The amido-pentadienoate-functionality of the rakicidins is a thiol reactive electrophile - Development of a general synthetic strategy

Article abstract of DOI:10.1039/c5cc04500b

We demonstrate that a unique class-defining functionality (mc-APD) found in macrocyclic natural products with potent anti-cancer activity, imparts these compounds with electrophilic reactivity. The mc-APD group represents an interesting structural hybrid between canonical biologically relevant Michael-acceptors. Further, a novel thiol-elimination method for preparation of the mc-APD group is outlined.

Full text of DOI:10.1039/c5cc04500b

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
The amido-pentadienoate-functionality of the
rakicidins is a thiol reactive electrophile –
development of a general synthetic strategy†
Cite this: Chem. Commun., 2015,
1, 12427
5
Received 1st June 2015,
Accepted 30th June 2015
a
a
a
b
Lise L. Clement, Michail Tsakos, Eva S. Schaffert, Carsten Scavenius,
bc
a
Jan J. Enghild and Thomas B. Poulsen*
DOI: 10.1039/c5cc04500b
www.rsc.org/chemcomm
We demonstrate that a unique class-defining functionality (mc-APD)
found in macrocyclic natural products with potent anti-cancer activity,
imparts these compounds with electrophilic reactivity. The mc-APD
group represents an interesting structural hybrid between canonical
biologically relevant Michael-acceptors. Further,
a novel thiol-
elimination method for preparation of the mc-APD group is outlined.
1
Electrophilic organic compounds are a source of both fascination
2
and concern within molecular pharmacology. Traditionally, small
molecule electrophiles have been viewed as problematic due to their
(
increased potential for) off-target effects compared to unreactive
3
compounds. For some classes of electrophiles such concern is
certainly legitimate, however, many examples demonstrate that
4
highly specific interactions between proteins and electrophilic
ligands are also possible, thus forming the basis for a number of
1,5
useful chemical probes. Indeed several approved drugs exploit
covalent mechanisms underscoring that reactive compounds can be
3
safe. Covalent modulators often show increased target residence
3
time and high ligand efficiency.
Michael acceptors constitute a major class of electrophiles
and although there is significant variation within the carbon
frameworks and electron withdrawing groups, privileged struc- Fig. 1 (A) Privileged Michael-acceptor functionalities (blue) found within natural
6
tures have emerged, as illustrated in Fig. 1A. Microcystin-LR and products. (B) Chemical structures of the natural compound rakicidin A, and
7
its synthetic surrogate 1 used in this study to test the electrophilic nature of
the mc-APD functionality (red).
syringolin A are classical examples of reactive natural products
containing a dehydroalanine and an endocyclic a,b-unsaturated
amide, respectively.
9
Rakicidin A (RakA, Fig. 1B) is a depsipeptide natural product handful of natural products has been found to possess. The
with selective toxicity towards hypoxic and stem-like cancer APD-group appears to be stable only within macrocyclic settings
8
cells. A defining characteristic of this compound is the presence (mc-APD) and even under these conditions its preparative hand-
1
0
of an unusual vinylogous dehydroalanine functionality, 4-amido- ling is non-trivial.
