Total Synthesis of Trioxacarcins DC-45-A1, A, D, C, and C7″-epi-C and Full Structural Assignment of Trioxacarcin C

Article abstract of DOI:10.1021/jacs.5b12687

Trioxacarcins DC-45-A2, DC-45-A1, A, D, C7″-epi-C, and C have been synthesized through stereoselective strategies involving BF₃·Et₂O-catalyzed ketone-epoxide opening and gold-catalyzed glycosylation reactions, and the full structural assignment of trioxacacin C was deciphered via the syntheses of both of its C7″ epimers. The gathered knowledge sets the foundation for the design, synthesis, and biological evalution of analogues of these natural products as potential payloads for antibody-drug conjugates and other delivery systems for targeted and personalized cancer chemotherapy.

Full text of DOI:10.1021/jacs.5b12687

Article
pubs.acs.org/JACS
Total Synthesis of Trioxacarcins DC-45-A1, A, D, C, and C7″-epi-C and
Full Structural Assignment of Trioxacarcin C
K. C. Nicolaou,* Quan Cai, Hongbao Sun, Bo Qin, and Shugao Zhu
Department of Chemistry, BioScience Research Collaborative, Rice University, 6100 Main Street, Houston, Texas 77005, United
States
S
* Supporting Information
ABSTRACT: Trioxacarcins DC-45-A2, DC-45-A1, A, D, C7″-
epi-C, and C have been synthesized through stereoselective
strategies involving BF3·Et2O-catalyzed ketone−epoxide open-
ing and gold-catalyzed glycosylation reactions, and the full
structural assignment of trioxacacin C was deciphered via the
syntheses of both of its C7″ epimers. The gathered knowledge
sets the foundation for the design, synthesis, and biological
evalution of analogues of these natural products as potential
payloads for antibody−drug conjugates and other delivery
systems for targeted and personalized cancer chemotherapy.
1. INTRODUCTION
The trioxacarcin family of naturally occurring compounds
comprises a number of highly potent cytotoxic agents whose
challenging molecular structures and potential in targeted
cancer chemotherapies have elevated them to attractive
synthetic targets.1,2 Isolated from Streptomyces bottropensis
DO-45, these molecules are characterized by a modified
anthraquinone scaffold,
a polyoxygenated 2,7-
dioxabicyclo[2.2.1]heptane core carrying a spiroepoxide moi-
ety, and a varying number of carbohydrate units. Their
mechanism of action as cytotoxic agents involves binding to
double-stranded DNA followed by covalent modification of the
genetic material through proximity-facilitated nucleophilic
attack from a DNA base3 on the trioxacarcin epoxide moiety.
X-ray crystallographic analyses of complexes of trioxacarcin A
(3) (Figure 1)4,5 and related compounds6 with an 8-mer duplex
DNA segment or guanine revealed intercalation of the aromatic
structural domain of the molecule into DNA. A covalent bond
between the N7 atom of a nearby guanine base and the
terminal carbon of the spiroepoxide unit formed through
nucleophilic attack of the former upon the latter was also
observed.5 It is this DNA-damaging event that leads to the
lethal potencies of these molecules, making them desirable
payloads for antibody−drug conjugates (ADCs)7 and other
drug delivery systems.8
Because of their structural novelty and desirability as
potential payloads, the trioxacarcins have been the subject of
several synthetic efforts,9−11 notably by the Myers group10 and
our laboratory.11 Myers reported the total synthesis of
trioxacarcins DC-45-A2 (1), DC-45-A1 (2), and A (3) based
on an elegant 1,3-dipolar cycloaddition strategy,10 while we
recently described the total synthesis of DC-45-A2 (1) through
an alternative strategy relying on a BF3·Et2O-catalyzed
Figure 1. Molecular structures of trioxacarcins DC-45-A2 (1), DC-45-
A1 (2), A (3), D (4), and C (originally assigned as 5a or 5b).
epoxyketone rearrangement/ring closure to afford the 2,7-
dioxabicyclo[2.2.1]heptane structural motif of the molecule.11
In this article we describe the total synthesis of trioxacarcins
DC-45-A2 (1), DC-45-A1 (2), A (3), and D (4) as well as
trioxacarcins C (5b) and C7″-epi-C (5a). The latter syntheses
enabled the full stereochemical assignment of trioxacarcin C as
depicted by structure 5b [i.e., C7″-(S)].
