Ergosteroids VII: perchloric acid-induced transformations of 7-oxygenated steroids and their bio-analytical applications--a liquid chromatographic-mass spectrometric study.

Article abstract of DOI:10.1016/S0045-2068(02)00010-X

Sulfate esters of 7-oxo-delta(5)-steroids can be selectively and quantitatively hydrolyzed to the corresponding free steroids in the presence of carboxylic acid esters by solvolysis with perchloric acid in ethyl acetate at room temperature. Sulfates as well as carboxylic acid esters, methyl ethers, and ketals can be quantitatively converted to the corresponding 3,5-diene-7-one derivatives by heating with perchloric acid in methanol at 65 degrees C. The dienes have a strong UV absorption with maximum centered around 284 nm. These reactions have been used for the characterization and structural elucidation of 7-oxygenated-delta(5)-steroids that are present in complex biomatrices and can also be used for the quantitative estimation of total 7-oxo-delta(5)-steroids (free as well as conjugated) in biological matrices.

Full text of DOI:10.1016/S0045-2068(02)00010-X

BIOORGANIC
CHEMISTRY
Bioorganic Chemistry 30 (2002) 233–248
www.academicpress.com
Ergosteroids VII: perchloric acid-induced
transformations of 7-oxygenated steroids
and their bio-analytical applications—a
liquid chromatographic-mass
q
spectrometric study
Ashok Marwah, Padma Marwah, and Henry Lardy*
Department of Biochemistry, Institute for Enzyme Research, University of Wisconsin,
Madison, WI 53726, USA
Received 12 December 2001
Abstract
5
Sulfate esters of 7-oxo-D -steroids can be selectively and quantitatively hydro-
lyzed to the corresponding free steroids in the presence of carboxylic acid esters by
solvolysis with perchloric acid in ethyl acetate at room temperature. Sulfates as well
as carboxylic acid esters, methyl ethers, and ketals can be quantitatively converted to
the corresponding 3,5-diene-7-one derivatives by heating with perchloric acid in
methanol at 65 °C. The dienes have a strong UV absorption with maximum centered
around 284 nm. These reactions have been used for the characterization and struc-
5
tural elucidation of 7-oxygenated-D -steroids that are present in complex biomatrices
5
and can also be used for the quantitative estimation of total 7-oxo-D -steroids (free
as well as conjugated) in biological matrices. Ó 2002 Elsevier Science (USA). All
rights reserved.
q
Abbreviations: APCI, Atmospheric pressure chemical ionization; DAD, diode array detector; ESI,
electrospray ionization; GC, gas chromatography; HPLC, high performance liquid chromatography; LC-
MS, liquid chromatography mass spectrometry; SIM, selected ion monitoring; UV, ultraviolet. DHEA,
dehydroepiandrosterone, 3b-hydroxyandrost-5-en-17-one.
*
Corresponding author. Fax: 608-265-2904.
E-mail address: halardy@facstaff.wisc.edu (H. Lardy).
0045-2068/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S0045-2068(02)00010-X

234
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
5
Keywords: 7-Oxo-dehydroepiandrosterone; 7-Oxygenated-D -steroids; Sulfate conjugates;
Perchloric acid; LC-MS
1
. Introduction
Hydrolysis of steroid conjugates has been of major interest to steroid investigators
because only a tiny fraction of steroids and their metabolites are excreted unchanged
and because of difficulties associated with the analysis of intact steroid conjugates [1].
A quantum leap in the field of bio-analytical chemistry, such as the advent of bench-
top mass spectrometric detectors, which can be coupled to GC and HPLC, and ready
availability of HPLC columns, which can withstand the deleterious effects of extreme
pH (1–11) mobile phases, has alleviated several of the difficulties associated with the
analysis of intact steroid conjugates. But hydrolysis of these conjugates, prior to
analysis, continues to be the most precise and direct method available. Steroids such
as dehydroepiandrosterone (DHEA) are known to exist in several forms in tissues
and blood (free steroid, sulfate conjugate, fatty acid esters, and glucuronides) [2,3],
making it difficult to measure the total steroid concentration in the biomatrix, and
leaving the scientist with no other option but to resort to hydrolysis, chemically or
otherwise.
Traditionally steroid conjugates have been hydrolyzed using strong inorganic
acids or enzymatically [1,4–9]. Both approaches have their advantages and limita-
tions. The enzymatic hydrolysis is mild, but may be time-consuming or may even be
incomplete in some matrices due to inhibition of enzymes [10]. Alterations in
structure and artifact formation in the case of labile steroid molecules during solv-
olytic procedures involving the use of strong acids cannot be ruled out. Perchloric
acid has been used for the hydrolysis of organic sulfates; and steroid sulfates, dis-
solved in organic solvents, have been solvolyzed by catalytic amount of perchloric
acid. Other polar derivatives of sterols such as phosphates and glucuronides were
unaffected [5]. A search of the literature revealed that the sulfate conjugate (I) of
7
-oxo-DHEA (II, 3b-hydroxyandrost-5-ene-7,17-dione) was resistant to hydrolysis
under various conditions (continuous ether extraction of acidified solution, enzy-
matic cleavage with a sulfatase-containing preparation, and acetic ester solvolysis,
etc.) and a complex hydrolysis could be achieved [11] only at a high concentration of
sulfuric acid, which also resulted in the partial formation of androsta-3,5-diene-7,
1
7-dione (VII).
During our ongoing investigation of the metabolism of DHEA and its 7-oxy-
genated derivatives in search of active metabolites that are responsible for the
enhancement of the formation of liver mitochondrial sn-glycerol 3-phosphate de-
hydrogenase and cytosolic malic enzyme [12–14], we expected and encountered the
problem of poor and variable recoveries of various 7-oxygenated DHEA derivatives
whenever acidic conditions were employed. The hydrolysis of sulfate conjugates of
7
-oxygenated DHEA derivatives was found to be highly dependent on the nature of
the solvent, reaction temperature, and time. Yields were variable and often a mixture
of products was obtained.

