Selective manipulation of steroid hydroxyl groups with boronate esters: Efficient access to antigenic C-3 linked steroid-protein conjugates and steroid sulfate standards for drug detection

Article abstract of DOI:10.1039/b610499a

The temporary protection of 17α-alkyl-5α-androstane-3β, 16β,17β triols as boronate esters is an efficient method for their regioselective functionalisation. This has been applied to the synthesis of protein-steroid conjugates 7-10 suitable for the develop

Full text of DOI:10.1039/b610499a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Selective manipulation of steroid hydroxyl groups with boronate esters:
efficient access to antigenic C-3 linked steroid–protein conjugates and steroid
sulfate standards for drug detection†‡
Natasha L. Hungerford,a Andrew R. McKinney,b Allen M. Stenhouseb and Malcolm D. McLeod*a
Received 2nd August 2006, Accepted 13th September 2006
First published as an Advance Article on the web 22nd September 2006
DOI: 10.1039/b610499a
The temporary protection of 17a-alkyl-5a-androstane-3b,16b,17b triols as boronate esters is an efficient
method for their regioselective functionalisation. This has been applied to the synthesis of
protein–steroid conjugates 7–10 suitable for the development of immunoassays targeting classes of
steroids banned from competition in Australian horse racing and other sports. The synthesis of steroids
sulfate conjugates 42 and 44 for use as reference standards is also reported.
metabolites in urine. Steroids with 17a-alkyl substituents are
Introduction
common agents of abuse in horse racing and other sports, owing
Anabolic androgenic steroids are an important class of perfor-
to the oral bioavailability imparted by the alkyl substituent. In
mance enhancing drugs with potential for misuse in horse racing
horses, these 17-alkyl anabolic steroids are extensively metabolised
prior to excretion3,4 and our approach was to target the D-ring
structures common to many equine metabolites of these steroids.
The application of this assay to the detection of equine stanozolol
metabolites is depicted in Scheme 1. Stanozolol, 2, administration
leads to excretion of the 16b-hydroxystanozolol glucuronide
conjugate 3 as the major phase II metabolite.3 Enzyme hydrolysis
liberates the 16b-hydroxystanozol 4 which is then detected in
the raw hydrolysed urine by an ELISA employing polyclonal
antibodies targeting the D-ring of the steroid.2 The assay has
been deployed by the Australian Racing Forensic Laboratory as
and other sport. As a result, the integrity of sporting contests
relies on stringent doping control measures targeting these agents.
Despite ongoing research, the detection of illicit steroid use
presents significant challenges due to a range of complicating
factors. Among these, the administration of anabolic steroids
frequently results in little or no excretion of the parent steroid in the
urine and instead, the steroid is converted into more hydrophilic
metabolites. The detection of steroid abuse therefore requires
appropriate reference materials and methods for the detection
of the metabolites derived from known steroidal agents. Of
greater concern, in 2003, the previously unknown anabolic steroid
tetrahydrogestrinone 1 (THG, Fig. 1)1 was uncovered as a doping
agent at the highest levels of world sport. This so-called ‘designer’
steroid was conceived to evade standard methods of selected ion
monitoring mass spectrometric detection and highlights the need
for broader screening methods for the detection of drugs in sport.
Fig. 1
Recently, we described2 an enzyme-linked immunosorbent assay
(ELISA) based screen for the detection of 17a-methyl steroid
aSchool of Chemistry, F11, The University of Sydney, NSW, 2006, Australia.
E-mail: m.mcleod@chem.usyd.edu.au; Fax: +61-2-9351-6650; Tel: +61-2-
9351-5877
bAustralian Racing Forensic Laboratory, P.O. Box 528, Kensington, NSW,
1465, Australia
† Electronic supplementary information (ESI) available: General experi-
mental details together with experimental procedures and spectroscopic
data for compounds 15, 19, 20, 26, 30–40, 44 and 47. See DOI:
10.1039/b610499a
‡ The HTML version of this article has been enhanced with colour images
Scheme 1
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3951–3959 | 3951
©

View Article Online
a primary screen for the detection of stanozolol metabolites in
race day samples.
antibodies suitable for the detection of prohibited steroids such
as ethylestrenol 11, norethandrolone 12 (Fig. 2) and their 16b-
hydroxylated metabolites.3 The 17-ethynyl antigens 7 and 8 would
afford antibodies suitable for the detection of danazol 13 and
its putative metabolites. The antibodies could also show some
selectivity for the detection of 18-homo steroids such as THG 1
and gestrinone 14.
The ELISA-based methods of drug detection have a number of
positive attributes. The assays do not routinely require sample
extraction and derivatisation procedures and are readily con-
ducted in array format leading to significant efficiencies. More
importantly, ELISA is a broad screen with the potential to detect
families of steroid metabolites that contain a common structural
motif (such as the hydroxylated D-ring of metabolite 4). ELISA
thus has significant promise as a forensic tool for the early
detection of previously unknown steroids or their metabolites.
The success of the reported ELISA targeting 17a-methylsteroids
suggested a number of avenues for further research. Our first
goal was the development of assays targeting other commonly
occurring D-ring substitution patterns associated with World
Anti-Doping Agency (WADA) and International Federation of
Horseracing Authorities (IFHA) prohibited anabolic steroids
such as 17a-ethyl and 17a-ethynyl steroids. This required the
development of polyclonal antibodies raised against these D-ring
motifs, which in turn required extension of the previously reported
methods for the synthesis of protein–steroid antigens 5 and 62 to
derivatives 7–10 bearing different D-ring substituents (Scheme 2).
