Efficient synthesis of 3-TBDMS-11α,25-dihydroxyvitamin D3 and D2 ethers

Article abstract of DOI:10.1016/j.jsbmb.2020.105638

Vitamin D deficiency might cause a wide variety of human disorders. As a prerequisite for appropriate diagnosis and therapy, medicinally relevant vitamin D metabolites have to be assayed most accurately and with high specificity. It has been demonstrated, that vitamin D conjugates, linked via a hydroxyl group at C11, might be promising for the development of highly specific antibodies to be employed in competitive protein binding assays. The connective synthesis of 3-TBDMS-11α,25-dihydroxyvitamin D₃ and D₂ ethers in 500 mg scale, starting from vitamin D₂, is described. For installation of a hydroxyl group at C11 a sequence of Pd(OAc)₂ mediated oxidation of an enone, epoxidation and subsequent epoxide ring opening was applied to obtain a suitable CD-ring precursor, that was connected with an A-ring diphenylphosphine oxide by Wittig-Horner reaction. Finally, an appropriate side chain was installed, respectively.

Full text of DOI:10.1016/j.jsbmb.2020.105638

Journal Pre-proof
Efficient synthesis of 3-TBDMS-11␣,25-dihydroxyvitamin D3 and D2
ethers
Lars Kattner, Erik Rauch
PII:
S0960-0760(19)30548-5
DOI:
https://doi.org/10.1016/j.jsbmb.2020.105638
SBMB 105638
Reference:
To appear in:
Journal of Steroid Biochemistry and Molecular Biology
Received Date:
Revised Date:
Accepted Date:
12 September 2019
27 January 2020
18 February 2020
Please cite this article as: Kattner L, Rauch E, Efficient synthesis of
3-TBDMS-11␣,25-dihydroxyvitamin D3 and D2 ethers, Journal of Steroid Biochemistry and
Molecular Biology (2020), doi: https://doi.org/10.1016/j.jsbmb.2020.105638
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2019 Published by Elsevier.

Efficient synthesis of 3-TBDMS-11,25-dihydroxyvitamin D3 and D2 ethers
Lars Kattner* and Erik Rauch
∗ Corresponding author at: Endotherm Life Science Molecules, 66123 Saarbruecken,
Germany, E-mail address: lars.kattner@endotherm.de (L. Kattner), Tel.: +49 681
3946-7570; Fax: +49 681 3946-7571.
Graphical abstract
OTos
S(O)2Ph
TMSO
OH
TMSO
7
O
8
O
OH
+
OH
P(O)Ph2
HO
Vitamin D2
10
TBDMSO
TBDMSO
9
OTos
S(O)2Ph
TMSO
TMSO
5
3
TBDMSO
TBDMSO
O
6
OTMS
4
Br
OH
HO
HO
OH
OH
TBDMSO
TBDMSO
1
2
Highlights:
 Formal synthesis of 11,25-dihydroxyvitamin D3. First formal synthesis of
11,25-dihydroxyvitamin D2, by application of a SmI2 mediated epoxide ring
opening to introduce an 11-hydroxy group at the CD ring moiety.
1

Abstract
Vitamin D deficiency might cause a wide variety of human disorders. As a
prerequisite for appropriate diagnosis and therapy, medicinally relevant vitamin D
metabolites have to be assayed most accurately and with high specificity. It has been
demonstrated, that vitamin D conjugates, linked via a hydroxyl group at C11, might
be promising for the development of highly specific antibodies to be employed in
competitive protein binding assays. The connective synthesis of 3-TBDMS-11,25-
dihydroxyvitamin D3 and D2 ethers in 500 mg scale, starting from vitamin D2, is
described. For installation of a hydroxyl group at C11 a sequence of Pd(OAc)2
mediated oxidation of an enone, epoxidation and subsequent epoxide ring opening
was applied to obtain a suitable CD-ring precursor, that was connected with an A-ring
diphenylphosphine oxide by Wittig-Horner reaction. Finally, an appropriate side chain
was installed, respectively.
Keywords: Vitamin D deficiency; vitamin D related diseases; biomarkers; vitamin D
metabolites; assay development; protein binding assays; ELISA; vitamin D synthesis
Introduction
In recent years it has become apparent, that vitamin D deficiency is connected to a
wide variety of disorders, beyond causing bone diseases [1]. Consequently, the
development of appropriate assays to measure the status of medicinally relevant
vitamin D metabolites in human blood, serum or relevant tissue is continuously being
improved [2]. A rapid way for assessment of vitamin D deficiency is the measurement
of 25-hydroxyvitamin D3 (25OHD3) concentration in blood or serum by immunological
2