2
,4-pentadienoate (APD) which, aside from the rakicidins, only a
Although the mechanistic foundation for the biological
activity of RakA is still obscure, the mc-APD has recently been
demonstrated to be a critical functionality. For this reason,
and due to its unusual chemical structure, it is of high interest to
clarify the eventual reactive potential of this group. The mc-APD is
expected to act as a Michael acceptor at both the b- and d-carbons
a
11
Department of Chemistry, Aarhus University, Langelandsgade 140, 8000 Aarhus C,
Denmark. E-mail: thpou@chem.au.dk
b
Department of Molecular Biology and Genetics, Aarhus University,
Gustav Wieds Vej 10, 8000 Aarhus C, Denmark
c
Interdisciplinary Nanoscience Center – iNANO, Aarhus University, Denmark
Electronic supplementary information (ESI) available: Experimental proce-
(Fig. 1B), albeit addition at the a- or g-position (due to acyl
†
dures, spectroscopic data and NMR spectra. See DOI: 10.1039/c5cc04500b
imine tautomerism) cannot be ruled out. From the synthetic
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 12427--12430 | 12427

View Article Online
Communication
ChemComm
1
4
point of view, we have previously demonstrated that a serinol- phosphonate 4 and protected racemic cysteinal 5. Following
derived amido-enal precursor could undergo Wittig-type trans- some optimization, we found that the desired HWE-reaction
1
0
formations to prepare the APD-group in non-cyclic form. This could be carried out at low temperature to yield a 1 : 1 cis/trans
highly unstable functionality could only be handled reproducibly mixture, and that the undesired cis-isomer 6a could be easily
under dilute, neutral/slightly basic conditions but was stabilized equilibrated in the presence of Bu P. Ester 13 was made in 5
3
upon subsequent macrocyclization. In their impressive recent steps from alcohol 8 and following removal of the t-Bu group,
synthesis of rakicidin A, Chen and co-workers showed that the acid was coupled to amine 7 using EDCI/HOAt in 63% yield
dehydration – through the intermediate mesylate – could be over 2 steps. Proceeding towards macrocyclization, the allyl and
1
2
postponed to the last synthetic step, thereby mimicking the Fmoc-groups were then sequentially removed and gratifyingly we
1
3
presumed final biosynthetic operation. In our hands, however, found that the macrolactamization could be effected by DEPBT to
related transformations have turned out to be highly substrate- afford 18a in a yield of 46% over the three steps. Several canonical
dependent and rather unpredictable.
coupling-systems proved ineffective or failed completely when tried
Thus, we hypothesized that a cysteine-residue might confer in this transformation. Having achieved the desired elimination
enhanced chemoselectivity as a mc-APD precursor, and there- precursor, the stage was set to test the thiol-elimination conditions
0
fore we developed a convergent synthesis of an appropriately using Ellman’s disulfide reagent (5,5 -dithiobis-(2-nitrobenzoic
protected macrocyclic precursor (Scheme 1). Boc-protected acid) or DTNB), originally developed for dehydroalanine-syntheses
15
amine 6b was constructed through a HWE-reaction between (Table 1). The trityl-group could be readily cleaved under standard
Scheme 1 Synthesis of macrocycle 18a. Reagents and conditions. (a) BrCH
2
C(O)Br (1.5 equiv.), NEt
(1.1 equiv.), 85 1C, 24 h; (d) n-BuLi (4.2 equiv.), THF, ꢀ78 1C, 60 min, then 5
P (1.5 equiv.), toluene, rt, 19 h; (f) 15% TFA/CH Cl , rt, 45 min; (g) TBDPSCl (1.35 equiv.), imidazole (3.1 equiv.), CH Cl , rt,
O (2.5 equiv.), THF/H O, rt, 57 h; (i) Boc O (2.0 equiv.), DMAP (0.3 equiv.), t-BuOH, rt, 90 min; (j) TBAF (2.0 equiv.), THF, rt, 4 h; (k) Fmoc-
Gly-OH (2.5 equiv.), HATU (3.0 equiv.), HOAt (3.0 equiv.), NEt(i-Pr) (5.0 equiv.), DMF, 0 1C, 90 min; (l) 15% TFA/CH Cl , rt, 16 h; (m) EDCI (1.8 equiv.), HOAt
Cl /DMF (4/1), 0 1C to rt, 19 h; (n) Pd(PPh (0.4 equiv.), NMA (2.5 equiv.), CH Cl , 3.5 h;
(8.0 equiv.), CH Cl /DMF (4/1), 0 1C to rt, 48 h. 1,2-DCE = 1,2-dichloroethane, THF =
3
(3.0 equiv.), CH
2
Cl
2
, ꢀ78 1C to ꢀ5 1C, 3 h;
(
b) P(OEt)
1.0 equiv.), ꢀ78 1C, 18 h; (e) Bu
0 h; (h) LiOHꢁH
3 2
(1.5 equiv.), 1,2-DCE, reflux, 22 h; (c) allyl alcohol, OSnBu
(
3
2
2
2
2
2
2
2
2
2
2
2
(
1.8 equiv.), 14 (1.5 equiv.), 7 (1.0 equiv.), NEt(i-Pr)
o) NHEt /MeCN (1/1), rt, 80 min; (p) DEPBT (6.0 equiv.), NEt(i-Pr)
2
(4.0 equiv.), CH
2
2
)
3 4
2
2
(
2
2
2
2
tetrahydrofuran, TFA = trifluoroacetic acid, DMAP = 4-dimethylaminopyridine, TBDPSCl = tert-butyldiphenylchlorosilane, TBAF = tetrabutylammonium
fluoride, HOAt = 1-hydroxy-7-azabenzotriazole, DMF = N,N-dimethylformamide, EDCI = 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide, NMA =
N-methylaniline, BOP-Cl = Bis(2-oxo-3-oxazolidinyl)phosphonic chloride, TFFH = tetramethylfluoroformamidinium hexafluorophosphate.
