Aldosterone Glucuronide, an Important Biomarker: Synthesis and Structure Elucidation of Novel Isomers

Article abstract of DOI:10.1002/chem.202004154

Aldosterone 1 is a mineralocorticoid, it has great influence on the blood pressure and its glucuronide is an important marker for the detection of several diseases. Here, we describe the chemical synthesis of different aldosterone-18- and 20-glucuronides. Reaction of trimethylsilyl 2,3,4-tri- acetyl-1-β-glucuronic acid methyl ester 5 b and aldosterone diacetate 11 in the presence of TMSOTf gave the 18-α-glucuronide 9 a. The 18-β-glucuronide 15 b and the 20-β-glucuronide 16 b could be obtained by reaction of methyl 2,3,4-tri-O-isobutyryl-1α-glucuronate trichloroacetimidate 14 and aldosterone 21-acetate 8 in the presence of TMSOTf or BF₃?OEt₂. Finally, reaction of aldosterone 21-acetate 8 and methyl 2,3,4-triacetyl-1α-glucuronate trichloroacetimidate 19 in the presence of TMSOTf gave the corresponding methyl 18-β-triacetylglucuronate 9 b, which was transformed into the desired aldosterone-18-β-glucuronide 3 by two enzyma- tic transformations.

Full text of DOI:10.1002/chem.202004154

Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202004154
&
Steroids
Aldosterone Glucuronide, an Important Biomarker: Synthesis and
Structure Elucidation of Novel Isomers
[
a]
Somraj Guha, Soundararasu Senthilkumar, Edgar Voß, and Lutz F. Tietze*
In memory of Rolf Huisgen
Abstract: Aldosterone 1 is a mineralocorticoid, it has great
influence on the blood pressure and its glucuronide is an
important marker for the detection of several diseases. Here,
we describe the chemical synthesis of different aldosterone-
O-isobutyryl-1a-glucuronate trichloroacetimidate 14 and al-
dosterone 21-acetate 8 in the presence of TMSOTf or
BF ·OEt . Finally, reaction of aldosterone 21-acetate 8 and
methyl 2,3,4-triacetyl-1a-glucuronate trichloroacetimidate 19
in the presence of TMSOTf gave the corresponding methyl
18-b-triacetylglucuronate 9b, which was transformed into
the desired aldosterone-18-b-glucuronide 3 by two enzyma-
tic transformations.
3
2
1
8- and 20-glucuronides. Reaction of trimethylsilyl 2,3,4-tri-
acetyl-1-b-glucuronic acid methyl ester 5b and aldosterone
diacetate 11 in the presence of TMSOTf gave the 18-a-glu-
curonide 9a. The 18-b-glucuronide 15b and the 20-b-glucur-
onide 16b could be obtained by reaction of methyl 2,3,4-tri-
Introduction
The main metabolic pathways of 1 are the conversion into the
dihydro- and tetrahydro derivatives followed by glucuronida-
tion in the liver and independently, by the direct glucuronida-
Aldosterone 1 (Figure 1) is the most important physiological
mineralocorticoid being formed in the cortex of adrenal glands
from cholesterol. By retention of water, tubular reabsorption of
sodium and secretion of potassium into the urine, it has a
[
3]
tion of 1 in the kidneys. The latter process is believed to
form the 18-b-glucuronide 3 (Figure 1). Although 1 exists as
equilibrium with 2 so far, no glucuronides from the C-20-hemi-
acetal have been found or discussed. Alteration in the glucuro-
nidation of aldosterone 1 results in a change of its concentra-
tion in the blood with significant consequences for the body.
For instance, inhibition of the glucuronidation by non-steroi-
[1]
major impact on the blood pressure. Plasma levels of 1 are
strictly regulated by a fine balance between its liberation con-
[
2]
trolled by renin as well as angiotensin and its metabolism.
[
4]
dal anti-inflammatory drugs (NSAID) potentiates the adverse
[
5]
cardiovascular effects of aldosterone .The measurement of
the concentration of 3 in the urine is therefore of high diag-
nostic value. However, establishing a high throughput quantifi-
cation of 3 either by LC-MS or by an immunoassay would re-
quire an analytical standard based on 3, but at the moment, a
satisfying access to 3 does not exist.
Aldosterone-18-b-d-glucuronide 3 belongs to the group of
[
6]
acetalglycosides. They are different from the usual glycosides
where an alcohol is connected to a sugar moiety by a glycosi-
dic bond, whereas in the acetalglycosides a hemi-acetal is
bound to a sugar moiety. These types of compounds are not
uncommon in nature and the largest group with such a func-
tionality are the iridoids as loganin and the secoiridoids as se-
Figure 1. Aldosterone 1 as well as 2 and aldosterone-18-b-d-glucuronide 3.