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
235
This aspect of hydrolysis of 7-oxygenated steroids has nowbeen investigated in
detail. We studied the hydrolysis and/or dehydration of 7-oxo-DHEA (II), its sulfate
and carboxylic esters, and of other related 7-oxygenated compounds such as 7-
hydroxy-DHEA, 3b,17b-dihydroxyandrost-5-en-7-one, androst-5-en-3b,7,17b-triol,
7
-oxo-cholesterol, etc. so as to find appropriate reaction conditions, leading to the
selective and exclusive formation of either hydrolyzed product or hydrolyzed and
dehydrated product. The reaction conditions were developed with microgram and
sub-microgram quantities so that these reactions were directly applicable to the
5
study, in biomatrices, of 7-oxygenated derivatives of 3-hydroxy-D -steroids, in
general, and 7-oxo-DHEA (II), in particular. It may be mentioned that 7-oxo-
DHEA (II) is a metabolite [15–17] of DHEA, once termed ‘‘fountain of youth’’ [18].
7
-Oxo-DHEA (II) is more active than DHEA in two biological assays [19–21] but
unlike DHEA is not converted to either androgens or estrogens [19,22]. It was iso-
lated from human urine [15,23] and peripheral venous plasma [24] more than four
decades ago but only a fewmethods are available for its analysis in biological fluids
or tissues [25]. We have reported [26] a validated HPLC method for the determi-
nation of 7-oxo-DHEA-3b-sulfate (I) in human plasma. In the present paper, we are
reporting for the first time a method, which can be used for the analysis of total
7
-oxo-DHEA (II) that is present in biomatrices. The method has also been used to
study the hydrolysis of fatty esters of 7-oxo-DHEA by rat brain homogenate and in
the characterization of the metabolites of DHEA produced in vitro by rat liver [27].
2
. Materials and methods
All reactions were carried out at nanomole (ꢀ0.1–40 nmol) scale. The reaction
mixture was analyzed by HPLC and LC-MS and yields were calculated using ex-
ternal standard methods. The identity of the products was established by comparison
with authentic samples (retention times and chromatography of spiked samples),
online diode array UV spectrum, and mass spectrum analysis. For bio-analytical
studies, pooled blood plasma samples from healthy human volunteers, obtained
from University of Wisconsin-Madison Hospital, were used.
The chromatographic system consisted of an Agilent 1100 series LC-MSD system,
which was composed of a quaternary pump (G1311A) with online degasser, ther-
mostated column compartment (G1316A), autosampler (G1313A), diode array UV
detector (DAD, G1315A), and an online mass detector (G1946A) that was coupled
with atmospheric pressure electrospray (ESI) or atmospheric pressure chemical
ionization device (APCI), as needed. Data were acquired and processed using
HPLC/MSD Chemstation version A.08.03 software from Hewlett–Packard.
Chromatography of the chemical samples was performed on a Zorbax C18 ana-
lytical column (3.5 lm, 4.6 ꢁ 75 mm) that was protected with a C guard column
18
(
Zorbax SB-C18, 4.6 ꢁ 12.5 mm, 5l). The column temperature was maintained at
0 ꢂ 0:5 °C. The flow rate was set at 1.0 mL/min and the eluent was simultaneously
monitored at three wavelengths, viz., 205 nm (bandwidth 10 nm), 240 nm (bandwidth
nm), and 284 nm (bandwidth 4 nm), respectively. The reference wavelength was set
at 360 nm for 205 and 240 nm wavelengths and 450 nm for 284 nm wavelength. The
4
4

