediate 30 (1H NMR analysis) with
sodium borohydride is very slow, requiring a large excess
of sodium borohydride with overnight stirring at room
temperature. Oxidative cleavage of the triol 31 with
sodium periodate led to keto alcohol 32 but an attempt
to directly derivatize the latter with tosyl chloride failed.
Selective monotosylation of 31 at the C-11 hydroxyl
followed by oxidative glycol cleavage was then considered
since tosylation of the 8,9-glycol moiety might be expected
to be significantly slower than the isolated primary
alcohol. Such an approach would have shortened the
eventually successful sequence leading to 37. Surpris-
ingly, no reaction was detected when 31 was treated with
a large excess of tosyl chloride (5 equiv) in pyridine, even
at room temperature for 20 h. Under stronger conditions
(Et3N, DMAP, CH2Cl2) monotosylation appeared to have
occurred, but apparently not selectively at the C-11
hydroxyl. The putative monotosylate under oxidative
cleavage conditions (NaIO4/aqueous dioxane) did not
afford 37. This approach was abandoned.
The crude oxidation product 32 earlier obtained from
31 was subjected to direct sodium borohydride reduction,
affording a mixture of two diastereomeric diols 33 and
34 (2:1 ratio). The stereochemistry of the two isomers
follows from the subsequent transformation and analysis.
Tosylation of the mixture of 33 and 34 with tosyl chloride
at -25 °C afforded an easily separable mixture of
monotosylate 35 and pyran 36. In a separate experiment,
33 was transformed only to tosylate 35 and 34 only to
cyclized product 36 under the tosylation conditions. In
the latter case, the initially formed monotosylate is
apparently not stable and is considered to be displaced
by the hydroxyl group in an intramolecular SN2 manner.
The stereochemistry of 34 was correlated to that of 36,
which was assigned on the basis of an observation of a
large trans-diaxial coupling (9.8 Hz) between H-8 and
H-14 (while decoupling the 13C and H’s of the *CH3
group). According to a Felkin-Ahn analysis (Chart 2),
(8S) product 34 should be the major isomer. Here,
however, a remote hydroxyl group chelation effect might
be involved in the reduction of 32, leading to a slight
preference for the observed 33 rather than the expected
34. As to why 33 affords the normal hydroxy-tosylate 35,
an intramolecular SN2 process involving the latter would
suffer a serious 1,3-diaxial interaction between the 18-
methyl group and the isotopically labeled methyl group
(i.e., the resulting product would be the C8 epimer of 36).
Oxidation of 35 with PCC afforded ketotosylate 37,
which upon treatment with either lithium diisopropyla-
mide (LDA) or lithium or sodium bistrimethylsilylamide
described above, we have found that this same methyl
ether undergoes osmylation only across the ∆10,19 double
bond (unpublished observation). From our experience, the
polarity of the desired product 14 is less than that
i
expected of a pentanol (Rf ) 0.53 in 1:10 PrOH/CH2Cl2);
its surprisingly moderate solubility in the aqueous phase
facilitated the work up. Oxidative cleavage of the pen-
tanol 14 with sodium periodate readily afforded Grund-
mann’s ketone 12 and keto aldehyde 11, the latter being
the key intermediate for preparing the 13C-labeled A-ring
fragment. Variable amounts of hydroxyaldehyde 15 were
also isolated from the reaction mixture and can be further
transformed to Grundmann’s ketone using prolonged
reaction times. However, more optimal yields of keto
aldehyde 11 (together with higher percentages of 15
relative to 12) are obtained at shorter reaction times.
The unlabeled and labeled A-ring fragments were
prepared as outlined in Scheme 3. Keto aldehyde 11 was
transformed into enyne 27 by the Corey-Fuchs proce-
dure26 superimposed on a normal Wittig process. Treat-
ment of 11 with at least 3.5 equiv of CBr4-based reagent
afforded a quantitative yield of 16. The use of less reagent
(2.5 equiv) afforded only a 64% yield of product, but there
is the potential of more economically utilizing [13C]-CBr4.
The second Wittig reaction using unlabeled reagent on
16 then afforded an 80% yield of products, which upon
HPLC separation afforded diene dibromide 18 (56%) and
bromoenyne 19 (17%). Treatment of either 18 or 19 or a
mixture with butyllithium afforded 24, which upon
quenching afforded the previously reported, highly vola-
tile enyne 27.22 The losses incurred because of this
volatility prompted the direct use of the labeled lithium
salts 25 and 26 in the coupling reactions described below.
The [19-13C] monolabeled alkynyllithium salt 25 was
prepared directly from a crude, dry mixture of 20 plus
21, which were in turn prepared from unlabeled 16. The
[7,19-13C2] di-labeled alkynyllithium salt 26 was likewise
prepared from a mixture of bromides 22 and 23. In an
analogous sequence, the latter mixture was prepared
from the crude monolabeled ketone 17.
(27) (a) Okamura, W. H.; Aurrecoechea, J . M.; Gibbs, R. A.; Norman,
A. W. Synthesis and Biological Activity of 9,11-Dehydrovitamin D3
Analogues: Stereoselective Preparation of 6â-Vitamin D Vinylallenes
and a Concise Enynol Synthesis for Preparing the A-Ring. J . Org.
Chem. 1989, 54, 4072-4083. (b) Daniel, D.; Middleton, R.; Henry, H.
L.; Okamura, W. H. Inhibitors of 25-Hydroxyvitamin D3-1R-Hydroxy-
lase: A-Ring Oxa Analogs of 25-Hydroxyvitamin D3. J . Org. Chem.
1996, 61, 5617-5625.
The synthesis of the requisite 13C labeled CD-ring
moiety is summarized in Scheme 4. The known Grund-
(26) Corey, E. J .; Fuchs, P. L. A Synthetic Method for the Formyl
to Ethynyl Conversion. Tetrahedron Lett. 1972, 3769-3772.

1642 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
Sch em e 4
Ch a r t 2
sitized irradiation (9-acetylanthracene, tert-butyl methyl
ether, Pyrex, Hanovia 450 W medium-pressure Hg
lamp)22 in satisfactory overall yield. An analogous se-
quence starting from the protected 25-hydroxy CD frag-
ment 43, which was derived from direct 25-hydroxylation
of 3929,30 followed by TMS protection, and double labeled
A-ring fragment 26 produced the triply 13C labeled 25-
hydroxy metabolite 58 and 5,6-trans isomer 59. Likewise,
the latter could be converted into the former by photo-
sensitized irradiation. In repeated trials, 25-hydroxyla-
tion of Grundmann’s ketone by RuCl3 proved erratic, but
acceptable results were obtained(39 to 42: 40-45% yield;
52-56% based on recycled starting material) by using
increased amounts of RuCl3 catalyst (up to 20 mol %) and
shorter reaction times (1-2 days) when compared to that
previously reported.29 In a second more efficient proce-
dure 25-hydroxylation was carried out with methyl-
(trifluoromethyl)dioxirane (39 to 42: 71-90% yield).30
Dimethyldioxirane oxidation30 was slower and gave yields
similar to those obtained using RuCl3.29
The double-labeled vitamin D metabolites 60 and 62
and their 5,6-trans isomers 61 and 63, respectively, were
prepared according to the same protocol as described for
the triple-labeled materials except that mono-13C-labeled
A-ring fragment 25 was used.
The triple-labeled form of the hormone 3, namely 68,
was synthesized from 45 in a manner parallel to that
described by Wilson20 as modified by the Roche group22
(Scheme 6). Allylic oxidation followed by TBDPS protec-
[LiN(TMS)2 or NaN(TMS)2] failed to give cyclized
product, leading only to tosylate cleavage product 32. It
should be noted that the tosylate 37 is relatively stable
in THF, hexane, ethyl acetate and benzene, but in
chloroform it readily produces 32. When 37 was trans-
formed into iodide 38, a highly labile compound, cycliza-
tion by treatment with LiN(TMS)2 at -78 °C with
quenching at -50 °C afforded 86% of the desired [9-13C]-
Grundmann’s ketone (39) and no detectable epimeriza-
tion at C-14. The use of NaN(TMS)2 led to small amounts
of C-14 epimer. Finally, it should be mentioned that
several alternative schemes directed toward preparing
39, including a more direct method reported by DeNin-
no28 using intermediates such as 29, were unsuccessful.