Received: December 10, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
2. RESULTS AND DISCUSSION
Article
Scheme 2. Construction of Glycosyl o-Alkynylbenzoate
a
Donor 12
2.1. Retrosynthetic Analysis and Rationale. The main
synthetic challenges posed by the trioxacarcins are the
a
Reagents and conditions: (a) LiTMP (1.2 equiv), −78 °C, 2 h, then 7
(1.3 equiv), THF, −100 °C, 4 h, 52% for 8 plus 40% total for the other
isomers; (b) MeNHOMe·HCl (3.0 equiv), n-BuLi (5.0 equiv), −50 to
−30 °C, 1 h, then 8, THF, −78 to −50 °C, 3 h; (c) MeLi (4.0 equiv),
THF, −78 to −50 °C, 5 h; (d) Ac2O (2.0 equiv), DMAP (0.2 equiv),
Et3N (3.0 equiv), CH2Cl2, 2 h, 56% over the three steps; (e) 3:1
AcOH/H2O, 80 °C, 3 h; (f) 11 (1.2 equiv), EDC (1.4 equiv), DMAP
(1.0 equiv), DIPEA (1.8 equiv), CH2Cl2, 6 h, 67% over the two steps.
LiTMP = lithium tetramethylpiperidide; EDC = 1-ethyl-3-(3-
Figure 2. Retrosynthetic analysis of the trioxacarcins defining aglycon
common precursor II and glycosyl donors III and IV as key building
blocks.
(dimethylamino)propyl)carbodiimide; DMAP
= 4-(N,N-
dimethylamino)pyridine, DIPEA = N,N-diisopropylethylamine.
construction of the polyoxygenated 2,7-dioxabicyclo[2.2.1]-
heptane structural motif (for 1−5) and the α-glycosidic bonds
(for 2−5) linking the carbohydrate moieties to the aglycon of
the molecule, as indicated in Figure 2. While the former
challenge was met through the aforementioned BF3·Et2O-
catalyzed epoxyketone rearrangement/cyclization,11 the latter
task remained, particularly because of the steric demands of the
carbohydrate donors and their stereoselective attachment to the
equally hindered aglycon. Our approach for solving this
problem relied on the recently developed gold-catalyzed
glycosylation reactions utilizing glycosyl o-alkynylbenzoates as
donors.12 Thus, retrosynthetic disconnection of the carbohy-
drate units from the aglycon of trioxacarcins A, D and C (3−5)
revealed key intermediates II−IV as potential precursors.
Scheme 1. Summary of the Synthesis of Common Trioxacarcin Precursor II11
B
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 3. Synthesis of Glycosyl o-Alkynylbenzoate Donors
15 and 16
Scheme 4. Total Synthesis of Trioxacarcins A (3) and D (4)
a
a
Reagents and conditions: (a) O3, 0.5 h, then Me2S (10 equiv),
MeOH, 12 h; (b) 11 (1.0 equiv), EDC (1.4 equiv), DMAP (1.0
equiv), DIPEA (1.8 equiv), CH2Cl2, 4 h, 50% over two steps (α:β
anomers ca. 1:4.3 dr); (c) Ac2O (2.0 equiv), DMAP (0.2 equiv),
pyridine, 25 °C, 3 h, 59%; (d) TMSCl (1.5 equiv), imidazole (3.0
equiv), CH2Cl2, 0 °C, 1 h, 70% (β-anomer exclusively). TMS =
trimethylsilyl.
Equipped with hydroxyl groups that are well-differentiated by
reactivity and appropriate protecting groups, aglycon inter-
mediate II11 was well-suited for manipulation so as to serve as a
common precursor to all of the targeted trioxacarcins. On the
other hand, the carbohydrate donors III and IV could be fine-
tuned by changes in their protecting groups (R1, R2, and R3)
and substituents (X and Y) to respond to any reactivity or
stereochemistry issues that might arise at the glycosylation
stage. The choice of the gold-catalyzed glycosylation reaction
and glycosyl donors III and IV was based on the previous
studies by Yu et al.12
2.2. Construction of Key Intermediates. Scheme 1
summarizes the construction of the common key precursor II
for the synthesis of all of the targeted trioxacarcins [i.e., DC-45-
A2 (1), DC-45-A1 (2), A (3), D (4), and C (5a and 5b)]. The
details of the synthesis of II and the conversion of this
precursor to trioxacarcin DC-45-A2 (1) were described
previously11 and will not be elaborated further here.