236
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
slit width was kept at 10 nm for detection at 205 nm, 4 nm for detection at 240 nm,
and 16 nm for detection at 284 nm. A methanol–water linear gradient (mobile phase
A) was used to analyze the samples. The linear gradient started from 30% methanol
to 96% methanol in 33 min and was brought back to an initial concentration of
30% in 3 min. A 10 min post-run was used to re-equilibrate the column and for the
base line to return to the normal. For the analysis of sulfate conjugates, acetic acid
(
(
1.0% v/v) was added to methanol and to water and a slightly different gradient
15% methanol to 60% methanol in 30 min) was used (mobile phase B). 7-Oxo-
DHEA-3b-stearate (III) and cholesterol (VI) were analyzed using a higher percent-
age of methanol (85–100% in 15 min; mobile phase C). The testosterone derivatives
(X–XVI) were analyzed using acetonitrile–water gradient (20–45% acetonitrile in
25 min; mobile phase D). The products obtained by perchloric acid treatment were
also analyzed using LC-MS-ESI in positive mode. Operating conditions were opti-
mized by flowinjection analysis (FIA) of 7-oxo-DHEA (II) and 7 b-hydroxy-DHEA
(
XVII) and the final parameters used were determined as: drying gas (N ) 12 L/min
2
at 350 °C; nebulizer pressure 40 psi; capillary voltage 5000 V; and gain 5. The samples
were run in scan mode for the mass range 200–600 and the fragmentor was set at
1
ical ionization (APCI) mode. Operating conditions were as follows: drying gas (N
00 V. 7-Oxo-DHEA-3b-stearate (III) and cholesterol (VI) were analyzed in chem-
2
)
5
.0 L/min at 350 °C; nebulizer 60 psi; capillary voltage 4000 V; gain 5; vaporizer at
4
and fragmentor was set at 70 V.
00 °C and corona at 4 lA. The samples were run in scan mode (mass range 200–600)
A different HPLC system was used for the analysis of plasma samples after per-
chloric acid treatment to convert 7-oxo-DHEA (II) to androsta-3,5-diene-7,17-dione
(
VII). It consisted of a Zorbax-SB column (C , 3.5l, 4.6ꢁ50 mm) and methanol–
8
water gradient containing 1% of 0.1 M phosphoric acid (50–70% methanol in 10 min)
at 1.0 mL flow rate (mobile phase E). The monitoring was at 284 nm with a band-
width of 4 nm. Plasma samples were also analyzed by LC-MS in ESI mode. Non-
volatile buffers and acids are deleterious to the mass detector; hence, LC-MS was
performed using methanol–water (pH 4.50 with acetic acid) gradient (65–100%
methanol at 0.5 mL/min flowrate, mobile phase F). Mass detector conditions were:
nebulizer 30 psi; N gas flow10 L/min at 350 °C; voltage 4500 V; fragmentor 90 V;
2
gain 5. Detection was made in selected ion monitoring mode (SIM) at m=z ¼ 285
þ
þ
ðM þ HÞ and 307 (M + Na) .
2.1. Hydrolysis of sulfate esters: reaction of 7-oxo-DHEA-3b-sulfate(I) with
perchloric acid in ethyl acetate
Perchloric acid (10 lL, 70%) was added to 7-oxo-DHEA-3b-SO (I, 1.0 lg in 2 mL
4
ethyl acetate) in a screw-capped tube and allowed to stand at 25 °C after cyclo-
mixing. After 0.25 h, the ethyl acetate layer was washed with 5% aq. bicarbonate
(
1 mL) and water (2 ꢁ 1 mL) the organic layer was separated by snap-drying in an
acetone–dry ice bath and evaporated under nitrogen. The residue was dissolved in
methanol–water (1:1, 200 lL) and a 10 lL sample was analyzed by LC-MS (Table 1,
entry 1). The product was identified as 7-oxo-DHEA (II) by comparison with an
R
authentic sample: retention time (t ) 11.71 min (mobile phase B); yield 99.5%

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
237
Table 1
The perchloric acid-induced transformations of 7-oxygenated steroids
a
S. No. Compound Solvent Temperature (°C) Time (h) Product composition (%)
a
b
c
1
2
3
4
5
6
7
8
9
I
EtOAc
EtOAc
Ether
25
40
25
0.25
8.0
0
>99
1
<1
99
I
0
I
0.75
0.5
88
0
12
0
0
I
CH
3
OH 65
100
0
II
III
III
IV
V
VI
VI
X
EtOAc
EtOAc
25
25
0.25
0.25
0.7
–
100
0
100
0
0
CH
CH
CH
3
3
3
OH 65
OH 65
OH 65
0
100
100
99.8
1
0.7
0.7
0
0
0
0.2
99
0.1
0
1
0
1
2
EtOAc
CH OH 65
EtOAc 40
25
0.5
0.7
–
1
1
3
–
99.9
99
1.0
1
I, 7-oxo-DHEA-3b-sulfate; II, 7-oxo-DHEA; III, 7-oxo-DHEA-3b-stearate; IV, 3b, 17b-di-
hydroxyandrost-5-en-7-one; V, 17b-hydroxy-3b-methoxyandrost-5-en-7-one; VI, 7-oxo-cho-
lesterol; X, 7-hydroxytestosterone, 3-ethylene glycol ketal.
a
a, Unreacted starting steroid; b, free steroid; c, hydrolyzed and dehydrated steroid.
þ
þ
þ
(
2
HPLC); LC-MS ðm=zÞ: 325 (M + Na) ; 303 ðM þ HÞ ; 285 (M–H O + H) ; k
2
max
42 nm (DAD).
2.2. Solvolysis followed by dehydration: reaction of 7-oxo-DHEA-3b-SO
perchloric acid in methanol
4
(I) with
Perchloric acid (10 lL, 70%) was added to a methanolic solution of 7-oxo-DHEA
100 ng in 2 mL methanol) in a screw-capped pyrex tube. The tubes were closed and
(
the contents were heated in a water bath at 65 °C. After 1 h, water (2 mL) was added
and methanol was removed under nitrogen at 40 °C. The remaining aqueous solution
was diluted with water (1 mL) and extracted with ether (3 mL). The ether layer was
washed with water (2 ꢁ 1 mL) and the ether was separated by snap-drying in
acetone–dry ice bath and evaporated under nitrogen. The residue was dissolved
in methanol–water (1:1, 200 lL) and a 10 lL aliquot was injected onto the chro-
matogram (Table 1, entry 4). The product was identified as androsta-3,5-diene-7,17-
dione (VII) by comparison with a sample synthesized in the laboratory: retention
time (t ) 19.76 min (mobile phase A); yield >99.8% (HPLC); LC-MS ðm=zÞ: 307
R
þ
þ
(
M + Na) ; 285 ðM þ HÞ ; k
284 nm (DAD).
max
2
.3. Extraction procedure for plasma samples and perchloric acid treatment
Plasma sample (1 mL each) was applied to Oasis solid-phase extraction cartridge
3
(
tridge was washed with water (1 mL) and 50% aqueous methanol containing 2%
Waters, 3 cm ), preconditioned with methanol (1 mL) and water (1 mL). The car-
acetic acid (1 mL) and dried under water pump suction (ꢀ10–15 in. of mercury) for

