Synthesis of the Aminovinylcysteine-Containing C-Terminal Macrocycle of the Linaridins

Article abstract of DOI:10.1021/acs.orglett.0c00218

N-Phthalimido-d-cysteine allyl ester was S-alkylated with 2-iodoethanol. The derived β-thioaldehyde was condensed with Nα-tetrachlorophthalimidovalinamide to afford a Z-thioenamide. Removal of the tetrachlorophthalimido protecting group and homologation with N-Boc-l-leucine afforded the linear tripeptide. Removal of the Boc and allyl protecting groups, followed by carbodiimide-mediated cyclization, led to the 13-membered ring with the aminovinylcysteine moiety embedded. This constitutes the C-terminal macrocycle of all known members of the linardin family of peptides, including the antileukemia agent, cypemycin.

Full text of DOI:10.1021/acs.orglett.0c00218

pubs.acs.org/OrgLett
Letter
Synthesis of the Aminovinylcysteine-Containing C‑Terminal
Macrocycle of the Linaridins
Joshua A. Lutz and Carol M. Taylor*
Cite This: https://dx.doi.org/10.1021/acs.orglett.0c00218
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: N-Phthalimido-D-cysteine allyl ester was S-alkylated with 2-
iodoethanol. The derived β-thioaldehyde was condensed with Nα-
tetrachlorophthalimidovalinamide to afford a Z-thioenamide. Removal of
the tetrachlorophthalimido protecting group and homologation with N-
Boc-L-leucine afforded the linear tripeptide. Removal of the Boc and allyl
protecting groups, followed by carbodiimide-mediated cyclization, led to
the 13-membered ring with the aminovinylcysteine moiety embedded. This
constitutes the C-terminal macrocycle of all known members of the linardin family of peptides, including the antileukemia agent,
cypemycin.
he linaridins1 are a family of peptides that characteristi-
cally feature a macrocycle, containing an aminovinylcys-
T
teine residue, at their C-terminus (Figure 1), and an N,N-
dimethylated-alanine residue at the N-terminus. The proto-
typical linaridin is the antileukemia agent, cypemycin (1).2
Grisemycin (2) is three residues shorter but highly
homologous.3 Recently, the salinipeptins were reported by
Fenical and co-workers.4 Salinipeptin A (3) contains a single
Val8 → Ala8 substitution relative to cypemycin; however, nine
of the amino acids are of the D-configuration (bold/italicized
in Figure 1). Salinipeptins B−D have variations at the N-
terminus, and D contains a Dhb7 → Thr7 substitution. All
contain the common macrocycle depicted in Figure 1.
Figure 1. Selected linaridins [Dhb = E-dehydrobutyric acid; D-
residues bold/italicized].
There has been considerable interest in the biosynthesis of
cypemycin, notably the recent work of Zhang and co-workers.5
Biosynthesis of the AviCys unit in lantibiotics has been
reviewed.6,7 A chemical synthesis of the macrocycle would
provide a common advanced intermediate to the synthesis of
the entire linaridin family. Moreover, the chemistry would
ultimately be applicable to other AviCys-containing peptides,
including members of the lantibiotic family.8 With regard to
the chemical synthesis of the AviCys motif, VanNieuwenhze
and co-workers reported a decarbonylative elimination in
peptide substrates9 and applied this to the synthesis of the C-
terminal ring of the lantibiotic, mersacidin.10 Inspired by
thioviridimide,11,12 Castle and co-workers reported the
addition of thiyl radicals to ynamides13 but have yet to report
the synthesis of a true AviCys residue. We recently described
the Lewis-acid-promoted condensation of acetamide with an
acetal, or aldehyde, to produce a Z-thioenamide.14 Calculations
predicted, and experiment confirmed, that stereoelectronic
effects favored formation of the Z-thioenamide. Specifically,
aldehyde 5 could be condensed with acetamide to give
thioenamide 6 in reasonable yield and with acceptable
stereoselectivity for the Z-double bond (Scheme 1).
Throughout our studies on the mechanism and optimization
of this reaction, we yearned to apply our knowledge to a
peptidyl amide and thus the construction of the AviCys motif.