It was anticipated that 17a-ethyl antigens 9 and 10 would afford
Fig. 2 WADA and IFHA prohibited anabolic steroids.
A second and related goal of this study was to improve synthetic
access to the antigenic steroid–protein conjugates 5, 7 and 9.
The 3-keto steroids 15–17 were key intermediates required for
conjugate synthesis and could be derived from the steroid triols
18–20 by selective oxidation of the C3-hydroxyl group. In the
reported synthesis of 15, the multi-step oxidation sequence (18 →
15) suffered from the formation of isomeric steroids, necessitating
HPLC separation to afford pure 3-keto steroid 15.2 In addition to
the development of antibodies already reported, the conjugate 5
was required as a reagent for the ELISA and this bottleneck to
the large-scale synthesis potentially limited the future application
of the successful assay. Furthermore, these difficulties also posed
problems for the synthesis of the new steroid antigens 7 and 9.
This paper reports an improved synthesis of protein–steroid
conjugates 7–10 suitable for the development of ELISA targeting
17a-ethyl and 17a-ethynyl steroids. The synthesis hinged on the
development of effective boronate ester-mediated strategies for
the selective functionalisation to afford steroid ketones 15–17. As a
further demonstration of the scope of this chemistry, the boronate
ester methodology is also employed for the selective synthesis
of steroid sulfate reference materials. Such steroid sulfates are
of interest as reference materials for phase II anabolic steroid
metabolites commonly observed following steroid administration.
Results and discussion
Synthesis of 16b,17b-dihydroxy-17a-alkyl-3-keto-steroids for
ELISA development
Synthesis of these 17a-alkyl-5a-androstane-3b,16b,17b-triols
(17a-alkyl-triols, 18–20) was achieved starting from commercially
available epiandrosterone 21. The diacetoxyketone 22 was syn-
thesised in two steps by enol acetate formation followed by lead
tetraacetate oxidation. The addition of excess methylmagnesium
bromide afforded 17a-methyl-triol 18 as previously described
Scheme 2
3952 | Org. Biomol. Chem., 2006, 4, 3951–3959
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Scheme 3 Reagents and conditions: (i) isopropenyl acetate, cat. H2SO4, 80% (ii) Pb(OAc)4, AcOH, Ac2O, 70% (iii) MeMgBr, Et2O, then H+, 97%
(iv) EtMgBr, Et2O, then H+ (v) HC≡CLi·en, THF, then H+, 71% (vi) Pd/C, MeOH, H2, 96%.
(Scheme 3).5,6 This route was then adapted for the synthesis
of ethynyl- and ethyl-substituted steroids 19 and 20. Attempted
treatment of ketone 22 with ethylmagnesium bromide afforded
the desired 17a-ethyl-triol 16 together with the reduction product
23, arising from hydride transfer from the Grignard reagent b-
carbon.7 The triols 20 and 23, were not readily separable by flash
chromatography. A convenient alternative was found in the addi-
tion of commercially available lithium acetylide–ethylenediamine
complex to give the 17a-ethynyl-triol 19, as a single diastereomer.
Catalytic hydrogenation then gave access to 17a-ethyl-triol 20 in
96% yield, as shown in Scheme 3.
Elaboration of these triols 18–20 to 16b,17b-dihydroxy-5a-
androstan-3-ones (17a-alkyl-3-ketones, 15–17) proved more chal-
lenging. This required selective manipulation of the C-3 hydroxyl
group in 17a-alkyl-triols 18–20. Selective protection of the vicinal
diol in steroid 18 as the acetonide was followed by oxidation to
ketone 24 (Scheme 4). Subsequent access to 17a-methyl-ketone
15 required acetonide deprotection under acidic conditions to
furnish the 16b,17b-diol array.2 Unfortunately, the acid promoted
deprotection of acetonide 24 under a wide range of conditions
provided mixtures of epimers 15 and 25. Separation of epimers
was only partially achieved by flash chromatography, and HPLC
separation was required.
To increase efficiency of the synthesis and to avoid the acidic
deprotection conditions and C17 epimerisation, exchange of the
acetonide protecting group for a benzylidene acetal protecting
group was investigated. The 17a-methyl-triol 18 was treated with
benzaldehyde dimethyl acetal in the presence of concentrated sul-
Scheme 4 Reagents and conditions: (i) 2,2-dimethoxypropane, pTsOH,
CH2Cl2, 74% (ii) CrO3, py, 70% (iii) Dowex 50W-X8, MeOH–H2O (1 : 1),
70%.
furic acid to give the acetal 26, as a single diastereomer (Scheme 5).
NOESY experiments were used to assign the stereochemistry at
the newly formed acetal stereocentre. Parikh–Doering8 oxida-
tion conditions gave higher yields (78%) for the conversion of
the 3-hydroxyl group to the corresponding ketone. Benzylidene
deprotection was then achieved, without epimerisation, with
Pd(OH)2 under a hydrogen atmosphere, to give the desired triol
15 exclusively, which was isolated in 93% yield.
be appropriate for the conversion of 17a-ethynyl-triol 19 to the
corresponding 17a-ethynyl-ketone 16, as the hydrogenolysis would
be incompatible with an alkyne. Hence, a more general method was
sought.
The use of boronate esters as protecting groups for diols
has been reviewed,9 with many applications to carbohydrate
synthesis and the regioselective manipulation of hydroxyl groups.