238
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
1
min. It was eluted with methanol (1 mL), transferred to a screw-capped glass tube,
and heated at 65 °C for 35 min after the addition of perchloric acid (70%, 5 lL).
Reaction mixture was diluted with water (4 mL) and extracted with ether (3 mL).
Ether layer was separated by snap-drying in an acetone–dry ice bath, washed with
water (1 mL), and evaporated under nitrogen at 40 °C. The residue was dissolved in
methanol–water (1:1 v/v, 300 lL) and 40 lL was injected into the HPLC system for
detection with DAD-UV at 284 nm (mobile phase E) and 15 lL for detection by LC-
þ
þ
MS (mobile phase F) in ESI-SIM mode at m=z ¼ 285 ðM þ HÞ or at 307 (M + Na) .
.4. Hydrolysis of 7-oxo-DHEA-3b-palmitate by rat brain homogenate
2
7
-Oxo-DHEA-3b-palmitate (10 lg) was incubated at 37 °C with rat brain ho-
mogenate (4 mL containing 1 g brain). Samples (1 mL each) were withdrawn at every
hour, diluted with acetonitrile (1 mL), and centrifuged (1500g) for 5 min. The su-
pernatant was concentrated (0.5 mL) under nitrogen and diluted with water (2 mL).
Aqueous layer was extracted with ether (3 mL). The ether layer was concentrated
under nitrogen and the residue was dissolved in methanol–water (60:40, 200 lL) and
subjected to LC-MS (methanol–water gradient, 30–100% methanol in 35 min, fol-
lowed by 100% methanol for 5 min). The extracted ion analysis showed the presence
þ
of 7-oxo-DHEA {t
R
11.8 min, m=z 325 (M + Na) } and 7-oxo-DHEA-3b-palmitate
þ
{t
R
37.2 min, m=z 563 (M + Na) }. The sample was analyzed again and fractions
corresponding to retention times of 7-oxo-DHEA and 7-oxo-DHEA-3b-palmitate
were collected separately, evaporated under nitrogen, and subjected to perchloric
acid reaction in methanol as mentioned earlier. The products were identified
as androsta-3,5-diene-7,17-dione (VII), as discussed above {t 19.76 min (mobile
R
þ
þ
phase A), 307 ðM þ NaÞ ; 285 ðM þ HÞ ; kmax 284 nm (DAD)}.
3
. Results and discussion
3
.1. Chemical degradations
It has been reported [4,5] that the sulfate conjugates of steroids can be solvolyzed
by treatment with perchloric acid using a variety of organic solvents. However, we
5
realized that the solvolysis of sulfate esters of 7-oxo-3-hydroxy-D -steroids was not a
straightforward reaction. The solvolysis was highly dependent on the nature of the
solvent, reaction temperature, time (Table 1, and Figs. 1 and 2), and mode of
workup. Use of polar solvents invariably led to the formation of diene(s). Even a
trace amount of residual acidity caused the dehydration during concentration, re-
sulting in the formation of complex mixtures, the degree of complexity depending
5
upon the nature of the 7-oxygenated-D -steroid and reaction workup conditions. We
were able to achieve a highly selective hydrolysis of sulfate esters in ethyl acetate
(Scheme 1) at ambient temperature in 15 min to yield the corresponding free steroids
in near-quantitative yield (Table 1, entry 1). These conditions did not alter the free
5
7
(
-oxo-3-hydroxy-D -steroids (Table 1, entry 5). Replacing ethyl acetate with ether
Table 1, entry 3) or methanol (Fig. 1), or a change in temperature (Fig. 2) or

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
239
Fig. 1. The effect of reaction time on the hydrolysis of 7-oxo-DHEA-3b-sulfate using
perchloric acid in methanol (0.5% v/v) at room temperature. (I) 7-Oxo-DHEA-3b-sulfate;
(
II) 7-oxo-DHEA; (VII) androsta-3,5-diene-7,17-dione.
Fig. 2. The effect of temperature on the hydrolysis of 7-oxo-DHEA-3b-sulfate using
perchloric acid in ethyl acetate (0.5% v/v). Reaction time 15 min. (I) 7-Oxo-DHEA-3b-sulfate;
(
II) 7-oxo-DHEA; (VII) androsta-3,5-diene-7,17-dione.