Coupling of the CD-ring fragment 39 with the doubly
labeled A-ring enyne anion 26 gave a nearly quantitative
yield of triple labeled adduct 44. Reduction of propargyl
alcohol 44 with a 1:1 mixture of lithium aluminum
hydride and sodium methoxide resulted in formation of
trans-∆6,7 olefin 50 in 84% yield. In the absence of sodium
methoxide, lower yields were obtained. Acid-catalyzed
solvolysis of 5020,22 provided the desired triply 13C labeled
vitamin D3 56 as well as its 5,6-trans isomer 57. The
latter could be transformed into the former by photosen-
(29) (a) Kiegiel, J .; Wovkulich, P. M.; Uskokovic, M. R. Chemical
Conversion of Vitamin D3 to Its 1, 25-Dihydroxy Metabolite. Tetrahe-
dron Lett. 1991, 32, 6057-6060. (b) Sicinski, R. R.; DeLuca, H. F.
Ruthenium Tetroxide Oxidation of Grundmann’s Ketone Derived from
Vitamin D3. Bioorg. Med. Chem. Lett. 1995, 5, 159-162.
(30) (a) Bovicelli, P.; Lupattelli, P.; Mincione, E. Oxidation of Natural
Targets by Dioxiranes. 2. Direct Hydroxylation at the Side-Chain C-25
of Cholestane Derivatives and of Vitamin D3 Windaus-Grundmann
Ketone. J . Org. Chem. 1992, 57, 5052-5054. (b) Curci, R.; Detomaso,
A.; Prencipe, T.; Carpenter, G. B. Oxidation of Natural Targets by
Dioxiranes. 3. Stereoselective Synthesis of (all-R)-Vitamin D3 Tri-
epoxide and of Its 25-Hydroxy Derivative. J . Am. Chem. Soc. 1994,
116, 8112-8115. (c) Rossella, M.; Fiorentino, M.; Fusco, C.; Curci, R.
Oxidations by Methyl(trifluoromethyl)dioxirane. Oxyfunctionalization
of Saturated Hydrocarbons. 2. J . Am. Chem. Soc. 1989, 111, 6749-
6757.
(28) (a) DeNinno, M. P. “Anomalous” Ozonolysis of Cyclic Allylic
Alcohols: Mechanism and Synthetic Utility. J . Am. Chem. Soc. 1995,
117, 9927-9928. (b) DeNinno, M. P.; McCarthy, K. E. The 14C
Radiolabelled Synthesis of the Cholesterol Absorption Inhibitor CP-
148, 623. A Novel Method for the Incorporation of a 14C Label in
Enones. Tetrahedron 1997, 53, 11007-11020.

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1643
Sch em e 5
Sch em e 6
yield of a 7:3 mixture of 68 and 70. It is noteworthy that
in Wilson’s original report, this same sovolysis carried
out under conditions similar to ours, only 68 was reported
although in higher yields (67%).
NMR Stu d ies of Tr ip le-La beled 25-Hyd r oxyvita -
m in D3 (25-D3). It is well-known from NMR studies that
the A-ring of vitamin D is a dynamic mixture of two
rapidly equilibrating chair conformers, the R- and â-forms
shown in Figure 1. For D3 (1) and 25-D3 (2), the
equilibrium shifts toward the more stable R-chair (73)
with its equatorially oriented hydroxyl as the solvent
polarity increases from cyclohexane (R/â ) 52/48) to
chloroform (R/â ) 58/42) to methanol (R/â ) 76/24).2b,7,9e
These conformational ratios were calculated using a
standard Karplus correlation based on the 3J trans
couplings (J a,a∼11.1 Hz and J e,e∼2.7 Hz)9 between H-3R
and H-4â of 7.1, 7.6 and 9.1 Hz, respectively. The
analogous chair/chair ratio of ∼1:1 for the hormone 1,25-
tion of 45 afforded 65, which was then reduced and then
deprotected to the triol 67. Solvolysis of 67 (to afford 68)
or its protected precursor 66 (to afford 69) proceeded
poorly and somewhat irreproducibly. Under our best
conditions, the triple-labeled hormone 68 was prepared
in ∼40% yield via the procedure described in the Experi-
mental Section. In our hands, only trace amounts of the
5,6-trans isomers (70 or 71) could be detected upon
sovolysis of 67 or 69, respectively. In the Roche report22
(unlabeled series), sovolysis of 67 carried out on a larger
scale and for a prolonged reaction time afforded a 75%
F igu r e 1. Chair conformers of vitamin D. In solution, the
A-ring of vitamin D rapidly equilibrates between the R-chair
(73) with an equatorial 3â-OH and the â-chair (72) possessing
an axial 3â-OH. The figure depicts this equilibrium for 1 (D3,
72a and 73a ), 2 (25-D3, 72b and 73b), and the hormone 1,25-
D3 (3 or 72c and 73c). In each case, the intercyclic diene (C-
5,6,7,8) possesses a near-planar 6-s-trans conformation with
the exocyclic 10(19) double bond twisted out of the C-6,5,10
plane by ∼(50°.

1644 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
3
D3 (3) is less solvent dependent (the J trans couplings
short mixing times. The reference distance (rref) is often
that between a pair of protons whose internuclear
distance is fixed and can be determined fairly precisely
(e.g., that between the geminal protons of a methylene
group). A series of NOESY NMR experiments were
conducted using a modified version of the double half-
filter pulse sequence described by Wider et al.15 In the
double half-filter variant of the standard NOESY experi-
ment, only those cross-peaks of the protons attached to
pairs of labeled carbons (on 13C-13C pairs, namely the
five protons attached to C-7, -9, and -19) appear, thus
eliminating cross-peaks resulting from all other proton
pair combinations (those attached to 12C-12C and 12C-
13C pairs). Thus, background NOESY cross-peaks (due
to protons on natural abundance 13C nuclei) are mini-
mized so that accurate cross-peak intensity measure-
ments can be made at very short mixing times.
ranged from 6.7 to 6.8 Hz for methanol and chloroform),
a not surprising result since its A-ring possesses a trans-
1,3-diol moiety. In the solid state (X-ray), 2 and 3
crystallize in opposite chair forms, the R-(73b) and
â-(72c) chair forms, respectively, but the parent D3 (1)
crystallizes as a 1:1 complex of both chair forms, â-(72a )
and R-(73a ).8 Interestingly, the X-ray structure of the
hormone 1,25-D3 (3) bound to the ligand binding domain
of its nuclear receptor as discussed in the Introduction
has revealed that this hormone, as in the crystal struc-
ture of the free ligand,8c resides as the â-conformer
72c.8a-e It was of some interest therefore that the pro-
hormone 25-D3 (2), according to a conformational analy-
sis of its A-ring (by a standard Karplus correlation based
3
on the J trans coupling of H-3R to H-4â), exists strictly
in the opposite R-conformer 73b when dissolved in SDS-
micelles, a membrane mimic.2b,7 With labeled material
in hand, NMR studies were initiated to illustrate the
feasibility of obtaining more detailed structural informa-
tion regarding ligand in aqueous micellar solutions.
The 13C NMR studies described herein were conducted
on the relatively more accessible triple labeled 25-D3 (58)
and the results follow. The chemical shifts of the C-7, -9,
and -19 signals of 58 in chloroform-d were virtually
invariant (δ 117.5, 29.0, and 112.4 ppm, respectively) at
concentrations between 0.1 and 40 mM. Saturated solu-
tions of 58 in D2O exhibited no detectable 13C NMR
signals, but it was determined that 0.1 mM aqueous
solutions of triple- labeled 58 in a solvent mixture of 60%
methanol-d4 and 40% D2O did exhibit easily detectable
signals at δ 117.9, 29.6, and 112.8 ppm. In light of the
low solubility of cholesterol in water,31 a medium in which
58 would be expected to be only slightly more soluble,
attention was next directed toward the use of solubilizing
agents. Experiments also revealed that 0.1 mM aqueous
solutions of 58 containing dimethylcyclodextrin32 (DMCD)
afforded weak, but acceptable signals at δ 118.0, 29.7,
and 112.6 ppm.
The NOE cross-peaks of proton pairs 9R-9â, 19E-19Z,
7-19Z, and 7-19E were screened at mixing times of 25,
50, 100, 200, 300, 600, and 800 ms to determine the
mixing time that provided optimal volume cross-peaks
for the distance calculations. The NOESY spectrum for
a 50 ms mixing time, optimized for the 9R-9â pair, is
given in Figure 2 and additional details are presented
in the Experimental Section and in the Supporting
Information section. Using the 9R-9â pair as rref [1.778
Å from molecular mechanics optimization (PC Model,
Serena Software) of published X-ray crystallographic
data8b] and the experimental Iref value of 8.39 (from
Figure 2), the 19Z-7 and 19Z-19E distances are calcu-
lated to be 2.2 and 1.9 Å, respectively. The latter 19Z-
19E distance is in excellent agreement with that calcu-
lated from the X-ray crystallographic data (1.896 Å).8b It
should be noted that the absence of an H19E-H7 cross-
peak at short mixing times precluded a reliable distance
measurement of this pair while cross-peaks involving H9R
and H9â, except for the geminal pair, were altogether
absent.
