Total Synthesis of ent-Pregnanolone Sulfate and Its Biological Investigation at the NMDA Receptor

Article abstract of DOI:10.1021/acs.orglett.7b03838

A unique asymmetric total synthesis of the unnatural enantiomer of pregnanolone, as well as a study of its biological activity at the NMDA receptor, is reported. The asymmetry is introduced by a highly atom-economic organocatalytic Robinson annulation. A new method for the construction of the cyclopentane D-ring consisting of Cu^I-catalyzed conjugate addition and oxygenation followed by thermal cyclization employing the persistent radical effect was developed. ent-Pregnanolone sulfate is surprisingly only 2.6-fold less active than the natural neurosteroid.

Full text of DOI:10.1021/acs.orglett.7b03838

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Total Synthesis of ent-Pregnanolone Sulfate and Its Biological
Investigation at the NMDA Receptor
Vojtech Kapras,† Vojtech Vyklicky,‡ Milos Budesinsky,† Ivana Cisarova,§ Ladislav Vyklicky,‡
Hana Chodounska,† and Ullrich Jahn*,†
†Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Flemingovo nam. 2, 166 10, Prague 6, Czech
Republic
‡Institute of Physiology of the Czech Academy of Sciences, Videnska 1083, 142 20 Prague 4, Czech Republic
§Department of Inorganic Chemistry, Faculty of Science, Charles University, Hlavova 8, 128 43 Prague 2, Czech Republic
S
* Supporting Information
ABSTRACT: A unique asymmetric total synthesis of the
unnatural enantiomer of pregnanolone, as well as a study of its
biological activity at the NMDA receptor, is reported. The
asymmetry is introduced by a highly atom-economic organo-
catalytic Robinson annulation. A new method for the
construction of the cyclopentane D-ring consisting of CuI-
catalyzed conjugate addition and oxygenation followed by
thermal cyclization employing the persistent radical effect was developed. ent-Pregnanolone sulfate is surprisingly only 2.6-fold
less active than the natural neurosteroid.
teroids are stereochemically the most conservative class of
signaling molecules and mediators in Eukaryota, serving as
S
hormones over long distances between tissues. In contrast,
neurosteroids are locally biosynthesized in nervous tissue and
directlyinfluencetheactivityofabroadvariety ofionchannels.1−3
Among these, glutamate receptors are the most prominent in the
brain, being present at ca. 80−90% of synapses.4 An especially
important subclass is the N-methyl-D-aspartate receptors
(NMDAR) modulating calcium transport into neuronal cells.
Pathological processes associated with overactivation of these
receptors contribute to a plethora of neurodegenerative diseases
and conditions.4,5 Neurosteroids, such as pregnanolone sulfate,
modulate the activity of NMDAR, and therefore, understanding
thesteroid−NMDARinteractionsiscrucialfortherationaldesign
ofneuroprotectivecompounds, whichcouldsignificantlyslowthe
progress of neurodegenerative conditions. However, little is
known about the mode andplace oftheir interaction. Onthe basis
of recently published X-ray structures of the NMDAR,6,7 single-
point mutations, as well as molecular modeling, it was suggested
that inhibitory steroids bind to the ion channel vestibule located
just above the cytoplasmic membrane. However, interactions via
the membrane cannot be excluded because of the difficult
characterization of binding at transmembrane proteins (Figure
1).8
Figure 1. Schematic depiction of the NMDA receptor in the cytoplasmic
membrane. A hypothetical binding site of inhibitory neurosteroids is
located in an extracellular vestibule. The membrane serves as a reservoir.
