Studies toward the Total Synthesis of the Marine Macrolide Salarin C

Article abstract of DOI:10.1021/acs.orglett.8b03422

A convergent strategy for the synthesis of dideoxysalarin C (3) as a potential intermediate for the total synthesis of the marine macrolide salarin C (1) is described. The macrolactone core of 3 was assembled by Suzuki coupling between alkyl iodide 9 and

Full text of DOI:10.1021/acs.orglett.8b03422

Letter
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Studies toward the Total Synthesis of the Marine Macrolide
Salarin C
Raffael Schrof and Karl-Heinz Altmann*
Department of Chemistry and Applied Biosciences, Institute of Pharmaceutical Sciences, ETH Zu
CH-8093 Zurich, Switzerland
̈
rich, Vladimir-Prelog-Weg 4,
̈
*
S Supporting Information
ABSTRACT: A convergent strategy for the synthesis of
dideoxysalarin C (3) as a potential intermediate for the total
synthesis of the marine macrolide salarin C (1) is described. The
macrolactone core of 3 was assembled by Suzuki coupling
between alkyl iodide 9 and vinyl iodide 8 and Shiina
macrolactonization as key transformations. All macrocyclic
intermediates were found to be of low stability.
alarin C (1) (Figure 1) is a marine macrolide of mixed
and two appended side chains of different sizes at C14 and
C15. The compound incorporates an unusual bis-epoxide
motif, with one of the epoxide moieties being adjacent to a
double bond. In chloroform solution, salarin C (1) is slowly
converted into salarin A (2) (ca. 50% conversion within 3
days), and it has been suggested that this transformation
involves the reaction of 1 with singlet oxygen in a Wasserman-
S
polyketide/nonribosomal peptide synthase origin that was
isolated in 2008 by Kashman and co-workers from the sponge
Fascaplysinopsis sp., collected in waters off the coast of
1
Madagascar. It is the most potent member of a small family
of structurally related, sponge-derived nitrogenous macrolides
(
salarins A−J and tausalarin C), which exhibit different degrees
1−4
1a
of antiproliferative activity.
type rearrangement.
In spite of its intriguing structural properties and its
appealing biological activity, no total synthesis of salarin C
(
1) has been reported so far. In fact, apart from the preparation
of a series of semisynthetic derivatives of salarin C (1) and A
5
(
2), only a single publication on synthetic work in relation to
any salarin can be found in the literature at this point. Thus,
Schackermann and Lindel have recently described the
̈
synthesis of the C1−C12 segment of salarin C (1) (i.e., 4,
Figure 2), and they showed that treatment of this compound
with singlet oxygen in methanol leads to oxazole ring opening
6
and formation of the corresponding N-triacylamine.
This transformation reflects the conversion of salarin C (1)
into salarin A (2) as previously observed by Kashman and co-
1a
workers (vide supra).
In this paper, we describe the synthesis of dideoxysalarin C
(
3) as a potential precursor for the synthesis of 1. In addition,
we considered 3 as an interesting synthetic target in its own
Figure 1. Molecular structures of salarin C (1), salarin A (2), and
dideoxysalarin C (3).
Salarin C (1) has been reported to inhibit the growth of the
human leukemia cell lines UT-7 and K562 with approximate
IC50 values of 1 and 0.1 μM, respectively; the growth of the
murine pro B cell line Ba/F3 was completely inhibited by 0.1
Figure 2. Structure of the C1−C12 segment of salarin C (1) prepared
6
1b
by Schackermann and Lindel.
̈
μM salarin C (1). The compound has also been found to
4
induce apoptosis in K562 cells.
Structurally, salarin C (1) features an 18-membered
Received: October 25, 2018
macrolactone core structure with an embedded oxazole moiety
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b03422
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
right, as it has been shown for other natural products that the
replacement of an epoxide moiety by a C−C double bond of
the corresponding geometry does not significantly impair
Scheme 2. Synthesis of Oxazole Building Block 8
7
8
biological activity; examples are epothilones A/B or rhizoxin.
If so, dideoxysalarin C (3) would provide a structurally less
complex starting point for SAR studies and optimization.
Our final retrosynthesis of 3 is depicted in Scheme 1.
According to this strategy, macrocycle formation was to be
Scheme 1. Retrosynthesis of Dideoxysalarin (3)
rt gave 17 in significantly lower yield. The TBS ether in 17 was
then selectively cleaved with aq HF/CH CN, and the resulting
3
hydroxy amide was treated with DAST to provide oxazoline 18
1
4
in excellent yield (83% for the two-step sequence from 17).