[
a] Dr. S. Guha, Dr. S. Senthilkumar, Dr. E. Voß, Prof. Dr. L. F. Tietze
Institute of Organic and Biomolecular Chemistry
Georg-August University of Gçttingen
[
7]
cologanin. The latter compound is a key intermediate in the
biosynthesis of indole, cinchona, ipecacuanha and pyrroloqui-
noline alkaloids.
Tammannstr. 2, 37077 Gçttingen (Germany)
E-mail: ltietze@gwdg.de
In general, the desired glucuronide 3 can be formed by two
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under:
[6]
different mechanisms. First the hydroxy group of the hemi-
acetal moiety in 1 can undergo a nucleophilic attack at C-1 of
a oxycarbenium ion 4 (Figure 2) formed from the correspond-
ing bromo or trichloroacetimidato glucuronate. In a second ap-
https://doi.org/10.1002/chem.202004154.
Part of a Special Issue celebrating the 1000th Issue of Chemistry—A Euro-
pean Journal.
Chem. Eur. J. 2020, 26, 15733 – 15737
15733
ꢀ 2020 Wiley-VCH GmbH

Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202004154
Figure 2. Mechanisms for the synthesis of aldosterone-18-b- and 18-a-glu-
curonides.
proach a nucleophilic attack of the oxygen of the OSiR group
3
Figure 3. Koenigs–Knorr Reaction of 7 and 8.
of a silyl glucuronate 5a–c could take place at carbocation 6
(
Figure 2) formed from the hemiacetal 1 with retention of con-
figuration at C-1 of the sugar moiety (formalistic view).
Acetalglycosides are usually more acid labile than the
common glycosides. Thus, the glucuronide 3 is hydrolyzed at
pH 1–2.5; on the other hand, it shows increased stability
against b-d-glucuronidase.
An enzymatic glucuronidation of aldosterone 1 to give 3 has
[8]
been accomplished by C. Girard et al. and K. M. Knights et
[4]
al using cell lysates; however, 3 has not been isolated due to
the very small amount formed in this process and was only de-
termined by LC-MS/MS. Moreover, its chemical synthesis has
also caused severe problems so far. The first synthetic investi-
Figure 4. Aldosterone derivatives.
[
9]
gations have been performed by Underwood and Frye using
a Koenigs-Knorr reaction of aldosterone 21-acetate 8, which
however did not give the desired derivative of compound 3.
treatment with acetic anhydride and pyridine in 90% yield. Re-
action of 5b and 11 in the presence of TMSOTf at À108C in
dry dichloromethane gave the aldosterone-18-a-glucuronide
9a in 38% yield and the known 18,21-anhydroaldosterone 12
in 5% yield; however, aldosterone-18-b-glucuronide 9b and
the also possible aldosterone-20-glucuronides 13 were not
found. It can be assumed that the formation of aldosterone-
18-a-glucuronide 9a is due to an isomerization of the b-trime-
thylsilyl sugar 5b under the reaction conditions to give the a-
trimethylsilyl sugar 5a. To avoid such an isomerization, we pre-
pared the b-t-butyldimethylsilyl sugar 5c from the correspond-
ing b-hydroxy sugar using TBDMSCl in 56% yield at 208C. Un-
fortunately, the reaction of 5c with aldosterone diacetate 11 in
the presence of TMSOTf did not lead to the desired com-
pound; the major product with 60% yield, besides a very small
amount of the a-glucuronide 9a, was 18,21-anhydroaldoster-
[10]
Later, Carpenter and Mattox again used a Koenigs-Knorr re-
action of 8 with a twentyfold excess of methyl acetobromoglu-
curonate 7 in the presence of a twentyfold excess of Ag CO3
2
via the formal carbocation 4 as a proposed intermediate. They
obtained a maximum of 10% of the desired aldosterone-18-b-
d-glucuronate 9b and in addition the 18-a-d-glucuronate 9a
as well as the 1-acetoxy-glucuronic acid methyl ester derivative
10. The authors mentioned in their publication that the repro-
ducibility of their process is low. We have repeated the work of
Carpenter and Mattox using a synthetically acceptable ap-
proach with only a twofold access of 7 and were not able to
isolate any of the desired aldosterone-18-b-d-glucuronate 9b.