240
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
Scheme 1. Hydrolysis of sulfate esters with perchloric acid (0.5% v/v) in ethyl acetate at room
temperature. (I) 7-Oxo-DHEA-3b-sulfate; (II) 7-oxo-DHEA.
5
Scheme 2. Transformation of 7-oxo-D -steroids into 3,5-dien-7-one derivatives using
perchloric acid in methanol at 65 °C. (I) 7-Oxo-DHEA-3b-sulfate; (II) 7-oxo-DHEA; (III) 7-
oxo-DHEA-3b-stearate; (IV) 3b,17b-dihydroxyandrost-5-en-7-one; (V) 17b-hydroxy-3b-
methoxyandrost-5-en-7-one; (VI) 7-oxo-cholesterol; (VII) androsta-3,5-diene-7,17-dione;
(
VIII) 17b-hydroxyandrosta-3,5-dien-7-one; (IX) cholesta-3,5-dien-7-one.
reaction time (Figs. 1 and Table 1, entry 2), resulted in the partial hydrolysis and/or
the formation of the dienone product. Carboxylic acid esters remained unaffected
5
(
Table 1, entry 6) in ethyl acetate. However, free 7-oxo-3-hydroxy-D -steroids
(Table 1, entries 8 and 11) as well as their carboxylic esters (Table 1, entry 7), sulfate
conjugates (Table 1, entry 4), and methyl ether (Table 1, entry 9) were all quanti-
tatively converted to corresponding 3,5-dien-7-one derivatives in methanol at 65 °C
(Scheme 2). The unconjugated 17-hydroxyl group present in the starting steroid
remained unaffected (Table 1, entries 8, 9, and 12).
4
This reaction was applicable with equal ease to 7-hydroxy-D -3-oxo steroids
Table 1, entry 12), the product being a 3-oxo-4,6-diene derivative (XI; Scheme 3).
(
We next studied the behavior of 7-oxo-testosterone as its 3-ketal derivative (XII).
This compound readily underwent deketalization in ethyl acetate/perchloric acid at
4
5
ambient temperature (Scheme 4) to yield a mix of D and D products (XIII and
XIV). However, upon heating in methanol/perchloric acid, XII yielded a mixture of

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
241
Scheme 3. Perchloric acid-induced transformation of 7-hydroxytestosterone, ethylene glycol
ketal (X) in ethyl acetate at ambient temperature; (XI) 17-hydroxyandrosta-4,6-dien-3-one.
Scheme 4. Perchloric acid-induced transformation of 7-oxotestosterone, ethylene glycol ketal
(XII) in methanol at 65 °C; (XIII) 7-oxo-testosterone; (XIV) 17-hydroxyandrost-5-ene-3,7-
dione; (XV) 17-hydroxy-3-methoxyandrosta-3,5-dien-7-one; (XVI) 17-hydroxy-7-methoxy-
androsta-4,6-dien-3-one.
1
7b-hydroxy-7-methoxyandrosta-4,6-dien-3-one (XV) and 17b-hydroxy-3-methoxy-
androsta-4,6-dien-7-one (XVI). The formation of these products demonstrated that
the keto groups (3 and 7) underwent enolization, followed by etherification. The
driving force for this reaction is provided by the formation of extended conjugation
ðk
322 nm). The 17-hydroxyl group remained unaffected. Therefore, perchloric
max
acid may be used for the selective alkylation of a conjugated hydroxyl group in the
presence of unconjugated alcohols. A partial formation of corresponding ethyl
ethers of XV and XVI was seen, even when, reaction time was extended during the
deketalization process in ethyl acetate.

242
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
7
-Hydroxy-DHEA (XVII) as well as androst-5-ene-3b,7,17b-triol (XVIII) lose up
to two molecules of water (resulting from loss of the 3-OH and 7-OH groups) on
treatment with perchloric acid (Scheme 5), yielding a mixture of several dienes and
trienes; the major products (>70%) were androsta-2,4,6-trien-17-one {XIX, m=z 269
þ
þ
þ
ðM þ HÞ and 291 (M + Na) } and androsta-2,4,6-trien-17-ol {XX, m=z 271
þ
1
ðM þ HÞ and 253 (M + H-H
2
OÞ }. Their structures were confirmed further by H
13
NMR and C NMR spectral analyses and k
(DAD) at 306 nm (calculated
max
303 nm). Table 2 lists the LC-MS data (kmax as seen in diode array UV spectra,
retention times, and the most abundant ions) of various 7-oxygenated steroids and
the products obtained by their perchloric acid treatment.
The reactions studied in the present investigation were carried out at nanomole and
sub-nanomole concentrations for LC-MS studies. Their progress and product com-
position were studied by HPLC and LC-MS. The products were identified, on the
basis of retention times, by comparison with known standards, elution behavior in
HPLC as well as LC-MS, UV spectra (obtained from an online diode array detector),
co-chromatography, derivatization to known compounds, and last but not the least
with the help of mass spectrometry by running the samples under different ionization
modes, i.e., positive vs. negative and electrospray vs. chemical ionization, fragmen-
tation patterns, and by comparison of mass spectra with those of known standards.
7
-Oxo-DHEA (I) has a UV maximum near 240 nm; therefore, all samples were
monitored at 240 nm for the presence of 7-oxo-DHEA (II) and related compounds.
The chromatograms were also monitored at 205 nm for the presence of 7-hydroxy
compounds. Lastly, the samples were simultaneously monitored at 284 and 310 nm
for the presence of conjugated dienes, trienes, and dienones, particularly after per-
5
chloric acid treatment. The products formed by the dehydration of 7-OH-D -steroids
are comparatively non-polar compounds containing two or more double bonds and
the absence of polar groups makes them poor substrates for electrospray mass
spectrometry. Consequently, they were also subjected to chemical ionization mass
spectrometry.
Scheme 5. Perchloric acid-induced dehydration of 7-hydroxy-DHEA (XVII) and androst-
5
-ene-3b, 7,17b-triol (XVIII) in methanol at 65 °C; (XIX) androsta-2,4,6-trien-17-one; (XX)
androsta-2,4,6-trien-17b-ol.