15
Finally, oxidation of 18 with MnO2 cleanly furnished the
desired oxazole in excellent overall yield (26%) for the nine-
step sequence from L-ethyl lactate. Initial attempts to oxidize
14
1
8 with BrCCl3 and DBU were unsuccessful, as these
conditions led to elimination of HI and alkyne formation.
As depicted in Scheme 3, the synthesis of iodide 9
commenced with the elaboration of S-glycidol (19) into allylic
alcohol 20 by TBS protection and subsequent one-carbon
16
homologation with dimethylsulfonium methylide. Olefin 20
was then submitted to cross-metathesis with PMB-protected 3-
butenol (21) in the presence of a Hoveyda−Grubbs second-
17
generation catalyst; the reaction produced a mixture of the
desired E olefin and homocoupling products together with
unreacted starting material that was not separated but directly
reacted with SEMCl to furnish SEM ether 22 in 43% overall
yield from 20 after purification by flash chromatography.
Treatment of 22 with buffered TBAF led to selective cleavage
of the primary TBS ether, and the ensuing free alcohol was
oxidized to the corresponding aldehyde 23 with Dess−Martin
performed by macrolactonization of seco acid 7; conversion of
the cyclization product 6 into 3 would then involve
deprotection/esterification of the primary hydroxy group,
followed by deprotection of the secondary hydroxy group at
9
C14 and final carbamoylation with acetylisocyanate. Seco acid
18
7
was to be formed by Suzuki cross coupling between a
periodinane (DMP); in the best case, 23 was obtained in
97% yield (73% for the two-step sequence from 22), but yields
for this oxidation varied significantly, mostly due to
boronate derived from alkyl iodide 9 and vinyl iodide 8,
followed by elaboration of the terminal triple bond into the
α,β,γ,δ-unsaturated dienoate moiety. Alkyl iodide 9 was
envisaged to be accessed from enyne 10 and aldehyde 11 in
19
decomposition of the product aldehyde 23. Other methods
investigated for the oxidation of deprotected 22 (Parikh−
10
20
21
22
a stereoselective Carreira alkynylation reaction, followed by
reduction of the triple bond.
Doering oxidation, TEMPO, PDC ) did not offer any
advantages and uniformly gave lower yields than DMP.
In one of the key steps of the synthesis, aldehyde 23 was
then to be joined with terminal enyne 24 in a stereoselective
The synthesis of oxazole 8 is summarized in Scheme 2 and
departed from L-ethyl lactate (12), which was converted into
aldehyde 13 by TBS protection and subsequent reduction of
10
Carreira alkynylation to produce intermediate 25, which
comprises more than half of the heavy atom framework of the
macrolactone core of dideoxysalarin C (3).
11
the ester group with DIBAL; addition of lithiated TMS-
acetylene to the crude aldehyde then provided propargylic
11
alcohol 14 in 85% overall yield from 12.
Alcohol 14 was further elaborated into the corresponding
As depicted in Scheme 4, the terminal enyne 24 was
prepared in a four-step sequence from but-3-yn-1-ol (27),
which was first converted into vinyl iodide 28 by radical
1
2
amine 15 via a two-step Mitsunobu/Gabriel sequence (72%)
13
followed by HATU-mediated amide bond formation with
Z)-3-iodobut-2-enoic acid (16) at −40 °C to furnish amide
7 in 70% yield. Notably, carrying out the coupling reaction at
hydrostannylation with Bu SnH/AIBN and subsequent tin−
3
(
1
iodine exchange, according to the methodology developed by
23
de Meijere and co-workers.
B
DOI: 10.1021/acs.orglett.8b03422
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 3. Synthesis of Iodide 9
group. The conversion of diol 26 into the primary tosylate
followed by S 2 displacement with iodide and protection of
N
the secondary hydroxy group as a TES ether then gave the
desired iodide building block 9 in excellent overall yield (87%
for the three-step sequence from free trienediol 26). The total
yield for the synthesis of 9 from S-glycidol (11 steps) was 8%.
Initial experiments on the Suzuki coupling between vinyl
iodide 8 and the ate complex derived from iodide 9 via
iodine−lithium exchange and reaction with 9-methoxy-BBN at
−
78 °C (Scheme 5) were conducted with Cs CO as the base
2 3
Scheme 5. Building Block Assembly and Endgame for the
Synthesis of 3
Scheme 4. Synthesis of Enyne 24
The reaction yielded a mixture of double bond isomers that
was isomerized to the pure E isomer by treatment with sodium
in refluxing methanol. Vinyl iodide 28 was then submitted to
24
Sonogashira coupling with TMS-acetylene to produce the
free enynol 29 in 74% yield; subsequent TBS protection
followed by cleavage of the TMS group from the terminal
triple bond with K CO in methanol then gave the desired
2
3
enyne 24.