One of our early experiments involved condensation of acetal 4
with alaninamide 7. Instead of the anticipated thioenamide 9,
imidazolidinone 8 was formed, via trapping of the putative
carbocationic intermediate by the carbamate nitrogen. This key
experiment signaled that the α-amine of our amino amide
reaction partner needed to be emasculated as a nucleophile, via
double protection. After much experimentation, we arrived at
the tetrachlorophthalimido (Tcp) group to play this role, viz.,
amide 10. Unfortunately, during the execution of these
experiments, we realized that our cysteine building blocks 4
and 5 were essentially racemized, giving rise to diastereomeric
mixtures of imidazolidinone 8 and thioenamide 11.
Received: January 15, 2020
https://dx.doi.org/10.1021/acs.orglett.0c00218
© XXXX American Chemical Society
Org. Lett. XXXX, XXX, XXX−XXX
A

Organic Letters
pubs.acs.org/OrgLett
Letter
find that 2-bromo-1,1-dimethoxyethane was unreactive. Since
our penultimate goal was an aldehyde in the side chain, we
invoked 2-iodoethanol as the electrophile to afford 16 that was
readily oxidized to give aldehyde 5. 2-Iodoethanol was more
reactive than 2-bromo-1,1-dimethoxyethane, and the primary
alcohol facilitated solubility in the biphasic reaction mixture.
Alcohol 16 was determined to be 84% ee by comparison with
racemic material prepared via a different route. With
enantioenriched aldehyde 5 in hand, thioenamide formation
proceeded as expected to give 11. The olefinic protons in 11
have a relatively small coupling constant of 7.4 Hz, reminiscent
of those on a double bond in a 5-membered ring (5−7 Hz)17
and consistent with our earlier natural bond orbital (NBO)
analysis that showed a partial positive charge on the sulfur and
electrostatic interactions with the nitrogen and oxygen of the
amide.14 While the signal for the proton proximal to sulfur is
obscured by those of the allyl group, the CHNH signal is a
sharp doublet of doublets at δ 6.78 ppm, exhibiting additional
splitting by the amide proton (J = 10.8 Hz). The 13C NMR
spectrum showed signals at δ 102.8 and 134.6 ppm, for the
SCH and NHCH carbons, respectively.
Scheme 1
While thioenamides have been isolated from nature, next to
nothing is known about their tolerance to manipulations
during synthesis, so subsequent steps were undertaken with
this uncertainty, particularly under the specter of β-elimination
of the α-proton and enethiolate. For example, during
attempted Fmoc-deprotection of thioenamide 17, methylene-
bridged dimer 18 was isolated from the reaction mixture
(Scheme 3). Apparently, the enethiolate anion arising from
elimination reacts with dichloromethane (solvent) twice to
produce 18.
Careful evaluation of each step in our previous synthesis of
aldehyde 514 informed us that two steps were accompanied by
substantial racemization at Cα: phthalimido protection
(phthalic anhydride, Et3N, toluene, Δ) and alkylation of the
side chain thiol (2-bromo-1,1-dimethoxyethane, Cs2CO3,
DMF). The production of enantioenriched material required
̀
substantial experimentation vis-a-vis alternative reaction
conditions and the order of events (Scheme 2).
Scheme 3
Scheme 2
With a view to extrapolating thioenamide 11 to the 13-
membered macrocycle of the linaridins, it was necessary to
chemoselectively remove the tetrachlorophthalimido group.
Fraser-Reid and co-workers had employed the Tcp group in
the carbohydrate arena and had demonstrated that it could be
removed, in the presence of a regular phthalimido (Pht)
group,18 using ethylenediamine (pKa of conjugate acid is
9.98).19 Not surprisingly, treatment of 11 with ethylenedi-
amine led to complex mixtures.20 However, we found that
treatment with a single equivalent of hydrazine (pKa of
N-Protection of H-D-Cys(Trt)-OH (12) with methyl 2-
[(succinimidooxy)carbonyl]-benzoate (MSB) according to
Casimir et al.15 required a free carboxylic acid. We hypothesize
that the acid reacts with the N-hydroxysuccinimide ester to
give a mixed anhydride that then delivers the acylating agent
intramolecularly. Formation of the allyl ester 14 was followed
by cleavage of the trityl thioether. Recognizing that alkylation
of the thiol 15 might be achieved, with retention of
stereochemical integrity, under the mild phase-transfer
conditions of Zhu and Schmidt,16 we were disappointed to
H2NNH3 8.10)19a,21 led to a single product within 10 min.