1
2428 | Chem. Commun., 2015, 51, 12427--12430
This journal is ©The Royal Society of Chemistry 2015

View Article Online
ChemComm
Communication
Table 1 Thiol-elimination to form the mc-APD system
Scheme 2 Successful thiol elimination via the unsymmetrical disulfide
intermediate 19.
a
b
Entry
Conditions
Yield (%)
1
2
3
4
5
DTNB, NEt
DTNB, NEt
3
, DMF then HMPT
, CH Cl /DMF then HMPT
0
0
different conditions and subsequently alkylated with iodoacet-
amide. Following tryptic proteolysis, cysteine-residues modified with
1 were identified by mass spectrometry. Although BSA contains a
single free cysteine residue, no modified peptides were detected
following a 2 hour exposure of the native protein to 100 mM of 1. We
therefore denatured and reduced the protein to expose all disulfide-
bound cysteines before incubation with 1 which resulted in
the detection of several modified peptides (Fig. 2A and B and
Table S1, ESI†). In this experiment all modified peptides were
identified based on their mass and sequences obtained by ms/ms
3
2
2
DTNB, DMF then DBU
DTNB, CH Cl /DMF then DBU
Npys-Cl, CH Cl then HMPT
11
12
2
2
c
2
2
0
a
All reactions were carried out in a one pot operation at ambient
b
c
temperature. Isolated yield. Only 18b obtained. HMPT = hexa-
methylphosphorous triamide, DBU = 1,8-diazabicyclo[5.4.0]undec-7-
ene, TES = triethylsilane.
conditions and the free thiol 18b was purified by a simple
precipitation. Several attempts were made with and without ancillary
base to first generate the initial mixed disulfide, which by TLC
analysis appeared to proceed smoothly. However, subsequent
attempts at inducing reduction-elimination under neutral condi-
tions with HMPT (Table 1, entries 1 and 2) were unsuccessful, and
base-mediated elimination employing DBU only resulted in small
quantities of the product (Table 1, entries 3 and 4).
It should be noted that we intentionally prepared the macrocycle
as a mixture of C9-epimers [the additional stereogenic center at C14
was arbitrarily chosen as (R)] in order to accommodate the possibility
that the efficiency of thiol-elimination might differ for the two
diastereomers. Following these surprising shortcomings with DTNB,
we turned our attention to alternative reagents and indeed found
that efficient disulfide formation could also be achieved using
(Fig. S1 and Table S1, ESI†).
Of the 35 cysteine residues in BSA, 19 were identified carrying the
macrocyclic modification (Table S1, ESI†). We also studied the
reactivity of 1 towards thiol nucleophiles under preparative condi-
tions with methyl thioglycolate as a model nucleophile (Scheme 3).