in 16% yield over 10 steps using a CD → BCD → ABCD ring
formation strategy.16
We opted for a new approach to ent-pregnanolone sulfate (ent-
1) in which the AB ring system would be installed first, followed
by annulation of rings C and D (Scheme 1). This provides the
additional advantage to introduce various side chains, in this case
that of ent-progesterone (ent-2). The desired side chain of ent-1 is
identical to that of ent-progesterone (ent-2). Disconnection of the
C-18 methyl group leads to 3, in which the D-ring and the oxygen
function in position C-20 are ideally positioned to develop a new
tandem process consisting of copper-catalyzed conjugate
addition to tricyclic enone 5 and a radical 5-exo cyclization with
oxygenative termination, linked by single-electron transfer (SET)
oxidationto4.17 Enone5canbetracedtoalternativetricycle6and
We hypothesized that ent-pregnanolone sulfate (ent-1) would
be an excellent tool to shed light on the action and binding mode
of inhibitory steroids on the NMDAR since it may show distinct
effects on direct binding but negligible differences by mere
membrane perturbation.9−14 The only previous total synthesis of
ent-1 was accomplished by Covey et al. by a 10-step modification
of ent-testosterone in 10% yield.15 The latter was in turn
synthesized according to an approach developed by Rychnovsky
Received: December 8, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b03838
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Retrosynthetic Analysis and Numbering of ent-1
Scheme 3. Preparation of Enone 19 and Initial Conjugate
Addition
Scheme 2. Robinson Annulation Approach to Tricycle 16
Figure 2. X-ray crystal structure of 19 and the product of oxidative
deprotection of 32.
Scheme 4. Tandem Conjugate Addition/Oxygenation
Reaction of Enone 19
subsequently to Wieland−Miescher ketone ent-7 by enantio- and
diastereoselective Robinson annulations.
The synthesis commenced with Et3N-catalyzed Michael
addition of 8 to methyl vinyl ketone in quantitative yield (Scheme
2).18 In contrast to the Hajos−Parrish ketone, proline is a
relatively inefficient catalyst for the preparation of octaline 7,19−22
but proline anilides are known to give better yield and
enantioselectivity.23,24 However, the asymmetric Robinson
annulation of 9 using the most efficient anilide 10a25 reported
so far gave suboptimal results. In contrast, the new SF5-
substituted anilide 10b proved to be superior with respect to
enantioselectivity and reactiontimes andisthus toour knowledge
thebestavailablecatalystforpreparationoftheWieland-Miescher
ketone 7 and potentially other aldol-type products.
The selective protection of conjugated enone 7 succeeded
under nonequilibrating conditions without migration of the
double bond, provided that the temperature was kept at −10 °C
or lower. These conditions are superior to Noyori’s ketalization
employing bis(trimethylsilyl) ethylene glycol, which requires a
week at −78 °C to go to completion.26 The subsequent substrate-
controlled Robinson annulation was performed in three steps via
enol 12 and aldehyde 13 affording a thermodynamic 3:1 mixture
of separable double bond isomers 14 and 15. The undesired
minor isomer 15 was recycled under basic conditions giving 14 in
a combined 80% yield. Reduction of 14 by lithium in liquid
ammonia furnished tricycle 16 with the desired trans−B-C ring
junction. Some deconjugation of the Δ9(11) double bond was
observed, but the byproduct was removed in the next step.
Ketone 16 was selectively converted to the kinetic
trimethylsilyl enolate 18 with lithium tetramethylpiperidide as a
base, and the crude 1:10 mixture of regioisomers 17/18 was
subjected to the Larock modification of the Saegusa oxidation
(Scheme3).27 Theconfigurationofenone19wasprovedbyX-ray
crystallography (Figure 2). The conjugate addition step was
tested with 3-butenyl Grignard reagent 20. Dilithium tetrachlor-
ocuprate, formed insitu from CuCl2 and LiCl, was found to bethe
optimal precatalyst giving good yields and reasonable diaster-
eoselectivity.28,29 For analytical purposes, hydrolysis of the ketal
unit to diones 21 and 22 was performed.