+
This turned out to be the ring-opened acylhydrazide
intermediate 21 (Scheme 4). Gentle treatment with acid was
required to expel 5,6,7,8-tetrachloro-2,3-dihydro-1,4-phthalazi-
nedione. Following much experimentation and optimization of
these conditions, direct acylation with Boc-Leu-OH gave a
B
https://dx.doi.org/10.1021/acs.orglett.0c00218
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
pubs.acs.org/OrgLett
Letter
leading to 23 with the naturally occurring D-configuration
exclusively.
Scheme 4
Interestingly, the chemical shifts and coupling constants for
the thioenamide functional group did not change much, upon
cyclization. Calculations and observations reported by Shang et
al.4 concur with a rigid conformation of the AviCys residue, in
which one of the diastereotopic Hβ protons is coupled to Hα
and the other coupling is not observed. In the case of 23, Hβ′
(δ 3.12) is coupled to Hα (3J = 5.2 Hz), while Hβ exhibits only
geminal coupling to Hβ′ but a strong correlation to the S−
CH proton in the NOESY spectrum. All three amide
protons were slow to exchange in CD3OD, with Val NH
exhibiting cross peaks with both Leu NH and AviCys NH in
the NOESY spectrum. No correlation was observed between
Leu NH and AviCys NH. These insights into the conformation
in the region of the AviCys residue are depicted in Figure 2.
moderate yield of the tripeptide 22 over two steps. At this
point, the diastereopurity had regressed to a 75:25 mixture vis-
̀
a-vis Cα (*) of the Cys residue, as determined by integration
of well-resolved signals in the proton NMR spectrum (viz., Val
Hα and the thioenamide NH).
What remained was deprotection of the N- and C-termini
and cyclization. Both deprotections required modifications to
standard reaction conditions due to the sensitivity of the
thioenamide. Liberation of the N-terminus, with TFA and
triethylsilane as the scavenger, led to partial reduction of the
thioenamide double bond, as evidenced by the disappearance
of the characteristic thioenamide signals in the proton NMR
spectrum and the observation of an (M + 2) peak in the mass
spectrum of crude product mixtures. Substituting thioanisole
for triethylsilane circumvented this side reaction. Deallylation,
catalyzed by palladium(0), with barbituric acid as the allyl
acceptor resulted in ∼40−50% of the sulfur in the thioenamide
being converted to the sulfoxide oxidation state. Instead,
thiosalicylic acid was employed as both the allyl acceptor22 and
a sacrificial reductant. It proved advantageous to perform the
Boc deprotection ahead of the allyl ester cleavage. Fortuitously,
the linear peptidyl amino acid was water soluble, and the
reaction byproducts could be extracted into chloroform.
Lyophilization of the aqueous layer gave the semipure
precursor to the macrocycle. We anticipated that the Z-
configuration of the thioenamide would favor conformations of
the linear peptide that would be amenable to cyclization.23
Cyclization was induced by EDC/HOBt with a peptide
concentration of ∼2 mM in DMF. Following reaction
overnight, the mixture was diluted with water and lyophilized
and the product isolated by RP-HPLC in 20% overall yield.
Remarkably, macrocycle 23 was obtained as a single species
by NMR, despite the fact that the linear precursor 22 was a
75:25 mixture of D/L at Cα of Cys. It has long been recognized
that the cyclization of linear tetra-, penta-, and hexapeptides is
often sufficiently slow and that the rate of epimerization of the
C-terminal residue is competitive. Indeed, Kessler and
Kutscher reported complete epimerization of the tyrosine
residue upon cyclization of H-Arg(NO2)-Lys(Z)-Asp(OBn)-
Val-Tyr-OH with DCC/DMAP to produce cyclo[Arg(NO2)-
Lys(Z)-Asp(OBn)-Val-D-Tyr], a thymopentin analogue.24
Attempted cyclization of H-Pro-Val-Pro-Tyr-opfp, by Schmidt
and Langner, led to a 31% yield of cyclo(Pro-Val-Pro-D-Tyr).25
It seems likely, therefore, that something akin to dynamic
kinetic resolution is taking place during the cyclization of 22,
Figure 2. (A) Correlations observed in the NOESY spectrum of
compound 23 (red of special interest). (B) Newman projection,
looking down the Cα−Cβ bond of the AviCys residue.