Scheme 2 depicts the construction of glycosyl o-alkynylben-
zoate carbohydrate donor 12 from readily available ethyl ester
acetonide 613 and dimethoxy acetal aldehyde 7.14 Addition of
the anion generated from 6 and LiTMP to 7 proceeded
smoothly to afford a mixture of four diastereoisomers from
which the desired isomer 8 was chromatographically isolated in
52% yield. The rest of the chromatographically obtained
materials consisted of a mixture of the remaining three isomers
of 8 (40% combined yield). The stereochemical identity of 8
was based on its conversion to a literature-known compound10b
(i.e., the β-anomeric acetate of compound 10; see the
Supporting Information (SI) for details). Conversion of ethyl
ester 8 to methyl ketone 9 was achieved through a three-step
sequence involving Weinreb amide generation (MeNHOMe·
HCl, n-BuLi), ketone formation (MeLi), and acetylation in 56%
overall yield. Acetonide methyl ketone 9 was then exposed to
AcOH/H2O to afford, upon cleavage of the acetonide and ring
closure, carbohydrate 10 as a mixture of anomers. After a short
column filtration, the latter was reacted with o-alkynylbenzoic
acid 11 in the presence of EDC to furnish the desired glycosyl
o-alkynylbenzoate ester 1212c in 67% overall yield for the two
steps). Glycosyl donor 12 was formed diastereoselectively as
a
Reagents and conditions: (a) II (1.0 equiv), Ph3PAuOTf (0.2 equiv),
12 (10 equiv), 4 Å molecular sieves, CH2Cl2, −20 °C, 5 min, 91%, α:β
> 20:1; (b) DDQ (2.0 equiv), 4:1 CH2Cl2/H2O (pH 7.0 buffer), 25
°C, 87%; (c) Ph3PAuNTf2 (0.2 equiv), 15 (2.0 equiv), 4 Å molecular
sieves, CH2Cl2, 0 °C, 5 min, 71%; (d) K2CO3 (2.0 equiv), MeOH, 0
°C, 40 min; (e) Et3N·3HF (20 equiv), CH3CN, 25 °C, 12 h, 64% over
the two steps; (f) Ph3PAuNTf2 (0.2 equiv), 16 (2.0 equiv), 4 Å
molecular sieves, CH2Cl2, 0 °C, 5 min, 92%, α:β > 20:1; (g) K2CO3
(2.0 equiv), MeOH, 0 °C, 40 min; (h) Et3N·3HF (20 equiv), CH3CN,
25 °C, 12 h, 70% over the two steps. DDQ = 2,3-dichloro-5,6-dicyano-
1,4-benzoquinone.
the β-anomer (separated by chromatography from trace
amounts of the α-anomer), as confirmed by the observed H
NMR coupling constants of its anomeric proton (δH‑1 6.29
1
3
1
ppm, dd, J = 10.4, 2.5 Hz; see the SI for the expanded H
NMR region of this signal). The preference for the formation of
the β-anomer is presumed to be due to steric hindrance
provided by the bulky o-alkynylbenzoate residue in the α-
anomer, as seen with manual molecular models.
The synthesis of β-anomeric glycosyl donors 15 and 16 is
shown in Scheme 3. Ozonolysis of the known olefinic triol
1310c followed by treatment of the resulting ozonide with Me2S
resulted in the formation of the corresponding pyranose, whose
exposure to o-alkynylbenzoic acid 11 in the presence of EDC
led to the desired building block 14 (50% overall yield for the
two steps; inseparable mixture of anomers, β:α = 4.3:1 dr).
Reaction of 14 with Ac2O in the presence of DMAP and
pyridine led to the corresponding acetoxy derivatives, whose
anomers were separated by chromatography to furnish β-
C
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
1
see the SI for the expanded H NMR region of this signal).