Appealing features include the ease of protection/deprotection
and in situ derivatisation which minimise the number of reaction
and purification steps. Such features would be well suited to
steroid hydroxyl group manipulation. Notably, Harmatha and
While the benzylidene protecting group conveniently gave
access to 17a-methyl-ketone 15, and would be suitable for the
preparation of 17a-ethyl-ketone 17, the methodology would not
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3951–3959 | 3953
©

View Article Online
Scheme 5 Reagents and conditions: (i) C6H5CH(OCH3)2, conc. H2SO4,
˚
Scheme 6 Reagents and conditions: (i) PhB(OH)2, DMF, CH2Cl2, 4 A
mol. sieves (ii) PCC/Al2O3 (iii) H2O2, NaOH(aq), THF.
˚
DMF, CH2Cl2, 95% (ii) Py·SO3, Et3N, 4 A mol. sieves, DMSO, CH2Cl2,
78% (iii) H2, Pd(OH)2, THF, 93%.
The application of phenylboronate ester chemistry to trihy-
droxylated steroids provided regioselective access to ketones 15–
17, ready for elaboration to conjugates 5–9. Efficient protection,
in situ and regioselective oxidation, followed by deprotection
rapidly afforded the desired dihydroxylated 3-keto-steroids. The
chemistry was readily scaleable and did not result in the production
of stereoisomeric steroids nor require HPLC to afford pure
compound.
co-workers10 utilised phenylboronic acid to selectively protect an
ecdysteroid side-chain vicinal diol in the presence of a more rigid
A-ring 2b,3b-diol. This enabled subsequent derivatisation of the
remaining secondary hydroxyl groups as acetate esters.
In the present case, the use of phenylboronic acid to protect the
16b,17b-diol unit in 17-alkyl-triols 18–20 would require boronate
ester formation on the rigid D-ring allowing for selective oxidation
of the 3-hydroxyl to give ketones 15–17. The sequential boronate
ester formation and in situ oxidation followed by oxidative removal
of the boronate ester, would provide an expedient alternative to
the step-wise strategies, and associated purifications, described
above (Schemes 4 and 5) for 17a-methyl-ketone 15. In the event,
treatment of 18 with phenylboronic acid (Scheme 6), in DMF–
dichloromethane in the presence of molecular sieves, followed by
direct oxidation at C-3, by addition of pyridinium chlorochromate
(PCC) on alumina,11 gave intermediate 27. Oxidative removal of
the boronate ester with sodium hydroxide and hydrogen peroxide,
then gave 15 exclusively, in 76% yield over the 3 steps. The basic
deprotection conditions avoided any epimerisation problems and
16b-hydroxymestanolone 15 was prepared in an extremely clean,
rapid, and efficient sequence that compares favourably to the 69%
yield obtained for the three step sequence via benzylidene acetal
26 (Scheme 5).
This approach was then applied to the conversion of 17a-
ethynyl-triol 19 and 17a-ethyl-triol 20, to ketones 16 and 17
respectively. In each case, intermediates 28 and 29 were formed
by treatment with phenylboronic acid and in situ oxidation
(PCC/Al2O3). Deprotection then (NaOH–H2O2) afforded ketones
16 and 17, in 76% and 71% yield (over the 3 steps), respectively
(Scheme 6). Again the deprotection conditions caused no epimeri-
sation, particularly in the case of sensitive 17a-ethynyl-ketone 19.12
Synthesis of 17b-hydroxy-17a-alkyl-3-keto-steroids for ELISA
development
A major finding of previous studies was that antibodies generated
against 16b,17b-dihydroxy-17a-alkyl-antigens such as 5 were more
selective and sensitive than antibodies generated against the
parent 17b-hydroxy-17a-alkyl-antigens such as 6 (Scheme 2).2 This
provides significant advantages in the development of ELISA
for drug detection. Antibodies raised against antigen 5 were
used for the detection of stanozolol metabolites in raw enzyme
hydrolysed urine. In contrast, antibodies raised against antigen 6
are used for the detection of methandriol metabolites but required
solid phase extraction to remove interferences presumed to arise
from endogenous compounds. To explore the generality of this
phenomenon the synthesis of additional mono-hydroxylated 17a-
ethynyl and 17a-ethyl-antigens 8 and 10 was required.
Epiandrosterone 21 was treated with excess lithium acetylide–
ethylenediamine complex, giving alkyne 32 in 94% yield
(Scheme 7). Oxidation of the 3-hydroxyl group was achieved
cleanly using Parikh–Doering conditions8 to give target 17a-
ethynyl-ketone 30 in 75% yield. Catalytic hydrogenation (Pd/C,
H2, NaHCO3) afforded corresponding 17a-ethyl-ketone 31, in 65%
yield.
3954 | Org. Biomol. Chem., 2006, 4, 3951–3959
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Scheme 9
Scheme 7 Reagents and conditions: (i) HC≡CLi·en, THF, then H+, 94%
˚
conjugates.14 These were identified by a process of conjugate
hydrolysis to give the free steroid, followed by derivatisation and
GCMS analysis against synthetically derived reference standards.
Of relevance to this work, the 17a-methyl triol 18 was identified as
a major steroidal metabolite occurring as the sulfate conjugate.
The routine observation of sulfate esters as metabolites makes
the synthesis of these compounds highly desirable as reference
standards that allow each step of the analytical procedure,
including sulfate ester hydrolysis, to be monitored. In the case of
17a-methyl triol 18, this also raised the question of which hydroxyl
group of the steroid carried the sulfate ester residue. Evidence
based on the configurational stability of the C17 stereogenic centre
suggested the sulfate was not conjugated to the tertiary hydroxyl
group.14
The ability to regioselectively manipulate the hydroxyl groups
in steroid triols, through boronate ester protection of the steroid
16,17-diol, would also enable selective and rapid access to steroid
sulfate derivatives. Previously, Vasella and co-workers15 employed
phenylboronic acid to protect the 4,6-diol of a methyl glycoside,
with subsequent stannylidene activation of the remaining diol
enabling regioselective sulfation.