In summary, we have constructed the aminovinylcysteine
(AviCys) motif in the center of a tripeptide. Deprotection of
the N- and C-termini and cyclization led to the C-terminal
macrocycle of cypemycin and other linaridins. The cyclization
favors the D-configuration of the AviCys residue, as is observed
in nature. Nuclear magnetic resonance experiments indicate a
well-defined, single conformation for the macrocycle. These
accomplishments set the stage for further investigation of
AviCys and peptides into which it is incorporated.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.orglett.0c00218.
Detailed experimental procedures for the synthesis of
compounds 5, 8, 10, 11, 13−18, 22, and 23. 1H and 13
C
NMR spectra for those compounds, plus 2D NMR
spectra for macrocycle 23. HPLC chromatograms for
compound 16 (ee determination) and the deallylation of
compound 22 (PDF)
AUTHOR INFORMATION
■
Corresponding Author
Carol M. Taylor − Department of Chemistry, Louisiana State
University, Baton Rouge, Louisiana 70810, United States;
orcid.org/0000-0002-6262-0796; Email: cmtaylor@
lsu.edu
C
https://dx.doi.org/10.1021/acs.orglett.0c00218
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Author
pubs.acs.org/OrgLett
Letter
(16) Zhu, X.; Schmidt, R. R. Eur. J. Org. Chem. 2003, 2003, 4069−
4072.
Joshua A. Lutz − Department of Chemistry, Louisiana State
University, Baton Rouge, Louisiana 70810, United States
(17) Field, L. D.; Sternell, S.; Kalman, J. R. Organic Structures from
Spectra, 2nd ed.; Wiley, 1995.
(18) (a) Debenham, J. S.; Madsen, R.; Roberts, C.; Fraser-Reid, B. J.
Am. Chem. Soc. 1995, 117, 3302−3303. (b) Debenham, J. S.;
Debenham, S. D.; Fraser-Reid, B. Bioorg. Med. Chem. 1996, 4, 1909−
1919. (c) Debenham, J. S.; Fraser-Reid, B. J. Org. Chem. 1996, 61,
432−433.
(19) (a) Hall, H. K. J. Am. Chem. Soc. 1957, 79, 5441−5444.
(b) Schwarzenbach, G. Helv. Chim. Acta 1933, 16, 522−528.
(20) Panza, L.; Compostella, F.; Imperio, D. Carbohydr. Res. 2019,
472, 50−57.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.orglett.0c00218
Author Contributions
All authors have given approval to the final version of the
manuscript.
Notes
(21) Schwarzenbach, G. Helv. Chim. Acta 1936, 19, 178−182.
(22) (a) Lemaire-Audoire, S.; Savignac, M.; Genet, J. P.; Bernard, J.
M. Tetrahedron Lett. 1995, 36, 1267−1270. (b) Baldwin, I. C.; Briner,
P.; Eastgate, M. D.; Fox, D. J.; Warren, S. Org. Lett. 2002, 4, 4381−
4384.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Science Foundation
[CH1464893]. J.A.L. was supported by the Louisiana Board of
Regents in the form of a Graduate Fellowship [LEQSF(2014-
19)-GF-02]. We thank the following staff in the Department of
Chemistry at LSU for their expertise: Connie David for mass
spectrometry and Thomas K. Weldeghiorghis and Fengli
Zhang for NMR.
(23) (a) White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509−523.
(b) De Leon Rodriguez, L. M.; Weidkamp, A. J.; Brimble, M. A. Org.
Biomol. Chem. 2015, 13, 6906−692.
(24) (a) Kessler, H.; Kutscher, B. Liebigs Ann. Chem. 1986, 1986,
869−892. (b) Kessler, H.; Kutscher, B. Liebigs Ann. Chem. 1986,
1986, 914−931.
