Function-Oriented Synthesis toward Peloruside A Analogues

Article abstract of DOI:10.1021/acs.orglett.9b00413

A convergent and rapid synthesis of original C2,C3-unsaturated, C11,C13-keto-enol macrocycles with a peloruside A skeleton has been developed. These original unsaturated macrocycles constitute valuable platforms to access peloruside A analogues with high

Full text of DOI:10.1021/acs.orglett.9b00413

Letter
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Function-Oriented Synthesis toward Peloruside A Analogues
†
†
†
†
†
Anne-Caroline Chany, Fred ic Legros, Heloua Haroun, Uday Kumar Kundu, Bohdan Biletskyi,
́ ́
er
Sergii Torlak, Monique Mathe-
Brigitte Renoux, Pascal Gosselin, Gilles Dujardin,* and Catherine Gaulon-Nourry*
†
‡
‡
§
§
́
Allainmat, Jacques Lebreton, Aurelie Mace, Bertrand Carboni,
́ ́
∥
∥
†
,†
,†
†
Institut des Molec
́
ules et Mater
́
iaux, IMMM UMR 6283 CNRS − Le Mans Universite,
́
Avenue Olivier Messiaen, Le Mans 72085
Cedex 9, France
‡Universite
France
́
de Nantes, Laboratoire CEISAM, UMR 6230 CNRS, Faculte des Sciences et des Techniques, Nantes, 44322 Cedex 3,
́
§
Univ Rennes, CNRS, ISCR [(Institut des Sciences Chimiques de Rennes)], UMR 6226, F-35000 Rennes, France
∥
∥
Institut de Chimie des Milieux et des Materiaux de Poitiers, IC2MP, Universite de Poitiers, UMR 7285 CNRS, 4 rue Michel
́ ́
Brunet, 86022 Poitiers, France
*
S Supporting Information
ABSTRACT: A convergent and rapid synthesis of original C2,C3-unsaturated, C11,C13-keto−enol macrocycles with a
peloruside A skeleton has been developed. These original unsaturated macrocycles constitute valuable platforms to access
peloruside A analogues with high diversity. The four-fragment strategy implemented features two aldol-type couplings with the
central C12−C14 building block TES-diazoacetone and a late-stage ring-closing metathesis. Enantiopure analogue 18ab showed
antiproliferative activity in the low micromolar range on NCI and MCF7 tumor cell lines.
5
eloruside A (PelA) is a highly cytotoxic 16-membered
cell lines. More recently, PelA proved highly efficient in
6
P
macrolide, extracted in 2000 from the New Zealand
targeting tumoral angiogenesis. Preliminary investigations on
its potential role on neurodegenerative and autoimmune
disorders also led to encouraging results. Despite its
1
marine sponge Mycale hentscheli (Figure 1). Natural congener
7
promising pharmacological profile, preclinical trials on PelA
7
,8
stagnate due to short supply of the natural product.
Six total syntheses have been achieved so far, providing
9
enantiopure peloruside A on a milligram scale. Around 20
analogues were reported, including natural congeners,
analogues produced by hemisynthesis or resulting from total
10
synthesis. Our previous work in this field had resulted in the
convergent total synthesis of a simplified C2,C3-unsaturated
analogue, displaying antiproliferative activity around 15 μM on
Figure 1. Structures of the marine sponge macrolides peloruside A
PelA) and peloruside B.
11
(
three tumor cell lines. The available data on the structure−
activity relationship (SAR) of PelA have been recently
10a
reviewed. Key structural elements have emerged, including
the primary alcohol of the lateral chain and the embedded
tetrahydropyran skeleton, though the influence of each
peloruside B was later discovered, displaying similar cytotox-
2
icity. Like paclitaxel, PelA acts as a potent microtubule
3
stabilizing agent, stopping cell division in the G2/M phase.
4
Due to its unique nontaxoid binding-site on tubulin β, PelA
displays exciting properties like synergistic effects with taxanes
and preserved efficiency in paclitaxel and epothilone-resistant
Received: January 31, 2019
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b00413
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
oxygenated substituent on this heterocycle is still unclear.
