2,2-Disubstituted analogues of the natural hormone 1α,25-dihydroxyvitamin D3: Chemistry and biology

Article abstract of DOI:10.1016/S0968-0896(02)00058-5

Six new 2,2-disubstituted analogues of the natural hormone calcitriol have been prepared. Chemical novelty includes (1) the first example of an inverse-electron-demand Diels-Alder cycloaddition using a pyrone diene and a difluorinated vinyl ether dienophi

Full text of DOI:10.1016/S0968-0896(02)00058-5

Bioorganic & Medicinal Chemistry 10 (2002) 2353–2365
2,2-Disubstituted Analogues of the Natural Hormone
1ꢀ,25-Dihydroxyvitamin D3: Chemistry and Biology
Gary H. Posner,a,* Benjamin T. Woodard,a Kenneth R. Crawford,a Sara Peleg,b
Alex J. Brown,c Patrick Doland and Thomas W. Kenslerd
aDepartment of Chemistry, Krieger School of Arts and Sciences, The Johns Hopkins University, Baltimore, MD 21218, USA
bDepartment of Medical Specialties, The University of Texas, M. D. Anderson Cancer Center, Houston, TX, 77030, USA
cRenal Division, Washington University School of Medicine, St. Louis, MO 63110, USA
dDepartment of Environmental Health Sciences, Division of Toxicological Sciences, Bloomberg School of Public Health,
The Johns Hopkins University, Baltimore, MD 21205, USA
Received 4 September 2001; accepted 2 January 2002
Abstract—Six new 2,2-disubstituted analogues of the natural hormone calcitriol have been prepared. Chemical novelty includes (1)
the first example of an inverse-electron-demand Diels–Alder cycloaddition using a pyrone diene and a difluorinated vinyl ether
dienophile, leading to difluorinated analogues 7 and (2) a conceptually streamlined approach to dimethylated 19-nor analogues 8.
Analogues 7a and 8a are similar to calcitriol in terms of in vitro antiproliferative activity, but they are different from calcitriol in
terms of transcriptional activity: difluorinated analogue 7a is 2–3 times more active transcriptionally than calcitriol, whereas
dimethylated analogue 8a is 7.5 times less active transcriptionally. Whereas the in vivo calcemic activity of difluorinated analogue
7a is similar to that of calcitriol, dimethylated analogue 8a is considerably less calcemic than calcitriol. Dimethylated analogue 8a
strongly suppresses parathyroid hormone (PTH) secretion. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
differentiation activities.3 Several of such designer ana-
The vitamin D family of steroid hormones is being
studied by diverse scientists interested in issues ranging
from fundamental organic and medicinal chemistry to
molecular biology, endocrinology, and new drug devel-
opment.1 Much of the motivation for this widespread
interest and effort is the growing appreciation that the
natural hormone 1a,25-dihydroxyvitamin D3 (calcitriol,
1) has many different human biological functions ran-
ging from regulating calcium and phosphorus homeo-
stasis to governing cell growth and differentiation.2
Medical use of the natural hormone calcitriol (1), how-
ever, is often risky due to the hypercalcemia typically
induced by the supraphysiological levels of this hor-
mone required to treat humans with diseases such as
cancer, psoriasis, and autoimmune disorders.1,2 Small
chemical changes at various positions in the structure
of the natural hormone have produced thousands of
synthetic analogues, some of which have desirably low
calcemic activity and high antiproliferative and pro-
logues are now undergoing preclinical evaluation for
chemotherapy of diverse human diseases, and some of
these synthetic analogues are U.S. FDA-approved drugs
(e.g. calcipotriol for treatment of psoriasis and 1a,25-
dihydroxy-19-nor-vitamin D2 for treatment of second-
ary hyperparathyroidism).2
Structural changes specifically at the 2-position of the
natural hormone in several cases produce analogues
having desirable physiological properties. For example,
the Chugai pharmaceutical company designed some
2-alkoxy and 2-alkyl analogues (e.g., 2 and 3), among
which ED-71 (2) is a leading drug candidate to treat
women who have osteoporosis.4 Also, we5,6 and
others7ꢀ9 designed a series of 2-alkyl analogues (4–6)
and showed that some of these 2-monosubstituted ana-
logues have strong affinity for the vitamin D receptor
(VDR). Additionally, 2b-fluoro-19-norcalcitriol10 and
2b-fluorocalcitriol11,12 have been prepared, and 4,4-
difluorocalcitriol has been prepared to probe A-ring
conformation.13 No 2,2-disubstituted analogue of calci-
triol (1), however, has been reported. Now, we describe a
series of six 2,2-disubstituted analogues, prepared using
*Corresponding author. Tel.: +1-410-516-7429; fax: +1-410-516-
8420; e-mail: ghp@jhunix.hcf.jhu.edu
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00058-5

2354
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
some novel organic chemistry. Strong electronic influ-
ence by two powerfully electronegative fluorine atoms14
at position-2 was expected to enhance the hydrogen-
bonding ability of the 1- and 3-OH groups and to afford
new analogues 7 with enhanced VDR affinity and lipo-
philicity. Considerable steric hindrance (but minimal
electronic influence) by two methyl groups at position-2
was expected to diminish the reactivity of the 1- and
3-OH groups (e.g., retarding 3-epimerization and A-ring
dehydration)15 in new 19-nor analogues 8 but probably
also to interfere with VDR binding.
proceeded smoothly to form polyfunctionalized cyclo-
hexene 10 without intermolecular or intramolecular
displacement of fluoride ion. Also, no loss of fluoride
was encountered during the several subsequent steps in
Scheme 1 using powerful nucleophiles (e.g., LiAlH4,
Ph2PLi) or using high temperature for Claisen rearran-
gement of the sulfoxide-containing vinyl ether derived
from allylic alcohol 12; conjugated dienyl ester 13 was
formed after in situ sulfoxide pyrolytic elimination.18
Standard protocol19 allowed smooth conversion of die-
nyl ester 13 into racemic 2,2-difluorinated A-ring phos-
phine oxide 14. Coupling of phosphine oxide (ꢁ)-14
with enantiomerically pure C,D-ring ketone (+)-15,
followed by desilylation, produced the desired 2,2-
difluorinated analogues (ꢀ)-7a and 7b in a 5:1 ratio.
Distinction between these two diastereomers was based,
Chemistry
We have extensively developed Diels–Alder cycloaddi-
tion reactions of heteroaromatic 2-pyrones with elec-
tronically matched dienophiles to form synthetically
versatile and stereochemically rich bicyclic lactone
adducts.16 We report here the first example of a suc-
cessful high pressure 4+2-cycloaddition between com-
mercial electron-poor 3-methoxycarbonyl-2-pyrone and
a new geminally difluorinated electron-rich vinyl ether
dienophile (Scheme 1); chromatographically pure race-
mic cycloadduct 9 was isolated in 73% yield, formed
regiospecifically and stereospecifically with the desired
1,3-trans dioxygenation pattern necessary for ultimate
transformation into calcitriol analogues 7. Nucleophilic
opening of the lactone in 9 using lithium allyloxide17
1
as in previous related cases, on their characteristic H
NMR (Table 1).19 Based on chromatographic compar-
ison and as expected based on the presence of two
fluorine atoms, difluorinated analogue 7a is noticeably
more lipophilic than calcitriol (1).
To simplify preparation of some 2,2-dimethyl analogues
of calcitriol (1), we chose to design analogues lacking
the 19-methylene group; DeLuca’s pioneering work has
shown that many 19-nor analogues of calcitriol (1)
retain much of the hormone’s desirable cell growth
regulatory properties while having low calcemic activ-
ities.20 Our streamlined synthetic approach, outlined in

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2355
Scheme 1.
Table 1. 1H NMR (d) and optical rotation characteristics of new
analogues
phosphine oxides 20 and 21. Each of these racemic A-
ring units underwent separate Horner–Wadsworth–
Emmons coupling with enantiomerically pure C,D-ring
ketone (+)-15 to form chromatographically separated
2,2-dimethyl-19-nor analogues 8a and 8b (in low yields)
and also 8c and 8d (in good yields). Distinction between
1,3-trans diastereomers 8a and 8b and between 1,3-cis
diastereomers 8c and 8d was based on optical rotations
and 1H NMR (Table 1)19 and also on analogy with
analogues of calcitriol (1) having similar 1,3-diol ste-
reochemical relationships.21
Analogue
C18-CH3
C6-H
C7-H
[a]2D5
7a
7b
0.55
0.54
6.45
6.47
6.02
6.00
8a
8b
8c
8d
0.54
0.55
0.54
0.55
6.29
6.30
6.31
6.33
5.84
5.83
5.85
5.86
+29
+21
+94
+42
Scheme 2 starting with symmetrical 2,2-dimethyl-1,3-
cyclohexanedione, involves a series of intermediates
grouped into two families: (1) a pair of diastereomeric
trans-1,3-dihydroxycyclohexanes 16 and (2) the corre-
sponding cis-1,3-dihydroxycyclohexane 17 having sym-
metrical substitution about a plane through C-2 and C-
5 of the cyclohexane ring and, therefore, having also a
simplified 1H NMR spectrum especially in the d 3.7
region (see Experimental). Noteworthy is the Michael
addition of a phenylthioacetate enolate that occurs
exclusively trans to the resident siloxy group leading to
two 3-siloxy-5-phenylthioacetate cyclohexanone pre-
cursors to adducts 16 and 17. In these two cyclohex-
anone precursors to adducts 16 and 17, proton NMR
shows clearly that the 3- and 5-substituents are trans to
each other, with the 3-siloxy group axial and with the
equatorial C-3 H appearing at d 3.8 as a doublet of
doublets with coupling constants of 2.2 and 4.0 Hz.
Sulfide oxidation of thioether adducts 16 and 17 into the
corresponding sulfoxides (as a mixture of diastereomers
that were not separated) and then pyrolysis gave the
pure a,b-unsaturated esters 18 and 19 in high yields.
Standard transformations19 then produced the A-ring
Biology
Motivation for interest in non-calcemic analogues of
calcitriol (1) rests not only on their potential chemo-
therapeutic applications but also on using these analo-
gues as sensitive probes of the fundamental molecular
mechanisms underlying the varied biological effects that
such analogues elicit. For example, we have discovered
that there is an inverse relationship between some ana-
logues’ transcriptional potencies and their interaction
specifically with AF-2 residues of VDR22 and that some
A-ring analogues with low affinity for VDR modulate
chondrocyte growth via membrane effects that are
dependent on the stage of cell maturation.23 New 2,2-
difluoro analogues 7 and 2,2-dimethyl analogues 8 now
provide some surprising observations; understanding
the molecular biology underlying these results may
advance this field in significant ways.
The antiproliferative activity, determined in vitro using
murine keratinocytes as previously described,24 of only
one of the four 2,2-dimethyl diastereomers (i.e., 8a with

2356
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
natural A-ring diol stereochemistry) is similar to that of
calcitriol (1, Fig. 1). Likewise, 2,2-difluoro diastereomer
7a with natural A-ring diol stereochemistry has anti-
proliferative activity similar to that of calcitriol (1, Fig.
1). The VDR-mediated transcriptional potencies of 2,2-
disubstituted analogues 7a and 8a, determined in vitro
using a previously described protocol in rat osteo-
sarcoma ROS 17/2.8 cells,25 are different from that of
calcitriol (1). The ED50 values for transcriptional activ-
ity are as follows: calcitriol (1), 0.4 nM; difluoro analo-
gue 7a, 0.15 nM; dimethyl analogue 8a, 3.0 nM.
Competition assays25 using the recombinant human
vitamin D receptor revealed the affinities of these two
analogues [relative to 100% binding of calcitriol (1)] to
be 9.6% for 7a and 1.3% for 8a. Thus, as rationally
designed, the presence of two methyl groups at the
2-position in analogue 8a strongly diminishes (by a fac-
tor of 75) binding of this analogue to the human VDR.
Remarkably however, this 2,2-dimethyl analogue 8a
is only 7.5-fold less active transcriptionally than cal-
citriol (1) in rat osteosarcoma cells. Likewise, although
2,2-difluoro analogue 7a binds to the human VDR only
about 1/10 as well as the natural hormone 1, 7a is 2–3
times more transcriptionally active in the rat osteo-
sarcoma cells than the natural hormone 1. Therefore, as
in some related instances,22,26 the levels of genomic
activity produced by analogues 7a and 8a are not
directly proportional to their affinities for the nuclear
VDR.
2,2-Dimethyl analogue 8a with natural 1a,3b-diol ste-
reochemistry (but not the diastereomeric 1a,3a-diol
analogue 8c) is a strong inhibitor in vitro of parathyroid
hormone (PTH) secretion (Fig. 2). Confluent cultures of
bovine parathyroid cells were incubated for 72 h with
various concentrations of dimethlylated analogues 8a,
8c or calcitriol (1). The steady state levels of PTH
secretion were then determined during a 3-h incubation
with fresh medium. As shown in Fig. 2, trans diol ana-
Figure 1.
Figure 2. Suppression of PTH secretion. Confluent cultures of bovine
parathyroid cells were treated for 72 h with the designated concentra-
tion of the vitamin D compounds. The cells were then incubated in
fresh medium, and the amount of PTH secreted during a 3-h period
was determined by RIA. The values are expressed as meanꢁSD
(n=6). *p<0.001.
Scheme 2.