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
243
Table 2
LC-MS analysis of 7-oxygenated steroids and their reaction products after perchloric acid
treatment
Ã
Compound
Retention UV kmax Mass sp. (ESI) ðm=zÞ
time (min) (nm)
a
b
c
À
I
20.2
11.7
14.7
242
242
236
242
242
238
284
284
280
234
381 ðM–HÞ
þ
þ
OÞþ
II
325 ðM þ NaÞ ; 303 ðM þ HÞ ; 285 ðM þ H–H
2
þd
III
IV
V
285 ðM þ H–C17
H
35COOHÞ
b
þ
þ
þ
þ
þ
þd
þ
þ
þ
9.8
16.3
631 ð2M þ NaÞ ; 327 ðM þ NaÞ ; 305 ðM þ HÞ
b
þ
341 ðM þ NaÞ ; 319 ðM þ HÞ
c
OÞþd
VI
VII
VIII
IX
X
8.8
401 ðM þ HÞ ; 383 ðM þ H–H
2
þ
b
b
c
19.8
19.4
13.4
24.6
307 ðM þ NaÞ ; 285 ðM þ HÞ
þ
þ
595 ð2M þ NaÞ ; 309 ðM þ NaÞ ; 287 ðM þ HÞ
383 ðM þ HÞ
e
þ
þ
371 ðM þ NaÞ ; 349 ðM þ HÞ ; 331 ðM þ H–H
2
OÞ ;
þ
2
87 ðM þ H–H
2
þ
þ
þ
þ
þ
þ
þ
O–OCH
2
CH
2
Þ
e
c
e
e
e
b
þ
þ
XI
XII
21.9
15.9
10.9
288
242
322
313
324
–
595 ð2M þ NaÞ ; 309 ðM þ NaÞ ; 287 ðM þ HÞ
þ
þ
þ
þ
369 ðM þ NaÞ ; 347 ðM þ HÞ
XIII + XIV
325 ðM þ NaÞ ; 303 ðM þ HÞ
XV and XVI 21.4
2.1
339 ðM þ NaÞ ; 317 ðM þ HÞ
2
13.2
339 ðM þ NaÞ ; 317 ðM þ HÞ
þ
OÞþ
XVII
XVIII
327 ðM þ NaÞ ; 287 ðM–H
2
OÞ ; 269 ðM–2H
2
b
þ
8.3
–
329 ðM þ NaÞ ; 289 ðM–H
2
OÞ ;
þ
þ
þ
þ
þ
2
71 ðM–2H
2
OÞ ; 253 ðM–3H
2
OÞ
b
b
XIX
XX
27.9
28.3
306
306
559 ð2M þ NaÞ ; 291 ðM þ NaÞ ;
þ
þ
2
69 ðM þ HÞ ; 241 ðM þ H–COÞ
þ
271 ðM þ HÞ ; 253 ðM þ H–H
2
OÞ
I, 7-Oxo-DHEA-3b-sulfate; II, 7-oxo-DHEA; III, 7-oxo-DHEA-3b-stearate; IV, 3b,17b-
dihydroxyandrost-5-en-7-one; V, 17b-hydroxy-3b-methoxyandrost-5-en-7-one; VI, 7-oxo-
cholesterol; VII, androsta-3,5-diene-7,17-dione; VIII, 17b-hydroxyandrosta-3,5-dien-7-one; IX,
cholesta-3,5-dien-7-one; X, 7-hydroxytestosterone, 3-ethylene glycol ketal; XI, 17-hydroxyan-
drosta-4,6-dien-3-one; XII, 7-oxotestosterone, ethylene glycol ketal; XIII, 7-oxo-testosterone;
XIV, 17-hydroxyandrost-5-ene-3,7-dione; XV, 17-hydroxy-3-methoxyandrosta-3,5-dien-7-one;
XVI, 17-hydroxy-7-methoxyandrosta-4,6-dien-3-one; XVII, 7-hydroxy-DHEA; XVIII, and-
rost-5-ene-3b,7,17b-triol; XIX, androsta-2,4,6-trien-17-one; XX, androsta-2,4,6-trien-17b-ol.
*
ESI mode unless specified otherwise.
Mobile phase B.
a
b
Mobile phase A.
Mobile phase C.
c
d
APCI mode.
Mobile phase D.
e
3
.2. Bioanalytical applications
A major constraint in the analysis of steroids in biomatrices is the presence of
steroids and their metabolites in nanogram quantities. The problem is complicated
further for steroid derivatives, which may be present as more than one conjugate, thus,