competitive protein binding assays, using specific antibodies. As a common
approach, an appropriate conjugate of a particular vitamin D metabolite is employed,
where the vitamin D skeleton is linked at its C3-OH-position of the A-ring to biotin or
bovine serum albumin (BSA), that competes with the antigen in the patient sample for
binding to a specific antibody [3]. As a major drawback, specific vitamin D antibodies
are challenging to develop. Unfortunately, using these conjugates, polyclonal
antibodies frequently lack sufficient specificity, particularly to distinguish between
metabolites bearing different substituents at the A-ring, such as 25OHD3 and
1,25(OH)2D3 (calcitriol). To overcome this problem, 11,25(OH)2D3 conjugates may
be used instead [4], in turn to be synthesized from a C3-tert-butyldimethylsilyl
(TBDMS) ether, that is suitable to be connected at its C11-OH-position via an
appropriate linker to biotin or BSA [5]. Furthermore, the corresponding vitamin D2
analogs have to be considered as well, because they also appear in human blood
and tissue and exhibit significant activity [6].
We describe a new efficient synthesis of 3-TBDMS-11,25-dihydroxyvitamin D3 (1)
and the first synthesis of its corresponding D2 analog (2) (Scheme 1), elaborated in
500 mg scale each.
3

OTMS
Br
4
25
OH
OTos
HO
11
TMSO
H
H
C
D
OTos
TMSO
H
H
H
H
O
A
1
7
TBDMSO
TBDMSO
3
P(O)Ph2
H
3
H
O
TBDMSO
OH
HO
6
9
S(O)2Ph
HO
TMSO
H
H
10
OH
S(O)2Ph
H
H
TMSO
H
H
O
TBDMSO
TBDMSO
2
5
8
Scheme 1: Retrosynthetic strategy for the synthesis of 3-TBDMS-11,25-
dihydroxyvitamin D3 (1) and 3-TBDMS-11,25-dihydroxyvitamin D2 (2).
Methods
In a retrosynthetic perspective, both target compounds (1 and 2) (Scheme 1) are
obtained by connecting a vitamin D skeleton, in turn already bearing a protected
hydroxyl group at C11 position, with an appropriate side chain. For the synthesis of
the vitamin D3 derivative (1), connection of a tosylate 3 with bromide 4 was planned,
for the corresponding D2 derivative (2) Julia olefination of sulfone 5 with aldehyde 6
was envisaged. CD-ring building blocks 7, in turn employed for the synthesis of 1,
and 8, for the synthesis of 2, were planned to be connected with A-ring
diphenylphosphine oxide 9, respectively. As well 9 as CD-ring building blocks 7 and 8
are accessible from readily available vitamin D2 10 by its oxidative degradation and
subsequent appropriate chemical modifications.
4

Results and Discussion
For the synthesis of 3-TBDMS-11,25-dihydroxyvitamin D3 (1) (Scheme 2), vitamin
D2 (10) was converted to diphenylphosphine oxide 9 and Inhoffen-Lythgoe diol 11
according established procedures [7]. Silylation of 11 with TBDMSCl and oxidation
afforded the corresponding ketone, that underwent a procedure described by
Takahashi et al. [8]. Thus, treatment with trimethylsilyl triflate and subsequent
oxidation by palladium(II) acetate gave enone 12. A following epoxidation with
hydrogen peroxide led to the corresponding epoxide 13. Ring opening with Me2CuLi
[9] gave 14, that was converted via 15 in tosylate 7. Wittig-Horner reaction of 7 with
diphenylphosphine oxide 9 gave 3. Bromide 4 was obtained from commercially
available 3-methyl-1,3-butandiol 16, that was tosylated and brominated to 17, and
finally silylated. Organomagnesium formed from 3 and subsequent nucleophilic
substitution on 4 provided 18. Due to their different susceptibility against oxalic acid,
the trimethylsilyl (TMS) groups could be removed in the presence of the TBDMS
group, leading directly to 1.
1) TBDMS-Cl, 99%
OTBDMS
Me2CuLi
OTBDMS
OH
2) PDC, 84%
3) TMS-OTf
4) Pd(OAc)2
OTBDMS
OTos
HO
HO
1) HF*Py, 81%
2) TosCl, 62%
H2O2,
NaOH
63%
O
55%
44%
O
O
OH
O
O
11
13
14
12
15
P(O)Ph2
OTos
TMSO
OTos
TMSO
9
TMS-Cl
98%
TBDMSO
HO
TMSO
3
O
BuLi
85%
7
OH
OH
Mg, CuBr*
DMSO
(COOH)2
MeOH
,
TBDMSO
73%
81%
1
18
OH
OH
1) pTosCl, Py, 75%
2) LiBr, THF, 87%
OTMS
TBDMSO
TBDMSO
TMS-Cl
83%
Br
HO
Br
16
17
4
Scheme 2. Synthesis of 3-TBDMS-11,25-dihydroxyvitamin D3 ether (1).
5