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2357
logue 8a appeared to be about 10 times less active than
calcitriol (1), whereas cis diol analogue 8c appeared to
be several hundred times less active than calcitriol (1).
Inhibition of PTH by calcitriol (1) is well established to
involve repression of PTH gene transcription27,28 and
vitamin D response elements (VDREs) have been iden-
tified in the promoters of the human,29 rat,30 and
chick31 PTH genes. Further characterization of these
negative VDREs and establishment of transcriptional
assays have been hampered by the lack of a parathyroid
cell line and the lack of activity of the PTH gene pro-
moter in non-parathyroid cells. The suppression of PTH
by vitamin D compounds is known to be VDR-depen-
dent, but the other components of the negative reg-
ulatory complex have not been characterized. In fact,
while VDR/RXR dimers appear to bind the rat and
chick VDREs,30,31 the dimerization partner for VDR
binding to the human PTH VDRE has not been identi-
fied.32 Thus, the molecular details of the mechanism for
transcriptional repression by liganded VDR are not
known. The VDR present in the parathyroid glands is
functionally indistinguishable from that in other tissues.
Studies examining structure–activity relationships for
the suppression of PTH by vitamin D analogues have
1.0 mg/kg body weight (Fig. 4) daily for 1 week. At a
20-fold higher dose than calcitriol (1), dimethyl analogue
8a caused only slight elevation of urinary calcium levels
(Fig. 3) but no slowing of animal weight gain (data not
shown). In sharp contrast, even at only a 2-fold higher
dose than calcitriol (1), difluoro analogue 7a elevated
urinary calcium levels more than did calcitriol (1, Fig.
4). This strong calcemic activity of difluoro analogue 7a
was not expected based on its relatively low VDR
binding affinity (9.6% that of the natural hormone 1) or
based on the low calcemic effect of the closely related
2b-fluorocalcitriol.12
In conclusion, rationally designed and newly synthe-
sized 2,2-disubstituted chemical entities 7a and 8a illus-
trate how even very small structural changes in large
molecules can produce powerful and different biological
effects. Specifically, the desirable characteristics of high
antiproliferative and PTH-suppression activities with
low calcemic activity make new dimethylated analogue
8a suitable for further in vivo testing as a potential
chemotherapeutic drug candidate. The new knowledge
presented here about the relationships between chemical
structure and biological activity may allow also other
analogues of calcitriol (1) to be prepared for medicinal
use and/or for use as sensitive probes of vitamin D
molecular biology and mechanism of action.34ꢀ36
revealed a close correlation between the VDR affinity
33
and suppressive activity.
The suppressive activity of
PTH by trans diol analogue 8a correlates well with the
transcriptional potency of this analogue in the rat
osteosarcoma cells, further substantiating that this sup-
pressive activity is mediated by the vitamin D receptor.
Experimental19
The in vivo calcemic activities of difluoro analogue 7a
and dimethyl analogue 8a were determined using our
standard protocol19 in which rats are treated daily with
calcitriol (1) at 0.5 mg/kg body weight and with our new
analogues at either 10.0 mg/kg body weight (Fig. 3) or
p-Methoxybenzyl 2,2,2-trifluoroethyl ether. Trifluoro-
ethanol (10 g, 0.10 mol) in THF (10 mL) was added
dropwise via addition funnel to a stirred suspension of
NaH (2.9 g, 0.12 mol) in THF (10 mL) at 0 ꢂC over a
period of 30 min. To the resulting mixture was added a
Figure 3. Effects of vitamin D3 analogues on urinary calcium excre-
tion in rats. Animals were treated with 0.5–10.0 microgram/kg body
weight of test compound po for 7 consecutive days, and urinary
excretion of calcium was measured during days 3–7. Values are
mean+SE from three animals in each group.
Figure 4. Effects of vitamin D3 analogues on urinary calcium excre-
tion in rats. Animals were treated with 0.5–10.0 microgram/kg body
weight of test compound po for 7 consecutive days, and urinary
excretion of calcium was measured during days 3–7. Values are
mean+SE from three animals in each group.

2358
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
solution of p-methoxybenzyl chloride (PMB-Cl, 19 g,
0.12 mol) in THF (10 mL) and N,N0-dimethylpropyle-
neurea (DMPU, 6 mL) via addition funnel over 1 h. After
this addition, tetrabutylammonium iodide (n-Bu4NI,
11 g, 0.03 mol) was added and the mixture warmed
slowly to rt overnight. After acidic workup, the crude
product was vacuum distilled to afford the desired ether
as a pale orange liquid (25 g, quantitative): bp 100–
Hz), 61.0 (d, J=6.1 Hz), 55.2, 53.2; 19F NMR (CDCl3,
CFCl3) dꢀ103.3 (ddd, J=244.9, 15.8, 5.1 Hz), ꢀ113.8
(d, J=244.9 Hz); IR (neat) 2954, 1780, 1751, 1613,
1515, 1340, 1295, 1250, 1171, 1131, 1103, 1049, 826.
Anal. Calcd for C17H16F2O6: C, 57.63; H, 4.55, found:
C, 57.64; H, 4.62.
Bis-silylated mixed malonate (ꢁ)-10. Asolution of
n-BuLi (1.6 M in hexanes, 0.900 mL, 1.44 mmol) was
added dropwise to freshly distilled allyl alcohol (5 mL)
at 0 ꢂC. Bicyclic lactone (ꢁ)-9 (300 mg, 0.847 mmol)
was then added in solution (5 mL CH2Cl2, 2.5 mL allyl
alcohol) over 10 min. The reaction was stirred at 0 ꢂC
for 3 h and quenched with saturated aq NH4Cl. The
aqueous phase was extracted with EtOAc (3ꢃ25 mL),
the combined organics dried over MgSO4, and the sol-
vents removed in vacuo to provide crude hydroxy mixed
malonate as a pale yellow oil requiring no further pur-
ification for the following deprotection.
101 ꢂC/0.5 mm Hg; H NMR (CDCl3) d 7.30–7.21 (m,
1
2H), 6.93–6.85 (m, 2H), 4.58 (s, 2H), 3.78 (s, 3H), 3.76
(q, J=9.0 Hz, 2H); 13C NMR (CDCl3) d 159.7, 129.6,
128.5, 124.1 (q, J=278.0 Hz), 113.9, 73.6, 66.6 (q,
J=33.7 Hz), 55.1; 19F NMR (CDCl3, CFCl3) d ꢀ74.2 to
ꢀ74.4 (m); IR (neat) 2937, 2837, 1613, 1515, 1280, 1250,
1164, 1117, 1034, 998, 965, 827, 665. Anal. Calcd for
C17H16F2O6: C, 54.55; H, 5.04, found: C, 54.51; H, 5.24.
p-Methoxybenzyl 2,2-difluorovinyl ether. Asolution of
n-BuLi (26 mL, 0.26 mol, 10 M in hexanes) was added
dropwise via addition funnel to a 1 L flask fitted with an
argon inlet and outlet bubbler charged with a cold
(ꢀ78 ꢂC), stirred solution of the aforementioned ether
(23 g, 0.10 mol) in Et2O (210 mL). The resulting solu-
tion was allowed to reach rt gradually overnight. Upon
cooling the mixture to 0 ꢂC, the reaction was quenched
with dropwise addition of saturated aq NH4Cl (50 mL).
The reaction mixture was then extracted with Et2O
(3ꢃ50 mL) and the combined organics filtered through
Celite, washed with H2O (2ꢃ250 mL), dried (Na2SO4),
filtered, and carefully concentrated to a brown oil.
Vacuum distillation gave the desired 2,2-difluorovinyl
ether as a pale yellow, somewhat unstable liquid (6.6 g,
32%) followed by a mixture of starting material and
desired product (3.4 g, ꢄ1:1). PMB-2,2-difluorovinyl
ether: bp 98 ꢂC/0.5 mm Hg; 1H NMR (300 MHz,
CDCl3) d 7.33–7.24 (m, 2H), 6.94–6.87 (m, 2H), 5.65
(dd, J=22.0, 3.8 Hz, 1H), 4.67 (s, 2H), 3.82 (s, 3H). Due
to its instability, this 2,2-difluorovinyl ether was used
immediately.
To the crude hydroxy mixed malonate was added
CH2Cl2, H2O (4.25 and 0.25 mL, respectively), and
DDQ (297 mg, 1.31 mmol). Additional DDQ (297 mg)
was added after 6 h of stirring to complete the depro-
tection. After 12 h, the reaction was cautiously quen-
ched with saturated aq NaHCO3, decanted into H2O
(50 mL), extracted with CH2Cl2 (3ꢃ20 mL), dried over
Na2SO4, filtered through silica (1 inch plug), and con-
centrated to an oil requiring no further purification for
the following bis-silylation.
The crude diol was dissolved in CH2Cl2 (8.0 mL),
cooled to ꢀ78 ꢂC, and treated with 2,6-lutidine (195 mL,
1.68 mmol) and TBSOTf (339 mL, 1.48 mmol). After
stirring for 3 h, additional 2,6-lutidine and TBSOTf
were added (same amounts as above). The reaction
mixture was allowed to warm to 0 ꢂC, at which time
additional 2,6-lutidine and TBSOTf were added (same
amounts as above) and the solution stirred at 0 ꢂC
overnight. Additional 2,6-lutidine and TBSOTf were
added (3ꢃs, same amounts as above) to complete the
reaction. After normal aqueous workup, the crude oil
was passed though flash silica gel (5% EtOAc/hexanes)
to afford (ꢁ)-10 as a pale yellow oil [323 mg, 92% from
4-Carbomethoxy-5-endo-[(p-methoxybenzyl)oxy]-6,6-di-
fluoro-3-oxo-2-oxabicyclo-[2.2.2]oct-7-ene (ꢁ)-9. Afive-
inch length of heat shrinkable Teflon tubing (Ace Glass
#12685–40) was heat sealed at one end with a stainless
steel plug. To the tube were added 3-carbomethoxy-2-
pyrone (1.0 g, 6.5 mmol), CH2Cl2 (10 mL), and PMB-
2,2-difluorovinyl ether (6.5 g, 32 mmol). The tube was
sealed at the other end with a stainless steel plug and
placed under high pressure (10–12 kbar) for 5 days.37
The tube was opened, the reaction mixture was trans-
ferred to a flask and concentrated in vacuo, and the
residue purified by silica gel chromatography (5–40%
EtOAc/hexanes). The fractions containing the desired
syn,endo isomer (ꢁ)-9 were combined and concentrated
to a faint yellow oil, which afforded white crystals on
1
(ꢁ)-9]: H NMR (CDCl3) d 6.02 (dq, J=10.4, 1.6 Hz,
1H), 5.85 (ddt, J=17.2, 10.4, 5.7 Hz, 1H), 5.73 (ddd,
J=10.4, 6.4, 2.4 Hz, 1H), 5.25 (dddd, J=10.4, 6.4, 1.6,
1.2 Hz, 2H), 5.05 (d, J=11.6 Hz, 1H), 4.61 (apparent
dddt, J=29.5, 13.1, 5.7, 1.2 Hz, 2H), 4.63–4.54 (m, 2H),
3.74 (s, 3H), 0.90 (s, 9H), 0.84 (s, 9H), 0.15 (d, J=2.8
Hz, 3H), 0.10 (s, 6H), 0.09 (s, 3H); 13C NMR (CDCl3) d
167.2, 166.3, 131.2, 129.4 (d, J=7.6 Hz), 123.6, 119.3
(t, J=249.4 Hz), 118.6, 71.8 (dd, J=24.7, 6.0 Hz),
66.4 (dd, J=19.8, 5.4 Hz), 63.0 (d, J=5.4 Hz), 53.1,
25.6, 18.2, 18.0, ꢀ4.5 (d, J=6.0 Hz), ꢀ4.9 (d, J=5.4
Hz), ꢀ5.3; 19F NMR (CDCl3, CFCl3) d ꢀ115.7 (d,
J=246.7 Hz),ꢀ124.9 (dd, J=246.7, 15.2 Hz); IR
(neat) 2954, 2931, 2858, 1749, 1255, 1221, 1200, 1180,
1135, 1111, 1040, 870, 838, 780; HRMS: calcd for
C24H48F2O6Si2+Na 453.2386, found 453.2386.
standing (1.7 g, 74%): mp 74–76 ꢂC; H NMR (CDCl3)
1
d 7.23–7.17 (m, 2H), 6.96–6.92 (m, 1H), 6.91–6.85 (m,
2H), 6.57–6.49 (m, 1H), 5.17–5.09 (m, 1H), 4.72 (d,
J=11.2 Hz, 1H), 4.49 (d, J=11.2 Hz, 1H), 4.36 (d,
J=8.0 Hz, 1H), 3.80 (s, 3H), 3.79 (s, 3H); 13C NMR
(CDCl3) d 165.5, 165.1, 159.7, 132.5, 130.3, 127.7, 126.6
(d, J=5.3 Hz), 118.6 (t, J=259.6 Hz), 113.7, 75.3 (t,
J=8.7 Hz), 74.9 (t, J=8.4 Hz), 74.7, 74.2 (d, J=2.2
Methyl ester (ꢁ)-11. Amixture of ( ꢁ)-10 (266 mg,
0.511 mmol), formic acid (24 mL, 0.638 mmol), Et3N (93