Alternatively, using the 19Z-19E Iref value of 5.66 at
the same 50 ms mixing time and the 19Z-19E calculated
X-ray distance of 1.896 Å (rref), the 9R-9â and 7-19Z
distances are calculated to be 1.8 (calculated 9R-9â X-ray
distance ) 1.778 Å8b) and 2.2 Å, respectively. Finally,
using the initial slopes of the build up curves over the
first 100 ms mixing times rather than the optimal volume
intensities at 50 ms, the corresponding calculated dis-
tances proved to be in excellent agreement (within <0.03
Å) with those given above.
Triple-labeled 25-D3 (58, 0.4 mM) in SDS (sodium
dodecyl sulfate, 40 mM) micelles (δ 117.4, 28.9 and 112.4
ppm) afforded excellent 13C NMR spectra.33 The 5,6-trans
isomer 59 under the latter conditions (0.4 mM in ligand
and 40 mM in SDS) afforded similar results with the
corresponding signals appearing at δ 116.4, 28.8, and
107.4 ppm (in chloroform-d, the corresponding signals
appear at δ 115.8, 29.0, and 108.1 ppm). Thus, these
labeled steroids exhibit only small differences in 13C NMR
chemical shifts of the labeled carbons in various media.
Internuclear distances (rij) within a molecule can be
evaluated from the NOE intensity (Iij) between two nuclei
Thus, the double half-filter NOESY data for 58 (2 mM)
in 20 mM SDS micellar solutions, which was evaluated
in four different ways using two different reference
distances (9R-9â or 19Z-19E), leads to what we believe
is a reliable 19Z-7 H-H distance of 2.2 Å. This distance
is shorter than the calculated X-ray distance of 2.54 Å.
The published X-ray crystallographic data8b for 25-D3
(58) reveals torsion angles of -176.9 and -56.7° for
C-5,6,7,8 and C-6,5,10,19, respectively. PC Model energy
minimization of the X-ray structure leads to torsion
angles of +169.2 and -48.2° for these same rotatable
i and j using the equation rij ) rref(Iref/Iij)1/6 34
. The precision
of the distance measurement is strongly dependent on
the reliability of the intensity measurements (Iij and Iref),
best achieved by carrying out the NOESY experiment at
(31) Haberland, M. E.; Reynolds, J . A. Self-Association of Cholesterol
in Aqueous Solution. Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 2313-
2316.
(32) Schneider, H.-J .; Hacket, F.; Rudiger, V.; Ikeda, H. NMR
Studies of Cyclodextrins and Cyclodextrin Complexes. Chem. Rev.
1998, 98, 1755-1785.
(33) (a) The CMC (critical micelle concentration) for SDS is quoted
as being in the range 8.0-8.4 mM. See Mukerjee, P.; Mysels, K. J .
Critical Micelle Concentrations of Aqueous Surfactant Systems.
NSRDS-NBS 36; U. S. Government Printing Office: Washington D.
C., 1977; 222 pp. (b) For leading references, see also: Mittal, K. L.;
Fendler, E. J ., Eds.; Solution Behavior of SurfactantssTheoretical and
Applied Aspects; Plenum Press: New York, 1982; 739 pp.
(34) (a) Oschkinat, H.; Muller, T.; Dieckmann, T. Protein Structure
Determination with Three- and Four-Dimensional NMR Spectroscopy.
Angew. Chem., Int. Ed. Engl. 1994, 33, 277-293. (b) Clore, G. M.;
Gronenborn, A. M. Assessment of Errors Involved in the Determination
of Interproton Distance Ratios and Distances by Means of One- and
Two-Dimensional NOE Measurements. J . Magn. Reson. 1985, 61, 158-
164.

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1645
This same 3J coupling is observed for the steroid without
the side-chain hydroxyl (namely 73a ), so presumably the
equatorial chair cyclohexanol portion of the steroid is
oriented with the polar hydroxyl oriented toward the
outer, polar surface of the micellar environment. At this
juncture, it is unclear whether the micellar environment
mechanically suppresses the chair-chair equilibrium
(Figure 1) so that the steroid A-ring is locked into only
the R-chair conformation 73b. Alternatively, the A-ring
may be rapidly equilibrating between both chair forms
(72b and 73b) with a strong bias toward the R-chair. As
indicated earlier, solvents of increasing polarity, espe-
cially those with hydrogen bonding ability, increases the
proportion of the equatorial hydroxyl conformer (73b) at
the expense of the axial form (72b). This question
remains unresolved.
In summary, the synthetic sequence developed for the
target isotopically labeled compounds is reasonably prac-
tical. The addition of isotopically enriched Grignard
reagent to an enone, oxidative cleavage and enolate
cyclization contributes a general procedure for C-9 isotope
labeling of the CD fragment. The requirement for *CBr4,
an expensive reagent for introducing label at C-7, is a
shortcoming in the synthesis. Some improvement utiliz-
ing a cheaper reagent would be desirable here. Using the
Wuthrich-type double half-filter pulse sequence, the
NOESY NMR experiment provided new steroid solution
structural information of a ligand associated with a
biological host system, namely in an SDS micellar
system, for the first time. The recent finding that LDL
(low-density lipoprotein) is a delivery system for certain
vitamin D metabolites renders the results described
herein of particular interest.35 Moreover, the micellar
constraining of the 25-D3 (2) molecule has recently had
practical value in studies of the photoisomerization of the
parent D3 (1).7 It remains for future studies to demon-
strate the use of the labeled ligands in biological host
systems, particularly proteins.
F igu r e 2. Double half-filter NOESY NMR spectrum of
[7,9,19-13C3]-25-hydroxyvitamin D3 (58) in sodium dodecyl
sulfate (SDS) micelles. A typical two-dimensional double half-
filter NOESY spectrum is shown. Using a modified version of
the pulse sequence reported by Wider et al.,15 this spectrum
was recorded at the optimum mixing time of 50 ms for a 2
mM sample of 58 in 20 mM SDS-d25 in D2O at 25 °C. The
bracketed H-H cross-peaks are for 9R-9â (1), 19Z-7 (2)
19E-7 (3) and 19Z-19E (4) pairs; the 19E-7 cross-peak (3),
which does not appear in this spectrum, emerges only at longer
mixing times. The data for the other mixing times (25, 100,
200, 300, 600, and 800 ms) as well as distance calculations
are summarized in the Supporting Information. The latter also
1
contains the one-dimensional H NMR spectrum of the same
sample used in this NOESY experiment. Additional details can
be found in the text. In another experiment, a sample for
recording the 13C NMR spectrum was prepared from 7,9,19-
(13C)3-25-D3 (58, 0.4 mg) and SDS (29 mg) in D2O (2.5 mL) in
a 10 mm NMR tube (26 °C, GN-500 spectrometer, 8,000 scans
and LB ) 3 Hz). Chemical shift data for this sample as well
as for 58 in other media are discussed in the text.
Exp er im en ta l Section
Gen er a l Meth od s. Spectral data are given in the Support-
ing Information unless otherwise noted. The purity of all major
isomers or intermediates was estimated by a combination of
HPLC and NMR analysis. The level of purity is indicated by
the inclusion of copies of 1H NMR spectra and 13C NMR spectra
in the Supporting Information.
bonds, including the19Z-7 H-H distance of 2.54 Å
mentioned above. It is apparent that simply manually
changing the C-5,6,7,8 angle from +169.2 to -165° can
easily account for the shortened 19Z-7 H-H distance
of 2.2 Å determined from the NOESY experiment. On
the other hand, constraint of the 19Z-7 H-H distance
to 2.2 Å followed by energy minimization leads to torsion
angles of +177.0 and -44.6°, respectively, of the C-5,6,7,8
and C-6,5,10,19 bonds. Thus, rotation of either of the
latter torsion angles in an appropriate manner can offer
a simple explanation for the shortening of the H7-H19Z
distance. The UV maximum of the micellar solutions of
25-D3 reveals an absorption at 265 nm, essentially the
same as that of the free ligand in methanol or hexane.
Thus, we cannot detect any unusual flattening or twisting
of the conjugated triene portion of the molecule from its
Tr a n sfor m a tion of Vita m in D3 (1) to Cyclovita m in 13.24
To a solution of vitamin D3 (50 g, 0.13 mol) in pyridine (300
mL) were added p-toluenesulfonyl chloride (recrystallized from
hexane, 49.5 g, 0.26 mol) and DMAP (2.0 g) at 0 °C under
argon. The reaction mixture was stirred at 0 °C for 1 h and at
room temperature for 15 h. An additional portion of p-
toluenesulfonyl chloride (10 g) was added at room temperature.
After being stirred for 8 h, the reaction mixture was poured
into a mixture of saturated aqueous NaHCO3 solution and ice.