Results and Discussion
[
12]
Since the Koenig–Knorr reaction of 7 and 8 gave only a low
yield of 9b (Figure 3) in a hardly reproducible way as described
one 12; here, one can assume that the system is too bulky
to give the glycosides. Finally, we treated 9a with BF3 OEt to
C
2
[
10]
by Carpenter et al. we have investigated new approaches for
the synthesis of aldosterone glucuronides. For this purpose,
we used a procedure being developed by us for the prepara-
tion of iridoid glucosides using a trimethylsilyl tetraacetyl-1b-d-
allow an isomerization to give at least a small amount of the
b-glucuronide 9b; however, 9b was not obtained but a small
amount of 10.
Since all attempts to obtain the b-glucuronide of aldoster-
one 9b using a silyl sugar had failed, we investigated the gly-
[
6]
glucoside. For the synthesis of 3 we employed the corre-
sponding trimethylsilyl 2,3,4-triacetyl-1b-d-glucuronic acid
methyl ester 5b, which is easily accessible from 2,3,4-triacetyl-
[
13]
cosidation of 8 using the trichloroacetimidate
of methyl
[
14]
2,3,4-tri-O-isobutyryl-d-glucuronate 14, since Brown et al.
[
15]
1
b-d-glucuronic acid methyl ester using HMDS and TMSCl in
and Stachulski et al.
have shown that by replacing acetyl
[11]
dichloromethane and pyridine at 08C.
with isobutyryl groups in the donor transacylations as a major
side reaction can be avoided and moreover the formation of
b-glycosides can be improved due to steric reasons. Indeed, re-
As a second reaction partner we used aldosterone diacetate
1 (Figure 4), which can be obtained from aldosterone 1 by
1
Chem. Eur. J. 2020, 26, 15733 – 15737
www.chemeurj.org
15734
ꢀ 2020 Wiley-VCH GmbH

Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202004154
action of 8 with 14 in the presence 20 mol-% of TMSOTf at
À408C in dry DCM under an argon atmosphere gave 14% of
the desired protected (18R)-aldosterone-18-b-glucuronic acid
methyl ester 15b (Figure 5). Moreover, 8% of the so far un-
known (18R, 20R)-aldosterone-20-b-glucuronic acid methyl
ester 16b was also obtained. In addition, the 1-glucuronic acid
methyl ester isobutyrate 17 was isolated in 12% yield and
2
5% of the starting material 8 could be recovered. The aldos-
terone-18-a-glucuronic acid methyl ester 15a and the aldoster-
one-20-a-glucuronic acid methyl ester 16a were not found. In-
terestingly, using TMSOTf at À108C gave a completely differ-
ent picture. Under these conditions b-glucuronides were not
obtained. The major product was the 18,21-anhydroaldoster-
one 12 with 50%; in addition, 12% of 17 and 3% of 15a were
found. Instead of TMSOTf also BF ·OEt could be used as a
Figure 6. Synthesis of the acid 18.
lowed by removal of the acetyl moieties and cleavage of the
methyl ester as well as final acid catalyzed hydrolysis of the
two semicarbazone moieties. Unfortunately, no NMR data of 3
prepared by this procedure have been reported; moreover,
NMR data of 3 do not exist in the literature at all.
3
2
Lewis acid at À108C. Here, 5.2% of 15b and 7.2% of 16b
were obtained. In addition, 10% of 17 and 50% of 8 were iso-
lated. The final steps in the preparation of 3 from 15b would
have been the hydrolysis of the methyl ester and the cleavage
of the three isobutyrate groups. Two major problems in these
transformations had to be considered. First, the elimination of
the isobutyrate group at C-4’ at the sugar and secondly the
Based on our comprehensive knowledge acquired in our
rather long enterprise of making (18R)-aldosterone-18-b-d-glu-
curonic acid 3 we finally came up with a straight forward syn-
thesis of 3, in which the final two steps namely the cleavage of
the methyl ester and the removal of the protecting groups at
the sugar and the acetate at the aldosterone moiety were ac-
complished using two enzymatic transformations. Moreover,
we also developed a reproducible glycosylation of aldosterone
21-acetate 8 using the trichloroacetimidate of methyl 2,3,4-tri-
acetyl-d-glucuronate 19 (Figure 7). Thus, reaction of 8 and 19
in the presence of 20 mol-% of TMSOTf at À408C in dry DCM
under an argon atmosphere gave the desired (18R)-aldoster-
one-18-b-d-glucuronic acid methyl ester 9b in 10% yield to-
gether with a mixture of 20-b- and 20-a-glucuronide 13b and
13a in 5% yield, which could not be separated. The rear-
ranged compound 10 was formed with 12% yield and 30% of
the starting material 8 could be recovered. Finally, 9b was
[16]
isomerization at C-17 in the aldosterone moiety.