(ii) Py·SO3, Et3N, DMSO, CH2Cl2, 4 A mol. sieves, 75% (iii) H2, Pd/C,
NaHCO3, 65%.
Synthesis of C-3 linked protein-steroid conjugates
Conversion of ketones 16, 17, 30 and 31 to steroid–protein
conjugates 7–10 is summarised in Scheme 8.2 Treatment of the
ketones with carboxymethoxylamine hemihydrochloride, gave the
corresponding carboxymethyloximes 33–36. Subsequent reaction
with N-hydroxysuccinimide and 1-[3-(dimethylamino)propyl]-3-
ethylcarbodiimide hydrochloride (EDC) yielded the activated
esters 37–40 respectively. Standard techniques then provided
access to steroid–protein conjugates 7–10, with the steroid linked
to the lysine residues of human serum albumin to provide the
antigenic material for antibody generation.2 The development of
antibodies and their application to the development of ELISA is in
progress using previously reported methods and will be reported
elsewhere.
Synthesis of steroid sulfate reference materials
Steroid sulfate esters are commonly observed phase II metabolites
that are derived from polyhydroxylated steroids.13 The charged
sulfate residue renders the hydrophobic steroid more soluble in
aqueous environments and more readily excreted in the urine.
An example of steroidal metabolism is given in Scheme 9. In
recent work, the administration of methyltestosterone 41 to a
thoroughbred gelding resulted in the excretion of a number of an-
drostane diol and triol metabolites as their sulfate or glucuronide
As shown in Scheme 10, the flexibility of the boronate ester ap-
proach, as applied to anabolic steroid metabolites, is demonstrated
by the selective conversion of triol 18 to 3-sulfate derivative 42.
Diol protection and sulfation are rapidly achieved in the one-pot,
giving intermediate 43. Oxidative phenylboronate deprotection
then provided the desired 3-sulfate 42 in 79% yield, over the 3
steps.
Scheme 8 Reagents and conditions: (i) (NH2OCH2CO2H)2·HCl, py (ii) EDC, 1,4-dioxane, CH2Cl2, N-hydroxysuccinimide.
This journal is The Royal Society of Chemistry 2006 Org. Biomol. Chem., 2006, 4, 3951–3959 | 3955
©

View Article Online
of the sulfate and retention of the TBS protecting group. However,
TBS deprotection was effected with 80% acetic acid in water, over
1 hour.17 These conditions were sufficiently mild to retain the acid
labile sulfate group, and 16-sulfate 44 was obtained in 69% yield.
Both sulfate standards were used in a recent study of the
phase II metabolites derived from methyltestosterone in a drug
administration trial. Both the C3-sulfate 42 and C16-sulfate
44 displayed retention time matches and mass spectrometric
behaviour consistent with the presence of significant sulfate
metabolites.18 Unfortunately, due to inadequate chromatographic
resolution and the likely presence of a number of isomeric sulfated
steroid triols that were not targeted for synthesis by this study, it
was not possible to unambiguously confirm the presence of these
sulfate compounds in urine. Nevertheless, the chemistry highlights
the application of boronate esters in the synthesis of regioisomeric
sulfate standards.
Experimental
General experimental details together with experimental proce-
dures and spectroscopic data for compounds 15, 19, 20, 26, 30–40,
44 and 47 have been deposited in ESI.†
Boronate ester mediated oxidation
˚
Scheme 10 Reagents and conditions: (i) PhB(OH)2, DMF, CH2Cl2, 4 A
mol. sieves (ii) SO3·py, DMF (iii) H2O2, aq. NaHCO3, THF, MeOH, 79%
16b,17b-Dihydroxy-17a-methyl-5a-androstan-3-one (b-hydroxy-
mestanolone) (15). Method 1. 17a-Methyl triol 182 (0.410 g,
1.27 mmol) was stirred (18 was partially insoluble) in a mixture
of DMF–CH2Cl2 (1 : 1, 2.4 mL). Phenylboronic acid (0.248 g,
(over 3 steps).
The alternative steroid 16-sulfate 44 could also be prepared
selectively as shown in Scheme 11. Treatment of triol 18 with
phenylboronic acid and in situ protection16 of the 3-hydroxyl
group by addition of tert-butyldimethylsilyl (TBS) chloride and
imidazole gave intermediate 45. Oxidative removal of the boronate
ester then gave diol 46 in 73% yield, over the 3 steps. Treatment of
diol 46 with excess sulfur trioxide–pyridine complex, and excess
pyridine, for 50 minutes, then resulted in regioselective sulfation of
the remaining secondary hydroxyl group at C16 to give protected
sulfate 47. The tetrabutylammonium fluoride (TBAF) mediated
silyl ether deprotection of sulfate 47 was unsuccessful, with excess
TBAF resulting only in formation of the tetrabutylammonium salt
˚
2.04 mmol) was added, followed by 4 A molecular sieves (20).