Additionally, the impact of each oxygenated functionality along
the macrolactone skeleton has not been systematically
explored. The crystallographic structure of a PelA/tubulin
Scheme 2. Synthesis of C3−C11 Fragments (± )-4a and
11
(5R,9S)-4b
4a
complex was resolved in 2014, as well as dually bound PelA/
4b
Taxol to microtubules, providing structural information on
the binding mode. Despite this propitious context, synthetic
studies toward new PelA analogues have been scarcely
1
0b,12
reported since 2014.
The lack of SAR study is
undoubtedly due to the structural complexity of PelA,
requiring challenging synthetic work for each single analogue.
In this context, the development of a convergent synthetic
strategy, allowing the late-stage functionalization of a synthetic
platform, would be of high value. In 2012, Taylor’s team
13
reported a pioneering work in this field. Their strategy relied
on a ring-closing olefin metathesis (RCM) step in order to
elaborate Z- and E-C12,C13 unsaturated PelA skeletons in 12
steps. Analogues resulting from functionalization of these
synthetic platforms were not biologically evaluated due to
chemical instability.
In the present work, we describe a four-fragment-based
approach for the function-oriented synthesis of PelA analogues
via the elaboration of stable C2,C3-unsaturated macrocyclic
platforms (platform approach). Our convergent strategy relies
on the five-step assembly of four fragments (Scheme 1). At
underlying the involvement of the C7-methoxy group in
stabilizing lateral contact interactions with a second tubulin
unit. The synthesis of aldehydic fragment (±)-4a could be
4
readily carried out in six steps (Scheme 2, eq 1). Convergent
elaboration of the THP moiety was achieved through Prins
15
Scheme 1. Four-Fragment Diazo Strategy toward C2,C3-
Unsaturated, C11,C13 Keto−Enol Macrocyclic Platforms
cyclization, induced by excess trifluoroacetic acid, between
the aldehyde derived from alcohol 5 and commercial bis-
homoallylic alcohol 6. Methanolysis of the trifluoroacetate
intermediate provided the racemic 7-hydroxytetrahydropyran
(
±)-7 in 65% yield as a single 5,7-cis/5,9-cis diastereoisomer.
Methylation of the resulting alcohol occurred in high yield,
leading to tetrahydropyran (±)-8, on which the equatorial
1
position of all substituents was clearly deduced from H NMR
3
3
3
analysis ( J
= J
= 11.5 Hz; J
= 11.7 Hz). Aldehyde
6
ax‑5
6ax‑7
8ax‑9
(
±)-4a resulted from a classical reduction/oxidation sequence
on methyl ester (±)-8. This straightforward synthesis could be
conducted on a multigram scale. Enantiopure aldehyde
(
11
5R,9S)-4b (Scheme 2, eq 2) was also selected as a valuable
simplified C3−C11 fragment. Indeed, when the PelA
polyoxygenated THP was replaced by the 7,8,9-trideoxy
THP moiety, the resulting simplified analogue retained a
first, C15−C20 fragment 1 and C12−C14 fragment
triethylsilyldiazoacetone 2 (TES-diazoacetone) would be
coupled by “methyl-side” aldolization (step 1), followed by
esterification of cinnamic acid (C1−C2 fragment 3) with the
resulting C-15 alcohol (step 2). Assembly of C3−C11
fragment 4 would be carried out by “diazo-side” addition
1
6
hundred nanomolar range cytotoxity.
Preparation of
enantiopure aldehyde (5R,9S)-4b was performed on a
multigram scale, via intermediate (R)-9, according to the 10-
1
1
step strategy developed in the course of our previous study.
17
pro-Peloruside C15−C20 fragment (R)-1a and nor-methyl
C15−C19 fragment (R)-1b were selected to implement our
strategy (Scheme 3). A robust nine-step synthesis of pro-
(
step 3), followed by dediazotation/1,2-H migration (step 4).