Analogously, 14 was reacted with TMSCl in the presence of
imidazole to afford TMS derivative 16 in 70% yield (pure β-
anomer after purification, anomeric proton δH1‑1 6.22 ppm, dd,
3J = 10.1, 2.4 Hz; see the SI for the expanded H NMR region
of this signal).
Scheme 5. Total Synthesis of Trioxacarcin DC-45-A1 (2)
2.3. Total Synthesis of Trioxacarcins A, D, and DC-45-
A1. With all of the building blocks available, we then turned
our attention to their assembly into the various naturally
occurring trioxacarcins. The first to be targeted were
trioxacarcins A (3) and D (4). Their stereoselective total
syntheses were accomplished as planned (Scheme 4). Coupling
of trioxacarcin carbohydrate acceptor II with glycosyl donor 12
in the presence of Ph3PAuOTf catalyst and 4 Å molecular
sieves in CH2Cl2 at −20 °C stereoselectively furnished the
desired and expected α-glycoside 17 (91% yield, α:β > 20:1,
anomeric proton δH‑1″ 5.82 ppm, dd, 3J = 4.2, 2.1 Hz; see the SI
for the expanded 1H NMR region of this signal). In preparation
for the attachment of the second sugar on the growing
molecule, the PMB protecting group was removed from 17
(DDQ, 87% yield), affording hydroxy compound 18 (see
Scheme 4) ready for the second glycosylation. The reaction of
carbohydrate acceptor 18 with carbohydrate donor 15 in the
presence of Ph3PAuNTf2 catalyst and 4 Å molecular sieves in
CH2Cl2 at 0 °C afforded the desired α-glycoside 19 (71% yield,
anomeric protons δH‑1″ 5.81 ppm, 3J = 3.9, 2.3 Hz and δH‑1′ 5.35
ppm, apparent dd, 3J = 3.1, 3.1 Hz; see the SI for the expanded
1H NMR region of these signals). Sequential exposure of the
latter to K2CO3 in MeOH (deacetylation) followed by Et3N·
3HF (desilylation) then furnished the coveted trioxacarcin A
a
Reagents and conditions: (a) TMSCl (2.0 equiv), imidazole (4.0
equiv), CH2Cl2, 0 °C, 95%; (b) DDQ (2.0 equiv), 4:1 CH2Cl2/H2O
(pH 7.0 buffer), 25 °C, 89%; (c) Ph3PAuNTf2 (0.2 equiv), 15 (2.0
equiv), 4 Å molecular sieves, CH2Cl2, 0 °C, 5 min, 69%; (d) Et3N·3HF
(20 equiv), CH3CN, 25 °C, 12 h, 89%.
Scheme 6. Synthesis of Anomerically Activated Derivatives
of 2[H]-Trioxacarcinoside B (25a and 25b)
a
(3)1f,10d in 64% overall yield. The H and 13C NMR spectral
1
data for synthetic 3 were identical to those reported for the
natural product10c (see the SI). On the other hand,
glycosylation of substrate 18 with glycosyl donor 16 under
the influence of the same catalyst (Ph3PAuNTf2, 0.2 equiv) and
conditions furnished the targeted bisglycosylated product 20 in
92% yield (α:β > 20:1, anomeric protons δH‑1″ 5.81 ppm, dd, 3J
3
= 4.1, 2.3 Hz and δH‑1′ 5.32, apparent dd, J = 3.1, 3.1 Hz; see
1
the SI for the expanded H NMR region of these signals).
Sequential and selective deprotection of the latter (K2CO3,
MeOH, then Et3N·3HF) finally yielded trioxacarcin D (4)1c in
1
70% overall yield for the last two steps. The H and 13C NMR
spectral data for synthetic 4 were identical to those reported for
the natural product1c (see the SI).
