Angewandte
Communications
Chemie
responsible for transcriptional initiation by RNA polymer-
ase II (RNAPII) as well as nucleoside excision repair.
Despite its great promise as an anticancer drug lead,
attempts to develop triptolide and its synthetic derivatives
over the past few decades have not succeeded.[5] Several
generations of clinical candidates have failed except for
Minnelideꢀ, the latest triptolide derivative that has entered
a phase I human clinical trial.[6] The major obstacles for
triptolide to becoming a clinically useful drug include its
general toxicity and lack of water solubility.[7]
date donor 13 was first converted into intermediate 15 upon
reaction with alcohol 14 in the presence of AgOTf, followed
by condensation with triptolide under the promotion of NIS
and TfOH to yield conjugate 16 (Figure 1c). Upon removal of
the benzoyl protecting groups with potassium carbonate in
aqueous methanol, 6 was obtained in 50% yield. The
structures of key reaction intermediates and all glutriptolides
1
(2–6) were confirmed by H and 13C NMR spectroscopy and
high-resolution mass spectrometry (see the Supporting Infor-
mation).
One strategy to reduce the toxicity of triptolide is to
deliver it selectively to tumor cells over their normal counter-
parts. A unique characteristic of tumor cells is that they have
a much higher demand for glucose than normal cells partially
owing to the Warburg effect.[8] Consequently, most cancer
cells overexpress one or more isoforms of glucose trans-
porters (GLUTs), particularly GLUT1 and GLUT3, to
sustain their growth and survival.[9] The overexpression of
GLUTs in tumor cells has been exploited to target tumor cells
selectively by conjugating cytotoxic drugs to glucose.[10]
Glucose conjugates have been made for a number of cytotoxic
drugs such as ifosfamide and taxol[11] and have been shown to
have lower toxicity and higher tumor cell selectivity in vitro.
We envisioned that conjugating triptolide with glucose could
also increase its selectivity for tumor cells, thereby decreasing
its toxicity. The resultant glucose–triptolide conjugates (glu-
triptolides) would enter cancer cells through GLUTs whereby
the linker would undergo cleavage to release triptolide,
allowing it to bind to XPB and block cell proliferation or
induce apoptosis.
We designed and synthesized five glutriptolide deriva-
tives, compounds 2–6 (Figure 1). We chose the C14 hydroxy
group as the site of linkage to glucose as it is the most reactive
functional group in triptolide and can undergo facile chemical
modification. Whereas there are multiple sites in glucose that
can be used to connect to triptolide, the C1 hydroxy group in
glucose has been successfully used to form active conjugates
with a number of known drugs.[10] The glutriptolides contain
one of three distinct linkages, namely a succinate (2, 3) or an
acetal (6) linker or a direct linkage through an ether bond (4,
5). Whereas the succinate linker is susceptible to hydrolysis by
cellular esterases, the acetal linker is more stable, and the
ether linkage likely remains intact inside cells.
The syntheses of glutriptolides 2 and 3 (Figure 1a)
commenced with acylation of the C14 hydroxy group of
triptolide with succinic anhydride (7) in the presence of
DMAP and pyridine to give intermediate 8.[12] Condensation
of 8 with protected glucose 9 in the presence of DMAP and
DCC yielded 10 as an anomeric mixture with a 1:2 ratio for
the a and b anomers. The benzyl protecting groups on the
glucose moiety were removed by hydrogenation in the
presence of palladium on charcoal, giving rise to glutripto-
lides 2 and 3, with 2 as the predominant product. Glutripto-
lides 4 and 5 were synthesized by a two-step sequence
including condensation of triptolide with compound 11 in the
presence of PPh3AuNTf2 and 4 ꢁ molecular sieves to give 12,
followed by removal of the benzyl protecting groups by
hydrogenolysis in the presence of palladium on charcoal
(Figure 1b). To prepare glutriptolide 6, the trichloroacetimi-
We determined the effects of the glutriptolides on the
DNA-dependent ATPase activity of TFIIH isolated from
HeLa cell nuclear extract as previously described.[3] Surpris-
ingly, none of the glutriptolide analogues showed appreciable
inhibitory activity (Figure 2a; see also the Supporting Infor-
mation, Figure S1). Based on the solution structure of 2
determined by 2D NMR spectroscopy, the lack of activity of 2
may be attributable to intramolecular interactions between
the succinate linker and the A and B rings of triptolide
(Figures S2 and S3), preventing the triptolide moiety from
interacting with XPB. The inability of 2 and 3 to inhibit XPB
renders them ideal prodrugs devoid of undesirable activity
prior to entry into cells where the linker is cleaved by cellular
esterases.
Next, we determined the effects of the glutriptolides on
the proliferation of HEK293T cells. Glutriptolides 2 and 3
retained considerable antiproliferative activity with IC50
values of 268 and 615 nm, respectively (Figure 2b). In
contrast, glutriptolides 4–6 had little effect on cell prolifer-
ation even at the highest concentrations. Those results are
likely due to the inability of glutriptolides 4–6 to undergo
linker cleavage inside cells, which is consistent with their lack
of activity towards XPB/TFIIH in vitro. Given that 2 had
a slightly higher antiproliferative activity than 3 (Figure 2b)
and was the major product of the synthesis (Figure 1a), we
decided to focus on 2 in ensuing studies.
To confirm that the antiproliferative activity of 2 was due
to inhibition of endogenous XPB, we determined its effect on
the stability of the catalytic subunit of RNAPII, which has
been shown to undergo degradation upon binding of tripto-
lide to XPB.[3] Glutriptolide 2 induced degradation of the
RPB subunit of RNAPII, although at higher concentrations in
comparison to triptolide (Figure 2c). Upon longer incubation
(72 h), the degradation of RPB became more pronounced
(Figure S4). In contrast, neither taxol nor doxorubicin
affected the stability of RPB (Figure S5), suggesting that the
induction of RPB degradation is specific for triptolide and 2.
To further assess whether the antiproliferative activity of 2
was mediated through inhibition of XPB, we determined the
sensitivity of an engineered XPB-C342T mutant cell line that
is resistant to triptolide as a result of a mutation of the active-
site cysteine that is covalently modified by triptolide.[4a]
Glutriptolide 2, like triptolide, had no inhibitory effect
whereas both taxol and doxorubicin retained their inhibitory
effects on the XPB-C342T cell line (Figure S6), indicating
that the antiproliferative activity of glutriptolide 2 is mediated
through inhibition of XPB.
To determine whether 2 depended on GLUT to enter
cells, we tested 2 in combination with a known inhibitor of
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 6
These are not the final page numbers!