After 15 minutes, no starting material remained insoluble, and
after 1 hour, TLC (ethyl acetate–hexane, 1 : 1) showed complete
conversion of starting material (Rf 0.2) to the corresponding
boronic ester (Rf 0.5). Additional CH2Cl2 (2 mL) was added,
followed by pyridinium chlorochromate (PCC)/alumina11 (3.18 g,
˚
25% w/w) and 4 A molecular sieves (10). The reaction mixture was
stirred for 20 hours after which time TLC (ethyl acetate–hexane,
1 : 1) showed complete conversion to intermediate product 27
(Rf 0.7). The reaction mixture was diluted with Et2O (20 mL)
and filtered through a pad of silica topped with celite, with Et2O
˚
Scheme 11 Reagents and conditions: (i) PhB(OH)2, DMF, CH2Cl2, 4 A mol. sieves (ii) TBSCl, imidazole (iii) H2O2, aq. NaOH, THF, 73% (over 3 steps)
˚
(iv) SO3·py, DMF, py, 4 A mol. sieves, 61% (v) 80% AcOH–H2O, 69%.
3956 | Org. Biomol. Chem., 2006, 4, 3951–3959
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
(500 mL) elution. Concentration in vacuo gave a colourless foam
which was directly dissolved in THF (12 mL) and treated with
H2O2 (30% aqueous solution, 3.2 mL) and NaOH (12% aqueous
solution, 2 mL) for 15 minutes. TLC (ethyl acetate–hexane, 1 : 2)
suggested the deprotection to be complete. The reaction mixture
was diluted with H2O (70 mL) and extracted into ethyl acetate
(4 × 30 mL). The combined ethyl acetate extracts were washed
with saturated Na2SO3 solution (20 mL), saturated NaCl solution
(20 mL), then dried (Na2SO4) and concentrated in vacuo. The
residue was dissolved in CH2Cl2–MeOH and pre-adsorbed onto
silica for flash chromatography (ethyl acetate–hexane, 1 : 2 then
1 : 1), which afforded b-hydroxymestanolone (15) (0.308 g, 76%).
requires 330.2195. 8%), 312 (15), 297 (38), 231 (54), 217 (62), 173
(65), 159 (56), 119 (55), 105 (63), 91 (100), 79 (82).
16b,17b-Dihydroxy-5a,17a-pregnan-3-one
(17). 17a-Ethyl
triol 20 (0.350 g, 1.04 mmol) was stirred (20 was partially
insoluble) in CH2Cl2 (3.0 mL). Phenylboronic acid (0.203 g,
˚
1.7 mmol) was added, followed by 4 A molecular sieves (20).
After 5 minutes, the starting material had dissolved, and after
1 hour, TLC (ethyl acetate–hexane, 1 : 1) showed complete
conversion of starting material (Rf 0.33) to the corresponding
boronic ester (Rf 0.6). Additional CH2Cl2 (2 mL) was added
followed by pyridinium chlorochromate (PCC)/alumina11 (3.3 g,
25% w/w). The reaction mixture was stirred for 3 hours after
which time TLC (ethyl acetate–hexane, 1 : 1) showed complete
conversion to intermediate product 29 (Rf 0.7). The reaction
mixture was diluted with Et2O (20 mL) and filtered through a
pad of silica topped with celite, with Et2O (100 mL) elution.
Concentration in vacuo gave a colourless foam which was directly
dissolved in THF (9 mL) and treated with H2O2 (30% aqueous
solution, 2.4 mL) and NaOH (12% aqueous solution, 1.5 mL)
for 1.25 hours. TLC (ethyl acetate–hexane, 1 : 2) suggested the
deprotection to be complete. The reaction mixture was diluted
with H2O (210 mL) and extracted into ethyl acetate (4 × 50 mL).
The combined ethyl acetate extracts were washed with saturated
Na2SO3 solution (40 mL), saturated NaCl solution (50 mL),
then dried (Na2SO4) and concentrated in vacuo. The residue was
dissolved in CH2Cl2–MeOH and pre-adsorbed onto silica for flash
chromatography (ethyl acetate–hexane, 1 : 3 then 1 : 2), which
afforded diol 17 (0.248 g, 71%). Rf 0.20 (ethyl acetate–hexane, 1 :
◦
Rf 0.2 (ethyl acetate–hexane, 1 : 1); mp 185–186 C (lit.2 185–
186 ◦C) [a]2D0 +8.4 (c 1.0, CH2Cl2) (lit.2 [a]2D0 +8.4 (c 1.0, CH2Cl2));
mmax/cm−1 (film) 3600–3000 (OH), 2939, 2920, 2854, 1713 (C O),
=
1447, 1381, 1356, 1271, 1219, 1175, 1124, 1067, 1040, 1024; dH
(200 MHz, CDCl3) 3.64 (1H, dd, J 7.9, 5.4 Hz, C16H), 3.00–2.80
(1H, br s, OH), 2.80–2.60 (1H, br s, OH), 2.48–1.92 (6H, m), 1.78–
0.63 (14H, m), 1.11 (3H, s, C20H), 1.01 (3H, s, C18H), 0.86 (3H, s,
C19H); dC (50 MHz, CDCl3) 212.0, 79.1, 77.6, 53.9, 46.9, 46.7, 44.9,
44.6, 38.5, 38.1, 35.7, 35.6, 34.8, 32.3, 31.5, 28.7, 23.7, 20.8, 13.5,
11.4; m/z (EI+) 320.2350 (M+, C20H32O3 requires 320.2351, 85%),
232 (60), 217 (100), 159 (45).
16b,17b-Dihydroxy-5a,17a-pregn-20-yn-3-one
(16). 17a-
Ethynyl triol 19 (0.107 g, 0.32 mmol) was stirred (19 was partially
insoluble) in CH2Cl2 (1.0 mL). Phenylboronic acid (0.062 g,
˚
0.51 mmol) was added, followed by 4 A molecular sieves (10).