Trioxacarcin DC-45-A1 (2) was synthesized from fragments
II and 15 as summarized in Scheme 5. Lactol substrate II was
protected (TMSCl, imidazole, 95% yield), and the resulting
TMS ether (21) was treated with DDQ to liberate the C4
hydroxyl group, furnishing carbohydrate acceptor 22 in 89%
yield. The latter was united with carbohydrate donor 15 in the
presence of Ph3PAuNTf2 catalyst and 4 Å molecular sieves in
CH2Cl2 at 0 °C to afford glycoside 23 (69% yield, anomeric
proton δH‑1′ 5.36 ppm, apparent dd, 3J = 2.9, 2.9 Hz; see the SI
a
1
Reagents and conditions: (a) NaBH4 (2.0 equiv), CeCl3·7H2O (1.2
for the expanded H NMR region of this signal). Concurrent
equiv), MeOH, −78 °C, 5 h, 88%; (b) triphosgene (3.0 equiv),
pyridine (5.0 equiv), CH2Cl2, 0 °C, 0.5 h, 86%; (c) K2CO3 (2.0 equiv),
MeOH, 0 °C, 15 min, 76%; (d) Li(t-BuO)3AlH (1.5 equiv), THF, 0
°C, 2 h, 53% over the two steps; (e) triphosgene (1.2 equiv), pyridine
(4.0 equiv), CH2Cl2, 0 °C, 0.5 h; (f) Ac2O (2.0 equiv), DMAP (0.5
equiv), pyridine (5.0 equiv), CH2Cl2, 25 °C, 0.5 h, 81% over two steps.
removal of the TMS and TBS groups from the latter (Et3N·
3HF) gave the coveted trioxacarcin DC-45-A1 (2) in 89% yield.
1
The H and 13C NMR spectral data for synthetic 2 were
identical to those reported for the natural product (see the
SI).1d,10d
2.4. Total Synthesis of Trioxacarcin C and C7″-epi-
Trioxacarcin C. The synthesis of trioxacarcins C (5a) and
C7″-epi C (5b) proved to be more challenging than those of
the other targeted trioxacarcins in that the two required 2[H]-
anomer 15 in 59% yield, as confirmed by NMR spectroscopic
3
analysis (anomeric proton δH‑1 6.26 ppm, dd, J = 9.8, 2.7 Hz;
D
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 7. Total Synthesis of C7″-epi-Trioxacarcin C (5a) and Trioxacarcin C (5b)
a
Reagents and conditions: (a) II (1.0 equiv), Ph3PAuOTf (0.3 equiv), 25a (5 equiv), 4 Å molecular sieves, CH2Cl2, 0 °C, 15 min, 92%, α:β > 20:1;
(b) NaH (10 equiv), 10:1 THF/ethylene glycol, 25 °C; (c) Ac2O (8.0 equiv), DMAP (1.0 equiv), pyridine (12 equiv), CH2Cl2, 25 °C, 3 h, 73% over
the two steps; (d) DDQ (2.0 equiv), 4:1 CH2Cl2/H2O (pH 7.0 buffer), 25 °C, 83%; (e) Ph3PAuNTf2 (0.2 equiv), 15 (2.0 equiv), 4 Å molecular
sieves, CH2Cl2, 0 °C, 5 min, 66%; (f) K2CO3 (2.0 equiv), MeOH, 0 °C, 1 h; (g) Et3N·3HF (20 equiv), CH3CN, 25 °C, 12 h, 68% over the two
steps; (h) II (1.0 equiv), Ph3PAuOTf (0.3 equiv), 25b (5 equiv), 4 Å molecular sieves, CH2Cl2, 0 °C, 15 min, 93%, α:β > 20:1; (i) NaH (10 equiv),
10:1 THF/ethylene glycol, 25 °C; (j) Ac2O (8.0 equiv), DMAP (1.0 equiv), pyridine (12 equiv), CH2Cl2, 25 °C, 3 h, 72% over the two steps; (k)
DDQ (2.0 equiv), 4:1 CH2Cl2/H2O (pH 7.0 buffer), 25 °C, 78%; (l) Ph3PAuNTf2 (0.2 equiv), 15 (2.0 equiv), 4 Å molecular sieves, CH2Cl2, 0 °C, 5
min, 72%; (m) K2CO3 (2.0 equiv), MeOH, 0 °C, 1 h; (n) Et3N·3HF (20 equiv), CH3CN, 25 °C, 12 h, 68% over the two steps.