The mixture was stirred at room temperature for 30 min to
quench the excess p-toluenesulfonyl chloride and then ex-
tracted with ether. The combined ether solution was washed
(35) Teramoto, T.; Endo, K.; Ikeda, K.; Kubodera, N.; Kinoshita, M.;
Yamanaka, M.; Ogata, E. Binding of Vitamin D to Low-Density-
Lipoprotein (LDL) and LDL Receptor-Mediated Pathway Into Cells.
Biochem. Biophys. Res. Commun. 1995, 215, 199-204.
(36) Baggiolini, E. G.; Iacobelli, J . A.; Hennessy, B. M.; Batcho, A.
D.; Sereno, J . F.; Uskokovic, M. R. Stereocontrolled Total Synthesis of
1R,25-Dihydroxycholecalciferol and 1R,25-Dihydroxyergocalciferol. J .
Org. Chem. 1986, 51, 3098-3108.
(37) VanAlstyne, E. M.; Norman, A. W.; Okamura, W. H. 7,8-Cis
Geometric Isomers of the Steroid Hormone, 1R,25-Dihydroxyvitamin
D3. J . Am. Chem. Soc. 1994, 116, 6207-6216.
3
electronic spectrum. The J coupling for 25D3 between
H6 and H7 remains at ∼11.1 Hz both in micelles and in
chloroform, suggesting that the C-5,6,7,8 diene portion
of the triene is nearly planar.
Finally, as mentioned earlier, the 3J coupling between
the 3R and 4â protons of ligand in micellar solutions is
characteristic of the presence of only the R-chair 73b.

1646 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
with 5% HCl and water and dried over anhydrous magnesium
sulfate. Flash chromatography of the crude material (10%
EtOAc in hexane) afforded cholecalciferyl tosylate (68.4 g,
98%)21 as a white solid, which was used directly in the
solvolysis reaction described next. To a solution of the tosylate
(34.2 g, 0.063 mol) in acetone (2.8 L) and water (280 mL) was
added powdered sodium bicarbonate (350 g) at room temper-
ature. The reaction mixture was refluxed with vigorous
mechanical stirring for 28 h and then was filtered. The filtrate
was concentrated under vacuum to about 300-500 mL. The
residual mixture was extracted with ether. The white solid
from the filtration was dissolved in water and extracted with
ether. The combined ether solution was washed with water
and dried over anhydrous magnesium sulfate. The crude
material was separated by flash chromatography (4% EtOAc
in hexane) to give the cyclovitamin 13 (15.0 g, 62%) as a
colorless oil.
(3R,5S,6S,7R,8R,10S)-3,5-Did eh yd r o-6,7,8,10,19-p en -
ta h yd r oxy-9,10-secoch olesta n e (14). To a solution of cy-
clovitamin 13 (19.8 g, 51.5 mmol) in a mixture of acetone (408
mL), pyridine (20.4 mL), and water (51 mL) was added osmium
tetraoxide (2.0 g, 7.8 mmol) and NMO (24.2 g, 0.206 mol) at
room temperature. The brown solution was purged with argon
and was then refluxed with stirring for 3.5 h. The black
reaction mixture was filtered through a short column of silica
gel covered with Celite (acetone). The filtrate was concentrated
under vacuum, and then a saturated aqueous NaCl solution
was added to the black residue. The mixture was extracted
with ether (4 × 400 mL), and then the combined ether solution
was dried over anhydrous MgSO4 and filtered through a short
silica gel column (EtOAc). Concentration of the filtrate afforded
pentaol 14 (light yellow foam, 18.7 g, 80%), which was used
directly in the following reaction. A sample for spectral
analysis was obtained by further purification by HPLC (silica
gel, 1:20 CH3OH/CH2Cl2).
chromatography (1:4 EtOAc/hexane) to give 16 (790 mg,
∼100%, colorless oil).
(1S,5R)-1-(2′,2′-Dibr om ovin yl)-2-m eth ylen ebicyclo[3.1.0]-
h exan e (18) an d (1R,5R)-1-(2′-Br om oeth yn yl)-2-m eth ylen e-
b icyclo[3.1.0]h exa n e (19). To a suspension of methyltri-
phenylphosphonium iodide (978 mg, 2.42 mmol) in THF (8 mL)
was added solid potassium tert-butoxide (271 mg, 2.42 mmol)
at room temperature. The yellow reaction mixture was stirred
at room temperature for 1 h, and a solution of 16 (340 mg,
1.21 mmol) in THF (6 mL) was then added. The reaction
mixture was stirred at room temperature for 2 h and was
quenched by addition of water. The mixture was extracted with
hexane (2 × 100 mL), and then the combined organic phase
was dried over MgSO4 and concentrated under vacuum. The
residue was dissolved in CH2Cl2 (10 mL), and pentane (200
mL) was introduced slowly. The precipitates were removed by
filtration through a short column (silica gel covered with Celite,
pentane). The filtrate was concentrated under vacuum, and
the residual liquid was separated by flash chromatography
(silica gel, hexane) (263 mg, colorless oil). This material was
normally used directly for the following coupling reaction
without further separation. Separation of the material by
HPLC (silica gel, hexane) afforded the less polar dibromide A
(18, 192.1 mg, 56%, colorless liquid) and the more polar
bromoenyne B (19, 41.1 mg, 17%, colorless liquid). Treatment
of either component A or B or a mixture with BuLi followed
by quenching gave the volatile enyne 27 as the only product
1
as determined by H NMR analysis.
13C-La beled A-Rin g F r a gm en ts 20-23. The monolabeled
A-ring enyne for introduction of 13C label at the C19 position
of the vitamin D metabolite was prepared from dibromoketone
16 (790 mg, 2.82 mmol in 8 mL of THF), [13CH3]-methyltri-
phenylphosphonium iodide (1.72 g, 4.23 mmol), and potassium
tert-butoxide (1 M in THF, 4.23 mL, 4.23 mmol) in 20 mL of
THF as described above for the unlabeled material. This
afforded a mixture (524 mg, 66%) of the less polar dibromide
A (20) and the more polar alkynyl bromide B (21) in a 4:1 ratio
(1H NMR analysis), which was used directly in the next step
as indicated at the end of this section.
(1S,5R)-1-For m ylbicyclo[3.1.0]h exan -2-on e (11), Gr u n d-
m a n n ’s Keton e (12), a n d 8â-F or m yl-d e-A,B-ch olesta n -8r-
ol (15). To a solution of pentaol 14 (10.41 g, 23 mmol) in
dioxane (162 mL) and water (54 mL) was added sodium
periodate (25.0 g, 116 mmol) at room temperature. The milky
white reaction mixture was stirred vigorously at room tem-
perature for 2.5 h and was then quenched by addition of
saturated aqueous NaCl solution. The mixture was extracted
with ether (4 × 500 mL), and then the combined ether solution
was dried over MgSO4. The solids were filtered off, and the
filtrate was concentrated carefully by fractional distillation
under a slight vacuum. The crude residue was separated by
flash chromatography (silica gel, 1:3 ether/pentane) to give in
order of elution Grundmann’s ketone 12 (6.0 g) and then keto
aldehyde 11 (2.8 g). The former was further purified by flash
chromatography (silica gel, 1:20 EtOAc/hexane) to give pure
12 (4.86 g, 90%, oil), and the latter was further purified by
HPLC (2:1 ether/hexane) to give 11 (2.0 g, 70%, oil). Small
amounts of residual solvent signals were invariably present
The double-labeled A-ring enyne for introduction of 13C label
at both the C7 and C19 positions of the vitamin D metabolite
was prepared as described for the unlabeled compound starting
from keto aldehyde 11 (300 mg, 2.42 mmol) in 8 mL of CH2Cl2
and labeled dibromo-Wittig reagent [prepared from [13C]-CBr4
(2.61 g, 7.89 mmol), triphenylphosphine (2.06 g, 7.89 mmol),
and Zn dust (511 mg, 7.89 mmol)]. There was obtained labeled
dibromoketone 17 (635 mg, 94%), which was used in the next
step without further examination. The Wittig reaction of the
crude labeled dibromoketone 17 (635 mg, 2.26 mmol) with
[13CH3]-methyltriphenylphosphonium iodide (1.83 g, 4.52 mmol)
and potassium tert-butoxide (507 mg, 4.52 mmol) in 25 mL of
THF as described above for the unlabeled material afforded a
mixture (442 mg, 70%) of the less polar dibromide A (22) and
the more polar alkynyl bromide B (23) in a 4:1 ratio (isolated
yield). The latter mixture was used directly in the next step
without further purification. Analytical samples of double
labeled materials were prepared by HPLC separation (silica
gel, hexane).