A final RCM step would lead to the expected C2−C3
unsaturated macrocycles (step 5). This “diazo route” illustrates
the synthetic potential of the sequential double cross-
aldolization methodology we had previously developed on
TES-diazoacetone. Both the C11,C13 keto−enol moiety and
the C2,C3 double bond constitute an attractive gateway to
diversity on the macrolactone skeleton. Moreover, diversity
could be generated through the modulation of the substituents
17
peloruside fragment (R)-1a was reported by Taylor’s team
based on a chiral oxazolidinone-controlled alkylation to
produce intermediate (R)-10 in five steps and a (Z)-selective
Still−Gennari olefination. The above strategy was carried out,
1
4
1
8a
using an alternative Ando-type procedure
phosphonate 12,
with diaryl
18b,c
to generate the trisubstituted alkene in
high yield with a unique (Z)-geometry (Scheme 3, eq 1).
C15−C20 fragment (R)-1a could thus be readily synthesized
on a multigram scale.
nor-Methyl fragment (R)-1b (Scheme 3, eq 2) was
elaborated in a context where structural modification on the
C-18 residue has no precedent in the literature. Original
aldehyde (R)-1b, while retaining a structure very close to the
natural side chain, could advantageously be more readily
7
8
9
18
R , R , and R on fragment 4 and R on fragment 1.
In order to enrich the SAR of PelA and to access potentially
active simplified analogues, our diazo strategy was imple-
7
mented with racemic C3−C11 fragment (±)-4a (R = OMe,
8
9
R = R = H) and enantiopure C3−C11 fragment (5R,9S)-
11
7
8
9
4
b
(R = R = R = H) (Scheme 2). Original fragment
(±)-4a was selected on the basis of crystallographic data,
B
DOI: 10.1021/acs.orglett.9b00413
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
17
Scheme 3. Synthesis of C15−C20 Fragment (R)-1a and
nor-Methyl C15−C19 Fragment (R)-1b
(15S)-14 and minor (15R)-14 diastereoisomers were
attributed on the basis of Mosher’s esters study on the
diastereoisomeric mixtures (see the SI). Pleasingly, ester-
ification of cinnamic acid 3 with alcohols 14a and 14b under
Steglich conditions allowed clean separation of the resulting
20
(
15S)-15 and (15R)-15 diastereoisomers in both series. The
cinnamic moiety constituted in our hands a stable surrogate to
the acrylic moiety, whose introduction failed whatever the
conditions used (acryloyl chloride/DIPEA or acrylic acid/
DCC/DMAP). Assembly of C3−C11 pyran fragments (±)-4a
and (5R, 9S)-4b with diazoketones (15S)-15a and (15S)-15b
was then carried out via low-temperature, LDA-induced, diazo-
21
side addition (Scheme 5). This step constituted a key feature
Scheme 5. Coupling of C3−C11 Fragments 4 and Ring-
Closing of the Macrocyclic Platforms
synthesized in only six steps from (S)-Roche ester. This
commercially available precursor was first transformed into the
corresponding TIPS-protected aldehyde (S)-13 in three steps.
The latter underwent Z-selective olefination with phosphonate
1
2 to provide exclusively the expected (Z)-trisubstituted
alkene. A reduction/oxidation sequence of the ester moiety
led to the targeted aldehyde (R)-1b. This synthesis could be
achieved on a multigram scale in 57% global yield over six
steps.
Coupling of C15−C20 fragments 1 with TES-diazoacetone
14
2
(C12−C14 fragment) and cinnamic acid 3 (C1−C2
fragment) was then performed (Scheme 4).
Scheme 4. Coupling between C12−C14, C15−C20, and
C1−C2 Fragments
in the establishment of our diazo strategy, as it was known to
be particularly suitable for hindered tetrahydropyranic
14
aldehydes. Reaction time and temperature were optimized
in order to prevent simultaneous elimination of the cinnamic
ester moiety (leading to the unwilled C14,C15-unsaturated,
highly conjugated system). The mixture of diastereoisomers
resulting from this coupling was directly subjected to a
dediazotation/1,2-H migration process catalyzed by
14,22
4,
Rh (OAc)
leading to a diastereomeric mixture of keto−
enols 16aa and 16ba and to enantiopure keto−enols 16ab and
6bb. The tautomeric structure of the sole keto−enols 16
observed in CDCl was deduced from HMBC experiments
2
1
3
(
see the SI). Ring-closing metatheses of dienes 16 were best
−
3
carried out using Grubbs II catalyst in toluene (10 M) at 50
C, providing macrocyclic platforms 17aa and 17ba as a 1:1
°
23
mixture of two diasteroisomers and enantiopure macrocycles
7ab and 17bb in good yields. The C2,C3 double bond was
1
3
generated with a unique (E)-geometry ( J = 15.8 Hz). The
2−3
convergent diazo strategy was thus validated with the
elaboration of stable macrocyclic platforms 17aa−17bb,
involving 12 or 15 steps for the longest linear sequence
(from a commercially available substrate), on a 180−500 mg
scale.