To avoid the elimination, we first hydrolyzed the methyl
ester using an excess of NaOH in dioxane/water at room tem-
perature for 30 min. The corresponding acid 18 was obtained
in 45% yield (Figure 6), which probably underwent an isomeri-
zation at C-17. However, compound 18 was not further investi-
gated, since the hydrolysis of the isobutyrate moieties in 18
could not be accomplished without destroying the molecule.
The isomerization at C-17 was also a major dilemma in the
[10]
synthesis of 3 by Carpenter et al. These authors solved the
problem in a rather complicated procedure by transforming
the obtained glucuronide into a double semicarbazone, fol-
Figure 5. Reaction of 8 and 14 to give aldosterone-18- and 20-b-glucuro-
nides 15b and 16b.
Figure 7. Synthesis of (18R)-aldosterone-18-b-d-glucuronic acid 3.
Chem. Eur. J. 2020, 26, 15733 – 15737
www.chemeurj.org
15735
ꢀ 2020 Wiley-VCH GmbH

Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202004154
transformed into the desired aldosterone-18-b-glucuronide 3
Conclusion
[17]
by treatment with PLE esterase to give the corresponding
[18]
acid 20 followed by reaction with lipase WG to remove the
acetyl moieties avoiding any isomerization at C-17 with 24%
overall yield.
In summary, a reproducible synthesis of the important bio-
marker (18R)-aldosterone-18-b-d-glucuronic acid 3 has been
developed via reaction of aldosterone 21-acetate 8 and the tri-
chloroacetimidate of the methyl 2,3,4-tri-O-acetyl-d-glucuronic
acid methyl ester 19 in the presence of TMSOTf at À408C. The
final cleavage of the methyl ester and the hydrolysis of the
acetyl moieties was achieved using PLE and Lipase WG in two
enzymatic reactions avoiding any isomerization at C-17. The so
far unknown configuration at C-18 in 3 has been determined
by NOESY experiments showing a significant NOE between 18-
and 8-H. Moreover, novel (18R, 20R)-aldosterone-d-glucuro-
nides have been prepared for the first time.
The structure elucidation of the new compounds was per-
formed using mass spectrometry and 2D NMR spectroscopy.
The b-configuration of the glycosidic bond in the b-glucuro-
nides 3, 9b, 15b and 16b (Figure 8) was deduced from the
coupling constant of the doublet for 1’-H of 7.9 Hz at d=4.39
for 3 and of 7.9–8.3 Hz of the signals at d=4.97 to 5.36 for 9b,
15b and 16b. The doublets at d=5.29–5.33 with J=4.3 to
4
.4 Hz clearly confirm the a-configuration at the sugar moiety
1
3
in the compounds 9a and 15a. In the C NMR spectra of the
8-glucuronides two signals for the carbonyl moieties C-3 and
C-20 are seen for example, at d=198.0 and 203.7 for 15a and
1
at d=198.0 and 204.4 for 15b, whereas in the 20-glucuronides Acknowledgements
as 16b only one signal for a carbonyl group at d=198.0 is ob-
served. In the latter compound also an upfield shift for the sig-
nals of the methylene group C-21 is observed from d=4.61
and 4.77 with J=16.0 Hz in 15b to d=4.07 and 3.90 with J=
We are very grateful to the Roche Diagnostics GmbH Penzberg
and Dr. Hans-Peter Josel for their financial support and the
supply of chemicals, especially a generous gift of aldosterone.
Moreover, we would like to thank Dr. Michael John at the
Georg-August-University for his extensive help in the structure
elucidation of the prepared glucuronides.
1
1.7 Hz in 16b. In the spectra of the final compound 3 two
doublets at d=4.44 and 4.24 with J=17.8 Hz are observed for
1-H . The CO-groups resonate at d=211.4 for C-20, d=198.2
2
2
for C-3 and d=170.3 for the glucuronic acid moiety. Clearly,
the signals for the acetates and the glucuronic acid methyl
ester at d=167.2 in the precursor 9b do not exist anymore.
Most difficult was the determination of the configuration at
C-18 and C-20 in 3, 9b, 15b and 16b as well as 9a and 15a,
which was finally achieved by extensive NOESY experiments.