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2359
mL, 0.664 mmol), Ph3P (27 mg, 0.102 mmol), and pal-
ladium (II) acetate (12 mg, 0.051 mmol) in 1,4-dioxane
(1.6 mL) was heated to 100 ꢂC in a sealed tube for 16 h.
Upon cooling, the tube was carefully purged under
argon and dilute HCl added (1 mL, 1 M aq). The con-
tents of the tube were transferred to a separatory funnel
and partitioned between H2O and CH2Cl2, extracted
(3ꢃ10 mL CH2Cl2), dried (MgSO4), filtered, con-
centrated, and passed through flash silica gel (5%
EtOAc/hexanes) to give (ꢁ)-11 (210 mg, 94%) as a
and pure (ꢁ)-13 (42 mg, 146 mg total, 79%) both as
1
clear oils. (ꢁ)-13: H NMR (CDCl3) d 5.72 (br s, 1H),
5.42 (t, J=1.6 Hz, 1H), 5.26 (br s, 1H), 4.59–4.47 (m,
1H), 4.21–4.04 (m, 1H), 4.13 (q, J=7.0 Hz, 2H), 2.53
(dt, J=13.6, 1.4 Hz, 1H), 2.33 (ddd, J=13.6, 4.8, 1.4
Hz, 1H), 1.23 (t, J=7.0 Hz, 3H), 0.92 (s, 9H), 0.88 (s,
9H), 0.11 (d, J=12.4 Hz, 6H), 0.10 (d, J=8.8 Hz, 6H);
13C NMR (CDCl3) d 165.4, 149.0, 141.1, 119.7, 119.6 (t,
J=250.2 Hz), 115.9, 72.6 (t, J=23.5 Hz), 69.4 (t,
J=25.8 Hz), 60.0, 41.4, 25.6, 25.5, 18.3, 18.0, 14.1,
ꢀ5.0, ꢀ5.1, ꢀ5.3; 19F NMR (CDCl3, CFCl3) d ꢀ120.7
(d, J=242.3 Hz), ꢀ123.7 (d, J=242.3 Hz); IR (neat)
2956, 2931, 2896, 2858, 1730, 1646, 1473, 1366, 1255,
1222, 1184, 1161, 1122, 1098, 1037, 1006, 940, 893, 838,
800; HRMS: calcd for C23H42F2O4Si2 476.2590, found
476.2587.
clear oil which solidified on standing: mp 42–43 ꢂC; H
1
NMR (CDCl3) d 6.91 (dd, J=4.8, 2.8 Hz, 1H), 4.69 (t,
J=6.4 Hz, 1H), 4.40–4.24 (m, 1H), 3.75 (s, 3H), 2.75–
2.60 (m, 1H), 2.45 (ddd, J=19.4, 9.0, 2.8 Hz, 1H), 0.91
(s, 9H), 0.85 (s, 9H), 0.17 (d, J=1.6 Hz, 3H), 0.14 (s,
3H), 0.09 (s, 3H), 0.08 (s, 3H); 13C NMR (CDCl3) d
165.6, 139.6, 130.5 (d, J=6.1 Hz), 119.3 (t, J=249.4
Hz), 67.8 (dd, J=87.9, 20.2 Hz), 64.8 (t, J=20.2 Hz),
51.8, 34.0, 25.7, 18.14, 18.11,ꢀ4.9,ꢀ5.0,ꢀ5.1; 19F NMR
(CDCl3, CFCl3) dꢀ121.6 (d, J=241.0 Hz),ꢀ130.4
(ddd, J=241.0, 25.1, 6.8 Hz); IR (neat) 2954, 2930,
2858, 1725, 1258, 1177, 1123, 1082, 900, 838, 781;
HRMS: calcd for C20H38F2O4Si2+Na 459.2174, found
459.2177.
A-ring phosphine oxide (ꢁ)-14. To a solution of (ꢁ)-13
(104 mg, 0.218 mmol) in PhCH3/CH2Cl2 (3.0 mL, 2:1)
at ꢀ78 ꢂC was slowly added a solution of DIBAL-H
(480 mL, 1.0 M in PhCH3, 0.480 mmol) via syringe. The
reaction was maintained at ꢀ78 ꢂC (1 h), then slowly
warmed to ꢀ50 ꢂC at which time the reaction was com-
plete by TLC analysis (25% EtOAc/hexanes). The
reaction was quenched with aqueous sodium potassium
tartrate (1 mL, 2 N), HCl (1 mL, 1 M aqueous) and H2O
(1 mL), the organic layer was separated and the aq layer
extracted with CH2Cl2 (3ꢃ6 mL), the combined organ-
ics dried (Na2SO4) and concentrated to give the desired
allylic alcohol (103 mg) as a colorless oil which was pure
enough to be carried forward directly.
Allylic alcohol (ꢁ)-12. To a solution of (ꢁ)-11 (200 mg,
0.458 mmol) in CH2Cl2 (5 mL) at ꢀ78 ꢂC was slowly
added diisobutylaluminum hydride (DIBAL-H, 1.01
mL, 1.01 mmol, 1.0 M in PhCH3). This mixture was
allowed to warm to rt and stirred (1 h) until the reaction
was complete by TLC analysis (20% EtOAc/hexanes).
The reaction was quenched with aqueous sodium
potassium tartrate (1 mL, 2 N), the mixture was extrac-
ted with CH2Cl2 (3ꢃ6 mL), the combined organic layers
were washed with H2O (4 mL), dried (Na2SO4), con-
centrated under reduced pressure, and passed through
flash silica gel (5–10% EtOAc/hexanes) to afford (ꢁ)-12
To a cold (ꢀ78 ꢂC) solution of the crude allylic alcohol
in THF (2.5 mL) was added a solution of n-BuLi (178
mL, 0.283 mmol, 1.53 M in hexanes) dropwise via syr-
inge. The resulting solution was warmed (0 ꢂC) for 30
min, cooled (ꢀ20 ꢂC), and treated with a precooled
solution of TsCl (62 mg, 0.327 mmol) in THF (0.5 mL).
After stirring for 30 min at 0 ꢂC, the reaction mixture
was cooled (ꢀ78 ꢂC) and treated with enough of a pre-
cooled solution of lithium diphenylphosphide [gener-
ated from the addition of n-BuLi (623 mL, 0.990 mmol,
1.53 M in hexanes) to a cold (ꢀ78 ꢂC) solution of Ph2PH
(172 mL, 0.990 mmol) in THF (2.6 mL)] until an orange
color persisted for 5 min (ꢄ2.1 mL). The reaction mix-
ture was quenched with H2O (three drops), warmed to
rt, the solvent removed in vacuo, the residual oil dis-
solved in CH2Cl2 (3.5 mL), treated with hydrogen per-
oxide (2 mL, 5% aqueous H2O2), and stirred vigorously
for 1 h. After general aqueous workup, the crude reac-
tion mixture was passed through flash silica (20–50%
EtOAc/hexanes) to afford 62 mg [46% from (ꢁ)-13] of
(ꢁ)-14 as a clear oil: 1H NMR (CDCl3) d 7.75–7.65 (m,
4H), 7.56–7.41 (m, 6H), 5.45 (apparent dd, J=14.6, 7.0
Hz, 1H), 5.36 (t, J=1.6 Hz, 1H), 4.96 (br s, 1H), 4.42–
4.30 (m, 1H), 4.10–3.99 (m, 1H), 3.35 (dt, J=14.2, 8.4
Hz, 1H), 3.17 (dt, J=15.6, 6.9 Hz, 1H), 2.45 (d, J=14
Hz, 1H), 2.31–2.21 (m, 1H), 0.91 (s, 9H), 0.81 (s, 9H),
0.10 (s, 3H), 0.07 (s, 3H), 0.05 (s, 3H), 0.02 (s, 3H); 13C
NMR (CDCl3) d 141.8 (d, J=2.2 Hz), 137.8 (d, J=11.4
Hz), 133.1 (d, J=33.4 Hz), 132.1 (d, J=33.4 Hz), 131.9
(t, J=2.2 Hz), 130.9 (dd, J=9.1, 3.8 Hz), 128.6 (dd,
J=11.4, 3.1 Hz), 120.0 (t, J=245.2 Hz), 117.5 (d,
(158 mg, 84%) as a white solid: mp 42–44 ꢂC; H NMR
1
(CDCl3) d 5.67 (m, 1H), 4.47 (t, J=9.2 Hz, 1H), 4.24–
4.06 (m, 1H), 4.11 (s, 2H), 2.58–2.40 (m, 1H), 2.34–2.17
(m, 1H), 1.63 (br s, 1H), 0.91 (s, 9H), 0.89 (s, 9H), 0.15
(s, 6H), 0.10 (s, 3H), 0.08 (s, 3H); 13C NMR (CDCl3) d
136.5 (m), 123.2, 120.1 (t, J=247.5 Hz), 68.7 (t, J=25.8
Hz), 67.0 (t, J=24.0 Hz), 63.8, 33.0 (m), 25.8, 25.6, 18.2,
18.1,ꢀ4.7, ꢀ4.9, ꢀ5.1; 19F NMR (CDCl3, CFCl3)
dꢀ122.5 toꢀ124.5 (m); IR (neat) 3601–3094 (br), 2954,
2930, 2859, 1472, 1463, 1256, 1186, 1126, 1102, 1080,
896, 878, 862, 837, 778. Anal. Calcd for C20H38F2O4Si2:
C, 55.84; H, 9.37, found: C, 56.10; H, 9.49.
Z-dienoate ester (ꢁ)-13. A25-mL hydrolysis tube con-
taining a solution of (ꢁ)-12 (158 mg, 0.387 mmol), 1-
phenylsulfinyl-2,2,2-triethoxyethane38 (289 mg, 1.01
mmol) and 2,4,6-trimethylbenzoic acid (TMBA, 6.8 mg,
0.041 mmol) in CH2Cl2 (2 mL) was heated to 115 ꢂC for
16 h. After cooling to rt, the solvent was removed in
vacuo and the resulting light yellow oil was purified by
flash silica gel chromatography (2–5% EtOAc/hexanes)
to afford a mixture of Z- and E-dienoate esters (64 mg,
ꢄ2.2:1.0) and pure Z-dienoate ester (ꢁ)-13 (104 mg).
Separation of the remaining Z/E mixture using pre-
parative plate chromatography (1000 mm, 5% EtOAc /
hexanes) afforded pure E-dienoate ester (14 mg, 8%)