At first, low-temperature LDA-induced methyl-side aldoliza-
tions were performed between aldehydes (R)-1a or (R)-1b and
TES-diazoacetone 2, prepared in three steps on a multigram
scale. Desilylation of any remaining C-protected diazoaldols
was ensured by methanolysis on the crude products, providing
aldols 14a (R = Et) and 14b (R = Me) as inseparable mixtures
of two diastereoisomers. Absolute configurations of major
1
9
Finally, the deprotection of the primary hydroxyl group on
the lateral side chain of macrocyclic platforms 17 was achieved
2
4
25
using Et N·3HF
or TBAF/AcOH,
leading to the
3
C
DOI: 10.1021/acs.orglett.9b00413
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
corresponding original peloruside A unsaturated analogues 18
ASSOCIATED CONTENT
Supporting Information
■
2
3
(
Scheme 6).
In vitro cytotoxic properties of these
*
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.9b00413.
Scheme 6. Deprotection of the Macrocyclic Platforms and
Significant in Vitro Cytotoxic Activities
Synthetic procedures and H and 13C NMR spectra for
1
all new compounds (PDF)
AUTHOR INFORMATION
Corresponding Authors
■
*
E-mail: catherine.gaulon@univ-lemans.fr.
E-mail: gilles.dujardin@univ-lemans.fr.
*
ORCID
Anne-Caroline Chany: 0000-0002-1063-6390
Jacques Lebreton: 0000-0003-2603-8982
Bertrand Carboni: 0000-0003-0529-715X
Catherine Gaulon-Nourry: 0000-0002-3601-8216
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are deeply grateful to the Cancer
■
́
opo
̂
le Grand-Ouest and
La Ligue Contre le Cancer for financial support (interregional
grants: CSIRGO 2013 - Ligues 72, 44 and 35; CSIRGO 2016
and CSIRGO 2018 - Ligue 72) and to Le Mans University for
A.-C.C.’s and U.K.K.’s postdoctoral fundings. We thank A.
Durand, C. Jacquemmoz, S. Bricaud, P. Gangnery, and E.
Mebold (IMMM, Le Mans Universite)
analyses and R. Le Guevel (ImPACcell-SFR BIOSIT, Rennes,
France) for biological tests.
́
for NMR and HRMS
́
compounds were evaluated on seven tumor cell lines: Huh7
and NCI (liver), Caco2 and HCT116 (lung), PC3 (prostate),
and MCF7 and MDA-MB-231 (breast). Enantiopure analogue
REFERENCES
■
(
(
1) West, L. M.; Northcote, P. T.; Battershill, C. N. J. Org. Chem.
2000, 65, 444−449.
2) Singh, A. J.; Xu, C.-X.; Xu, X.; West, L. M.; Wilmes, A.; Chan, A.;
Hamel, E.; Miller, J. H.; Northcote, P. T.; Ghosh, A. K. J. Org. Chem.
010, 75, 2−10.
(3) Hood, K. A.; West, L. M.; Rouwe,
1
8ab, featuring the ethyl-substituted lateral chain, was active in
a low micromolar range on two tumor cell lines (NCI: IC = 5
5
0
μM; MCF7: IC = 6 μM), while its congener with the methyl
5
0
2
chain 18bb was less active (NCI: IC = 30 μM; MCF7: IC =
50
50
́
B.; Northcote, P. T.; Berridge,
1
6 μM). Interestingly, the 7,8,9-trideoxy-THP moiety did not
M. V.; Wakefield, St. J.; Miller, J. H. Cancer Res. 2002, 62, 3356−
wipe out the cytotoxicity. Therefore, even though our
analogues proved only slightly active, this observation is in
3360.