244
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
making it difficult to determine their total concentration in the system. As discussed
5
above, the 3-hydroxy-D -steroids oxygenated at position 7 are prone to dehydration
and this tendency is particularly pronounced in the case of 7-oxo derivatives, which
were selectively and quantitatively converted into 3,5-dien-7-one derivatives, which
are stable and also have a better UV absorbance. The same is true for 7-hydroxy
4
derivatives of 3-oxo-D -steroids. These features were successfully exploited in devel-
oping a strategy for the estimation of total 7-oxo-DHEA (II) and its derivatives in
biological matrices (human plasma, rat liver, and brain) after their conversion to the
corresponding 3,5-diene-7-one. In a typical example, human plasma was subjected to
cleanup by solid-phase extraction and the total 7-oxo-DHEA was eluted with meth-
anol. The methanol solution was treated with perchloric acid to convert the total 7-
oxo-DHEA into dienone derivative (VII). After evaporation and dilution with water,
the dienone was extracted with ether and, subsequently, analyzed by HPLC on a short
8
C column using methanol–water gradient and detected at 284 nm. The method is
simple, fast (total run time 10 min), and sensitive (limit of detection 1 ng, for a signal-
to-noise ratio of 3). Fig. 3A shows a representative chromatogram of human plasma
blank that is treated with perchloric acid at 284 nm, while Fig. 3B shows a chro-
matogram of androsta-3,5-diene-7,17-dione (VII, 40 ng) in the same scale. The sen-
sitivity of the method was improved further by subjecting the samples to LC-MS in
electrospray mode on a C18 column using methanol–water gradient at pH 5 (with
acetic acid, mobile phase D). The detection was made in selected ion monitoring mode
þ
þ
at m=z 285 (M + HÞ and m=z 307 (M + Na) . The formation of sodium ion adducts in
electrospray LC-MS is well known and their use in quantitative analysis has been
documented [28]. Fig. 4A shows an LC-MS chromatogram of perchloric acid-treated
Fig. 3. (A) Chromatogram of human plasma blank (at 284 nm) after solid phase cleanup and
perchloric acid treatment in methanol at 65 °C. (B) Androsta-3,5-diene-7,17-dione (40 ng).
Chromatographic conditions: Zorbax-SB C
(
8
column (3.5 m, 4:6 ꢁ 50mm); methanol–water
1% of 0.1 M phosphoric acid) gradient (50–70% methanol in 10 min) at 1 mL/min flowrate.

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
245
Fig. 4. (A) Chromatogram of human plasma blank (LC-MS) after solid phase cleanup and
perchloric acid treatment in methanol at 65 °C. (B) Another treated plasma blank spiked with
androsta-3,5-diene-7,17-dione (2 ng). Chromatographic conditions: Zorbax-SB C18 column
(
3.5 m, 4:6 ꢁ 75mm); methanol–water (pH 4.50 with acetic acid) gradient (65–100% methanol
in 10 min at 0.5 mL/min flowrate; gas flow10 L/min at 350 °C; voltage 4500; fragmentor 90 V;
nebulizer 30 psi.

246
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
human plasma extract in SIM mode (ESI), while Fig. 4B shows a chromatogram of
perchloric acid treated human plasma extract spiked with 2 ng of 3,5-dien-7-one (VII).
This method is routinely used by us [27] to establish the structure of various
7
-oxygenated metabolites of dehydroepiandrosterone (DHEA) that are produced by
rat liver in vitro, by collecting the fractions of interest, then subjecting them to
perchloric acid treatment in methanol, and comparing the dienone derivative, thus,
formed, with authentic samples. In another study concerning hydrolysis of 7-oxo-
DHEA fatty esters by rat brain homogenate (Fig. 5), it was difficult to prove the
presence of unhydrolyzed ester solely on the basis of extracted ion chromatograms of
total-ion current (TIC) spectrum, which was quite complicated for the 7-oxo-
DHEA-3b-palmitate due to the presence of interfering peaks from the biomatrix
(brain). Therefore, the fractions corresponding to retention times of 7-oxo-DHEA
and 7-oxo-DHEA-3b-palmitate were collected and treated separately with perchloric
acid and the product was identified as androsta-3,5-dien-7,17-dione (VII). This
allowed us to prove beyond doubt the structures of desired compounds that are
present in this complex matrix system.
Fig. 5. Hydrolysis of 7-oxo-DHEA-3b-palmitate by rat brain homogenate. Each chromato-
gram in full scale. (1, blue) Total ion chromatogram of the brain homogenate extract, scale
þ
1
–3500 K. (2, red) Extracted ion chromatogram of 7-oxo-DHEA fm=z ¼ 325 ðM þ NaÞ g,
scale 1–120 K. (3, green) Extracted ion chromatogram of 7-oxo-DHEA-3b-palmitate
þ
fm=z ¼ 563ðM þ NaÞ g, scale 1–175 K. LC-MS conditions: methanol–water gradient
(
30–100% methanol in 35 min, followed by 100% for 5 min) on Zorbax-SB C
18
column (4:6 ꢁ 75mm) at 1 mL/min. N
000 V; and fragmentor 75 V.
2
flow12 L/min at 350 °C; nebulizer 40 psi; voltage
5