(25) Schmidt, U.; Langner, J. J. Pept. Res. 1997, 49, 67−73.
REFERENCES
■
(1) Claesen, J.; Bibb, M. Proc. Natl. Acad. Sci. U. S. A. 2010, 107,
16297−16302.
(2) (a) Komiyama, K.; Otoguro, K.; Segawa, T.; Shiomi, K.; Yang,
H.; Takahashi, Y.; Hayashi, M.; Otani, T.; Omura, S. J. Antibiot. 1993,
46, 1666−1671. (b) Minami, Y.; Yoshida, K.; Azuma, R.; Urakawa, A.;
Kawauchi, T.; Otani, T. Tetrahedron Lett. 1994, 35, 8001−8004.
(3) Claesen, J.; Bibb, M. J. J. Bacteriol. 2011, 193, 2510−2516.
(4) Shang, Z.; Winter, J. M.; Kauffman, C. A.; Yang, I.; Fenical, W.
ACS Chem. Biol. 2019, 14, 415−425.
(5) (a) Mo, T.; Liu, W.-Q.; Ji, W.; Zhao, J.; Chen, T.; Ding, W.; Yu,
S.; Zhang, Q. ACS Chem. Biol. 2017, 12, 1484−1488. (b) Ding, W.;
Yuan, N.; Mandalapu, D.; Mo, T.; Dong, S.; Zhang, Q. Org. Lett.
2018, 20, 7670−7673. (c) Mo, T.; Yuan, H.; Wang, F.; Ma, S.; Wang,
J.; Li, T.; Liu, G.; Yu, S.; Tan, X.; Ding, W.; Zhang, Q. FEBS Lett.
2019, 593, 573−580.
(6) Sit, C.; Yoganathan, S.; Vederas, J. Acc. Chem. Res. 2011, 44,
261−268.
(7) Repka, L. M.; Chekan, J. R.; Nair, S. K.; van der Donk, W. A.
Chem. Rev. 2017, 117, 5457−5520.
(8) De Leon Rodriguez, L. M.; Williams, E. T.; Brimble, M. A. Chem.
- Eur. J. 2018, 24, 17869−17880.
(9) García-Reynaga, P.; Carrillo, A. K.; VanNieuwenhze, M. S. Org.
Lett. 2012, 14, 1030−1033.
(10) Carrillo, A. K.; VanNieuwenhze, M. S. Org. Lett. 2012, 14,
1034−1037.
(11) Related to the isolation of thioviridiamide, see: (a) Hayakawa,
Y.; Sasaki, K.; Nagai, K.; Shin-ya, K.; Furihata, K. J. Antibiot. 2006, 59,
6−10. (b) Kawahara, T.; Izumikawa, M.; Kozone, I.; Hashimoto, J.;
Kagaya, N.; Koiwai, H.; Komatsu, M.; Fujie, M.; Sato, N.; Ikeda, H.;
Shin-ya, K. J. Nat. Prod. 2018, 81, 264−269.
(12) For selected anticancer studies on thioviridamide, see:
(a) Frattaruolo, L.; Lacret, R.; Cappello, A. R.; Truman, A. W. ACS
Chem. Biol. 2017, 12, 2815−2822. (b) Takase, S.; Kurokawa, R.;
Kondoh, Y.; Honda, K.; Suzuki, T.; Kawahara, T.; Ikeda, H.; Dohmae,
N.; Osada, H.; Shin-ya, K.; Kushiro, T.; Yoshida, M.; Matsumoto, K.
ACS Chem. Biol. 2019, 14, 1819−1828.
(13) Banerjee, B.; Litvinov, D. N.; Kang, J.; Bettale, J. D.; Castle, S.
L. Org. Lett. 2010, 12, 2650−2652.
(14) Lutz, J. A.; Subasinghege Don, V.; Kumar, R.; Taylor, C. M.
Org. Lett. 2017, 19, 5146−5149.
(15) Casimir, J. R.; Guichard, G.; Briand, J. P. J. Org. Chem. 2002,
67, 3764−3768.
D
https://dx.doi.org/10.1021/acs.orglett.0c00218
Org. Lett. XXXX, XXX, XXX−XXX