◦
2); mp 170–175 C; [a]2D2 +3.4 (c 0.71, CHCl3); mmax/cm−1 (KBr
After 5 minutes, the starting material had dissolved, and after
1 hour, TLC (ethyl acetate–hexane, 1 : 1) showed complete
conversion of starting material (Rf 0.3) to the corresponding
boronic ester (Rf 0.5). Additional CH2Cl2 (1 mL) was added,
followed by pyridinium chlorochromate (PCC)/alumina11 (1.2 g,
25% w/w). The reaction mixture was stirred for 3 hours after
which time TLC (ethyl acetate–hexane, 1 : 1) showed complete
conversion to intermediate product 28 (Rf 0.8). The reaction
mixture was diluted with Et2O (15 mL) and filtered through
a pad of silica topped with celite, with Et2O (80 mL) elution.
Concentration in vacuo gave a colourless foam which was directly
dissolved in THF (3 mL) and treated with H2O2 (30% aqueous
solution, 0.8 mL) and NaOH (12% aqueous solution, 0.5 mL)
for 15 minutes. TLC (ethyl acetate–hexane, 1 : 2) suggested the
deprotection to be complete. The reaction mixture was diluted
with H2O (70 mL) and extracted into ethyl acetate (5 × 20 mL).
The combined ethyl acetate extracts were washed with saturated
Na2SO3 solution (20 mL), saturated NaCl solution (20 mL),
then dried (Na2SO4) and concentrated in vacuo. The residue was
dissolved in CH2Cl2–MeOH and pre-adsorbed onto silica for flash
chromatography (ethyl acetate–hexane, 1 : 3 then 1 : 1), which
afforded diol 16 (0.◦081 g, 76%). Rf 0.5 (ethyl acetate–hexane,
1 : 1); mp 253–256 C; [a]2D2 −18.9 (c 0.19, CH2Cl2–MeOH, 4 :
1); mmax/cm−1 (film) 3550–3200 (OH), 3238 (≡C–H), 2918, 2106
=
disk) 3600–3000 (OH), 2927, 1710 (C O), 1447, 1055, 999; dH
(200 MHz, CDCl3) 3.86–3.73 (1H, m, H16), 2.99 (1H, br d, J 4.1,
16-OH), 2.76 (1H, br s, 17-OH), 2.48–1.93 (6H, m), 1.76–0.60
(16H, m), 1.00 (3H, s, CH3), 0.89 (3H, t, J 7.1, C21–H3), 0.84
(3H, s, CH3); dC (50 MHz, CDCl3) 212.2, 80.2, 72.8, 53.8, 46.6,
46.4, 45.8, 44.6, 38.5, 38.0, 35.7, 35.5, 35.2, 32.3, 31.6, 28.7, 27.1,
20.9, 14.2, 11.4, 7.2; m/z (EI+) 334.2515 (M+, C21H34O3 requires
334.2508, 25%), 247 (100), 229 (60), 217 (60), 159 (45), 91 (40).
Boronate ester mediated sulfation
Sodium 3b,16b,17b-trihydroxy-17a-methyl-5a-androstane 3-
sulfate (42). 17a-Methyl triol 182 (0.030 g, 0.093 mmol) was
stirred (18 was partially insoluble) in a mixture of DMF–CH2Cl2
(1 : 1, 0.3 mL). Phenylboronic acid (0.018 g, 0.15 mmol) was added,
˚
followed by 4 A molecular sieves (5). After 10 minutes, additional
CH2Cl2 (0.05 mL) was added, and after another 10 minutes,
TLC (ethyl acetate–hexane, 1 : 1) showed complete conversion
of starting material (Rf 0.2) to the corresponding boronic ester
(Rf 0.5). Additional DMF (1 mL) was added, followed by sulfur
trioxide–pyridine complex (0.044 g, 0.28 mmol). The reaction
mixture was stirred for 70 minutes after which time TLC (ethyl
acetate–hexane, 1 : 1) showed incomplete conversion. Additional
sulfur trioxide–pyridine complex (0.024 g, 0.15 mmol) was added
and stirring continued (30 minutes). TLC (CH2Cl2–MeOH, 4 : 1)
showed a single spot (Rf 0.4) due to intermediate product 43.
The molecular sieves were removed and the reaction mixture
was poured into saturated NaHCO3 solution (30 mL), which
was extracted with ethyl acetate (1 × 20 mL) and CHCl3–iPrOH
(3 : 1, 3 × 20 mL). The organic extracts were dried (Na2SO4)
=
(C≡C), 1703 (C O), 1150; dH (300 MHz, CDCl3–MeOD, 3 : 1)
3.93–3.85 (1H, m, H16), 2.39 (1H, s, ≡CH), 2.26–1.94 (4H, m),
1.87–1.76 (2H, m), 1.56–0.49 (14H, m), 0.80 (3H, s, CH3), 0.60
(3H, s, CH3); dC (75 MHz, CDCl3–MeOD, 3 : 1) 213.5, 85.7, 77.4,
76.5, 74.0, 53.2, 46.3, 45.7, 44.1, 38.1, 37.6, 35.3, 35.0, 34.0, 33.2,
31.0, 28.3, 20.4, 12.2, 10.8; m/z (EI+) 330.2187 (M+, C21H30O3
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3951–3959 | 3957
©

View Article Online
and concentrated. The residue obtained was directly dissolved
in a mixture of THF–MeOH (2 : 1, 1.5 mL) and treated with
H2O2 (30% aqueous solution, 0.23 mL) and saturated NaHCO3
solution (0.17 mL) for 10 minutes. TLC (CH2Cl2–MeOH, 4 : 1)
suggested the complete conversion to deprotected product (Rf 0.3).