10
The Carreira alkynylation of aldehyde 23 with enyne 24
gave the desired propargylic alcohol 25 in moderate isolated
yields of up to 50% but with excellent stereochemical purity
(
dr > 95:5) (Scheme 3). The reaction proved to be scale-
dependent, and yields declined significantly below 50% on
scales larger than 1.5 g of 23. Gratifyingly, the subsequent
PMB cleavage and syn-selective reduction of the triple bond
with a mixture of Zn(Cu/Ag), water, and methanol in the
presence of TMSCl proceeded smoothly to furnish the free
trienediol 26 in 64% overall yield (from 25). Care had to be
taken in the reduction step for the TMSCl to be of high quality
and free of traces of HCl in order to avoid loss of the TBS-
and gave the desired coupling product 30 in 66% yield. While,
in principle, this outcome appeared acceptable, further
investigation of the reaction conditions revealed that the use
25
26
of Tl CO instead of Cs CO led to improved yields of 85−
2
3
2
3
99%. Interestingly, TlOEt, which has been successfully
employed as a base for Suzuki cross-couplings in other
C
DOI: 10.1021/acs.orglett.8b03422
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
2
7
cases, delivered 30 only in 25% yield (still impure), although
TLC had indicated rapid, complete conversion of starting
material. Different methods were then investigated to induce
the selective cleavage of the secondary TES ether without loss
of the TBS group, including PPTS/MeOH, AcOH/THF/
and 9 by Suzuki coupling and macrocyclic ring closure by
Shiina macrolactonization provided efficient access to the
macrolactone core structure of 3. All subsequent steps were
hampered by the instability of intermediates, which also
prevented their rigorous purification by conventional silica gel
chromatography. Independent of the question if 3 could, in
fact, be converted into the natural product salarin C (1) in a
selective fashion, our data may indicate that any practical
approach toward the total synthesis of 1 should minimize the
number of steps that have to be carried out after macrocycle
formation.
31
H O, and HF·pyridine. Of these, treatment of 30 with PPTS
proved to be the most practical, since the reaction proceeded
at a reasonable rate and with good selectivity. While AcOH/
2
THF/H O was also selective, the reaction was slower and did
2
not reach full conversion; HF·pyridine led to partial
concomitant cleavage of the primary TBS ether. As the TMS
group on the triple bond was unaffected by PPTS/MeOH, this
group was removed prior to TES ether cleavage by treatment
of 30 with K CO in methanol. Subsequent TES ether cleavage
then provided the free alcohol 31 in 72% overall yield (from
0). The completion of the carbon framework of the salarin
macrocycle involved two-carbon homologation of 31 into ester
3 by stannylcupration of the terminal triple bond and
subsequent Stille coupling with (Z)-methyl-2-iodoacrylate
32). Ester 33 was obtained in 68% overall yield from 31
with an E/Z ratio about the C4−C5 double bond of ca. 18/1.
Ester saponification with TMSOK in Et O provided the
ASSOCIATED CONTENT
Supporting Information
■
*
S
2
3
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b03422.
3
28
3
1
Synthetic procedures, complete analytical data, and H
1
3
and C NMR spectra for all new compounds.
Description of the stability experiments with 35 and 3
and associated HPLC traces (PDF)
(
29
2
corresponding carboxylic acid (7), which proved to be unstable
AUTHOR INFORMATION
Corresponding Author
on silica gel and, therefore, was directly submitted to
■
3
0
macrocyclication under Shiina conditions. The macro-
cyclization gave a 65% yield of a ca. 9/1 mixture of the
desired macrolactone 34 and a major impurity, which we
assume to be the C2−C3 E isomer of 34. Unfortunately,
rigorous purification of this mixture was not possible due to
partial decomposition during purification by flash chromatog-
raphy. Stability problems were also encountered for all
products downstream of 34; as a consequence, these materials
were all obtained as isomeric mixtures, and these are what the
yields reported in Scheme 5 refer to. Partial deprotection of 34
with buffered TBAF followed by EDCI/DMAP-mediated
esterification of the resulting primary alcohol with caprylic
acid and subsequent removal of the SEM group with BCl3·
SMe2 then furnished 35 as the precursor for the final
carbamoylation reaction. Macrolactone 35 could be obtained
in pure form after purification by RP-HPLC in 14% yield for
the three-step sequence from 34. Finally, reaction of 35 (non-
*
E-mail: karl-heinz.altmann@pharma.ethz.ch.