All attempts to perform the initially envisaged tandem copper-
catalyzed conjugate addition/radical cyclization (cf. Scheme 1)
met with limited success. These results indicated that the
cyclization to the steroid skeleton 27 could not compete with
oxygenationto25and26(Scheme4)andthereforewedecidedto
develop the tandem conjugate addition/oxygenation reaction
further and promote the projected radical cyclization sub-
sequently under thermal conditions (vide infra). Under
optimized conditions,28 tricyclic oxygenated products 25 and
26 were obtained in combined 84% yield with reasonable 4.5:1
selectivity for 8,14-trans/cis-25a/b and excellent 13:1 13,14-
trans,cis-25/26 diastereoselectivity by performing the conjugate
addition with Grignard reagent 23 and direct enolate oxygenation
using N-oxoiminium salt 24.17 Byproducts 27, 28, and 29 were
formed only to a very small extent.
Taking advantage of the known thermal lability of alkoxy-
amines, 25/26 were subjected toa radical cyclization based onthe
persistent radical effect. Such reactions, in which reversible
generation and reaction of a transient radical are regulated by the
B
DOI: 10.1021/acs.orglett.7b03838
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 5. Radical Cyclization of Keto Alkoxyamines 25a or
26a Forming the D Ring
Scheme 6. Conversion of 27 to ent-Pregnanolone Sulfate
Figure 3. (A) Concentration−response curve for the effect of nat-1
(open circles) and ent-1 (filled circles) at GluN1/GluN2B receptors (n =
5 cells, mean SD). (B) Example of trace obtained by simultaneous
application of ent-1 (150 μM) with 1 mM glutamate (duration of
application is indicated by open and filled bars, respectively).
quantitative yield. The side chain of 39 was epimerized to provide
ent-progesterone (ent-2),37 which was thus en passant
obtained.38−40
Its hydrogenation stereoselectively afforded diketone ent-40.41
The sterically more accessible ketone function at C-3 was
subsequently stereoselectively reduced by NaBH4/CeCl3 in very
good yield. Treatment of alcohol ent-41 with the pyridine−sulfur
trioxide complex led to quantitative conversion to the target
pregnanolone sulfate (ent-1), totaling 18 steps and 5.5% overall
yield for the whole synthesis. Compounds ent-1, ent-2, ent-40, and
ent-41 may serve to access other ent-neurosteroids.
The biological activity of ent-1 was determined by whole cell
voltage clamp on recombinant HEK293 cells expressing GluN1/
GluN2B receptors (Figure 3). Currents induced by 1 mM
glutamate were recorded at a holding potential of −60 mV,
applying different concentrations of steroid sulfate nat-1 or ent-
1.28 The known neurosteroid nat-1 showed an IC50 = 36.4 3.9
μM (Hill coefficient h = 1.1 0.1) in accordance with previous
measurements. In contrast, the unnatural enantiomer ent-1
showed lower but significant inhibitory activity with IC50 = 94.4
15.7 μM (h = 1.2 0.1).
This difference in affinity toward NMDAR suggests a direct
involvement of inhibitory neurosteroids with the protein, in
contrast to an expected negligible difference for mere membrane
effect. Ontheotherhand, the2.6-timeslowerinhibitoryactivityof
unnaturalent-1rulesoutaspecificbindingpocketattheprotein. It
seems, therefore, more likely that inhibitory neurosteroids act
inside the channel as blockers, held in the pore mostly by
electrostatic interactions.8 A similar phenomenon was very
recently observed for nat- or ent-batrachotoxin inside a voltage-
gated Na+-channel.42 A dominant interaction of neurosteroids
with the protein−membrane interface seems to be unlikely but
cannot be completely ruled out.8 The lack of enantiodiscrimina-
tion described here, together with previous structure−activity
studies of neuroactive steroids43 suggest that simpler analogs with
absolute ent stereochemistry may serve well as channel blockers
since they will be less prone to metabolic degradation and will not
interfere with endocrine signaling mediated by native steroids.10
In conclusion, a conceptually new synthetic approach to ent-
pregnanolone sulfate (ent-1) was developed. The use of novel
catalyst 10b in the asymmetric Robinson annulation allowed for
an exceptionally effective synthesis of the widely used Wieland−
Miescher ketone (ent-7). The tandem conjugate addition/
oxygenation sequence with a TEMPO surrogate and the
subsequent thermal radical cyclization were effective for
annulation of the D-ring. A study of the generality of this reaction
sequence is underway. A remarkably selective hydrogenation of
presence of the persistent radical are an attractive alternative to
standard cyclizations.30−32 Their power has been rarely
demonstrated in total syntheses.33,34 Heating a mixture of all
alkoxyamines 25 and 26 to 100 °C triggered a smooth cyclization
of 25a and 26a to steroid derivative 27 in quantitative yield as a
5:5:2:1 diastereomeric mixture (Scheme 5). The ratio of epimers
at C-17 was found to be 3:1 after oxidative deprotection of the
alkoxyamine unit to triketones 30 and 31. Remarkably, the
undesiredC-14epimer25bremained unchanged, thusfacilitating
the otherwise difficult separation of both trans-alkoxyamines 25a
and 25b. The formation of diastereomers of 27 during the
cyclization is inconsequential, since they converge into the target
steroid ent-1 (vide infra).