The reaction mixture was diluted with saturated Na2SO3 solution
(30 mL) and extracted into CHCl3–iPrOH (3 : 1, 5 × 15 mL). The
combined organic extracts were dried (Na2SO4) and concentrated
in vacuo. The residue was dissolved in CH2Cl2–MeOH and pre-
adsorbed onto silica for flash chromatography (CH2Cl2–MeOH–
H2O, 60 : 15 : 1), which afforded 3-sulfate 42 (0.031 g, 79%). Rf 0.3
(CH2Cl2–MeOH–H2O, 60 : 15 : 1); mp 176–179 ◦C; [a]D23 −41.8 (c
0.22, MeOH); mmax/cm−1 (film) 3700–3100 (OH), 2943, 2914, 1641,
1447, 1379, 1223 (O–SO2), 1061; dH (300 MHz, CDCl3–MeOD,
1 : 1) 4.32–4.16 (1H, m, H3), 3.56 (1H, dd, J 8.0, 5.6, H16), 2.19–
1.94 (2H, m), 1.84–0.78 (17H, m), 1.07 (3H, s, CH3), 0.80 (3H, s,
CH3), 0.78 (3H, s, CH3), 0.68–0.56 (1H, m); dC (75 MHz, DMSO)
77.8, 76.2, 74.9, 53.8, 46.5, 44.5, 44.5, 36.6, 35.4, 35.2, 35.1, 34.9,
32.2, 31.6, 28.6, 28.3, 24.1, 20.2, 13.8, 12.0; m/z (ESI-) 401.1988
(C20H33O6S requires 401.1998, 100%).
requires 436.3373, 2%), 379 (M+-tBu, 100), 303 (10), 285 (26), 267
(30).
Conclusions
The selective manipulation of steroid hydroxyl groups in steroid
triols, using boronate ester protection of the vicinal diol, has
enabled the regioselective derivatisation of the remaining hy-
droxyl group. One-pot phenylboronate protection, followed by
regioselective oxidations, sulfations or silyl ether protections were
achieved, with subsequent oxidative deprotection of the boronate
ester rapidly providing the derivatised steroid. This methodology
has been applied to the efficient preparation of steroid antigens 5, 7
and 9 and sulfates 42 and 44. In the future, this methodology could
be extended to the preparation of further derivatised steroids, e.g.
glycosylated steroids.
A range of new steroid protein conjugates 7–10 have been
prepared. These are currently being applied to develop ELISAs as
screening tools for the detection of steroid metabolites. It is antic-
ipated based on previous results that these conjugates will provide
assays to detect D-ring structures associated with ethylestrenol
11, norethandrolone 12 and danazol 13, their metabolites and
structurally related steroids. The generation of antibodies and
the development of ELISAs is currently in progress and will be
reported elsewhere.
Boronate ester mediated alcohol protection
3b-(tert-Butyldimethylsilyloxy)-17a-methyl-5a-androstane-16b,
17b-diol (46). 17a-Methyl triol 182 (0.250 g, 0.78 mmol) was
stirred (18 was partially insoluble) in a mixture of DMF–CH2Cl2
(1.3 : 1, 1.6 mL). Phenylboronic acid (0.151 g, 1.24 mmol) was
Acknowledgements
˚
added, followed by 4 A molecular sieves (20). After 4 hours,
TLC (ethyl acetate–hexane, 1 : 1) showed complete conversion
of starting material (Rf 0.2) to the corresponding boronic ester (Rf
0.5). Additional DMF (0.7 mL) was added, followed by imidazole
(0.423 g, 6.21 mmol) and TBSCl. The reaction mixture was stirred
for 40 hours, when TLC (ethyl acetate–hexane, 1 : 4) showed a
single spot (Rf 0.7) due to intermediate product 45. The reaction
mixture was diluted with ethyl acetate (30 mL) and filtered through
celite. Saturated NaHCO3 solution (50 mL) was added, and the
layers separated. The aqueous portion was further extracted with
ethyl acetate (2 × 30 mL). The combined organic extracts were
washed with saturated NaCl solution (20 mL), dried over Na2SO4
and concentrated. The resulting residue was dissolved in THF
(7 mL) and treated with H2O2 (30% aqueous solution, 1.7 mL)
and NaOH (3 M, 1.1 mL) for 4 hours. TLC (ethyl acetate–hexane,
1 : 4) suggested the deprotection to be complete. The reaction
mixture was diluted with H2O (100 mL) and extracted into ethyl
acetate (3 × 40 mL). The combined ethyl acetate extracts were
washed with saturated Na2SO3 solution (30 mL), saturated NaCl
solution (30 mL), then dried (Na2SO4) and concentrated in vacuo.