(
4) (a) Prota, A. E.; Bargsten, K.; Northcote, P. T.; Marsh, M.;
16
Altmann, K.-H.; Miller, J. H.; Díaz, J. F.; Steinmetz, M. O. Angew.
Chem., Int. Ed. 2014, 53, 1621−1625. (b) Kellogg, E. H.; Hejab, N.
M. A.; Howes, S.; Northcote, P.; Miller, J. H.; Díaz, J. F.; Downing, K.
H.; Nogales, E. J. Mol. Biol. 2017, 429, 633−646. (c) Castro-Alvarez,
A.; Pineda, O.; Vilarrasa, J. ACS Omega 2018, 3, 1770−1782.
(5) (a) Hamel, E.; Day, B. W.; Miller, J. H.; Jung, M. K.; Northcote,
P. T.; Ghosh, A. K.; Curran, D. P.; Cushman, M.; Nicolaou, K. C.;
Paterson, I.; Sorensen, E. J. Mol. Pharmacol. 2006, 70, 1555−1564.
accordance with Altmann’s team results. Analogues 18aa and
8ba, although obtained as 1:1 diastereoisomeric mixtures,
were also tested, in order to gain a first-order approximation of
1
26
their cytotoxic properties. Only 18aa displayed a weak
27
activity (Caco2: IC50 = 15 μM).
To conclude, a function-oriented strategy toward original
PelA analogues has been developed. The convergent
elaboration of stable unsaturated macrocyclic platforms with
a PelA skeleton has been achieved, involving 12 or 15 steps in
the longest linear synthetic sequence. The four-fragment diazo
strategy implemented allows introduction of diversity by
modulating the substituents on the C3−C11 and C15−C20
fragments. The low-micromolar cytotoxicity effects of analogue
(
b) Wilmes, A.; O’Sullivan, D.; Chan, A.; Chandrahasen, C.; Paterson,
I.; Northcote, P. T.; La Flamme, A. C.; Miller, J. H. Cancer Chemother.
Pharmacol. 2011, 68, 117−126.
6) (a) Chan, A.; Singh, A. J.; Northcote, P. T.; Miller, J. H. Invest.
(
New Drugs 2015, 33, 564−574. (b) Ganguly, A.; Cabral, F.; Yang, H.;
Patel, K. D. Oncoscience 2015, 2, 585−595.
(
7) Kanakkanthara, A.; Northcote, P. T.; Miller, J. H. Nat. Prod. Rep.
2016, 33, 549−561.
8) Cao, Y.-N.; Zheng, L.-L.; Wang, D.; Liang, X.-X.; Gao, F.; Zhou,
X.-L. Eur. J. Med. Chem. 2018, 143, 806−828.
9) (a) Liao, X.; Wu, Y.; De Brabander, J. K. Angew. Chem., Int. Ed.
(
1
8ab is encouraging and seems to highlight the importance of
the ethyl residue on the lateral chain. Further functionalization
of the macrocyclic platforms is under investigation for
positions C2, C3, C11, and C13 to provide a library of PelA
analogues in order to enrich the SAR studies.
(
2
1
003, 42, 1648−1652. (b) Taylor, R. E.; Jin, M. Org. Lett. 2005, 7,
303−1305. (c) Ghosh, A. K.; Xu, X.; Kim, J.-H.; Xu, C.-X. Org. Lett.
2008, 10, 1001−1004. (d) Evans, D. A.; Welch, D. S.; Speed, A. W.
D
DOI: 10.1021/acs.orglett.9b00413
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
H.; Moniz, G. A.; Reichelt, A.; Ho, S. J. Am. Chem. Soc. 2009, 131,
3
840−3841. (e) Jacobsen, E. N.; McGowan, M. A.; Stevenson, C. P.;
Schiffler, M. A. Angew. Chem., Int. Ed. 2010, 49, 6147−6150.
f) Hoye, T. R.; Jeon, J.; Kopel, L. C.; Ryba, T. D.; Tennakoon, M. A.;
Wang, Y. Angew. Chem., Int. Ed. 2010, 49, 6151−6155.