A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
247
4
. Conclusions
In the present study, the reaction of 7-oxygenated D - and D -steroids with per-
4
5
chloric acid has been studied in detail. The sulfate esters could be selectively and
quantitatively solvolyzed in ethyl acetate at ambient temperature to the corre-
sponding free steroids under mild conditions; whereas sulfate and carboxylic acid
esters, ethers, ketals, etc. of these steroids were quantitatively converted to dienone
derivatives in methanol at elevated temperatures. The later reaction can be utilized
for the estimation of total 7-oxo-DHEA that is present in human plasma and to
5
verify the presence of 7-oxo-D -steroids in complex biomatrices.
Acknowledgment
This work was supported by Hollis Eden Pharmaceuticals, San Diego, CA, USA.
References
[
1] H.L. Bradlow, in: S. Bernstein, S. Solomon (Eds.), Chemical and Biological Aspects of
Hydrolysis, Springer, NewYork, 1970, pp. 131–171.
[
[
[
[
2] D.L. Jones, V.H.T. James, J. Steroid Biochem. 22 (1985) 243–247.
3] R. Roy, B. Belanger, Endocrinology 133 (1993) 683–689.
4] S. Burstein, S. Lieberman, J. Biol. Chem. 233 (1958) 331–335.
5] C. Mathur, V.V.K. Prasad, S.V. Raju, M. Welch, S. Lieberman, Proc. Natl. Acad. Sci.
USA 90 (1993) 85–88.
[
[
[
[
6] H.-Ch. Curtius, M. Muller, J. Chromatogr. 39 (1967) 410–427.
7] G. Messeri, G. Cugnetto, G. Moneti, M. Serio, J. Steroid Biochem. 20 (1984) 793–796.
8] A. Marwah, P. Marwah, H. Lardy, J. Chromatogr. A 935 (2001) 279–296.
9] J. Combie, J.W. Blake, T.E. Nugent, T. Tobin, Clin. Chem. 28 (1982) 83–86.
[
[
[
[
10] V. Graef, E. Furuya, O. Nishikaze, Clin. Chem. 23 (1977) 532–535.
11] K. Schubert, K. Wehrberger, G. Hobe, Endocrinol. Exp. 5 (1971) 205–210.
12] H. Lardy, N. Kneer, M. Bellei, V. Bobyleva, Ann. NY Acad. Sci. 774 (1995) 171–179.
13] V. Bobyleva, N. Kneer, M. Bellei, D. Battelli, H.A. Lardy, J. Bioenerg. Biomembr. 25
(
14] P. Marwah, A. Marwah, N. Kneer, H.A. Lardy, Steroids 66 (2000) 581–595.
1993) 313–321.
[
[
15] D.K. Fukushima, A.D. Kemp, R. Schneider, M. Stokem, T. Gallagher, J. Biol. Chem.
210 (1954) 129–137.
[
[
16] R. Hampl, L. Starka, Endocrinol. Exp. 1 (1967) 5–13.
17] J.L. Fitzpatrick, S.L. Ripp, N.B. Smith, W.M. Pierce Jr., R.A. Prough, Arch. Biochem.
Biophys. 389 (2001) 278–287.
[
[
18] E.-E. Baulieu, J. Clin. Endocrinol. Metab. 81 (1996) 3147–3151.
19] H. Lardy, B. Partridge, N. Kneer, Y. Wei, Proc. Natl. Acad. Sci. USA 92 (1995)
6617–6619.
[
[
[
20] H. Lardy, N. Kneer, B. Partridge, Y. Wei, P. Marwah, Steroids 63 (1998) 158–165.
21] J. Shi, S. Schulze, H. Lardy, Steroids 65 (2000) 124–129.
22] L. CÕedard, B. Fillmann, R. Knuppen, B. Lisboa, H. Breuer, Z. Physiol. Chem. 338 (1964)
89–99.

248
A. Marwah et al. / Bioorganic Chemistry 30 (2002) 233–248
[
[
[
23] A.M. Bongiovanni, J. Clin. Endocrinol. Metab. 26 (1966) 1240–1244.
24] E.-E. Baulieu, J. Clin. Endocrinol. Metab. 22 (1962) 501–510.
25] H. Ludwig-Koehn, A. Sziedat, D. Matthaei, H.V. Henning, Clin. Chim. Acta 12 (1982)
189–198.
[
[
[
26] A. Marwah, P. Marwah, H. Lardy, J. Chromatogr. B 721 (1999) 197–205.
27] A. Marwah, P. Marwah, H. Lardy, J. Chromatogr. B 767 (2002) 285–299.
28] A. Marwah, P. Marwah, H. Lardy, J. Chromatogr. B 757 (2001) 333–342.