ORCID
Karl-Heinz Altmann: 0000-0002-0747-9734
Present Address
(
R.S.) School of Chemistry, University of Bristol, Cantock’s
Close, Bristol BS8 1TS, UK.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the ETH Zu
KHA). The project was also conducted within the framework
of COST Action CM1407. We are indebted to Bernhard
Pfeiffer, Leo Betschart, and Philipp Waser for NMR support, to
Kurt Hauenstein (all ETHZ) for practical advice, and to the
ETHZ-LOC MS-Service for HRMS spectra.
■
̈
rich (base funding to
9
HPLC purified) with in situ generated acetylisocyanate
installed the acetylcarbamate side chain in high yield. Pure 3
was obtained by RP-HPLC; however, due to the stability
problems during the purification, only 0.2 mg of 3 was
ultimately obtained from 1.6 mg of the original mixture of
isomers.
REFERENCES
■
(
1) (a) Bishara, A.; Rudi, A.; Aknin, M.; Neumann, D.; Ben-Califa,
N.; Kashman, Y. Tetrahedron Lett. 2008, 49, 4355−4358. For correct
biological data, see: (b) Bishara, A.; Rudi, A.; Aknin, M.; Neumann,
D.; Ben-Califa, N.; Kashman, Y. Tetrahedron Lett. 2008, 49, 5437.
2) Bishara, A.; Rudi, A.; Aknin, M.; Neumann, D.; Ben-Califa, N.;
Kashman, Y. Org. Lett. 2008, 10, 153−156.
3) Bishara, A.; Rudi, A.; Aknin, M.; Neumann, D.; Ben-Califa, N.;
Using the HPLC-purified macrolactones 35 and 3, we have
attempted to assess the stability of these compounds in pH7
phosphate buffer solution by means of RP-HPLC (see the
Supporting Information). While the interpretation of these
experiments is not entirely unambiguous, due to the formation
of precipitates upon dilution of the respective DMSO stock
solutions with phosphate buffer (100-fold dilution; nominal
final concentration of 9 and 8 μM for 3 and 35, respectively),
the data strongly suggest that both compounds are unstable
and at rt disappear with a half-life of less than 24 h. If such
stability issues also exist for the natural product salarin C (1)
how this would affect the interpretation of the biological data
obtained for the compound remains to be determined.
In summary, we have elaborated a convergent strategy for
the synthesis of dideoxysalarin C (3) that is based on alkyl
iodide 9 and vinyl iodide 8 as key building blocks. Linking 8
(
(
Kashman, Y. Tetrahedron 2010, 66, 4339−4345.
(4) Ben-Califa, N.; Bishara, A.; Kashman, Y.; Neumann, D. Invest.
New Drugs 2012, 30, 98−104.
(
5) Zur, L. G.; Bishara, A.; Aknin, M.; Neumann, D.; Ben-Califa, N.;
Kashman, Y. Mar. Drugs 2013, 11, 4487−4509.
(
(
6) Schac
7) (a) Su, D.; Meng, D.; Bertinato, P.; Balog, A.; Sorensen, E. J.;
̈
kermann, J.-N.; Lindel, T. Org. Lett. 2017, 19, 2306−2309.
Danishefsky, S. J.; Zheng, Y.-H.; Chou, T.-C.; He, L.; Horwitz, S. B.
Angew. Chem., Int. Ed. Engl. 1997, 36, 757−759. (b) Nicolaou, K. C.;
Vourloumis, D.; Li, T. H.; Pastor, J.; Winssinger, N.; He, Y.; Ninkovic,
S.; Sarabia, F.; Vallberg, H.; Roschangar, F.; King, N. P.; Finlay, M. R.
V.; Giannakakou, P.; Verdier-Pinard, P.; Hamel, E. Angew. Chem., Int.
Ed. Engl. 1997, 36, 2097−2103.
D
DOI: 10.1021/acs.orglett.8b03422
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(
8) Takahashi, M.; Iwasaki, S.; Kobayashi, H.; Okuda, S.; Murai, T.;
Sato, Y.; Haraguchi-Hiraoka, T.; Nagano, H. J. Antibiot. 1987, 40, 66−
2.
9) Song, D.-Q.; Du, N.-N.; Wang, Y.-M.; He, W.-Y.; Jiang, E.-Z.;
7
(
Cheng, S.-X.; Wang, Y.-X.; Li, Y.-H.; Wang, Y.-P.; Li, X.; Jiang, J. D.
Bioorg. Med. Chem. 2009, 17, 3873−3878.
(10) (a) Frantz, D. E.; Fassler, R.; Carreira, E. M. J. Am. Chem. Soc.