No attempts were made to isolate and characterize the
labeled forms of hydrocarbon 27 because of volatility problems
encountered in handling 27 itself. The crude lithium salts 25
and 26 were used directly. For preparative purposes, crude
dry mixtures of 20 plus 21 or 22 plus 23 were transformed
directly to 25 and 26, respectively. A detailed analysis of NMR
data for separated and purified 22 and 23 (with a side by side
comparison to unlabeled materials 18 and 19, respectively),
but not for 20 and 21, is presented in the Supporting
Information.
1
in H and 13C NMR spectra of the keto aldehyde, but consider-
able loss of material due to its instability was encountered in
attempts at further purification. When reactions were carried
out at shorter reaction times, variable amounts (20-50%) of
the 8â-hydroxy-8R-carbaldehyde 15 (which eluted between
ketone 12 and keto aldehyde 11) were also isolated.
(1S,5R)-1-(2′,2′-Dib r om ovin yl)b icyclo[3.1.0]h exa n -2-
on e (16). To a suspension of zinc dust (780 mg, 12 mmol) and
triphenylphosphine (3.14 g, 12 mmol) in dry CH2Cl2 (15 mL)
was added a solution of carbon tetrabromide (3.98 g, 12 mmol)
in dry CH2Cl2 (7 mL) via cannula at room temperature. The
resulting yellow suspension was stirred at room temperature
for 22 h. The keto aldehyde 11 (350 mg, 2.82 mmol) in CH2Cl2
(6 mL) was then added to the red reaction mixture via cannula
at 0 °C. The reaction mixture was stirred at 0 °C for 1.5 h.
Pentane (200 mL) was slowly added to the reaction mixture
with stirring, and then the mixture was filtered through Celite.
The solid was dissolved in CH2Cl2 (30 mL) and was then
precipitated with pentane (200 mL). The solids were filtered
off through Celite. The combined filtrate was concentrated
under vacuum, and the residual oil was separated by flash
De-A,B-ch olest-9(11)-en -8-on e (28). This material was
prepared exactly as previously described by this laboratory.27
In a typical sequence, reaction of Grundmann’s ketone (12,
5.30 g, 18.9 mmol) with LDA [from 2.49 g (24.6 mmol) of
diisopropylamine and 24.6 mmol of n-BuLi] in THF (100 mL)
followed by PhSeCl (4.95 g, 25.8 mmol) treatment and then

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1647
m-CPBA oxidation [10.4 g (49.2 mmol) in CH2Cl2 (60 mL)]
afforded 3.18 g (60%) of enone 28.
gel) to give a minor nonpolar product 36 (1.45 g, 32%) and
hydroxy tosylate 35 (4.49 g, 60%). The analytical samples of
ether 36 and hydroxy-tosylate 35 were prepared by further
HPLC purification (silica gel, 2.5% EtOAc in hexane and 1:2
EtOAc/hexane, respectively).
[8R-13C]-9,11-Did eh yd r o-8r-m et h yl-d e-A,B-ch olest a n -
8â-ol (29). To magnesium turnings (360 mg, 15 mmol) covered
with ether (3 mL) was added a solution of [13C]-iodomethane
(1.95 g, 13.6 mmol, 99% isotopic purity, Cambridge Isotope
Laboratory) in ether (8 mL) at a rate slow enough to maintain
a gentle reflux of the reaction mixture. After the addition, the
reaction mixture was refluxed for 30 min. The Grignard
reagent was then transferred via cannula to a suspension of
CeCl3 (dried at 160 °C for 5 h in vacuo, 6.65 g, 27 mmol) in
THF (45 mL) at 0 °C. The slurry was stirred at 0 °C for 1 h,
and then a solution of enone 28 (1.42 g, 5.40 mmol) in THF (3
mL) was added. The reaction mixture was then stirred at 0
°C for 2 h and at room temperature for 1 h. Ether (100 mL)
was added to the reaction mixture followed by saturated NH4-
Cl. The separated aqueous phase was extracted with ether (2
× 150 mL) and then the combined ether solution was dried
over MgSO4 and concentrated in vacuo to give 29 (1.43 g, 94%)
as a colorless oil, sufficiently pure for the following reaction.
The analytical sample was prepared by further purification
by HPLC (1:10 EtOAc/hexane), but the product is unstable to
silica gel column chromatography.
[9-13C]-9,11-Seco-11-t osyloxy-d e-A,B-ch olest a n -8-on e
(37). To a solution of hydroxy-tosylate 35 (980 mg, 2.23 mmol)
in CH2Cl2 (30 mL) was added pyridinium chlorochromate (960
mg, 4.46 mmol) at room temperature. The dark brown mixture
was stirred at room temperature for 15 h and was then directly
loaded onto a short column and eluted (silica gel, 1:4.5 EtOAc/
hexane). Further purification by HPLC (1:4 EtOAc/hexane)
afforded ketone 37 as a liquid (860 mg, 88%).
[9-13C]-11-Iod o-9,11-seco-d e-A,B-ch olesta n -8-on e (38).
To a solution of keto-tosylate 37 (458 mg, 1.05 mmol) in THF
(20 mL) was added lithium iodide (421 mg, 3.14 mmol) at room
temperature. The reaction mixture was heated at 50 °C with
stirring for 2 h. Water (100 mL) was added and then the
mixture was extracted with ether (2 × 150 mL). The combined
ether solution was dried over anhydrous MgSO4 and concen-
trated under vacuum. The residual oil was separated by HPLC
(silica gel, 1:20 EtOAc/hexane) to give iodoketone 38 (319 mg,
78%) as a slightly yellow liquid, which was used immediately
without purification in the next step because of its instability
even at room temperature. Its crude 1H NMR spectrum was
consistent with the assigned structure.
[8r-13C]-9,11-Seco-8r-m eth yl-de-A,B-ch olestan e-8â,9,11-
tr iol (31). A solution of allylic alcohol 29 (1.42 g, 5.09 mmol)
dissolved in methanol (250 mL) and pyridine (8 mL) was
purged with nitrogen through a gas dispersion tube. The
solution cooled to -78 °C was then purged with ozone until
the solution turned blue (1.5-2 h). The ozone flow was stopped
and the reaction mixture was purged with N2 at -78 °C for 1
h. Sodium borohydride (2 g) was added to the solution at -78
°C in one portion and the resulting mixture was stirred at -78
°C for 1 h while a gentle flow of N2 was maintained. After
addition of another 2 g of NaBH4, the cooling bath was
removed. The mixture was stirred at room temperature for 1
h followed by slow addition of 2 g of NaBH4. After another 1
h stirring, a final portion of NaBH4 (2 g) was added and the
reaction mixture was allowed to stir at room temperature for
24 h. The mixture was concentrated in vacuo, saturated NaCl
solution was added to the remaining residue, and the mixture
was extracted with ether (4 × 150 mL). The combined ether
solution was dried over MgSO4 and concentrated in vacuo. The
crude product was separated by flash chromatography and
then subjected to HPLC (1:10 CH3OH/CH2Cl2) to give pure 31
(1.25 g, 78%).
[9-13C]-De-A,B-ch olesta n -8-on e (39). To a solution of LiN-
(TMS)2 (LHMDS, 1 M solution in THF, 1.12 mL, 1.12 mmol)
in THF (2 mL) was added a solution of iodoketone 38 (220
mg, 0.56 mmol) in THF (2 mL) by cannula at -78 °C. The
reaction mixture was stirred at -78 °C for 1 h at which time
ether (10 mL) was added. After stirring for 5 min at -78 °C,
H2O (5 mL) was then added to the mixture and the latter was
removed from the cold bath. The ether was decanted and the
remaining ice slush was washed with ether (1 × 10 mL). The
combined ether solutions were dried over MgSO4, filtered, and
concentrated. The residual oil was separated by flash chro-
matography (1:10 EtOAc/hexane) and was further purified by
HPLC (1:20 EtOAc/hexane) to give 9-13C-Grundmann’s ketone
39 (127 mg, 86%, colorless oil). Similar reaction conditions
using NaN(TMS)2 (NHMDS) resulted in a reduced yield of the
desired [9-13C]-Grundmann’s ketone 39 with varying amounts
of the C14 epimer. LDA was even less effective.
[9-13C]-25-Hyd r oxy-d e-A,B-ch olesta n -8-on e (42). To a
solution of [13C]-Grundmann’s ketone (39,128 mg, 0.483 mmol)
in carbon tetrachloride (3.6 mL) and acetonitrile (3.6 mL) in a
25 mL round-bottom flask was added a solution of a 1/1
mixture of 0.5 M KH2PO4/0.5 M NaOH (4.8 mL) and sodium
periodate (368 mg, 1.72 mmol). The mixture was stirred at
room temperature for 5 min, and then RuCl3-xH2O (12 mg)
was added.29 The resulting brown solution was purged with
argon, and after 10 min, the reaction mixture became yellow.