6
The reaction was carried out in DMSO-d and was monitored
1
periodically by H NMR spectroscopy. Curiously, after 24 hours we
observed predominant formation of a compound that we tentatively
assign to be 21 (see ESI†) as well as the 1,6-addition product 20
as a minor component. Compound 21 likely results from thiol-
addition to the g-carbon atom of the mc-APD, presumably
3-nitro-2-pyridinesulfenyl chloride (Npys-Cl, Table 1, entry 5). This
reagent was initially developed as a protecting group for alcohols
16
17
and amines, being later on extended to thiols, but to the best of
our knowledge has never been employed in thiol-eliminations. The
Npys-mixed disulfides are easily handled and can be purified by
chromatography. Exposure of the Npys-disulfide to HMPT resulted
cleanly in deprotection to form 18b (Table 1, entry 5) indicating that
the nucleophilic phosphine preferentially attacks the sulfur-atom
distal to the macrocycle. Encouraged, however, by the easy formation
of the Npys-disulfide, we took advantage of a further feature of this
reagent: its ability to directly exchange with the trityl protecting
18
group. Thus, exposure of compound 18a to NPys-Cl at ꢀ30 1C
overnight allowed for direct generation of the mixed disulfide 19 in
good yield (Scheme 2). Surveying conditions for the final elimina-
Fig. 2 Electrophilic reactivity of macrocycle 1. (A) and (B) LC-MS data of
tion, we found that exposure of the mixed Npys-disulfide to DBU at the BSA-derived tryptic peptide SLHTLFGDELCK modified with either
iodoacetamide (CA) or the macrocycle 1 (MC). The peptide is identified
as a triple charged species at (A) 473.9 m/z (CA) and (B) 557.9 m/z (MC)
ambient temperature furnished the mc-APD-containing macrocycle
1
in good yield after prep-TLC purification.
corresponding to a mass difference of 252 Da equivalent to the mass
difference between iodoacetamide and 1. The identity of the peptide and
position of the modification were confirmed by LC-MS/MS data (Fig. S1
Next we evaluated the potential reactivity of macrocycle 1
towards biological nucleophiles using bovine serum albumin
(
BSA) as a model protein. We incubated BSA with 1 under and Table S1, ESI†).
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 12427--12430 | 12429

View Article Online
Communication
ChemComm
TBP gratefully acknowledges the Danish Cancer Society,
the Lundbeck Foundation (Lundbeckfoundation Fellow grant),
and the Carlsberg Foundation for financial support. JJE
acknowledges support from the Danish National Research
Foundation and the Danish Research Council for Strategic
Research/Natural Sciences.
Notes and references
1
(a) C. Drahl, B. F. Cravatt and E. J. Sorensen, Angew. Chem., Int. Ed.,
005, 44, 5788–5809; (b) M. Gersch, J. Kreuzer and S. A. Sieber, Nat.
2
Prod. Rep., 2012, 29, 659–682.
D. C. Liebler and F. P. Guengerich, Nat. Rev. Drug Discovery, 2005, 4,
2
3
410–420.
J. Singh, R. C. Petter, T. A. Baillie and A. Whitty, Nat. Rev. Drug
Discovery, 2011, 10, 307–317.
4
5
J. Baell and M. A. Walters, Nature, 2014, 513, 481–483.
M. E. Bunnage, E. L. P. Chekler and L. H. Jones, Nat. Chem. Biol.,
2013, 9, 195–199.
6
7
C. MacKintosh, K. A. Beattie, S. Klumpp, P. Cohen and G. A. Codd,
FEBS Lett., 1990, 264, 187–192.
D. Krahn, C. Ottmann and M. Kaiser, Nat. Prod. Rep., 2011, 28,
1854–1867.
Scheme 3 Preparative reaction between macrocycle
1 and methyl
thioglycolate.
8 (a) K. D. McBrien, R. L. Berry, S. E. Lowe, K. M. Neddermann,
I. Bursuker, S. Huang, S. E. Klohr and J. E. Leet, J. Antibiot., 1995, 48,
1446–1452; (b) Y. Yamazaki, S. Kunimoto and D. Ikeda, Biol. Pharm.