trioxacarcinoside B donors 25a and 25b (Scheme 6) had to be
prepared selectively and attached to the growing molecule
through the desired α-glycosidic bond. While the latter
stereoselectivity was expected on the basis of the anomeric
effect, the diastereoselective reduction of hydroxyketo acetate
precursor 12 had to be investigated. Scheme 6 depicts the
selective synthesis of glycosyl donors 25a and 25b from 12 and
the mechanistic rationale for the stereoselectivity of the two
reductions involved. Exposure of hydroxyketo acetate 12 to
NaBH4 in the presence of CeCl3·7H2O furnished diol 24a
exclusively, presumably through transition state A (shown as a
Newman projection in Scheme 6b), in which the initially
formed Ce complex orients the carbonyl moiety in such a way
as to favor reduction from the Si face (front; the Re face is
blocked by the Me group on C5″, back). On the other hand,
treatment of hydroxyketo acetate 12 with K2CO3 (removal of
Ac) followed by Li(t-BuO)3AlH reduction led to the opposite
diastereoisomer 24b in 53% overall yield (single isomer). The
observed stereoselectivity in this reduction may be explained by
invoking transition state B (Scheme 6b), in which the newly
liberated secondary alcohol (C3″) participates in forming the
six-membered metallacycle with concurrent rotation of the
C4″/C5″ bond, orienting the carbonyl group to a position
favoring reduction from the Re face (front), with the C5″ Me
group no longer in the way. With the hydroxyl groups installed
stereoselectively, intermediates 24a and 24b were smoothly
converted to glycosyl donors 25a [(CCl3O)2CO, 86% yield]
and 25b [(CCl3CO)2CO, Ac2O, 81% overall yield] as shown in
Scheme 6. The structural assignments of compounds 25a and
25b were based on the observed nuclear Overhauser effects
3
(NOEs) as shown in Scheme 6a and the J coupling constants
of their anomeric protons (anomeric proton of 25a δH‑1 6.25
3
ppm, dd, J = 8.0, 2.7 Hz; anomeric proton of 25b δH‑1 6.21
ppm, apparent dd, 3J = 6.3, 6.3 Hz; see the SI for the expanded
1H NMR region of these signals). The carbonate derivatives of
these glycosyl donors were chosen on the basis of manual
molecular modeling and in order to rigidify them and cap both
hydroxyl groups, a tactic that resulted in high-yielding
glycosylation reactions (see below).
Scheme 7 summarizes the coupling of trioxacarcin aglycon
derivative II with glycosyl donors 25a (Scheme 7a) and 25b
(Scheme 7b) and the elaboration of the products to give the
E
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
two C7″ diastereomeric trioxacarins C (5a and 5b). Reaction of
II with excess 25a or 25b in the presence of Ph3PAuOTf
catalyst and 4 Å molecular sieves in CH2Cl2 at 0 °C furnished
glycoside 26a or 26b in 92% and 93% yield, respectively, and
α:β > 20:1 dr (anomeric proton of the α-anomer of 26a δH‑1″
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b12687.
■
S
Experimental procedures and compound characterization
data (PDF)
3
5.77 ppm, dd, J = 3.7, 2.0 Hz; anomeric proton of the α-
anomer of 26b δH‑1″ 5.79 ppm, d, 3J = 4.2 Hz; see the SI for the
1
expanded H NMR region of these signals). Reaction of 26a
AUTHOR INFORMATION
Corresponding Author
*kcn@rice.edu
Notes
■
with ethylene glycol and NaH resulted in the cleavage of both
the carbonate and acetate groups from the growing molecule,
liberating all three hydroxyl groups (see Scheme 7a).
Acetylation of the resulting compound (Ac2O, DMAP, py)
then led to bisacetate 27a in 73% overall yield. The latter
compound was treated with DDQ, causing cleavage of the PMB
group and furnishing intermediate glycosyl acceptor 28a (83%
yield). Reaction of 28a with glycosyl donor 15 in the presence
of Ph3PAuNTf2 catalyst and 4 Å molecular sieves in CH2Cl2
furnished protected trioxacarcin C (29a) (66% yield, anomeric
protons δH‑1″ 5.72 ppm, apparent dd, 3J = 2.5, 2.5 Hz and δH‑1′
5.35, apparent dd, 3J = 3.0, 3.0 Hz; see the SI for the expanded
1H NMR region of these signals), from which C7″-(R)-
trioxacarcin C (5a) was obtained upon sequential treatment
with K2CO3 in MeOH (selective deacetylation due to different
steric environments) and Et3N·3HF (desilylation) in 68%
overall yield. A similar sequence starting from the C7″-epimeric
substrate 26b and the same carbohydrate donor (i.e., 15) led to
the formation of C7″-(S)-trioxacarcin C (5b) via intermediates
26b−29b in similar yields, as shown in Scheme 7b. That this
series of compounds were formed predominantly as the α-
anomers (α:β > 20:1 anomeric ratio) as expected was
confirmed by their 1H NMR spectra, which exhibited the
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Cancer Prevention & Research
Institute of Texas (CPRIT), the Welch Foundation (Grant C-
1819), and Rice University. Q.C. and B.Q. gratefully acknowl-
edge Chongqing University for postdoctoral fellowships. H.S.