After the mixture was stirred under argon at 50 °C for 52 h,
the reaction was quenched by the addition of saturated
aqueous NaCl solution (30 mL). The mixture was extracted
with ether (3 × 120 mL), and then the combined ether solution
was dried over MgSO4 and concentrated. The crude residue
was separated by flash chromatography and was further
purified by HPLC (silica gel, 1:2 EtOAc/hexane) to give 42 (55
mg, 41%; 54% based on recovered starting material) and
starting material 39 (31 mg).
In a second more efficient procedure, a distilled solution of
methyl(trifluoromethyl)dioxirane (MTD) in 1,1,1-trifluoropro-
panone (TFP) was prepared according to the method of
Curci30a-c by oxidation of a solution of TFP (10 mL, 13.3 g,
119 mmol), NaHCO3 (10 g, 119 mmol), ethylenediaminetet-
raacetic acid-disodium salt dihydrate (0.33 g, 1 mmol, Ald-
rich), and doubly distilled water (15 mL) with potasssium
peroxomonosulfate (oxone, 16.7 g, 0.054 mmol). To the solution
of MTD at 0 °C was added a solution of labeled Grundmann’s
ketone 39 (150 mg, 0.56 mmol) in 2 mL of dry CH2Cl2 and
then the solution was stirred at this same temperature for 1
h. The reaction was quenched by addition of 10 mL of H2O
followed by extraction with ether (2 × 15 mL). The combined
organic layers were dried with MgSO4, filtered, and concen-
[9-13C]-9,11-Seco-d e-A,B-ch olesta n e-8,11-d iols (33 a n d
34). To a solution of 31 (1.20 g, 3.8 mmol) in a mixture of
dioxane (30 mL) and water (15 mL) was added sodium
periodate (1.63 g, 7.6 mmol) at room temperature. The reaction
mixture was stirred at room temperature for 3 h and was then
cooled to 0 °C. Sodium borohydride (2.0 g) was added slowly,
and then after the mixture was stirred at 0 °C for 1 h, the ice
bath was removed. The mixture was allowed to stir at room
temperature overnight, saturated aqueous NaCl solution was
added to the mixture, and then the latter was extracted with
ether. The combined ether solution was dried over MgSO4 and
concentrated in vacuo. The residual oil was purified by flash
chromatography (silica gel, 1:2 EtOAc/hexane) to give an
epimeric mixture of diols (1.11 g, ∼quantitative). This mixture
was used directly in the following reaction without further
separation. Analytical samples were prepared by HPLC sepa-
ration (silica gel, 1:1.2 EtOAc/hexanes): A (less polar, minor
isomer 34) and B (more polar, major isomer 33) in a ∼1:2 ratio.
[9-13C]-11-Tosyloxy-9,11-seco-de-A,B-ch olestan -8r-ol (35)
a n d [8-13C]- 8r-Meth yl-9-oxa -d e-A,B-ch olesta n e (36). To
a solution of the epimeric mixture of diols 34 and 33 (4.87 g,
17.1 mmol; ∼1:2 mixture of minor, less polar diol A and major,
more polar diol B) in pyridine (50 mL) was added p-toluene-
sulfonyl chloride (3.91 g, 20.5 mmol) at -40 °C. The reaction
mixture was stirred at -25 °C for 15 h and then at -20 °C for
6 h. Water (250 mL) was added at -20 °C, and then the
mixture at ambient temperature was extracted with ether (4
× 300 mL). The combined ether solution was dried over MgSO4
and filtered through a short column (silica gel, ether). The
crude product was separated by flash chromatography (silica

1648 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
trated under reduced pressure. The resulting crude oil was
separated by flash chromatography (silica gel, 1:2, EtOAc/
hexane) and further purified by HPLC (1:2 EtOAc/hexane) to
give 42 (111 mg, 0.40 mmol, 71%).
addition of water, the mixture was extracted with ether (2 ×
100 mL). The combined ether solution was dried over MgSO4
and concentrated under vacuum. The crude product was
separated by flash chromatography (1:20 EtOAc/hexane) and
was further purified by HPLC (1:20 EtOAc/hexane) to afford
47 (76.1 mg, 89%, oil).
[7,9,19-13C3]-(3R,5R,8R)-3,5-Did eh yd r o-8-h yd r oxy-9,10-
secoch olest-10(19)-en -6-yn e (44). To a solution of a mixture
of the double-labeled dibromide 22 and the alkynyl bromide
23 (160 mg, 0.57 mmol) in THF (8 mL) was added BuLi (2.5
M solution in hexane, 460 µL, 1.14 mmol) at -78 °C. The
solution was stirred at -78 °C for 0.5 h, and the cooling bath
was removed. The solution of 26 was stirred at room temper-
ature for 1 h and was then again cooled to -78 °C. A solution
of ketone 39 (100.7 mg, 0.38 mmol) in THF (4 mL) was then
added via cannula, and the reaction mixture was stirred at
-78 °C for 0.5 h and at room temperature for 0.5 h. After
addition of water, the mixture was extracted with ether (2 ×
100 mL). The combined ether solution was dried over MgSO4
and concentrated under vacuum. The crude product was
separated by flash chromatography (1:25 EtoAc/hexane) and
was further purified by HPLC (1:25 EtOAc/hexane) to afford
triple-labeled 44 (142.4 mg, 97%, oil).
[7,9,19-13C3]-(3R,5R,8R)-(6E)-3,5-Did eh yd r o-9,10-seco-
ch olesta -6,10(19)-d ien -8-ol (50). To a solution of propargyl
alcohol 44 (140 mg, 0.364 mmol) in THF (18 mL) was added
LiAlH4 (50 mg) and CH3ONa (50 mg) successively at room
temperature. The reaction mixture was refluxed for 1 h and
was then quenched at room temperature by addition of 0.1 M
aqueous HCl solution. The mixture was extracted with 1:1
ether/hexane, and then the combined organic phase was dried
over MgSO4 and concentrated under vacuum. The residual oil
was separated by flash chromatography (1:40 EtOAc/hexane)
and was further purified by HPLC (1:40 EtOAc/hexane) to give
50 (118 mg, 84%, oil).
[7,9,19-13C3]-(3R,5R,8R)-(6E)-3,5-Did eh yd r o-8,25-d ih y-
d r oxy-9,10-secoch olesta -6,10(19)-d ien e (51). To a solution
of propargyl alcohol 45 (160 mg, 0.34 mmol) in THF (18 mL)
were added LiAlH4 (50 mg) and CH3ONa (50 mg) successively
with stirring at room temperature. The reaction mixture was
refluxed for 1 h and was then quenched at room temperature
by addition of saturated aqueous NH4Cl solution. The mixture
was extracted with 1:1 ether/hexane. The combined organic
phase was dried over MgSO4 and concentrated under vacuum.
The residual oil was separated by flash chromatography (1:
20 EtOAc/hexane and then 1:4 EtOAc/hexane) to give the less
polar silyl ether of 51 (from 1:20 EtOAc/hexane elution) and
the more polar trimethylsilyl cleavage compound 51 (from 1:4
EtOAc/hexane elution). The former was dissolved in CH2Cl2
(2 mL), and pyridinium p-toluenesulfonate (20 mg) was added
to the solution. The solution was stirred at room temperature
for 5 h and was directly filtered through a column (silica gel,
1:4 EtOAc/hexane). The resulting alcohol was combined with
the earlier polar trimethylsilyl cleavage compound 51. The
combined alcohol fractions were further purified by HPLC (1:3
EtOAc/hexane) to afford a colorless oil (109.8 mg, 80%).
[9,19-13C2]-(3R,5R,8R)-(6E )-3,5-Did e h yd r o-9,10-se co-
ch olesta -6,10(19)-d ien -8-ol (52). To a solution of propargyl
alcohol 46 (73 mg, 0.19 mmol) in THF (10 mL) were added
LiAlH4 (25 mg) and CH3ONa (25 mg) successively at room
temperature. The reaction mixture was refluxed for 1 h and
was then quenched at room temperature by addition of 0.1 M
aqueous HCl solution. The mixture was extracted with 1:1
ether/hexane, and then the combined organic phase was dried
over MgSO4 and concentrated under vacuum. The residual oil
was separated by flash chromatography (1:20 EtOAc/hexane)
and was further purified by HPLC (1:50 EtOAc/hexane) to give
52 (64.9 mg, 88%, oil).