2360
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
J=7.6 Hz), 114.9, 72.7 (t, J=22.3 Hz), 69.5 (t, J=26.5
Hz), 40.9, 31.5 (d, J=70.6 Hz), 25.62, 25.57, 18.3, 18.0,
ꢀ5.0, ꢀ5.1; 19F NMR (CDCl3, CFCl3) d ꢀ118.0 to
ꢀ125.0 (m); IR (neat) 2952, 2929, 2856, 1472, 1438,
1255, 1166, 1120, 1093, 939, 892, 837, 780, 695, 552,
509; HRMS: calcd for C33H49F2O3PSi2 618.2926, found
619.3007.
(neat) 3636–3072 (br), 2937, 2869, 1377, 1215, 1156,
1077, 757; HRMS: calcd for C27H42F2O3+Na 475.3000,
found 475.3002.
5-(t-Butyldimethylsilyl)oxy-6,6-dimethyl-2-cyclohexe-
none. 3-Hydroxy-2,2-dimethylcyclohexanone (2.39 g, 16.8
mmol), prepared according to literature procedures,40
was dissolved in CH2Cl2 (50 mL), cooled to ꢀ78 ꢂC, and
treated with 2,6-lutidine (2.92 mL, 25.2 mmol) and
TBSOTf (3.84 mL, 16.8 mmol). The solution was stirred
for 30 min atꢀ78 ꢂC, quenched with H2O (30 mL), and
extracted with CH2Cl2. The combined organics were
dried over MgSO4 and concentrated in vacuo. Silica gel
chromatography (3% EtOAc/hexanes) provided 3-(t-
butyldimethylsilyl)oxy-2,2-dimethylcyclohexanone as a
General HWE Coupling: 2,2-difluorocalcitriol analogues
(ꢀ)-7a and 7b. Prior to reaction, phosphine oxide
(ꢁ)-14 and C,D-ring ketone (+)-1539 were
azeotropically dried with benzene and left under vacuum
for 48 h. To a solution of (ꢁ)-14 (63 mg, 0.102 mmol)
in THF (1.1 mL) at ꢀ78 ꢂC was added a solution of
n-BuLi (64 mL, 0.101 mmol, 1.59 M in hexanes). After
stirring in the dark for 1 h, a precooled (ꢀ78 ꢂC) solu-
tion of (+)-15 (32 mg, 0.08 mmol) in THF (0.3 mL) was
slowly cannulated into the flask containing the red-
orange ylide. The reaction mixture was maintained at
ꢀ78 ꢂC (3 h) then warmed to ꢀ40 ꢂC (2 h). During this
time the red-orange color faded to light yellow. The
reaction was quenched with H2O (3 mL), extracted with
Et2O (3ꢃ5 mL), dried (MgSO4), filtered, and con-
centrated to give a crude product which was purified by
flash silica gel chromatography (1% EtOAc/hexanes) to
give a mixture of silyl protected analogues (23 mg,
57%), followed by recovered (+)-15 (elution with 20%
EtOAc/hexanes, 7 mg, 37%) and (ꢁ)-14 (elution with
50% EtOAc/hexanes, 27 mg, 43%).
1
colorless oil: H NMR (CDCl3) d 3.64 (dd, J=7.2, 2.8
Hz, 1H), 2.34 (apparent t, J=6.8 Hz, 2H), 2.04–1.85 (m,
2H), 1.77–1.66 (m, 1H), 1.65–1.53 (m, 1H) 1.08 (s, 3H),
1.04 (s, 3H), 0.84 (s, 9H), 0.01 (s, 3H), 0.00 (s, 3H); 13C
NMR (CDCl3) d 214.6, 78.2, 51.4, 37.1, 29.2, 25.6, 23.1,
20.4, 17.9, ꢀ4.4, ꢀ5.2; IR (CHCl3) 2954, 2858, 1713,
1472, 1255, 1120, 1082, 1002, 869, 835, 775; HRMS (EI)
m/z (M+) calcd 256.1859 for C14H28O2Si, found 256.1856.
3-(t-Butyldimethylsilyl)oxy-2,2-dimethylcyclohexanone
(1.53 g, 5.96 mmol) in THF (15 mL) was cooled to
ꢀ78 ꢂC and then cannulated into a freshly prepared
solution of LDA(6.55 mL, 1 M in THF/hexanes). The
reaction mixture was stirred for 15 min and then cooled
in liquid nitrogen until the THF just begins to form a
slurry. Asolution of PhSeCl (3.30 g, 17.2 mmol) in
CH2Cl2 (15 mL) was added all at once. The reaction
temperature increased to ꢀ55 ꢂC upon addition. Dilute
HCl (6.5 mL, 1 N aq) was added and the reaction mix-
ture was warmed to rt, dried (MgSO4), and the organic
solvents removed in vacuo. The residue was taken up in
CH2Cl2 (40 mL), the mixture cooled to 0 ꢂC, and H2O2
(3 mL, 29% aqueous) was added over 1 h. The reaction
mixture was dried over MgSO4, the organic solvents
removed in vacuo, and the residue purified by silica gel
chromatography (3% EtOAc/hexanes) to provide the
desired 5-(t-butyldimethylsilyl)oxy-6,6-dimethyl-2-cyclo-
Half of the mixture of silyl-protected analogues was
dissolved in THF (1.5 mL), treated with tetra-
butylammonium fluoride (TBAF, 68 mL, 0.068 mmol,
1 M in THF), allowed to stir at rt overnight, and con-
centrated to an oil. Purification by flash silica gel
chromatography (40% EtOAc/hexanes) gave 4.2 mg
(58%) of a mixture of (ꢀ)-7a and 7b. The other half of
this mixture was dissolved in EtOH (3.0 mL), cooled to
0 ꢂC, and treated with a solution of hydrogen fluoride
(100 mL, 49% aq HF), warmed to room temperature
over 2 h, and treated with additional HF solution
(3ꢃ100 mL) to complete global deprotection. Upon
completion by TLC analysis (50% EtOAc/hexanes), the
reaction mixture was cautiously quenched with satu-
rated NaHCO3 solution and subjected to a general
extraction. Purification by flash silica gel chroma-
tography (40% EtOAc/hexanes) afforded 3.4 mg (47%)
of the diastereomeric mixture. The combined mix-
ture of diastereomers was subjected to HPLC
separation (CHIRALCEL1 OJ semipreparative column,
15% i-PrOH/hexanes, 3 mL/min) to afford only pure
diastereomer 7a. (ꢀ)-7a: (3 mg, 12%, 1a,3b, Rf 37.2
1
hexenone (757 mg, 50%) as a colorless oil: H NMR
(CDCl3) d 6.73 (ddd, J=10.0, 5.2, 3.2 Hz, 1H), 5.96
(ddd, J=10.0, 2.2, 1.6 Hz, 1H), 3.83 (dd, J=7.6, 4.8 Hz,
1H), 2.53 (ddt, J=18.8, 4.8, 1.6 Hz, 1H), 2.39 (dddd,
J=18.8, 7.6, 3.0, 2.6 Hz, 1H), 1.12 (s, 3H), 1.05 (s, 3H),
0.89 (s, 9H), 0.07 (s, 3H), 0.05 (s, 3H); 13C NMR
(CDCl3) d 204.1, 144.5, 128.4, 74.4, 48.6, 32.8, 25.7,
21.3, 18.2, 17.9, ꢀ4.3, ꢀ5.0; IR (CDCl3) 2956, 2931,
2857, 1681, 1253, 1099, 876, 837, 776; HRMS (EI) m/z
(M+) calcd 254.1702 for C14H26O2Si, found 254.1698.
1
min); [a]-12.6 (c 1.0, CHCl3); H NMR (CDCl3) d 6.45
(d, J=11.2 Hz, 1H), 6.02 (d, J=11.2 Hz, 1H), 5.62–5.55
(m, 1H), 5.26–5.20 (m, 1H), 4.60–4.46 (m, 1H), 4.28–
4.15 (m, 1H), 2.81 (dd, J=14.0, 4.8 Hz, 1H), 2.68 (d,
J=14.0 Hz, 1H), 2.45 (ddd, J=14.0, 4.8, 3.8 Hz, 1H),
1.22 (s, 6H), 0.94 (d, J=6.4 Hz, 3H), 0.55 (s, 3H); 13C
NMR (CDCl3) d 144.8, 141.4, 129.0, 126.9, 118.1 (t,
J=247.0 Hz), 116.7, 115.7, 71.9 (t, J=22.0 Hz), 71.1,
69.4–68.5 (m), 56.5, 56.4, 46.0, 44.4, 40.4, 39.3, 36.3,
36.1, 29.4, 29.2, 27.6, 23.7, 22.2, 20.8, 18.8, 12.0; 19F
NMR (CDCl3, CFCl3) d ꢀ121.6 to ꢀ126.0 (m); IR
1,3-trans Bis protected diols (ꢁ)-16. To ethyl phenyl-
thioacetate41 (3.23 g, 16.5 mmol) in THF (30 mL) at
ꢀ78 ꢂC was added a freshly prepared solution of LDA
(15.0 mL, 1 M in THF/hexanes) and the reaction mix-
ture was stirred for 30 min. Asolution of 5-( t-butyl-
dimethylsilyl)oxy-6,6-dimethyl-2-cyclohexenone (952 mg,
3.74 mmol) in THF (30 mL) was added, the reaction
was warmed to ꢀ30 ꢂC, stirred for 12 h, quenched with
H2O (30 mL), and extracted with CH2Cl2. The com-
bined organics were washed once with brine and dried