The synthesis of 3-TBDMS-11,25-dihydroxyvitamin D2 ether (2) is by far more
challenging comparing to the synthesis of its D3 counterpart (Scheme 3). A
methodology described for the synthesis of 1,25-(OH)2VD2 by Kutner et al. [10],
could finally be applied successfully. Key step is the introduction of the side chain via
Julia olefination of phenyl sulfone 5 with aldehyde 6. CD-ring ketone 8, bearing a
C11-OH group, was synthesized analogously to 7, by a procedure described by
Takahashi et al. [8]. The synthesis of 2 was carried out employing sulfone 19.
Treatment of 19 with trimethylsilyl triflate and subsequent oxidation by palladium(II)
acetate gave enone 20, that was epoxidized to 21. Reaction of 21 to produce 22
surprisingly failed with lithium dimethylcuprate. Alternatively, SmI2 treatment [11]
provided 22, that turned out to be very sensitive against epimerization, in 98% yield.
Subsequently, the OH group of 22 at C11 was protected as its TMS ether, giving
sulfone 8. The vitamin D skeleton was subsequently build up by coupling sulfone 8
with diphenylphosphine oxide 9, leading to 5. In the following step, the side chain
was connected by Julia olefination of 5 with aldehyde 6, in turn synthesized in two
steps from commercially available ester 23. Grignard reaction of 23 with
methylmagnesium chloride gave diol 24, that was oxidized to aldehyde 6.
Surprisingly, oxidation with Dess Martin periodinane, how it is described in the
literature, failed. By contrast, TEMPO oxidation with sodium hypochlorite as a co-
oxidizing agent was successful. In course of a Julia coupling of sulfone 5 with
aldehyde 6 the TMS protective group turned out to be unstable, so that 2 was
obtained in the same reaction step. Although the yield was surprisingly low (16%),
presumably due to simultaneous partial reduction of the free C25 hydroxyl function in
the side chain or reduction of the C11 hydroxyl group using Na/Hg. Since no less
polar UV-visible compound could be isolated, this explanation is hypothetical and the
last synthesis step could be further optimized.
6

S(O)2Ph
1) TMS-OTf
2) Pd(OAc)2
S(O)2Ph
S(O)2Ph
S(O)2Ph
HO
H2O2, NaOH
90%
SmI2
98%
O
87%
O
O
O
19
21
20
22
O
P(O)Ph2
S(O)2Ph
TMSO
S(O)2Ph
TMSO
TMS-Cl
88%
9
HO
O
8
5
TBDMSO
OH
BuLi
49%
1) LDA
2) Na/Hg
TBDMSO
16%
2
TBDMSO
MeMgCl
80%
NaOCl, KBr, TEMPO
61%
HO
OMe
HO
O
OH
OH
O
23
6
24
Scheme 3. Synthesis of 3-TBDMS-11,25 dihydroxyvitamin D2 ether (2).
1
The structures of target compounds 1 and 2 were assigned by H/13C-NMR and
HRMS [12].
Conclusion
3-TBDMS-11,25 Dihydroxyvitamin D3 and D2 ethers (1) and (2) could be prepared
in 500 mg scale each to explore their potential for the development of new
conjugates to discover specific antibodies for advanced immunological assays or
other applications.
Acknowledgement
This work was supported by Roche Diagnostics GmbH (Germany).
7