[7,9,19-13C3]-(3R,5R,8R)-3,5-Did eh yd r o-8-h yd r oxy-9,10-
seco-25-(tr im eth ylsilyloxy)ch olest-10(19)-en -6-yn e (45)
fr om 42 via 43 a n d 22-23. To a solution of hydroxyketone
42 (125.3 mg, 0.445 mmol) in CH2Cl2 (15 mL) was added
trimethylsilylimidazole (287 mg, 2.0 mmol) at room temper-
ature. The reaction mixture was stirred at room temperature
for 30 h and was directly subjected to a column chromatog-
raphy (silica gel, 1:10 EtOAc/hexane). The crude product was
further purified by HPLC (1:12 EtOAc/hexane) to give 43
(142.3 mg, 90%, colorless oil). The unlabeled 43 has been
previously reported.36,37
To a solution of a mixture of the dibromide 22 and the
alkynyl bromide 23 (181 mg, 0.64 mmol) in THF (8 mL) was
added BuLi (2.48 M solution in hexane, 520 µL, 1.28 mmol)
at -78 °C. The solution of 26 was stirred at -78 °C for 0.5 h,
and then the cooling bath was removed. The solution was
stirred at room temperature for 1 h and was then again cooled
to -78 °C. A solution of ketone 43 (134 mg, 0.38 mmol) in THF
(4 mL) was added via cannula. The reaction mixture was
stirred at -78 °C for 0.5 h and at room temperature for 0.5 h.
After addition of water, the mixture was extracted with ether
(2 × 100 mL). The combined ether solution was dried over
MgSO4 and concentrated under vacuum. The crude product
was separated by flash chromatography (1:20 EtOAc/hexane)
and was further purified by HPLC (1:20 EtOAc/hexane) to
afford 45 (161.9 mg, 90%, oil).
[9,19-13C2]-(3R,5R,8R)-3,5-Did eh yd r o-8-h yd r oxy-9,10-
secoch olest-10(19)-en -6-yn e (46). To a solution of a mixture
of the monolabeled dibromide 20 and alkynyl bromide 21 (100
mg, 0.36 mmol) in THF (6 mL) was added n-BuLi (2.45 M
solution in hexane, 290 µL, 0.72 mmol) at -78 °C. The solution
was stirred at -78 °C for 0.5 h, and then the cooling bath was
removed. The resulting solution of 25 was stirred at room
temperature for 1 h and was then again cooled to -78 °C. A
solution of ketone 39 (61.2 mg, 0.23 mmol) in THF (2 mL) was
added via cannula. The reaction mixture was stirred at -78
°C for 0.5 h and at room temperature for 0.5 h. After addition
of water, the mixture was extracted with ether (2 × 100 mL).
The combined ether solution was dried over MgSO4 and
concentrated under vacuum. The crude product was separated
by flash chromatography (1:20 EtOAc/hexane) and was further
purified by HPLC (1:20 EtOAc/hexane) to afford pure 46 (74.8
mg, 84%, oil).
[9,19-13C2]-(3R,5R,8R)-(6E )-3,5-Did eh yd r o-8,25-d ih y-
d r oxy-9,10-secoch olesta -6,10(19)-d ien e (53). The same pro-
cedure as for the preparation of triple-labeled material 51 was
used. Starting from 47 (75 mg, 0.16 mmol), 59.4 mg (92%) of
53 was obtained after the deprotection of the intermediate 25-
trimethylsilyloxy derivative of 53.
[7,9,19-13C3]-Vita m in D3 (56) a n d [7,9,19-13C3]-5,6-tr a n s-
Vita m in D3 (57). To a solution of 50 (117 mg, 0.30 mmol) in
dioxane (15 mL) and H2O (5 mL) was added p-toluenesulfonic
acid monohydrate (20 mg) at room temperature. The solution
was purged with argon then heated at 60 °C with stirring for
3 h. After cooling and the introduction of 1:1 ether/hexane (150
mL), the solution was washed with saturated aqueous NaH-
CO3 solution. The aqueous phase was extracted with 1:1 ether/
hexane (2 × 150 mL), and the combined organic phase was
dried over MgSO4 and concentrated under vacuum. The
residual material was separated by HPLC (1:5 EtOAc/hex-
ane): the first major fraction (white foam, 30.2 mg, 26%) was
characterized as triply 13C-labeled 5,6-trans-vitamin D3 (57);
the second major peak was triply 13C-labeled vitamin D3 (56,
white foam, 39 mg, 34%).
[9,19-13C2]-(3R,5R,8R)-3,5-Did eh yd r o-8-h yd r oxy-9,10-
seco-25-(tr im eth ylsilyloxy)ch olest-10(19)-en -6-yn e (47).
To a solution of a mixture of monolabeled dibromide 20 and
alkynyl bromide 21 (142 mg, 0.51 mmol) in THF (6 mL) was
added n-BuLi (2.45 M solution in hexane, 420 µL, 1.02 mmol)
at -78 °C. The solution was stirred at -78 °C for 0.5 h, and
the cooling bath was removed. The solution of 25 was stirred
at room temperature for 1 h and was then again cooled to -78
°C. A solution of ketone 43 (64 mg, 0.18 mmol) in THF (4 mL)
was added via cannula. The reaction mixture was stirred at
-78 °C for 0.5 h and at room temperature for 1 h. After
[7,9,19-13C3]-25-Hyd r oxyvita m in D3 (58) a n d [7,9,19-
13C3]-25-Hyd r oxy-5,6-tr a n s-vita m in D3 (59). To a solution
of 51 (109 mg, 0.27 mmol) in dioxane (15 mL) and water (5

Synthesis of 13C-Labeled Vitamin D Metabolites
J . Org. Chem., Vol. 67, No. 5, 2002 1649
mL) was added p-toluenesulfonic acid monohydrate (20 mg)
at room temperature. The solution was purged with argon and
was heated with stirring at 60 °C for 3 h. Water (50 mL) was
added, and the mixture was extracted with ether (2 × 100 mL).
The combined ether solution was dried over MgSO4 and
concentrated under vacuum. The residual material was sepa-
rated by HPLC (silica gel, 1:1.5 EtOAc/hexane) to afford two
major fractions: A, less polar (a mixture of triple labeled 25-
hydroxyvitamin D3 (95%) and 25-hydroxyprevitamin D3 (5%);
B, more polar [triple-labeled 5,6-trans-25-hydroxyvitamin D3,
(59, white powder, 23 mg, 21%)]. The vitamin-previtamin
mixture (fraction A) was then further separated by HPLC
(Microsorb Cyano column, 5 µm, 10 × 250 mm, Rainin
Instrument Co., 98:2:0.3 hexane/iPrOH/CH3CN) to afford the
desired labeled 25-hydroxyvitamin 58 (white powder, 37 mg,
34%).
under vacuum. The crude product was purified by HPLC
(Whatman Partisil 10) using 10% EtOAc/hexanes to yield the
pure trans alcohol 66 (4.3 mg, 72%), as a colorless oil.
[7,9,19-13C3]-(1S,3R,5R,8R)-3,5-Did eh yd r o-1,8,25-tr ih y-
d r oxy-9,10-seco-25-ch olesta -6,10(19)-d ien e (67). A solution
of 2.9 mg (0.0039 mmol) of TBDPS protected alcohol 66 in 0.5
mL of THF was stirred under argon at room temperature. To
this solution was added 12 µL of TBAF (1 M in THF, 0.012
mmol), and the solution was stirred in the dark at room
temperature for 16 h. The mixture was then diluted with 5
mL of EtOAc and then washed with water (3 × 5 mL). The
combined organic extracts were then dried over magnesium
sulfate, filtered, and concentrated under vacuum. The crude
product was purified by HPLC using a Whatman partisil 10
column eluting with 90:10 EtOAc/hexanes to afford 1.5 mg
(92% yield) of spectroscopically pure triol 67 as a colorless oil.
[7,9,19-13C3]-1r,25-Dih yd r oxyvita m in D3 (68). A solution
of 0.7 mg (0.0016 mmol) of triple-labeled starting material 67,
0.3 mg of p-TsOH, and 1 mL of dioxane-H2O (2:1) was stirred
at 75 °C for 1 h. The cooled solution was extracted with 5 mL
of EtOAc, and then the combined organic extracts were washed
successively with water (2 × 5 mL) and aqueous NaHCO3.
After drying over magnesium sulfate, the EtOAc solution was
filtered and concentrated under vacuum. The crude product
was purified by HPLC using EtOAc as the eluent on a
Whatman Partisil 10 column to afford spectroscopically pure
triple labeled hormone 68 (0.3 mg) as a colorless oil in 40%
yield.
[9,19-13C2]-Vita m in D3 (60) a n d [9,19-13C2]-5,6-tr a n s-
Vita m in D3 (61). Following the same procedure as for the
preparation of triple-labeled 56, 64 mg (0.165 mmol) of alcohol
52 afforded doubly 13C-labeled vitamin D3 60 (21.5 mg, 34%,
more polar) and 5,6-trans-vitamin 61 (17.4 mg, 27%, less
polar).