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2361
over MgSO4, then concentrated in vacuo. Silica gel
chromatography of the residue (3% EtOAc/hexanes)
provided two diastereomeric thioethers (609 mg, 36%;
Likewise, the more polar thioether (504 mg, 1.12 mmol),
MeOH (30 mL), and NaBH4 (13 mg, 1.5Hꢀ equiv)
provided two diastereomeric alcohols (257 mg, 51%;
162 mg, 31%) as colorless oils. Less polar, 1,3-cis alcohol:
1H NMR (CDCl3) d 7.46–7.22 (m, 5H), 4.19–4.05 (m,
2H), 3.72–3.68 (m, 1H), 3.61 (d, J=6.8 Hz, 1H), 3.53–
3.49 (m, 1H), 2.69–2.57 (m, 1H), 2.10–2.02 (m, 1H),
1.91–1.79 (m, 2H), 1.75–1.66 (m, 1H), 1.26 (t, J=7.2
Hz, 3H), 1.14 (s, 3H), 0.91 (s, 9H), 0.90 (s, 3H), 0.10 (s,
3H), 0.09 (s, 3H); 13C NMR (CDCl3) d 171.5, 134.7,
131.9, 128.9, 127.4, 77.5, 75.7, 61.1, 57.9, 37.5, 33.5,
32.2, 28.3, 25.7, 24.6, 24.3, 17.8, 14.0, ꢀ4.7, ꢀ5.3; IR
(CHCl3) 3512, 2954, 2930, 2858, 1733, 1472, 1257, 1150,
1095, 1052, 837, 777; HRMS (EI) m/z (M+H+) calcd
453.2495 for C24H40O4SSi, found 453.2489. More polar
1,3-trans alcohol: 1H NMR (CDCl3) d 7.46–7.21 (m,
5H), 4.17–4.04 (m, 2H), 3.79 (dd, J=11.6, 4.8 Hz, 1H),
3.62–3.58 (m, 1H), 3.55 (d, J=8.0 Hz, 1H), 2.49–2.37
(m, 1H), 1.98–1.91 (m, 1H), 1.83–1.76 (m, 1H), 1.58–
1.38 (m, 3H), 1.17 (t, J=7.2 Hz, 3H), 1.04 (s, 3H), 0.89
(s, 9H), 0.83 (s, 3H), 0.06 (s, 3H), 0.04 (s, 3H); 13C
NMR (CDCl3) d 171.7, 134.6, 132.0, 128.9, 127.4, 76.6,
72.5, 61.1, 57.7, 40.0, 34.4, 33.3, 32.6, 25.9, 24.4, 18.1,
18.0, 14.0, ꢀ4.4, ꢀ5.1; IR (CHCl3) 3428 (br), 2954,
2929, 2857, 1733, 1472, 1255, 1153, 1071, 980, 832, 775;
HRMS (EI) m/z (M+H+) calcd 453.2495 for
C24H40O4SSi, found 453.2585.
1
504 mg, 30%). Less polar thioether: H NMR (CDCl3)
d 7.48–7.24 (m, 5H), 4.22–4.05 (m, 2H), 3.82 (dd,
J=4.2, 2.2 Hz, 1H), 3.57 (d, J=7.6 Hz, 1H), 2.79
(ddddd, J=12.4, 12.4, 7.6, 4.8, 3.8 Hz, 1H), 2.70 (ddd,
J=14.6, 4.8, 2.0 Hz, 1H), 2.44 (dd, J=14.6, 12.4 Hz,
1H), 2.09 (ddd, J=13.6, 12.4, 2.2 Hz, 1H), 1.83 (dddd,
J=13.6, 4.2, 3.8, 2.0 Hz, 1H), 1.20 (t, J=7.2 Hz, 3H),
1.13 (s, 3H), 1.07 (s, 3H), 0.83 (s, 9H), 0.03 (s, 3H), 0.02
(s, 3H); 13C NMR (CDCl3) d 213.0, 171.1, 134.1, 132.2,
129.1, 127.8, 77.7, 61.4, 57.4, 50.1, 41.0, 34.8, 31.8, 25.8,
24.3, 21.3, 18.0, 14.0, ꢀ4.5, ꢀ5.1; IR (CHCl3) 2956,
2930, 2857, 1732, 1714, 1472, 1258, 1151, 1063, 1025,
832, 776; HRMS (EI) m/z (M+H+) calcd 451.2338 for
C24H38O4SSi, found 451.2343. More polar thioether: 1H
NMR (CDCl3) d 7.47–7.23 (m, 5H), 4.19–4.06 (m, 2H),
3.84 (dd, J=3.8, 2.2 Hz, 1H), 3.62 (d, J=6.8 Hz, 1H),
2.83 (ddddd, J=12.4, 12.0, 6.8, 4.8, 4.0 Hz, 1H), 2.68
(dd, J=14.4, 12.4 Hz, 1H), 2.37 (ddd, J=14.4, 4.8, 2.0 Hz,
1H), 2.12 (dddd, J=14.0, 4.0, 3.8, 2.0 Hz, 1H), 2.01 (ddd,
J=14.0, 12.0, 2.2 Hz, 1H), 1.19 (t, J=7.2 Hz, 3H), 1.14 (s,
3H), 1.07 (s, 3H), 0.85 (s, 9H), 0.06 (s, 3H), 0.04 (s, 3H);
13C NMR (CDCl3) d 212.9, 171.0, 133.6, 132.6, 129.0,
127.9, 77.7, 61.3, 57.0, 50.2, 40.9, 34.7, 32.4, 25.7, 24.3,
21.3, 18.0, 14.1, ꢀ4.5, ꢀ5.1; IR (CHCl3) 2956, 2930,
2857, 1732, 1714, 1472, 1257, 1154, 1102, 1063, 1027,
833, 776; HRMS (EI) m/z (M+H+) calcd 451.2338 for
C24H38O4SSi, found 451.2338.
Each of the 1,3-trans alcohols from above was protected
as follows: to a solution of alcohol (1 equiv) in CH2Cl2
at ꢀ30 ꢂC were added 2,6-lutidine (1.25 equiv) and
TBSOTf (1.1 equiv). The reaction mixture was stirred
for 30 min, quenched with H2O, and extracted with
CH2Cl2. The combined organics were dried over
MgSO4, the solvents removed in vacuo, and the residue
purified by column chromatography to provide two 1,3-
trans bis silyl ethers (ꢁ)-16 both as a colorless oils. First
To a solution of the less polar thioether (585 mg, 1.30
mmol) in MeOH (40 mL) was added NaBH4 (5 mg,
1.5Hꢀ equiv) in portions until starting material was
consumed. The reaction was quenched with H2O and
extracted with EtOAc. The combined organics were
washed with brine and dried over MgSO4. After
removal of the organic solvents in vacuo, silica gel
chromatography provided two diastereomeric alcohols
(162 mg, 27%; 362 mg, 62%) as colorless oils. Less
polar 1,3-cis alcohol: 1H NMR (CDCl3) d 7.47–7.22 (m,
5H), 4.19–4.02 (m, 2H), 3.95 (d, J=10 Hz, 1H), 3.70–
3.66 (m, 1H), 3.56 (d, J=8.0 Hz, 1H), 3.55–3.50 (m,
1H), 2.66–2.54 (m, 1H), 2.26–2.18 (m, 1H), 1.86–1.58
(m, 3H), 1.18 (t, J=7.2 Hz, 3H), 1.14 (s, 3H), 0.89 (s,
9H), 0.88 (s, 3H), 0.08 (s, 3H), 0.07 (s, 3H); 13C NMR
(CDCl3) d 171.5, 134.0, 132.5, 128.9, 127.6, 77.5, 75.7,
61.0, 57.2, 37.6, 33.4, 32.4, 28.1, 25.8, 24.7, 24.4, 17.8,
14.1, ꢀ4.7, ꢀ5.2; IR (CHCl3) 3511 (br), 2954, 2930,
2858, 1732, 1472, 1258, 1151, 1094, 1052, 838, 778;
HRMS (EI) m/z (M+H+) calcd 453.2495 for
C24H40O4SSi, found 453.2494. More polar 1,3-trans
1
1,3-trans bis silyl ether (197 mg, 347 mmol, 97%); H
NMR (CDCl3) d 7.47–7.20 (m, 5H), 4.20–4.10 (m, 1H),
4.09–3.99 (m, 1H), 3.75 (dd, J=7.2, 4.4 Hz, 1H), 3.56–
3.53 (m, 1H), 3.48 (d, J=8.0 Hz, 1H), 2.42–2.29 (m,
1H), 2.11–2.02 (m, 1H), 1.66–1.56 (m, 1H), 1.52–1.43
(m, 1H), 1.29–1.18 (m, 1H), 1.18 (t, J=7.2 Hz, 3H), 0.93
(s, 3H), 0.884 (s, 9H), 0.882 (s, 9H), 0.78 (s, 3H), 0.06 (s,
3H), 0.03 (s, 3H), 0.02 (s, 3H), 0.01 (s, 3H); 13C NMR
(CDCl3) d 171.6, 134.2, 132.4, 128.9, 127.5, 72.7, 61.0,
57.5, 42.6, 33.2, 32.4, 31.1, 25.9, 25.8, 18.1, 18.0, 14.0,
ꢀ4.0, ꢀ4.4, ꢀ5.0; IR (CHCl3) 2955, 2929, 2857, 1735,
1472, 1256, 1096, 1072, 833, 774; HRMS (EI) m/z
(M+H+) calcd 567.3360 for C30H54O4SSi2, found
567.3349. Second 1,3-trans bis silyl ether: (182 mg, 322
1
mmol, 93%); H NMR (CDCl3) d 7.45–7.20 (m, 5H),
1
alcohol: H NMR (CDCl3) d 7.46–7.20 (m, 5H), 4.19–
4.11 (ddd, J=14, 7.2, 1.6 Hz, 2H), 3.74 (dd, J=11.2, 4.8
Hz, 1H), 3.59–3.56 (m, 1H), 3.54 (d, J=8.0 Hz, 1H),
2.43–2.31 (m, 1H), 1.95–1.87 (m, 1H), 1.66–1.36 (m, 3H),
1.19 (t, J=7.2 Hz, 3H), 0.94 (s, 3H), 0.90 (s, 9H), 0.87
(s, 9H), 0.79 (s, 3H), 0.06 (s, 3H), 0.04 (s, 3H), 0.01 (s, 3H),
0.00 (s, 3H); 13C NMR (CDCl3) d 171.7, 134.8, 131.8,
128.9, 127.3, 76.7, 72.7, 61.0, 57.8, 40.5, 35.1, 33.3, 32.7,
25.91, 25.86, 25.2, 18.5, 18.1, 18.0, 14.1, ꢀ4.1, ꢀ4.3, ꢀ5.0,
ꢀ5.1; IR (CHCl3) 2929, 2857, 1737, 1472, 1368, 1256,
1150, 909, 873, 833, 775, 691; HRMS (EI) m/z (M+H+)
calcd 567.3360 for C30H54O4SSi2, found 567.3355.
4.00 (m, 2H), 3.80 (dd, J=11.4, 4.6 Hz, 1H), 3.59–3.55
(m, 1H), 3.52 (d, J=8.8 Hz, 1H), 2.47–2.35 (m, 1H),
2.26–2.18 (m, 1H), 1.68–1.50 (m, 3H), 1.17 (t, J=7.2
Hz, 3H), 1.03 (s, 3H), 0.88 (s, 9H), 0.82 (s, 3H), 0.03 (s,
3H), 0.01 (s, 3H); 13C NMR (CDCl3) d 171.5, 134.1,
132.3, 128.9, 127.5, 76.6, 72.5, 61.0, 57.4, 40.0, 34.4,
33.0, 32.9, 25.8, 24.4, 18.1, 18.0, 14.0, ꢀ4.5, ꢀ5.1; IR
(CHCl3) 3426 (br), 2954, 2930, 2857, 1732, 1472, 1256,
1150, 1070, 833, 775; HRMS (EI) m/z (M+H+) calcd
453.2495 for C24H40O4SSi, found 453.2503.