The residue was dissolved in CH2Cl2 and pre-adsorbed onto silica
for flash chromatography (ethyl acetate–hexane, 1 : 4 to 1 : 2),
which afforded diol 46 (0.248 g, 73%). Rf 0.5 (ethyl acetate–hexane,
1 : 1); mp 216–220 ◦C; [a]D22 −18.0 (c 0.51, CHCl3); mmax/cm−1 (film)
3600–3000 (OH), 2931, 1377, 1360, 1250, 1091, 1070, 1053; dH
(200 MHz, CDCl3) 3.64 (1H, dd, J 8.0, 5.4, H16), 3.60–3.44 (1H,
m, H3), 2.64 (1H, br s, OH), 2.57 (1H, br s, OH), 2.26–2.08 (1H,
m), 1.74–0.78 (18H, m), 1.12 (3H, s, CH3), 0.88 (9H, s, (CH3)3CSi),
0.83 (3H, s, CH3), 0.81 (3H, s, CH3), 0.66–0.50 (1H, m), 0.04 (6H, s,
(CH3)2Si); dC (50 MHz, CDCl3) 79.2, 77.8, 72.1, 54.6, 47.1, 45.1,
44.9, 38.6, 37.2, 35.8, 35.6, 34.9, 32.5, 32.0, 31.9, 28.6, 25.9, 23.8,
20.6, 18.2, 13.6, 12.4, −4.6; m/z (EI+) 436.3364 (M+, C26H48O3Si
This work was supported by the Australian Racing Forensic
Laboratory and the Australian Research Council (LP0211196).
References
1 D. H. Catlin, M. H. Sekera, B. D. Ahrens, B. Starcevic, Y.-C. Chang
and C. K. Hatton, Rapid Commun. Mass Spectrom., 2004, 18, 1245–
49; D. H. Catlin, B. D. Ahrens and Y. Kucherova, Rapid Commun.
Mass Spectrom., 2002, 16, 1273–75; M. H. Sekera, B. D. Ahrens, Y.-
C. Chang, B. Starcevic, C. Georgakopoulos and D. H. Catlin, Rapid
Commun. Mass Spectrom., 2005, 19, 781–84.
2 N. L. Hungerford, B. Sortais, C. G. Smart, A. R. McKinney, D. D.
Ridley, A. M. Stenhouse, C. J. Suann, K. J. Munn, M. N. Sillence and
M. D. McLeod, J. Steroid Biochem. Mol. Biol., 2005, 96, 317–34.
3 For recent studies see: M. C. Dumasia, Rapid Commun. Mass Spec-
trom., 2003, 17, 320–29; A. R. McKinney, C. J. Suann, A. J. Dunstan,
S. L. Mulley, D. D. Ridley and A. M. Stenhouse, J. Chromatogr., B:
Biomed. Appl., 2004, 811, 75–83; A. R. McKinney, D. D. Ridley and
C. J. Suann, J. Mass Spectrom., 2001, 36, 145–50; A. R. McKinney and
D. D. Ridley, Aust. J. Chem., 2001, 54, 757–61; S. Poelmans, K. De
Wasch, H. F. De Brabander, M. Van De Wiele, D. Courtheyn, L. A.
van Ginkel, S. S. Sterk, Ph. Delahaut, M. Dubois, R. Schilt, M. Nielen,
J. Vercammen, S. Impens, R. Stephany, T. Hamoir, G. Pottie, C. Van
Poucke and C. Van Peteghem, Anal. Chim. Acta, 2002, 473, 39–47.
4 A. R. McKinney, D. D. Ridley and P. Turner, Aust. J. Chem., 2003, 56,
829–38.
5 T. Takegoshi, Chem. Pharm. Bull., 1972, 20, 1260–71.
6 N. S. Leeds, D. K. Fukushima and T. F. Gallagher, J. Am. Chem.
Soc., 1954, 76, 2943–48; M. Numazawa, M. Shelangouski and M.
Nakakoshi, Steroids, 2001, 66, 743–48.
7 A similar result was obtained for the addition of ethylmagnesium iodide
to 3b,16b-diacetoxy-5a-estran-17-one (ref. 4).
8 J. R. Parikh and W. v. E. Doering, J. Am. Chem. Soc., 1967, 89, 5505–07.
9 P. J. Duggan and E. M. Tyndall, J. Chem. Soc., Perkin Trans. 1, 2002,
1325–39.
10 J. Pis, J. Hykl, M. Budesinsky and J. Harmatha, Collect. Czech. Chem.
Commun., 1993, 58, 612–18; J. Pis, J. Hykl, M. Budesinsky and J.
Harmatha, Tetrahedron, 1994, 50, 9679–90.
3958 | Org. Biomol. Chem., 2006, 4, 3951–3959
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
11 Y.-S. Cheng, W.-L. Liu and S.-H. Chen, Synthesis, 1980, 223–24; C.
Liljebris, B. M. Nilsson, B. Resul and U. Hacksell, J. Org. Chem., 1996,
61, 4028–4034.
12 On one occasion, treatment of steroid 16 with acidic CDCl3, resulted
in unwanted C17 epimerisation. To avoid this, storage of CDCl3 over
anhydrous K2CO3 and filtration through basic alumina is necessary.
13 F. Pu, A. R. McKinney, A. M. Stenhouse, C. J. Suann and M. D.
McLeod, J. Chromatogr., B: Biomed. Appl., 2004, 813, 241–6; W.
Scha¨nzer, Clin. Chem., 1996, 42, 1001–20.
14 A. R. McKinney, C. J. Suann and A. M. Stenhouse, Anal. Chim. Acta,
2006, accepted for publication (10.1016/j.aca.2006.08.025).
15 S. Langston, B. Bernet and A. Vasella, Helv. Chim. Acta, 1994, 77,
2341–53.
16 V. Bhaskar, A. D’Elia, P. J. Duggan, D. G. Humphrey, G. Y. Krippner,
T. T. Nhan and E. M. Tyndall, J. Carbohydr. Chem., 2003, 22, 867–79.
17 E. J. Corey and A. Venkateswarlu, J. Am. Chem. Soc., 1972, 94, 6190–
91.
18 F. Pu, personal communication.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3951–3959 | 3959
©