The steroid skeleton was completed by alkylation of the
thermodynamic enolate generated from 27 (Scheme 6). A
screening of bases singled out KH as the only viable option for the
synthesisof32. Theresultingdiastereomeric mixtureof32and33
was hard to separate and was therefore used directly in the next
step. The correct stereochemistry at C-13 after methylation was
confirmed by X-ray crystallography (Figure 2).
Removal of the C-12 keto group in 32 was unexpectedly
difficult, possibly because of steric hindrance caused by the
alkoxyamine unit. After much experimentation, conversion to
enol nonaflate by deprotonation with KH in the presence of a
catalytic amount of tBuOH and reaction with nonaflyl fluoride
(NfF), followed by hydrogenolysis proved to be optimal.35,36 In
contrast, LDA, LiHMDS, and KHMDS were less effective,
converting only one of the diastereomers. The four diastereomers
34/35 were isolated in 56% yield over two steps from 27. Direct
hydrogenation of 34/35 was unsuccessful, therefore the alkoxy-
amine was first oxidatively deprotected by buffered m-CPBA to
furnish two separable diastereomers 36 and 37 and a small
amount of unreacted 34 as a single diastereomer. Nonaflate 36
was selectively hydrogenated to ketal 38 in good yield. The solid-
supported Amberlite base was crucial for good conversion, as the
reaction seemed to be inhibited by coformed nonaflate salts in its
absence.28 Mild acidic deprotection of the labile ketal gave 39 in
C
DOI: 10.1021/acs.orglett.7b03838
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(7) Lee, C.-H.; Lu, W.; Michel, J. C.; Goehring, A.; Du, J.; Song, X.;
Gouaux, E. Nature 2014, 511, 191.
(8) Vyklicky, V.; Krausova, B.; Cerny, J.; Balik, A.; Zapotocky, M.;
Novotny, M.; Lichnerova, K.; Smejkalova, T.; Kaniakova, M.; Korinek,
M.; Petrovic, M.; Kacer, P.; Horak, M.; Chodounska, H.; Vyklicky, L. Sci.
Rep. 2015.
̈
an enol nonaflate serves as an excellent way to deoxygenate
sterically hindered ketones. The results of electrophysiological
testing clearly demonstrated that the purported steroid binding
pocket at the NMDAR is stereochemically relatively nonspecific
butruledoutneurosteroidactionmediatedsolelybythechangein
membrane physicochemical properties. Further synthetic effort is
directed toward truncated steroid congeners to study minimum
binding requirements for neuroactive steroid analogues.
(9) Covey, D. F. Steroids 2009, 74, 577.
(10) Biellmann, J.-F. Chem. Rev. 2003, 103, 2019.
(11)Alakoskela, J.-M.;Covey, D. F.;Kinnunen, P. K. J. Biochim. Biophys.
Acta, Biomembr. 2007, 1768, 131.