2362
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
1,3-trans unsaturated ester (ꢁ)-18. Each 1,3-trans bis
silyl ether (ꢁ)-16 was separately subjected to the fol-
lowing oxidation: to a solution of bis silyl ether (182 mg,
321 mmol) in CH2Cl2 (7–8 mL) at 0 ꢂC was added a
solution of m-CPBA(55 mg, 321 mmol) in CH2Cl2 (1
mL) dropwise, maintaining the solution at less than 5 ꢂC
at all times. Upon disappearance of the starting mate-
rial, the reaction was quenched with saturated NaHCO3
solution, dried over MgSO4, and chromatographed over
silica to provide two diastereomeric sulfoxides as color-
less oils. From first 1,3-trans bis silyl ether above: (less
polar sulfoxide, 52 mg, 90 mmol, 28%); 1H NMR
(CDCl3) d 7.70–7.43 (m, 5H), 3.88–3.74 (m, 3H), 3.63–
3.60 (m, 1H), 3.33 (d, J=6.8 Hz, 1H), 2.95–2.82 (m,
1H), 1.91–1.68 (m, 3H), 1.56–1.44 (m, 1H), 0.99 (t,
J=7.2 Hz, 3H), 0.96 (s, 3H), 0.93 (s, 9H), 0.87 (s, 9H),
0.79 (s, 3H), 0.09 (s, 3H), 0.05 (s, 3H), 0.03 (s, 3H), 0.02
(s, 3H); 13C NMR (CDCl3) d 167.1, 142.8, 131.8, 128.9,
125.5, 78.3, 76.7, 72.5, 61.1, 40.6, 34.1, 33.5, 31.1, 25.92,
25.88, 25.2, 18.5, 18.1, 18.0, 13.9, ꢀ4.0, ꢀ4.3, ꢀ5.1; IR
(CHCl3) 2955, 2930, 2885, 2857, 1726, 1472, 1252, 1098,
1083, 1070, 1053, 833, 775; HRMS (EI) m/z (M+H+)
calcd 583.3309 for C30H54O5Si2S, found 583.3313;
(more polar sulfoxide, 112 mg, 192 mmol, 60%); 1H
NMR (CDCl3) d 7.63–7.47 (m, 5H), 4.10–3.89 (m, 2H),
3.70 (dd, J=10.8, 4.4 Hz, 1H), 3.59–3.56 (m, 1H), 3.23
(d, J=8.0 Hz, 1H), 2.54–2.40 (m, 1H), 1.86–1.78 (m,
1H), 1.68–1.59 (m, 2H), 1.52–1.41 (m, 1H), 1.08 (t,
J=7.2 Hz, 3H), 0.92 (s, 3H), 0.86 (s, 9H), 0.82 (s, 9H),
0.78 (s, 3H), 0.08 (s, 6H), 0.02 (s, 3H), ꢀ0.01 (s, 3H);
13C NMR (CDCl3) d 166.2, 142.0, 131.5, 129.1, 124.7,
76.2, 75.5, 72.4, 61.2, 40.6, 34.4, 33.0, 29.7, 25.8, 25.8,
25.1, 18.5, 18.0, 14.0, ꢀ4.1, ꢀ4.4, ꢀ5.0, ꢀ5.2; IR
(CHCl3) 2955, 2929, 2857, 2855, 1732, 1472, 1255, 1094,
1057, 834, 775; HRMS (EI) m/z (M+H+) calcd
583.3309 for C30H54O5Si2S, found 583.3318. From sec-
ond 1,3-trans bis silyl ether above: (less polar sulfoxide,
at reflux for 36 h. After removing the solvent in vacuo,
silica gel chromatography (1–3% EtOAc/hexanes) pro-
vided (ꢁ)-18 (212 mg, 81%) as a colorless oil with the
1
following physical characteristics: H NMR (CDCl3) d
5.66 (br s, 1H), 4.18–4.06 (m, 2H), 3.69 (dd, J=8.0, 4.0
Hz, 1H), 3.67 (dd, J=6.4, 3.6 Hz, 1H), 3.03 (dd,
J=14.4, 6.4 Hz, 1H), 2.94 (ddd, J=14.4, 3.6, 1.6 Hz,
1H), 2.38 (ddd, J=14.0, 4.0, 0.6 Hz, 1H), 2.22 (ddd,
J=14.0, 4.0, 1.2 Hz, 1H), 1.25 (t, J=7.2 Hz, 3H), 0.94
(s, 3H), 0.93 (s, 3H), 0.87 (s, 9H), 0.85 (s, 9H), 0.07 (s,
3H), 0.04 (s, 3H), 0.018 (s, 3H), 0.016 (s, 3H); 13C NMR
(CDCl3) d 166.4, 157.7, 116.9, 75.6, 74.9, 59.5, 41.8,
40.8, 33.4, 25.82, 25.76, 22.3, 21.1, 18.03, 17.97, 14.3,
ꢀ4.1, ꢀ4.6, ꢀ5.0, ꢀ5.1; IR (CHCl3) 2954, 2929, 2894,
2857, 1716, 1652, 1472, 1251, 1162, 1091, 1045, 835,
776; HRMS (EI) m/z (M+H+) calcd 457.3169 for
C24H48O4Si2, found 457.3176.
1,3-trans phosphine oxide (ꢁ)-20. To a solution of
(ꢁ)-18 (120 mg, 263 mmol) in Et2O (5 mL) atꢀ40 ꢂC
was added a solution of lithium aluminum hydride
(LAH, 660 mL, 1 M in Et2O) and the reaction mixture
was warmed to 0 ꢂC over 2 h. The reaction was quen-
ched with a few drops of 1 N NaOH and the pre-
cipitated salts were washed with several portions of
EtOAc. The combined organics were dried over MgSO4
and the solvents removed in vacuo. Chromatographic
purification of the residue over silica gel (5–10%
EtOAc/hexanes) provided the corresponding allylic
alcohol (104 mg, 96%) as a colorless oil: 1H NMR
(CDCl3) d 5.49–5.39 (m, 1H), 4.11 (d, J=6.8 Hz,
2H), 3.66–3.59 (m, 2H), 2.42–2.30 (m, 2H), 2.22–2.08
(m, 2H), 0.91 (s, 6H), 0.88 (s, 9H), 0.87 (s, 9H), 0.05
(s, 3H), 0.04 (s, 3H), 0.02 (s, 3H), 0.01 (s, 3H); 13C
NMR (CDCl3) d 138.4, 124.4, 75.12, 75.09, 58.7, 41.1,
40.8, 32.9, 25.9, 25.8, 21.8, 21.7, 18.1, 18.0, ꢀ4.1, ꢀ4.2,
ꢀ5.0; IR (CHCl3) 3317 (br), 2956, 2930, 2887, 2857,
1472, 1252, 1090, 878, 836, 724; HRMS (EI) m/z
(M+NH+4 ) calcd 432.3329 for C22H46O3Si2, found
432.3338.
1
89 mg, 153 mmol, 44%); H NMR (CDCl3) d 7.70–7.45
(m, 5H), 3.96–3.80 (m, 3H), 3.64–3.60 (m, 1H), 3.45 (d,
J=5.6 Hz, 1H), 2.95–2.86 (m, 1H), 1.86–1.72 (m, 3H),
1.01 (t, J=7.2 Hz, 3H), 0.96 (s, 3H), 0.93 (s, 9H), 0.88
(s, 9H), 0.78 (s, 3H), 0.11 (s, 3H), 0.06 (s, 3H), 0.05 (s,
3H), 0.03 (s, 3H); 13C NMR (CDCl3) d 167.1, 142.6,
131.8, 128.9, 125.5, 77.7, 76.4, 72.6, 61.1, 40.6, 35.6,
31.0, 30.0, 25.92, 25.88, 25.2, 18.4, 18.1, 18.0, 14.0, ꢀ4.0,
ꢀ4.3, ꢀ5.0, ꢀ5.1; IR (CHCl3) 2955, 2930, 2857, 1725,
1250, 1074, 834, 775; HRMS (EI) m/z (M+H+) calcd
583.3309 for C30H54O5Si2S, found 583.3324; (more
polar sulfoxide, 81 mg, 140 mmol, 40%); 1H NMR
(CDCl3) d 7.63–7.47 (m, 5H), 4.00–3.90 (m, 1H), 3.86–
3.72 (m, 2H), 3.58–3.55 (m, 1H), 3.18 (d, J=9.6 Hz,
1H), 2.82–2.68 (m, 1H), 2.20–2.11 (m, 1H), 1.70–1.35
(m, 3H), 1.01 (t, J=7.6 Hz, 3H), 0.96 (s, 3H), 0.91 (s, 9H),
0.90 (s, 9H), 0.81 (s, 3H), 0.08 (s, 3H), 0.04 (s, 3H), 0.03
(s, 3H), 0.02 (s, 3H); 13C NMR (CDCl3) d 165.6, 141.9,
131.1, 128.9, 124.5, 76.4, 74.5, 72.2, 61.0, 40.6, 34.9, 33.2,
29.7, 25.94, 25.91, 25.1, 18.5, 18.1, 18.0, 13.9, ꢀ3.9, ꢀ4.3,
ꢀ5.17, ꢀ5.12; IR (CHCl3) 2955, 2928, 2881, 2855, 1722,
1471, 1257, 1072, 833, 774; HRMS (EI) m/z (M+H+)
calcd 583.3309 for C30H54O5Si2S, found 583.3318.
Following the procedure for the synthesis of (ꢁ)-14
above, this allylic alcohol (157 mg, 378 mmol) in THF (5
mL) was tosylated with n-BuLi (310 mL, 1.58 M in hex-
anes) and TsCl (97 mg, 510 mmol) in THF (1 mL),
treated with potassium diphenylphosphide (KPPh2, 1.2
mL, 0.5 M in THF) at 0 ꢂC, oxidized with H2O2 (200
mL, 29% aq), and purified upon passage through silica
gel (30–40% EtOAc/hexanes) to provide (ꢁ)-20 (161
mg, 269 mmol, 71%) as a white solid: 1H NMR (CDCl3)
d 7.75–7.40 (m, 10H), 5.26–5.17 (m, 1H), 3.54 (dd,
J=6.8, 3.6 Hz, 1H), 3.51 (dd, J=8.0, 4.0 Hz, 1H), 3.18–
2.97 (m, 2H), 2.31–2.22 (m, 1H), 2.06–1.96 (m, 1H),
1.81–1.72 (m, 1H), 0.84 (s, 9H), 0.83 (s, 3H), 0.81 (s,
3H), 0.809 (s, 9H), ꢀ0.026 (s, 3H), ꢀ0.029 (s, 3H),
ꢀ0.04 (s, 3H), ꢀ0.05 (s, 3H); 13C NMR (CDCl3) d 139.7
(d, J=12.1 Hz), 133.0 (d, J=26.6 Hz), 132.0 (d, J=27.4
Hz), 131.7, 131.0 (t, J=10.2 Hz), 128.4 (dd, J=11.4, 3.8
Hz), 112.7 (d, J=9.1 Hz), 75.2 (d, J=1.5 Hz), 74.5 (d,
J=1.5 Hz), 40.9, 40.6 (d, J=2.2 Hz), 33.0 (d, J=1.5
Hz), 30.2 (d, J=69.8 Hz), 25.8, 25.7, 22.0, 21.2, 18.0,
17.9, ꢀ4.2, ꢀ5.0, ꢀ5.1; IR (CHCl3) 3058, 2956, 2929,
2893, 2856, 1472, 1462, 1438, 1361, 1252, 1201, 1105,
Amixture of all four aforementioned sulfoxides (334
mg, 573 mmol total) was taken up in benzene and heated