Bull., 2007, 30, 261–265; (c) M. Takeuchi, E. Ashihara, Y. Yamazaki,
S. Kimura, Y. Nakagawa, R. Tanaka, H. Yao, R. Nagao, Y. Hayashi,
H. Hirai and T. Maekawa, Cancer Sci., 2011, 102, 591–596.
(a) G. Carr, M. Poulsen, J. L. Klassen, Y. Hou, T. P. Wyche,
T. S. Bugni, C. R. Currie and J. Clardy, Org. Lett., 2012, 14,
through intermediacy of the acyl-imine tautomer. After 72 hours,
however, adduct 20 was the sole product observed, perhaps indicat-
ing a reversibility of 21 back to its precursor. In accord, we were
unable to purify compound 21 when this reaction was stopped at
early time points, whereas compound 20 could be purified and
characterized. Acyl-imine tautomerism in dehydroalanine function-
alities is known, however (reversible) addition of simple thiols to
such systems is to the best of our knowledge unprecedented.
Collectively, these experiments demonstrate the reactivity of the
mc-APD group towards thiol-nucleophiles.
9
2
822–2825; (b) H. Nishioka, S. Nakajima, M. Nagashima, M. Kojiri
and H. Suda, JP 10147594 A 19980602, 1998; (c) M. Igarashi,
T. Shida, Y. Sasaki, N. Kinoshita, H. Naganawa, M. Hamada and
T. Takeuchi, J. Antibiot., 1999, 52, 873–879.
1
0 T. B. Poulsen, Chem. Commun., 2011, 47, 12837–12839.
11 N. Oku, S. Matoba, Y. M. Yamazaki, R. Shimasaki, S. Miyanaga and
19
Y. Igarashi, J. Nat. Prod., 2014, 77, 2561–2565.
1
2 F. Sang, D. Li, X. Sun, X. Cao, L. Wang, J. Sun, B. Sun, L. Wu,
G. Yang, X. Chu, J. Wang, C. Dong, Y. Geng, H. Jiang, H. Long,
S. Chen, G. Wang, S. Zhang, Q. Zhang and Y. Chen, J. Am. Chem.
Soc., 2014, 136, 15787–15791.
3 J. C. Albright, A. W. Goering, J. R. Doroghazi, W. M. Metcalf and
N. L. Kelleher, J. Microbiol. Biotechnol. Res., 2014, 41, 451–459.
In conclusion, we have reported a new method for the formation
of the mc-APD group found in a number of depsipeptide natural
products, including the rakicidins. Our discovery of the superior
1
ability of the Npys-group to facilitate thiol elimination to construct 14 T. Doi, Y. Numajiri, A. Munakata and T. Takahashi, Org. Lett., 2006,
8
, 531–534.
the mc-APD system is notable, and we anticipate that this reaction
will enable construction of focused mc-APD-containing compound
libraries. The versatility of Npys-Cl is currently underappreciated
and additional applications of this reagent are warranted. Finally,
we have characterized an inherent reactivity of the mc-APD group
1
5 J. M. Chalker, S. B. Gunnoo, O. Boutureira, S. C. Gerstberger,
M. Fern ´a ndez-Gonz ´a lez, G. J. L. Bernardes, L. Griffin, H. Hailu,
C. J. Schofield and B. G. Davis, Chem. Sci., 2011, 2, 1666–1676.
6 (a) R. Matsueda and K. Aiba, Chem. Lett., 1978, 951–952;
1
(
3
b) R. Matsueda and R. Walter, Int. J. Pept. Protein Res., 1980, 16,
92–401.
towards thiol nucleophiles, including protein-resident cysteine 17 R. Matsueda, T. Kimura, E. T. Kaiser and G. R. Matsueda, Chem.
Lett., 1981, 737–740.
8 R. Matsueda, S. Higashida, R. J. Ridge and G. R. Matsueda,
Chem. Lett., 1982, 921–924.
19 R. K. Olsen and A. J. Kolar, Tetrahedron Lett., 1975, 3579–3582.
residues and we speculate that electrophilic reactivity may play a
key role in the interactions between mc-APD-containing natural
products and their, as of yet, unknown biological targets.
1
1
2430 | Chem. Commun., 2015, 51, 12427--12430
This journal is ©The Royal Society of Chemistry 2015