(12) Katona, B. W.; Krishnan, K.; Cai, Z. Y.; Manion, B. D.; Benz, A.;
Taylor, A.; Evers, A. S.; Zorumski, C. F.; Mennerick, S.; Covey, D. F. Eur.
J. Med. Chem. 2008, 43, 107.
(13) Covey, D. F.; Nathan, D.; Kalkbrenner, M.; Nilsson, K. R.; Hu, Y.;
Zorumski, C. F.; Evers, A. S. J. Pharmacol. Exp. Ther. 2000, 293, 1009.
(14)Li, W.;Jin, X.;Covey, D. F.;Steinbach, J. H. J. Pharmacol. Exp. Ther.
2007, 323, 236.
(15) Nilsson, K. R.; Zorumski, C. F.; Covey, D. F. J. Med. Chem. 1998,
41, 2604.
(16) Rychnovsky, S. D.; Mickus, D. E. J. Org. Chem. 1992, 57, 2732.
ASSOCIATED CONTENT
■
S
* Supporting Information
TheSupportingInformationisavailablefreeofchargeontheACS
Publications website at DOI: 10.1021/acs.orglett.7b03838.
Experimental details, characterization data of all com-
pounds, and electrophysiological experiments (PDF)
Accession Codes
CCDC 1048883 and 1048885−1048888 contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge via www.ccdc.cam.ac.uk/data_request/
cif, or by emailing data_request@ccdc.cam.ac.uk, or by
contacting The Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
(17) Kafka, F.; Holan, M.; Hidasova,
Klepetar va, B.; Jahn, U. Angew. Chem., Int. Ed. 2014, 53, 9944.
(18) Bradshaw, B.; Etxebarría-Jardi, G.; Bonjoch, J.; Viozquez, S. F.;
Guillena, G.; Najera, C. Adv. Synth. Catal. 2009, 351, 2482.
(19) Bui, T.; Barbas, C. F., III Tetrahedron Lett. 2000, 41, 6951.
́ ̌ ́
D.; Pohl, R.; Císarova, I.;
́ ̌ ́
o
́
́
(20) Canellas, S.; Ayats, C.; Henseler, A. H.; Pericas
̀
, M. A. ACS Catal.
̃
2017, 7, 1383.
(21) Gallier, F.; Martel, A.; Dujardin, G. Angew. Chem., Int. Ed. 2017, 56,
12424.
AUTHOR INFORMATION
■
(22) Guo, H.-C.; Ma, J.-A. Angew. Chem., Int. Ed. 2006, 45, 354.
(23) Bradshaw, B.; Bonjoch, J. Synlett 2012, 23, 337.
(24) Zlotin, S. G.; Kucherenko, A. S.; Beletskaya, I. P. Russ. Chem. Rev.
2009, 78, 737.
Corresponding Author
*E-mail: jahn@uochb.cas.cz.
ORCID
́
́ ́ ́
(25) de Arriba, A. L. F.; Simon, L.; Raposo, C.; Alcazar, V.; Moran, J. R.
Ullrich Jahn: 0000-0001-8386-1348
Notes
Tetrahedron 2009, 65, 4841.
(26) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21,
1357.
(27) Larock, R. C.; Hightower, T. R.; Kraus, G. A.; Hahn, P.; Zheng, D.
Tetrahedron Lett. 1995, 36, 2423.
(28) See the Supporting Information.
(29) Reetz, M. T.; Kindler, A. J. Organomet. Chem. 1995, 502, C5.
(30) Fischer, H. Chem. Rev. 2001, 101, 3581.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Generous support by the Gilead Sciences and IOCB Research
Center, IOCB CAS (RVO 61388963), the Institute of Physiology
CAS (RVO 67985823), the Technology Agency of the Czech
Republic (TE01020028), the Grant Agency of the Czech
Republic (P304/12/G069), the Ministry of Health of the
Czech Republic (NV15-29370A), and CAS (MSM200111601)
is gratefully acknowledged. I.C. thanks the Ministry of Education,
Youth and Sports of the Czech Republic (MSM0021620857) for
financial support. We thank Dr. P. Beier and Dr. G. Iakobson
(both IOCB CAS) for a generous gift of SF5-substituted anilines.