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2363
1076, 1054, 877, 862, 838, 820, 804, 775; HRMS (EI)
m/z (M+H+) calcd 599.3506 for C34H55O3Si2P, found
599.3513.
1.74–1.60 (m, 2H), 1.46–1.37 (m, 1H), 1.13 (t, J=7.2
Hz, 3H), 0.96 (s, 3H), 0.90 (s, 9H), 0.85 (s, 9H),
0.81 (s, 3H), 0.09 (s, 3H), 0.03 (s, 3H), 0.01 (s, 3H),
ꢀ0.05 (s, 3H); 13C NMR (CDCl3) d 172.1, 133.3, 133.0,
128.9, 128.1, 73.0, 72.8, 61.1, 54.1, 42.6, 33.2, 32.4, 31.1,
25.9, 25.8, 18.1, 18.0, 14.0, ꢀ4.0, ꢀ4.4, ꢀ5.0; IR
(CHCl3) 2955, 2929, 2857, 1738, 1472, 1256, 1150, 1074,
872, 837, 774; HRMS (EI) m/z (M+H+) calcd 567.3360
for C30H54O4SSi2, found 567.3366.
2,2-Dimethyl-19-norcalcitriol analogues (+)-8a and
(+)-8b. Following the general procedure for HWE
coupling found above for (ꢀ)-7a and 7b, phosphine
oxide (ꢁ)-20 (92 mg, 0.154 mmol) in THF (1.5 mL),
PhLi (95 mL, 1.68 M solution in cyclohexane/Et2O), and
(ꢁ)-15 (54 mg, 0.154 mmol) in THF (0.75 mL) provided
11 mg (10%) of a mixture of silyl protected analogues as
a colorless oil after column chromatography (0–10%
EtOAc/hexanes). This mixture was immediately taken
up in EtOH (1.5 mL) and 2 mL of 20% HF in EtOH/
H2O was added. After 48 h at rt, water was added and
the mixture was extracted with EtOAc. The organic
fractions were combined, washed with brine, dried over
MgSO4, and concentrated in vacuo. Silica gel chroma-
tography (3–6% EtOH/CH2Cl2) provided a mixture of
(+)-8a and (+)-8b as a white solid. Further purification
by chiral HPLC (Daicel-Chiralpak AS, 5% EtOH/hex-
anes) provided the more polar analogue (+)-8a [1.0 mg,
1.5% from (+)-15] followed by analogue (+)-8b [1.9
mg, 2.8% from (+)-15] both as white solids. (+)-8a:
[a]+29 (c 0.7, EtOH); 1H NMR (CDCl3) d 6.30 (d,
J=11.0 Hz, 1H), 5.83 (d, J=11.0 Hz, 1H), 3.68–3.60
(m, 2H), 2.83–2.74 (m, 1H), 2.66 (dd, J=13.6, 3.8 Hz,
1H), 2.56 (dd, J=13.6, 4.0 Hz, 1H) 2.35 (dd, J=14.0,
7.6 Hz, 1H), 2.23 (dd, J=14.0, 6.8 Hz, 1H), 1.22 (s, 6H),
1.06 (s, 3H), 1.04 (s, 3H), 0.93 (d, J=6.4 Hz, 3H), 0.55
(s, 3H); 13C NMR (CDCl3) d 143.0, 131.5, 123.7, 115.2,
77.2, 75.4, 75.0, 71.1, 56.5, 56.3, 45.8, 44.4, 40.8, 40.6,
40.3, 36.4, 36.1, 32.7, 31.6, 29.4, 29.2, 28.9, 23.5, 22.6,
22.3, 21.2, 20.8, 20.7, 18.8, 14.1, 12.1; UV (EtOH) lmax
251 nm (e 17600). IR (CHCl3) 3460 (br), 3021, 2962,
2928, 2854, 1602, 1458, 1377, 1261, 1097, 1031, 811;
HRMS (EI) m/z (M+) calcd 432.3603 for C28H48O3,
found 432.3597. (+)-8b: [a]+21 (c 1.4, EtOH); 1H
NMR (CDCl3) d 6.29 (d, J=11.2 Hz, 1H), 5.84 (d,
J=11.2 Hz, 1H), 3.70–3.58 (m, 2H), 2.84–2.74 (m, 1H),
2.70 (dd, J=14.0, 4.0 Hz, 1H), 2.59 (dd, J=14.0, 3.4
Hz, 1H), 2.29 (dd, J=13.6, 9.0 Hz, 1H), 2.21 (dd,
J=14.0, 6.4 Hz, 1H), 1.22 (s, 6H), 1.07 (s, 3H), 1.03 (s,
3H), 0.94 (d, J=6.4 Hz, 3H), 0.54 (s, 3H); 13C NMR
(CD3OD) d 143.0, 131.6, 123.7, 115.1, 77.2, 75.6, 74.8,
71.1, 60.4, 56.5, 56.3, 45.8, 44.4, 40.7, 40.4, 40.3, 36.4,
36.1, 32.9, 29.4, 29.2, 28.9, 27.6, 23.5, 22.3, 20.8, 18.8,
14.2, 12.0; UV (EtOH) lmax 252 nm (e 33,000). IR
(CHCl3) 3021, 2927, 1602, 1457, 1377, 1032; HRMS
(EI) m/z (M+) calcd 432.3603 for C28H48O3, found
432.3593.
1,3 cis Unsaturated ester (ꢁ)-19. Following the oxi-
dation procedure of (ꢁ)-16 above, ether (ꢁ)-17 (654 mg,
1.15 mmol) and m-CPBA(198 mg, 1.15 mmol) in
CH2Cl2 (35 mL total) provided two diastereomeric
sulfoxides as colorless oils: (less polar sulfoxide, 271 mg,
464 mmol, 40%); 1H NMR (CDCl3) d 7.70–7.40 (m,
5H), 3.81–3.72 (m, 2H), 3.66 (d, J=10.4 Hz, 1H), 3.58
(dd, J=10.0, 4.0 Hz, 1H), 3.43 (dd, J=10.0, 4.0 Hz,
1H), 2.86–2.77 (m, 1H), 2.21–2.13 (m, 1H), 1.84 (ddd,
J=13.8, 10.0, 5.0 Hz, 1H), 1.73 (ddd, J=14.0, 10.0, 5.2
Hz, 1H), 1.52–1.44 (m, 1H), 0.98 (t, J=7.2 Hz, 3H),
0.96 (s, 3H), 0.91 (s, 9H), 0.86 (s, 12H), 0.13 (s, 3H),
0.09 (s, 3H), 0.00 (s, 3H), ꢀ0.05 (s, 3H); 13C NMR
(CDCl3) d 167.3, 141.8, 132.0, 128.8, 125.8, 75.9, 73.1,
72.9, 61.3, 42.0, 33.5, 32.2, 31.3, 25.9, 25.8, 25.7, 18.1,
18.0, 13.7, ꢀ4.1, ꢀ4.4, ꢀ4.98, ꢀ5.03; IR (CHCl3) 2955,
2930, 2885, 2857, 1726, 1472, 1298, 1253, 1113, 1074,
872, 837, 775; HRMS (EI) m/z (M+H+) calcd 583.3309
for C30H54O5SSi2, found 583.3302; (more polar sulfox-
ide, 349 mg, 598 mmol, 52%); 1H NMR (CDCl3) d 7.58–
7.47 (m, 5H), 3.84–3.74 (m, 1H), 3.71–3.60 (m, 1H), 3.48
(dd, J=11.6, 4.0 Hz, 1H), 3.40 (d, J=12.4 Hz, 1H), 3.34
(dd, J=11.4, 4.2 Hz, 1H), 2.85–2.76 (m, 1H), 2.23–2.15
(m, 1H), 1.86–1.75 (m, 2H), 1.41–1.32 (m, 1H), 0.99 (s,
3H), 0.89 (s, 9H), 0.85 (t, J=7.6 Hz, 3H), 0.83 (s, 3H),
0.82 (s, 9H), 0.10 (s, 3H), 0.08 (s, 3H), ꢀ0.05 (s, 3H),
ꢀ0.13 (s, 3H); 13C NMR (CDCl3) d 165.5, 140.9, 131.2,
129.0, 124.2, 72.9, 72.8, 69.7, 61.1, 43.0, 33.1, 30.6, 29.4,
25.8, 25.71, 25.66, 18.0, 17.8, 13.5, 12.4, ꢀ4.1, ꢀ4.5,
ꢀ4.8, ꢀ5.1; IR (CHCl3) 2955, 2930, 2885, 2857, 1733,
1473, 1253, 1115, 1073, 871, 837, 775; HRMS (EI) m/z
(M+H+) calcd 583.3309 for C30H54O5SSi2, found
583.3296.
Amixture of these diastereomeric sulfoxides (594 mg,
1.02 mmol total) was taken up in benzene and heated at
reflux for 6 h. After removing the solvent in vacuo, silica
gel chromatography (1–3% EtOAc/hexanes) provided
(ꢁ)-19 (489 mg, 100%) as a colorless oil with the fol-
lowing physical characteristics: 1H NMR (CDCl3) d
5.67 (t, J=1.6 Hz, 1H), 4.15 (q, J=7.2 Hz, 2H), 3.82
(ddd, J=13.6, 4.8, 1.6 Hz, 1H), 3.22 (ddd, J=11.6, 7.0,
4.6 Hz, 1H), 2.39–2.30 (m, 1H), 2.13 (ddd, J=13.2, 4.8,
1.6 Hz, 1H), 2.04–1.95 (m, 1H), 1.28 (t, J=7.2 Hz, 3H),
0.97 (s, 3H), 0.90 (s, 9H), 0.89 (s, 9H), 0.88 (s, 3H), 0.12
(s, 3H), 0.05 (s, 3H), 0.04 (s, 3H), 0.03 (s, 3H); 13C
NMR (CDCl3) d 166.2, 155.1, 116.2, 75.8, 75.5, 59.7,
42.6, 42.5, 34.3, 25.82, 25.79, 25.4, 18.02, 18.00, 14.3,
12.1, ꢀ4.1, ꢀ5.0, ꢀ5.1; IR (CHCl3) 2956, 2930, 2894,
2858, 1717, 1652, 1472, 1362, 1253, 1241, 1161, 1140,
1112, 1079, 1051, 871, 837, 804, 775; HRMS (EI) m/z
(M+H+) calcd 457.3169 for C24H48O4Si2, found
457.3178.
Leading to 1,3-cis compounds
1,3-cis Bis protected diol (ꢁ)-17. Each of the 1,3-cis
alcohols from above [see (ꢁ)-16 experimental] was
protected separately in an analogous manner to the
1,3-trans alcohols to afford an identical bis silyl ether
(ꢁ)-17 (from first 1,3-cis alcohol: 397 mg, 701 mmol,
88%; from second 1,3-cis alcohol, 257 mg, 453 mmol,
1
80%) as a colorless oil: H NMR (CDCl3) d 7.49–7.28
(m, 5H), 4.19–3.95 (m, 2H), 3.71 (d, J=12.0 Hz, 1H),
3.50–3.33 (m, 2H), 2.34–2.24 (m, 1H), 2.14–2.06 (m, 1H),