Although 3 is a known compound, the configuration at C-18
had not been determined so far. In the 18-glucuronides the
Conflict of interest
The authors declare no conflict of interest.
Keywords: aldosterone · enzymatic hydrolysis · glucuronides ·
Koenigs–Knorr reaction · silyl glycosides
(
18R)-configuration was proven by a strong interaction of 18-H
with 8-H, a moderate with 15b-H and a weak interaction with
[
[
[
1] J. M. C. Connell, E. Davies, J. Endocrinol. 2005, 186, 1–20.
2] M. Briet, E. Schiffrin, Nat. Rev. Nephrol. 2010, 6, 261–273.
3] T. Bledsoe, G. W. Liddle, A. Riondel, D. P. Island, D. Bloomfield, B. Sinclair-
Smith, J.Clin.Invest. 1966, 45, 264–269.
4] K. M. Knights, L. K. Winner, D. J. Elliot, K. Bowalgaha, J. O. Miners, Br. J.
Clin. Pharmacol. 2009, 68, 402–412.
5] Deleterious effects of elevated aldosterone and impaired glucuronida-
tion: a) M. Crilly, A. Mangoni, Br. Med. J. Open 2011, 1, e000076; b) R. W.
Schrier, A. Masoumi, E. Elhassan, Clin. J. Am. Nephrol. 2010, 5, 1132–
the 19-Hs of the methyl group. In the 20-glucuronides the
(
20R)-configuration was confirmed by a NOE-signal between
the 21-Hs and 17a-H, whereas these glucuronides have also
the R-configuration at C-18, showing the same NOE signals for
[
[
1
8-H as in the case of the 18-glucuronides. Finally, we have
also determined the so far unknown configuration at C-18 of
the known 18,21-anhydroaldosterone 12 as (18S) using NOESY
experiments.
1140.
[
[
[
[
6] L. F. Tietze, R. Fischer, Angew. Chem. Int. Ed. Engl. 1983, 22, 888–888;
Angew. Chem. 1983, 95, 902–903.
7] L. F. Tietze, Angew. Chem. Int. Ed. Engl. 1983, 22, 828–841; Angew. Chem.
1
983, 95, 840–853.
8] C. Girard, O. Barbier, G. Veilleux, M. El-Alfy, A. Belanger, Endocrinology
003, 144, 2659–2668.
9] R. H. Underwood, N. L. Frye, Steroids 1972, 20, 515–529.
2
[
[
[
10] P. C. Carpenter, V. R. Mattox, Biochem. J. 1976, 157, 1–14.
11] L. F. Tietze, R. Seele, Carbohydr. Res. 1986, 148, 349–352.
12] M. Harnik, Y. Kashman, M. Cojocaru, S. Lewicka, P. Vecsei, Steroids 1989,
5
4, 11–19.
13] R. R. Schmidt, K.-H. Jung, Carbohydrates in Chemistry and Biology, Vol. 1
Eds.: B. Ernst, G. W. Hart, P. Sinay), Wiley-VCH, Weinheim 2000, pp. 5–
9.
14] R. T. Brown, N. E. Carter, S. P. Mayalarp, F. Scheinmann, Tetrahedron
000, 56, 7591–7594.
15] P. M. O’Neill, F. Scheinmann, A. V. Stachulski, J. L. Maggs, B. K. Park, J.
Med. Chem. 2001, 44, 1467–1470.
16] D. N. Kirk, B. W. Muller, J. Steroid Biochem. 1982, 16, 269–276.
[
(
5
[
[
[
2
Figure 8. Stereochemical assignment of 18- and 20-glucuronides by NOE sig-
nals.
Chem. Eur. J. 2020, 26, 15733 – 15737
www.chemeurj.org
15736
ꢀ 2020 Wiley-VCH GmbH

Full Paper
Chemistry—A European Journal
doi.org/10.1002/chem.202004154
[
17] a) A. Boumendjel, M. Adeleeieblanc, G. Williamson, D. Barrron, J. Agric.
[18] A. G. Gum, T. Kappes-Roth, H. Waldmann, Chem. Eur. J. 2000, 6, 3714–
Food Chem. 2009, 57, 7264–7267; b) A. Baba, T. Yoshioka, Org. Biomol.
Chem. 2006, 4, 3303–3310.
3721.
Manuscript received: September 11, 2020
Version of record online: November 19, 2020
Chem. Eur. J. 2020, 26, 15733 – 15737
www.chemeurj.org
15737
ꢀ 2020 Wiley-VCH GmbH