(31) Studer, A. Chem. - Eur. J. 2001, 7, 1159.
(32) Focsaneanu, K.-S.; Scaiano, J. C. Helv. Chim. Acta 2006, 89, 2473.
(33) Xu, J.; Caro-Diaz, E. J. E.; Lacoske, M. H.; Hung, C.-I.; Jamora, C.;
Theodorakis, E. A. Chem. Sci. 2012, 3, 3378.
(34) Amatov, T.; Pohl, R.; Císaro
2015, 54, 12153.
̌ ́
va, I.; Jahn, U. Angew. Chem., Int. Ed.
(35) Cacchi, S.; Morera, E.; Ortar, G. Tetrahedron Lett. 1984, 25, 4821.
(36) Cacchi, S.; Morera, E.; Ortar, G. Org. Synth. 2011, 88, 260.
(37) Butenandt, A.; Schmidt-Thome,
B 1939, 72, 1112.
́
J.; Paul, H. Ber. Dtsch. Chem. Ges.
(38) Johnson, W. S.; Gravestock, M. B.;McCarry, B. E. J. Am. Chem. Soc.
1971, 93, 4332.
REFERENCES
■
(1) Giatti, S.; Romano, S.; Pesaresi, M.; Cermenati, G.; Mitro, N.;
Caruso, D.; Tetel, M. J.; Garcia-Segura, L. M.; Melcangi, R. C. Steroids
2015, 103, 23.
(2) Korinek, M.; Kapras, V.; Vyklicky, V.; Adamusova, E.; Borovska, J.;
Vales, K.; Stuchlik, A.; Horak, M.; Chodounska, H.; Vyklicky, L. Steroids
2011, 76, 1409.
(3) Baulieu, E. E.; Robel, P.; Schumacher, M. Neurosteroids and Brain
Function.InInternationalReviewofNeurobiology;Biggio,G.,Purdy, R.H.,
Eds.; Academic Press: San Diego, CA, 2001; Vol. 46, pp 1−32.
(4) Traynelis, S. F.; Wollmuth, L. P.; McBain, C. J.; Menniti, F. S.;
Vance, K. M.; Ogden, K. K.; Hansen, K. B.; Yuan, H.; Myers, S. J.;
Dingledine, R. Pharmacol. Rev. 2010, 62, 405.
(39) Auchus, R. J.; Kumar, A. S.; Boswell, C. A.; Gupta, M. K.; Bruce, K.;
Rath, N. P.; Covey, D. F. Arch. Biochem. Biophys. 2003, 409, 134.
(40) Stork, G.; McMurry, J. E. J. Am. Chem. Soc. 1967, 89, 5464.
(41) Tsuji, N.; Suzuki, J.; Shiota, M.; Takahashi, I.; Nishimura, S. J. Org.
Chem. 1980, 45, 2729.
(42) Logan, M. M.; Toma, T.; Thomas-Tran, R.; Du Bois, J. Science
2016, 354, 865.
(43) Slavikova, B.; Chodounska, H.; Nekardova, M.; Vyklicky, V.;
Ladislav, M.;Hubalkova, P.;Krausova, B.;Vyklicky, L.;Kudova, E. J. Med.
Chem. 2016, 59, 4724.
(5) Vyklicky, V.; Korinek, M.; Smejkalova, T.; Balik, A.; Krausova, B.;
Kaniakova, M.; Lichnerova, K.; Cerny, J.; Krusek, J.; Dittert, I.; Horak,
M.; Vyklicky, L. Physiol. Res. 2014, 63 (Suppl 1), S191.
(6) Karakas, E.; Furukawa, H. Science 2014, 344, 992.
D
DOI: 10.1021/acs.orglett.7b03838
Org. Lett. XXXX, XXX, XXX−XXX