2364
G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
1,3-cis Phosphine oxide (ꢁ)-21. To a solution of (ꢁ)-19
(460 mg, 1.01 mmol) in Et2O (15 mL) at ꢀ78 ꢂC was
added a solution of DIBAL-H (6 mL, 1 M in hexanes)
and the reaction mixture was warmed to room tem-
perature over 2 h. After stirring at room temperature
overnight, the reaction was quenched with dilute HCl
and extracted with several portions of EtOAc. The
combined organics were washed with brine, dried over
MgSO4, and the solvents removed in vacuo. Chroma-
tographic purification of the residue over silica gel pro-
vided the target allylic alcohol (194 mg, 468 mmol, 46%)
(+)-8c and (+)-8d as a white solid. Further purification
by C-18HPLC (69% MeCN/H2O) provided the more
polar analogue (+)-8c [5.8 mg, 30% from (ꢁ)-21], fol-
lowed by analogue (+)-8d [5.7 mg, 30% from (ꢁ)-21]
1
both as white solids. (+)-8c: [a] +94 (c 5.8, EtOH); H
NMR (CDCl3) d 6.31 (d, J=11.2 Hz, 1H), 5.85 (d,
J=11.2 Hz, 1H), 3.53–3.43 (m, 2H), 2.88–2.75 (m, 1H),
2.64–2.46 (m, 3H), 2.30 (dd, J=14.0, 6.8 Hz, 1H), 1.22
(s, 6H), 1.09 (s, 3H), 1.05 (s, 3H), 0.93 (d, J=6.4 Hz,
3H), 0.54 (s, 3H); 13C NMR (CD3OD) d 142.6, 133.0,
122.5, 117.1, 77.2, 76.7, 71.6, 58.1, 57.6, 47.0, 45.4, 42.6,
42.3, 42.0, 37.9, 37.6, 34.1, 30.9, 30.0, 28.9, 25.3, 24.7,
23.4, 22.0, 19.5, 13.1; UV (EtOH) lmax 250 nm (e
344,000). IR (CHCl3) 3362, 2927, 2872, 1616, 1468,
1378, 1214, 1146, 1082, 1052, 1015, 994, 934, 880;
HRMS (EI) m/z (M+) calcd 432.3603 for C28H48O3,
found 432.3601. (+)-8d: [a] +42 (c 5.0, EtOH); 1H
NMR (CDCl3) d 6.33 (d, J=11.6 Hz, 1H), 5.86 (d,
J=11.6 Hz, 1H), 3.58–3.48 (m, 2H), 2.86–2.77 (m, 1H),
2.66–2.51 (m, 3H), 2.30 (dd, J=14.0, 5.6 Hz, 1H), 1.22
(s, 6H), 1.11 (s, 3H), 1.05 (s, 3H), 0.94 (d, J=6.8 Hz,
3H), 0.55 (s, 3H); 13C NMR (CD3OD) d 142.6, 133.0,
122.5, 117.1, 77.1, 76.7, 71.6, 58.1, 57.6, 47.0, 45.4, 42.6,
42.3, 42.0, 37.9, 37.6, 35.0, 30.9, 29.4, 29.3, 28.9, 25.3,
24.7, 23.4, 22.0, 19.5, 13.3, 12.6; UV (EtOH) lmax 250
nm (e 33,000). IR (CHCl3) 3374, 2925, 1618, 1467, 1378,
1082, 1052, 1015, 934, 864; HRMS (EI) m/z (M+) calcd
432.3603 for C28H48O3, found 432.3597.
1
as a colorless oil: H NMR (CDCl3) d 5.46 (tt, J=6.8,
1.6 Hz, 1H), 4.14 (d, J=6.8 Hz, 1H), 3.19 (dd, J=11.4,
5.0 Hz, 1H), 3.13 (dd, J=11.6, 4.8 Hz, 1H), 2.47 (ddd,
J=13.4, 4.6, 1.4 Hz, 1H), 2.28–2.18 (m, 1H), 2.08 (ddd,
J=13.2, 4.8, 1.6 Hz, 1H), 2.00–1.90 (m, 1H), 0.95 (s,
3H), 0.90 (s, 9H), 0.89 (s, 9H), 0.86 (s, 3H), 0.06 (s, 3H),
0.04 (s, 3H), 0.030 (s, 3H), 0.028 (s, 3H); 13C NMR
(CDCl3) d 136.9, 123.9, 76.1, 75.9, 58.6, 42.8, 41.6, 33.8,
25.8, 25.6, 18.02, 17.99, 11.9, ꢀ4.06, ꢀ4.10, ꢀ5.0; IR
(CHCl3) 3313 (br), 2956, 2930, 2887, 2857, 1472, 1463,
1362, 1255, 1078, 1050, 1006, 874, 837, 803, 774, 671;
HRMS (EI) m/z (M+H+) calcd 415.3064 for
C22H46O3Si2, found 415.3058.
In a manner analogous to the synthesis of (ꢁ)-20, this
allylic alcohol (194 mg, 468 mmol), n-BuLi (385 mL,
1.58 M in hexanes), TsCl (120 mg, 632 mmol), KPPh2
(1.4 mL, 0.5 M in THF), and H2O2 (250 mL, 29% aqu-
eous) afforded (ꢁ)-21 (225 mg, 80%) as an amorphous
solid after silica gel chromatography (30–40% EtOAc/
hexanes): 1H NMR (CDCl3) d 7.75–7.42 (m, 10H),
5.37–5.27 (m, 1H), 3.06 (dd, J=14.8, 7.6 Hz, 1H), 3.00
(dd, J=11.0, 5.0 Hz, 1H), 2.85 (dd, J=11.4, 4.6 Hz,
1H), 2.19–2.01 (m, 2H), 1.60–1.49 (m, 1H), 0.87 (s, 9H),
0.86 (s, 12H), 0.75 (s, 3H), 0.06 (s, 6H), 0.04 (s, 3H),
0.03 (s, 3H); 13C NMR (CDCl3) d 138.0 (d, J=11.4 Hz),
132.9 (d, J=47.9 Hz), 131.9 (d, J=47.8 Hz), 131.8 (d,
J=5.4 Hz), 131.1 (t, J=9.1 Hz), 128.5 (d, J=11.4 Hz),
112.8 (d, J=8.4 Hz), 75.7 (d, J=3.1 Hz), 75.2 (d, J=2.3
Hz), 42.6, 41.6 (d, J=2.2 Hz), 33.7, 30.4 (d, J=69.1
Hz), 25.74, 25.71, 25.4, 17.94, 17.91, 11.7, ꢀ4.0, ꢀ4.1,
ꢀ4.9, ꢀ5.0; IR (CHCl3) 3058, 2956, 2929, 2893, 2856,
1472, 1438, 1361, 1252, 1201, 1105, 1076, 1054, 877,
862, 838, 820, 804, 775; HRMS (EI) m/z (M+H+) calcd
599.3506 for C34H55O3Si2P, found 599.3501.
PTH cell culture. Bovine parathyroid glands were
obtained from a local slaughterhouse and transported
to the laboratory in cold PBS. The glands were digested
with collagenase as previously described42 and seeded at
a density of 80,000 cells/cm2 in DMEM–Ham’s F-12
(1:1) containing 4% heat-inactivated newborn calf
serum, 15 mM HEPES, 100 IU/mL penicillin, 100 mg/
mL streptomycin, 5 mg/mL insulin, 2 mM glutamine, 5
mg/mL holo-transferrin and 1% non-essential amino
acids. After 24 h cells were placed in medium containing
0.1% bovine serum albumin in place of the serum.
Except for the initial 24 h, the cells were grown to con-
fluency (6 days) in serum-free medium.
Analysis of parathyroid hormone secretion. Parathyroid
cell cultures were prepared as described above. On the
third day of culture the cells were treated with the vita-
min D compounds (calcitriol or its analogues) at con-
centrations ranging from 10 pM to 100 nM with daily
changes of the medium for 3 days. Steady state PTH
secretion was determined by washing the cells three
times with Dulbecco’s PBS and then placing them in
treatment mediaꢂ for 3 h. The media were collected,
centrifuged at 4 C and analyzed for PTH using CH9
antibody as described previously.43
2,2 - Dimethyl - 19 - norcalcitriol Analogues (+) - 8c and
(+)-8d. Following the general procedure for HWE
coupling found above for (ꢀ)-7a and 7b, (ꢁ)-21 (62 mg,
0.104 mmol), PhLi (62 mL, 1.68 M solution in cyclohex-
ane/Et2O), and (+)-15 (40 mg, 0.114 mmol) gave 32 mg
(42%) of a mixture of silyl protected analogues as a
colorless oil after silica gel chromatography (0–10%
EtOAc/hexanes). The mixture of diastereomers was
then deprotected without separation. The mixture was
taken up in THF (1.5 mL) and Et3N (35 mL) and a
solution of TBAF (250 mL, 1 M in THF) added. After
48 h at room temperature, H2O was added and the
mixture was extracted with EtOAc (3ꢃ). The organic
fractions were combined, washed with brine, dried over
MgSO4 and concentrated in vacuo. Silica gel chroma-
tography (3–6% EtOH/CH2Cl2) provided a mixture of
Acknowledgements
We thank the NIH for support of this work (CA44530 to
G.H.P. and TWK and DK 50583 to S.P.). We thank Prof.
K. Afarinkia (Kings College, UK) for early work at Johns
Hopkins on the Diels–Alder cycloaddition in Scheme 1,
and Dr. Milan Uskokovic for a generous gift of 15.

G. H. Posner et al. / Bioorg. Med. Chem. 10 (2002) 2353–2365
2365
References and Notes
22. Peleg, S.; Nguyen, C.; Woodard, B. T.; Lee, J.-K.; Posner,
G. H. Mol. Endocr. 1998, 12, 525.
1. Feldman, D.; Glorieux, F. H.; Pike, J. W. Vitamin D; Aca-
demic Press: San Diego, CA, 1997.
2. Holick, M. F. Vitamin D; Humana: Totowa, 1999.
3. Bouillon, R.; Okamura, W. H.; Norman, A. W. Endocrine
Rev. 1995, 16, 200.
23. Greising, D. M.; Schwartz, Z.; Posner, G. H.; Sylvia, V. L.;
Dean, D. D.; Boyan, B. D. J. Cell. Phys. 1997, 171, 357.
24. Posner, G. H.; Nelson, T. D.; Guyton, K. Z.; Kensler,
T. W. J. Med. Chem. 1992, 35, 3280.
25. Posner, G. H.; Lee, J. K.; Wang, Q.; Peleg, S.; Burke, M.;
Brem, H.; Dolan, P.; Kensler, T. W. J. Med. Chem. 1998, 41,
3008.
4. Nishii, Y.; Okano, T. Steroids 2001, 66, 137, and references
therein.
5. Posner, G. H.; Johnson, N. J. Org. Chem. 1994, 59, 7855.
6. Posner, G. H.; Cho, C.-G.; Anjeh, T. E. N.; Johnson, N.;
Horst, R. L.; Kobayashi, T.; Okano, T.; Tsugawa, N. J. Org.
Chem. 1995, 60, 4617.
7. Suhara, Y.; Nihei, K.; Tanigawa, H.; Fujishima, T.;
Konno, K.; Nakagawa, K.; Okano, T.; Takayama, H. Bioorg.
Med. Chem. Lett. 2000, 10, 1129.
8. Fujishima, T.; Zhaopeng, L.; Konno, K.; Nakagawa, K.;
Okano, T.; Yamaguchi, K.; Takayama, H. Bioorg. Med.
Chem. 2001, 9, 525.
26. Somjen, D.; Waisman, A.; Lee, J.-K.; Posner, G. H.;
Kaye, A. M. J. Steroid Biochem. Mol. Biol. 2001, 77, 205.
27. Silver, J.; Naveh-Many, T.; Mayer, H.; Schmeltzer, H. J.;
Popovtzer, M. M. J. Clin. Invest. 1986, 78, 1296.
28. Russell, J.; Sherwood, L. M. Trans. Assoc. Amer. Physi-
cians 1987, 100, 256.
29. Demay, M. B.; Kiernan, M. S.; DeLuca, H. F.; Kronen-
berg, H. M. Proc. Nat. Acad. Sci. 1992, 89, 8097.
30. Russell, J.; Ashok, S.et al. J. of Bone Mineral Res. 1999,
14, 1828.
9. Sicinski, R. R.; Prahl, J. M.; Smith, C. M.; DeLuca, H. F.
J. Med. Chem. 1998, 41, 4662.
10. Mikami, K.; Ohba, S.; Ohmura, H.; Kubodera, N. Chir-
ality 2001, 366.
31. Liu, S. M.; Koszewski, N.; Lupez, M.; Malluche, H. H.;
Olivera, A.; Russell, J. Mol. Endocrin. 1996, 10, 206.
32. Mackey, S. L.; Heymont, J. L.; Kronenberg, H. M.;
Demay, M. B. Mol. Endocrin. 1996, 10, 298.
11. DeLuca, H. F., Schnoes, H. K., Ikekawa, N., Tanaka, Y.,
Morisaki, M., Oshida, J. US Patent 4,307,025, 1981.
33. Brown, A. J. Unpublished results.
34. Jones, G.; Strugnell, S. A.; DeLuca, H. F. Physiolog. Revs.
1998, 78, 1193.
35. Jones, G. Principles of Bone Biology; Raisz, L., Belizikian,
J., Roden, G., Eds.; Academic Press, San Diego, CA; 2002;
Vol. 2, Chapter 81, 1377.
12. Scheddim, D.; Moyer, H.; Schonecker, B.; Gliesling, S.;
¨
¨
Reichenbacher, M. Steroids 1998, 63, 633.
13. Schimizu, M.; Twasaki, Y.; Yamada, S. Tetrahedron Lett.
1999, 40, 1699.
14. O’Hagan, D.; Rzepa, H. S. Chem. Commun. 1997, 645.
15. Brown, A. J.; Ritter, C.; Slatopolsky, E.; Muralidharan,
K. R.; Okamura, W. H.; Reddy, G. S. J. Cell. Biochem. 1999,
73, 106.
16. Woodard, B. T.; Posner, G. H. Adv. Cycloadd. 1999, 5, 47,
and references therein.
17. Posner, G. H.; Dai, H.; Bull, D. S.; Lee, J. K.; Eydoux, F.;
Ishihara, Y.; Welsh, W.; Pryor, N.; Petr, S., Jr. J. Org. Chem.
1996, 61, 671.
18. Posner, G. H.; Kinter, C. M. J. Org. Chem. 1990, 55,
3967.
19. Posner, G. H.; Wang, Q.; Han, G.; Lee, J. K.; Crawford,
K.; Zand, S.; Brem, H.; Peleg, S.; Dolan, P.; Kensler, T. W. J.
Med. Chem. 1999, 42, 3425.
20. Perlman, K. L.; Sicinski, R. R.; Schnoes, H. K.; DeLuca,
H. F. Tet. Lett. 1990, 31, 1823.
21. Muralidharan, K. R.; de Lera, A. R.; Isaeff, S. D.;
Norman, A. W.; Okamura, W. H. J. Org. Chem. 1993, 58,
1895.
36. Yamada, S.; Yamamoto, K.; Masuno, H. Curr. Pharma-
ceut. Des. 2000, 6, 733.
37. High Pressure Generator model #PG-200-HPC, LECO
Corporation/Tem-Press Division, PO Box 390, Bellefonte, PA,
16823, USA.
38. Posner, G. H.; Crouch, R. D.; Kinter, C. M.; Carry, J.-C.
J. Org. Chem. 1991, 56, 6981.
39. Baggiolini, E. G.; Iacobelli, J. A.; Hennessy, B. M.;
Batcho, A. D.; Sereno, J. F.; Uskokovic, M. R. J. Org. Chem.
1994, 51, 3098.
40. Mori, K.; Hideto, M. Org. Syn. 1990, 68, 56.
41. Ethyl phenylthioacetate was prepared by the condensation
of thiophenol with ethyl bromoacetate. For a related prepara-
tion, see Hamed, E. A.; Elbardan, A. A.; Moussa, A. M.
Phosphorus, Sulfur, Silicon and Rel. Elem. 1991, 62, 269.
42. Brown, A. J.; Berkoben, M.; Ritter, C. S.; Slatopolsky, E.
Endocrinology 1992, 130, 276.
43. Hruska, K. A.; Kopelman, R.; Rutherford, W. E.; Klahr,
S.; Slatopolsky, E. J. Clin Invest. 1975, 56, 39.