[9,19-13C2]-25-Hyd r oxyvita m in D3 (62) a n d [9,19-13C2]-
25-Hyd r oxy-5,6-tr a n s-vita m in D3 (63). Following the same
procedure for the preparation of triple-labeled 58, 18 mg (0.045
mmol) of 53 afforded 7.3 mg of doubly 13C-labeled vitamin D
62 (40%, less polar) and 4.4 mg of 5,6-trans-vitamin 63 (24%,
more polar).
[7,9,19-13C3]-(1S,3R,5R,8R)-3,5-Dideh ydr o-1,8-dih ydr oxy-
9,10-seco-25-(t r im et h ylsilyloxy)ch olest -10(19)-en -6-yn e
(64). To a suspension of 7.1 mg (0.064 mmol) of selenium
dioxide in 5 mL of dry CH2Cl2 was added 0.21 mL of a tert-
butyl hydroperoxide solution (3.0 M in 2,2,4-trimethylpentane,
0.64 mmol), and the resulting mixture was stirred at room
temperature for 1 h. The solution was then cooled to 0 °C, and
100 mg (0.21 mmol) of propargyl alcohol 45 in 2 mL of CH2Cl2
was added. The reaction mixture was kept at 4 °C for 20 h,
whereupon 5 mL of 20% aqueous Na2S2O3 was added. After
the mixture was stirred vigorously at room temperature for
0.5 h, the separated aqueous phase was extracted with ether
(2 × 100 mL). The organic extracts were combined, dried over
magnesium sulfate, filtered, and concentrated under vacuum.
The residue was dried under high vacuum for 2 h, dissolved
in 8 mL of dry THF, and cooled to -50 °C. A lithium
triethylborohydride solution (0.64 mL, 1 M in THF) was added
dropwise over 5 min. The solution was then allowed to warm
to 0 °C and quenched by the addition of 50 mL of H2O. The
mixture was extracted with 2 × 100 mL of ether, and then
the organic extracts were combined, dried over magnesium
sulfate, and concentrated under vacuum. The crude product
was purified by silica gel chromatography (1:4 EtOAc/hexanes)
to give 66 mg (64%) of allylic alcohol 64.
Gen er a l P r oced u r e for P h otoisom er iza tion of 5,6-
tr a n s-Vita m in D to th e Cor r esp on d in g Vita m in D. Argon
was bubbled through a solution of the mixture of the 5,6-cis-
and 5,6-trans-vitamin D compounds (20 mg) and 9-acetylan-
thracene (10 mg) in tert-butyl methyl ether (12 mL) in a 3/4 ×
5 in. Kimax test tube for 10 min. The solution was irradiated
under argon using a Hanovia 450 W medium-pressure lamp
filtered through a Pyrex, water jacketed condenser well for 1
h. The solvent was removed, and the residual oil was separated
by flash chromatography to afford the corresponding vitamin
(∼20 mg, ∼100%).
2D NOE Dou ble-Ha lf-F ilter ed NMR Stu d y of [7,9,19-
13C3]-25-D3 (58) in SDS-d 25 Micelles. To a solution of sodium
dodecyl sulfate-d25 (SDS-d25, MW ) 313.50, Cambridge Isotope
Laboratories, 3.14 mg, 0.0100 mmol) in ethanol (0.75 mL),
prepared by warming the mixture over a steam bath until it
dissolved, was added an ethanolic solution of [7,9,19-13C3]-25-
D3 (58, 0.890 mL, 0.001124 M, 0.0010 mmol). The concentra-
tion was established from its UV-based extinction coefficient
of ꢀ ) 18 300 at λmax 265 nm (a value of ꢀ ∼12 000 in SDS was
estimated experimentally, but the weight of steroid used was
quite small). The solvent was then removed under reduced
pressure on a rotary evaporator. The resulting white residue
was dried under high vacuum with protection from light for
12 h at room temperature. To the white residue was added
deuterium oxide (Cambridge Isotope Laboratories, 99.9%,
0.500 mL) and the resulting suspension subjected to sonication
for 5 min (270 W, 43 kHz; Fisher Ultrasonic Cleaner, model
no. FS14). The resulting clear colorless solution (20 mM in
SDS-d25; 2.0 mM in [7,9,19-13C3]-25-D3) was transferred to a
Shigemi NMR tube and the plunger inserted so as to eliminate
any bubbles present.
Using the above sample, NOESY NMR experiments were
run using a variant of the pulse sequence described by Wider
et al.15 in which the phase cycling was modified to produce
the desired sub-spectrum directly. A build up curve was
created using the volumes for the cross-peaks observed for the
proton pairs 9R-9â, 19E-19Z, 7-19Z, and 7-19E at mixing
times of 25, 50, 100, 200, 300, 600, and 800 milliseconds to
determine the appropriate mixing time that provides the
optimal volume cross-peak intensities for the distance calcula-
tions. The results are discussed in the text (including Figure
2), and additional details are given in the Supporting Informa-
tion.
[7,9,19-13C3]-(1S ,3R ,5R ,8R )-1-(t er t -Bu t yld ip h e n ylsi-
lyloxy)-3,5-d id eh yd r o-8-h yd r oxy-9,10-seco-25-(tr im eth yl-
silyloxy)ch olest-10(19)-en -6-yn e (65). A solution of 20 mg
(0.04 mmol) of propargyl alcohol 64, 7 mg (0.1 mmol) of
imidazole and 21 mg (0.07 mmol) of TBDPS-Cl in 1 mL of CH2-
Cl2 was stirred under argon at room temperature for 12 h.
After addition of 2 mL of H2O, the mixture was extracted with
3 × 5 mL of hexanes. The combined organic extracts were dried
over magnesium sulfate, filtered, and concentrated under
vacuum. The residue was purified by preparative thin-layer
chromatography, eluting with 10% EtOAc/hexanes to afford
25 mg (0.034 mmol, 85% yield) of 65 as a colorless oil.
[7,9,19-13C3]-(1S ,3R ,5R ,8R )-1-(t er t -Bu t yld ip h e n ylsi-
lyloxy)-3,5-d id eh yd r o-8-h yd r oxy-9,10-seco-25-(tr im eth yl-
silyloxy)-ch olesta -6,10(19)-d ien e (66). To a solution of
propargyl alcohol 65 (6 mg, 0.008 mmol) in 1 mL of THF was
added NaOMe (3 µL of a 25% NaOMe in methanol solution)
and LiAlH4 (140 µL of a 1.0 M solution in THF). The reaction
mixture was stirred under argon at room temperature for 2
h. The reaction mixture was then quenched by the addition of
5 mL of saturated NH4Cl solution and extracted with 3 × 10
mL of diethyl ether. The organic extract was washed with
brine, dried over magnesium sulfate, filtered, and concentrated
In another experiment, a solution of sodium dodecyl sulfate-
d25 (Cambridge Isotope Laboratories, 29 mg) in ethanol (10
mL), prepared by warming the mixture until dissolved, was

1650 J . Org. Chem., Vol. 67, No. 5, 2002
Okamura et al.
added a solution of [7,9,19-13C3]-25-D3 (58, 0.4 mg) in ethanol
(400 µL). The solvent was removed under vacuum on a rotary
evaporator. The resultant free flowing white powder was dried
under high vacuum for 24 h at room temperature. Deuterium
oxide (CIL, 99.9% D, 2.5 mL) was added and the resulting
suspension was subjected to sonication for 5 min (270 W, 43
kHz; Fischer Ultrasonic Cleaner, model no. FS14). Solid
material was removed by filtration (Millex-SR Millipore filter,
0.5 µm). The resulting clear solution was used for recording
the 13C NMR spectrum. Additional details for the NMR
measurements are given in the caption to Figure 2.
NMR Sa m p le P r ep a r a tion for 58 in Aqu eou s Dim eth -
ylcyclod extr in (DMCD) Solu tion . To a solution of dimeth-
ylcyclodextrin (DMCD, 10 mg, Aldrich; also referred to as
methyl-â-cyclodextrin) in D2O (2.5 mL) was added a solution
of triply labeled 25-D3 (0.3 mg) in CD3OD (30 µL) with stirring.
The clear solution was lyophilized (cooled with dry ice and
evacuated with vacuum pumping) at room temperature over-
night to give a white powder. Approximately 1 mg of the
powder was dissolved in D2O (0.7 mL) and used for the NMR
study (5 mm tube).
Ack n ow led gm en t. This study was generously sup-
ported by grants from the NIH, NSF, the Cambridge
Isotope Laboratories, and the Committee on Research
of the University of California, Riverside. We also
acknowledge generous supplies of starting materials
from Solvay, Hoffmann-LaRoche, and Leo Pharmaceu-
ticals. K.R. also acknowledges the DFG (Germany) for
a postdoctoral stipend.
Su p p or tin g In for m a tion Ava ila ble: Spectral and ana-
lytical data. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O011096Y