̈
2
000, 122, 1806−1807. (b) Boyall, D.; Frantz, D. E.; Carreira, E. M.
Org. Lett. 2002, 4, 2605−2606.
11) (a) Song, L.; Yao, H.; Zhu, L.; Tong, R. Org. Lett. 2013, 15, 6−
(
9
́
. (b) de Leseleuc, M.; Godin, E.; Parisien-Collette, S.; Levesque, A.;
́ ́ ́
Collins, S. K. J. Org. Chem. 2016, 81, 6750−6756.
(
(
(
12) Ma, S.; Wu, B.; Jiang, X. J. Org. Chem. 2005, 70, 2588−2593.
13) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397−4398.
14) Phillips, A. J.; Uto, Y.; Wipf, P.; Reno, M. J.; Williams, D. R.
Org. Lett. 2000, 2, 1165−1168.
15) Aoyama, T.; Sonoda, N.; Yamauchi, M.; Toriyama, K.; Anzai,
M.; Ando, A.; Shioiri, T. Synlett 1998, 1998, 35−36.
16) Alcaraz, L.; Harnett, J. J.; Mioskowski, C.; Martel, J. P.; Le Gall,
T.; Shin, D.-S.; Falck, J. R. Tetrahedron Lett. 1994, 35, 5449−5452.
17) Chatterjee, A. K.; Choi, T. L.; Sanders, D. P.; Grubbs, R. H. J.
(
(
(
Am. Chem. Soc. 2003, 125, 11360−11370.
(
(
18) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155−4156.
19) Parameters such as the time of exposure to base upon workup,
reaction time, and the contact time with silica gel during purification
were carefully optimized for the reaction. The optimized procedure
included a fixed reaction time of 30 min, filtration of the reaction
mixture through Celite, and rapid purification on a short flash column
(after solvent removal). Aldehyde 23 was then immediately used in
the next step. On the basis of NMR analysis of decomposed batches
of 23, the double bond appeared to have migrated into conjugation
with the carbonyl group, and this may have been associated with E/Z
isomerization. However, the exact nature of the decomposition
products has not been investigated. We have also prepared the TBS
(
instead of SEM)-protected variant of 22, but oxidation of this alcohol
delivered the corresponding aldehyde only in low yields, presumably
due to TBS migration to the primary hydroxy group before oxidation.
(
5
(
20) Parikh, J. R.; Doering, W. V. E. J. Am. Chem. Soc. 1967, 89,
505−5507.
21) de Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. Synthesis
1
(
(
996, 1996, 1153−1176.
22) Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 20, 399−402.
23) Pilli, R. A.; De Andrade, C. K. Z.; Souto, C. R. O.; de Meijere,
A. J. Org. Chem. 1998, 63, 7811−7819.
24) Trend, R. M.; Ramtohul, Y. K.; Stoltz, B. M. J. Am. Chem. Soc.
005, 127, 17778−17788.
25) (a) Mohamed, Y. M. A.; Hansen, T. V. Tetrahedron 2013, 69,
872−3877. (b) Boland, W.; Pantke, S. J. J. Prakt. Chem./Chem.-Ztg.
994, 336, 714−715.
26) (a) Uenishi, J.; Beau, J. M.; Armstrong, R. W.; Kishi, Y. J. Am.
(
2
(
3
1
(
Chem. Soc. 1987, 109, 4756−4758. (b) Sato, M.; Miyaura, N.; Suzuki,
A. Chem. Lett. 1989, 18, 1405−1408.
(27) Chemler, S. R.; Trauner, D.; Danishefsky, S. J. Angew. Chem.,
Int. Ed. 2001, 40, 4544−4568.
28) Betzer, J.-F.; Delaloge, F.; Muller, B.; Pancrazi, A.; Prunet, J. J.
Org. Chem. 1997, 62, 7768−7780.
29) Laganis, E. D.; Chenard, B. L. Tetrahedron Lett. 1984, 25,
831−5834.
30) Shiina, I.; Ibuka, R.; Kubota, M. Chem. Lett. 2002, 31, 286−
87.
31) Interestingly, Schac
(
(
5
(
2
(
̈
kermann and Lindel noted that their salarin
C fragment 4 was stable in chloroform solution in the presence of air
and light, in contrast to salarin C (1) itself (ref 6). They conclude that
this “makes it likely that additional structural features of salarin C (1)
contribute to its instability”. This is in line with our observation that
the pronounced instability of intermediates only surfaced after
macrolactone formation.
E
DOI: 10.1021/acs.orglett.8b03422
Org. Lett. XXXX, XXX, XXX−XXX