gratefully acknowledges the China Scholarship Council for
financial support. We thank Drs. L. B. Alemany and Q.
Kleerekoper (Rice University) for NMR spectroscopic
assistance and Drs. C. Pennington (Rice University) and I.
Riddington (University of Texas at Austin) for mass
spectrometric assistance.
REFERENCES
■
(1) (a) Tomita, F.; Tamaoki, T.; Morimoto, M.; Fujimoto, K. J.
Antibiot. 1981, 34, 1519. (b) Tamaoki, T.; Shirahata, K.; Iida, T.;
Tomita, F. J. Antibiot. 1981, 34, 1525. (c) Maskey, R. P.; Helmke, E.;
Kayser, O.; Fiebig, H. H.; Maier, A.; Busche, A.; Laatsch, H. J. Antibiot.
2004, 57, 771. (d) Shirahata, K.; Kaisha, T. U.S. Patent 4,459,291,
1984. (e) Tomita, F.; Tamaoki, T.; Shirahata, K.; Iida, T.; Morimoto,
M.; Fujimoto, K. EP 0,029,309, 1984. (f) Tomita, F.; Tamaoki, T.;
Shirahata, K.; Iida, T.; Motimoto, M.; Fujimoto, K. U.S. Patent
4,511,560, 1985.
(2) (a) Fujimoto, K.; Morimoto, M. J. Antibiot. 1983, 36, 1216.
(b) Maiese, W. M.; Labeda, D. P.; Korshalla, J.; Kuck, N.; Fantini, A.
A.; Wildey, M. J.; Thomas, J.; Greenstein, M. J. Antibiot. 1990, 43, 253.
(c) Cassidy, J.; Graham, M. A.; Huinink, W. T. B.; McDaniel, C.;
Setanoians, A.; Rankin, E. M.; Kerr, D. J.; Kaye, S. B. Cancer
Chemother. Pharmacol. 1993, 31, 395.
(3) Sun, D.; Hansen, M.; Clement, J. J.; Hurley, L. H. Biochemistry
1993, 32, 8068.
́
(4) (a) Maskey, R. P.; Sevvana, M.; Uson, I.; Helmke, E.; Laatsch, H.
Angew. Chem., Int. Ed. 2004, 43, 1281. (b) Fitzner, A.; Frauendorf, H.;
Laatsch, H.; Diederichsen, U. Anal. Bioanal. Chem. 2008, 390, 1139.
(5) Pfoh, R.; Laatsch, H.; Sheldrick, G. M. Nucleic Acids Res. 2008, 36,
3508.
3
anticipated J coupling constants for their anomeric protons
(26b δH‑1″ 5.79 ppm, d, 3J = 4.2 Hz; 29b δH‑1″ 5.73 ppm, d, 3J =
4.0 Hz and δH‑1′ 5.35, apparent dd, 3J = 3.1, 3.1 Hz).
Comparison of the 1H and 13C NMR spectroscopic data for the
two synthetic trioaxacarcins C (5a and 5b) revealed the identity
of the naturally occurring trioxacarcin C as structure 5b
1
possessing the C7″-(S)-configuration (see the SI for H and
13C NMR spectral comparisons).1e This finding leads to the
designation of 5a as C7″-epi-trioxacarcin C. Synthetic
trioxacarcin C (5b) exhibited 1H and 13C spectral data identical
to those reported for the natural product (see the SI).
3. CONCLUSION
(6) Propper, K.; Dittrich, B.; Smaltz, D. J.; Magauer, T.; Myers, A. G.