References and notes
[1] Feldman D, Pike JW, Bouillon R, Giovannucci E, Goltzman D, Hewison M, editors.
Vitamin D. 4th ed. Volume 2: Health, disease and therapeutics. London, San Diego,
Cambridge, Oxford: Elsevier/Academic Press, 2018. ISBN 978-0-12-809963-6
[2] Farrell J, Herrmann M: Determination of vitamin D and its metabolites: Best Pract
Res Clin Endocrinol Metab. 2013;27(5):675-688.
[3] Armbruster FP et al. 1999; EP1097132B1.
[4] Kobayashi N, Hisada A, Shimada K: Specifity of the polyclonal antibodies raised
against a novel 25-hydroxyvitamin D3-bovine serum albumin conjugate linked through
the C-11 position: J Steroid Biochem Molec Biol. 1994;48(5/6):567-572; J Steroid
Biochem Mol Biol. 1997;62(1):79-87.
[5] Kattner, L, Bernardi D: An efficient synthesis of 1,25-dihydroxyvitamin D3 LC-
biotin: J Steroid Biochem Mol Biol. 2017;173:89-92. DOI: 10.1016/j.jsbmb.
2017.01.025; Armbruster FP, Friedl S, Karmatschek M, Heckl-Ostreicher B, Reichel
H, Woloszczuk: Development of a novel ELISA for 1,25-dihydroxyvitamin D: Clin Lab.
2000;46:165-166. ISSN: 1433-6510.
[6] Houghton LA, Vieth R: The case against ergocalciferol (vitamin D2) as a vitamin
supplement: Am J Clin Nutr. 2006;84: 694-697.
[7] Lythgoe B: Synthetic approaches to vitamin D and its relatives: Chem Soc Rev.
1980;109:449-475.
[8] Doi T, Hijikuro I, Takahashi T: An efficient solid-phase synthesis of the vitamin D3
system: J Am Chem Soc. 1999;121:6749-6750; Hijikuro I, Doi T, Takahashi T:
8

Parallel synthesis of a vitamin D3 library in the solid-phase. J Am Chem Soc.
2001;123:3716-3722.
[9] Zhu GD, Haver DV, Jurriaans H, De Clercq PJ: 11-Fluoro-1α-hydroxyvitamin D3:
The quest for experimental evidence of the folded vitamin D conformation.
Tetrahedron 1994;50(24):7049-7060.
[10] Pietraszek A, , Malińska M, , Chodyński M, Krupa M, Krajewski K, Cmoch P,
Woźniak K, Kutner A: Synthesis and crystallographic study of 1, 25-
dihydroxyergocalciferol analogs: Steroids 2013;78:1003-1014.
[11] Molander GA, Hahn G: Lanthanides in organic synthesis. 4. Reduction of ,-
epoxy ketones with samarium diiodide. A route to chiral, nonracemic aldols: J Org
Chem. 1986;51:2596-2599.
1
[12] 1: HNMR (D6-acetone, 500 MHz):  0.106 (s, 3xH-TBDMS), 0.119 (s, 3xH-
TBDMS), 0.599 (s, 3xH18), 0.913 (s, 9xH-TBDMS), 1.013 (d, J=6.5 Hz, 3xH21),
1.167 (s, 6xH26/27), 2.800 (s, 1xOH), 3.068 (s, 1xOH), 3.160 (dd, J=4.5, 13.0 Hz,
1xH), 3.587 (d, J=5.5 Hz, 1xH), 3.810 (m, 1xH11), 3.957 (m, 1xH3), 5.065 (d, J=1.5
Hz), 6.122 (d, J=11.0 Hz, 1xH7), 6.287 (d, J=11.0 Hz, 1xH6). 13CNMR (D6-
acetone,125 MHz):  -4.53, -4.47, 13.44, 18,62, 19.22, 21.51, 22.42, 26.20, 28.81,
32.99, 36.92, 37.04, 37.32, 39.79, 45.27, 45.97, 47.40, 50.91, 56.83, 56.92, 68.53,
70.07, 70.93, 112.35, 119.90, 122.13, 137.76, 138.63, 146.75. HRMS (ESI+) calcd for
C33H58O3Si 530.4155, found 530.4146.
2: 1HNMR (D6-acetone, 500 MHz):  0.105 (s, 3xH-TBDMS), 0.118 (s, 3x H-TBDMS),
0.610 (s, 3x H18), 0.913 (s, 9xH-TBDMS), 1.017 (d, J=7.0 Hz, 3xH21), 1.088 (d, J=7
Hz, 3xH28), 1.105/1.140 (2xs, 3xH26/C27), 2.799 (s, 1xOH), 2.999 (s, 1xOH), 4.772
(d, J=2.5 Hz, 1xH19), 5.061 (s, 1xH19), 5.306 (dd, J=9.0, 15.5 Hz, 1xH23), 5.447 (dd,
9

J=9.0, 15.5 Hz, 1xH22), 6.118 (d, J=11.0 Hz, 1xH6), 6.285 (d, J=11.0 Hz, 1xH7).
13CNMR (D6-acetone, 125 MHz):  -4.53, -4.47, 13.69, 15.58, 18.62, 21.36, 22.36,
26.20, 28.98, 32.98, 37.02, 39.79, 41.20, 45.86, 47.38, 48.83, 50.70, 56.66, 56.88,
68.52, 70.90, 71.93, 112.34, 119.90, 122.11, 129.74, 131.60, 137.38, 137.79,
138.54, 146.73. HRMS (ESI+) calcd for C34H58O3Si 542.4155, found 542.4144.
10