10) (a) For a review of reported PelA analogues and SAR studies,
see: Brackovic, A.; Harvey, J. E. Chem. Commun. 2015, 51, 4750−
765. (b) Van Der Eycken, J.; Smans, G.; Cornelus, J.; Van Den
(
(
4
Bossche, D.; Jacobs, N. PCT/EP2014/075903, 2014. (c) Hong, S.
W.; Singh, A. J.; Patel, V.; Russell, E. R.; Field, J. J.; Miller, J. H.;
Northcote, P. J. Nat. Prod. 2018, 81, 2125−2128.
(
11) Zimmermann, N.; Pinard, P.; Carboni, B.; Gosselin, P.; Gaulon-
Nourry, C.; Dujardin, G.; Collet, S.; Lebreton, J.; Mathe-
M. Eur. J. Org. Chem. 2013, 2303−2315.
12) Kennington, S. C. D.; Romo, J. M.; Romea, P.; Urpi, F. Org.
́
Allainmat,
(
Lett. 2016, 18, 3018−3021.
(
(
13) Zhao, Z.; Taylor, R. E. Org. Lett. 2012, 14, 669−671.
14) Lancou, A.; Haroun, H.; Kundu, U. K.; Legros, F.;
Zimmermann, N.; Mathe-
Gaulon-Nourry, C.; Gosselin, P. Tetrahedron 2012, 68, 9652−9657.
15) Sabitha, G.; Reddy, N. M.; Prasad, M. N.; Yadav, J. S. Helv.
Chim. Acta 2009, 92, 967−976.
16) Wullschleger, C. W.; Gertsch, J.; Altmann, K.-H. Chem. Eur. J.
́
Allainmat, M.; Lebreton, J.; Dujardin, G.;
(
(
2
(
(
013, 19, 13105−13111.
17) Taylor, R. E.; Jin, M. Org. Lett. 2003, 5, 4959−4961.
18) (a) Ando, K. J. Org. Chem. 1998, 63, 8411−8416. (b) Sabitha,
G.; Gurumurthy, Ch.; Yadav, J. S. Synthesis 2014, 46, 110−118.
(
c) Touchard, F. P. Eur. J. Org. Chem. 2005, 1790−1794.
19) Abid, I.; Gosselin, P.; Mathe-Allainmat, M.; Abid, S.; Dujardin,
(
́
G.; Gaulon-Nourry, C. J. Org. Chem. 2015, 80, 9980−9988.
(
(
20) Dale, J. S.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512−519.
21) Minor diastereoisomers (15R)-15a and (15R)-15b, produced
in significant amounts, will constitute valuable C15−C20 fragments
for the future elaboration of a library of PelA analogues in order to
complete the SAR studies.
(
22) Pellicciari, R.; Fringuelli, R.; Sisani, E.; Curini, M. J. Chem. Soc.,
Perkin Trans. 1 1981, 2566−2569.
23) Diastereoisomeric mixtures of macrocycles 17aa, 17ba, 18aa,
(
and 18ba could not be efficiently separated by silica gel column
chromatography; only partial separation was achieved.
(
24) (a) Pirrung, M. C.; Shuey, S. W.; Lever, D. C.; Fallon, L. Bioorg.
Med. Chem. Lett. 1994, 4, 1345−1346. (b) Tius, M. A.; Hu, H.;
Kawakami, J. K.; Bush-Peterson, J. J. Org. Chem. 1998, 63, 5971−
5
(
976.
25) Smith, A. B., III; Chen, S. S.-Y.; Nelson, F. C.; Reichert, J. M.;
Salvatore, B. A. J. Am. Chem. Soc. 1995, 117, 12013−12014.
Degradation occurred when no acetic acid was added to TBAF.
(
26) Ongoing synthetic efforts focus on the elaboration of
enantiopure C3-C11 fragment 4 in order to access the corresponding
stereopure macrocyclic platforms and to provide accurate biological
evaluation of analogues 18aa and 18ba.
(
27) All the IC50 values and the protocol used for in vitro assays of
cell proliferation inhibition are described in the SI. The compounds
were first tested at a unique dose of 25 μM, and IC50 was then
calculated for the active ones.
E
DOI: 10.1021/acs.orglett.9b00413
Org. Lett. XXXX, XXX, XXX−XXX