̈
The herein-described total syntheses of trioxacarcins DC-45-A1
(2), A (3), D (4), C (5b), and C7″-epi-C (5a) provide access
to these potent antitumor agents and open the way for the
synthesis of other members of the class, natural or designed.
The high yields and stereoselectivities of the gold-catalyzed
glycosylation reactions are particularly impressive given the
complexity and sensitivity of the substrates employed,
suggesting their potential for applications in other challenging
situations. The described chemistry is expected to facilitate the
design and synthesis of more or less complex analogues of the
trioxacarcins aiming at highly potent cytotoxic agents as
required for ADCs and other drug delivery systems for targeted
cancer chemotherapy. Our studies toward this goal will be
reported in due course.
Bioorg. Med. Chem. Lett. 2014, 24, 4410.
(7) (a) Sievers, E. L.; Senter, P. D. Annu. Rev. Med. 2013, 64, 15.
(b) Perez, H. L.; Cardarelli, P. M.; Deshpande, S.; Gangwar, S.;
Schroeder, G. M.; Vite, G. D.; Borzilleri, R. M. Drug Discovery Today
2014, 19, 869. (c) Chari, R. V. J.; Miller, M. L.; Widdison, W. C.
Angew. Chem., Int. Ed. 2014, 53, 3796.
(8) (a) Polu, K. R.; Lowman, H. B. Expert Opin. Biol. Ther. 2014, 14,
1049. (b) Cohen, R.; Vugts, D. J.; Visser, G. W. M.; Stigter-van
Walsum, M.; Bolijn, M.; Spiga, M.; Lazzari, P.; Shankar, P. S.; Sani, M.;
Zanda, M.; van Dongen, G. A. M. S. Cancer Res. 2014, 74, 5700.
(9) (a) Suami, T.; Nakamura, K.; Hara, T. Bull. Chem. Soc. Jpn. 1983,
56, 1431. (b) Konig, C. M.; Harms, K.; Koert, U. Org. Lett. 2007, 9,
̈
4777.
̌
(10) (a) Svenda, J.; Hill, N.; Myers, A. G. Proc. Natl. Acad. Sci. U. S. A.
2011, 108, 6709. (b) Magauer, T.; Myers, A. G. Org. Lett. 2011, 13,
̌
5584. (c) Smaltz, D. J.; Svenda, J.; Myers, A. G. Org. Lett. 2012, 14,
F
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1812. (d) Magauer, T.; Smaltz, D. J.; Myers, A. G. Nat. Chem. 2013, 5,
886.
(11) Nicolaou, K. C.; Cai, Q.; Qin, B.; Petersen, M. T.; Mikkelsen, R.
J.; Heretsch, P. Angew. Chem., Int. Ed. 2015, 54, 3074.
(12) (a) Li, Y.; Yang, Y.; Yu, B. Tetrahedron Lett. 2008, 49, 3604.
(b) Yang, Y.; Li, Y.; Yu, B. J. Am. Chem. Soc. 2009, 131, 12076. (c) Li,
Y.; Yang, X.; Liu, Y.; Zhu, C.; Yang, Y.; Yu, B. Chem. - Eur. J. 2010, 16,
1871. (d) Zhang, Q.; Sun, J.; Zhu, Y.; Zhang, F.; Yu, B. Angew. Chem.,
Int. Ed. 2011, 50, 4933. (e) Tang, Y.; Li, J.; Zhu, Y.; Li, Y.; Yu, B. J. Am.
Chem. Soc. 2013, 135, 18396. (f) Nie, S.; Li, W.; Yu, B. J. Am. Chem.
Soc. 2014, 136, 4157.
(13) (a) Gardiner, J. M.; Panchal, N. R.; Stimpson, W. T.; Herbert, J.
M.; Ellames, G. J. Synlett 2005, 17, 2685. (b) Zhang, Z.; Ding, Y.; Xu,
J.; Chen, Y.; Forsyth, C. J. Org. Lett. 2013, 15, 2338.
(14) Bi, L.; Zhang, Y.; Zhao, M.; Wang, C.; Chan, P.; Tok, J. B.-H.;
Peng, S. Bioorg. Med. Chem. 2005, 13, 5640.
G
DOI: 10.1021/jacs.5b12687
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX