10.1002/anie.201814519
Angewandte Chemie International Edition
COMMUNICATION
(Figure S12), indicating the incomplete conversion of Cu(DTC)2.
Thus, UV-Vis spectroscopy was used to probe the reactions by
monitoring the characteristic absorption band in the visible region
of copper(II) (Figure S17). Cu(DTC)2 shows much slower reaction
speeds relative to CuCl2 (Figure 5a, Figure S17, S19).
Additionally, the reaction with R=Et (ca. 64% conversion) appears
several times faster than that with R=Ph (ca. 25% conversion) in
the initial 10 minutes, consistent with the results of the competition
experiments. After 3 hours, the conversion of Cu(DTC)2 is ca.
78% for TpMe,MeZnSEt and ca. 48% for TpMe,MeZnSPh,
respectively. PhSSPh is nonvolatile and can thus be conveniently
isolated by thin-layer chromatography for yield analysis. A molar
amount equal to a half of the consumed Cu(DTC)2 was observed
(Figure S18), confirming the same redox stoichiometry as simple
CuX2 (Figure 2). Further kinetic studies indicate that the reaction
is 2.57 ± 0.33 order in Cu(DTC)2 and 1.62 ± 0.03 order in
TpMe,MeZnSEt (Figure S24), which seems to be consistent with a
dimeric mechanism in step 1 (vide supra). The consistent
oxidation of zinc(II) thiolates by various copper(II) complexes to
form disulfides suggests that the intracellular clustering of NPL4
proteins induced by Cu(DTC)2[5] might potentially involve a similar
chemical process. In contrast to the hard Cl/O donors in CuX2,
DTC has two soft sulfur donors conjugated with one lone-pair-
donating nitrogen. As such, the coordination bonds in Cu(DTC)2
are more covalent and its copper center is less electron-deficient
than those in CuX2, which presumably makes Cu(DTC)2 less
oxidizing relative to CuX2. The minor negative shift of the redox
potential for Cu(DTC)2 (0.02V vs Fc+/Fc) (Figure S16) relative to
CuCl2 (0.09V vs Fc+/Fc) implies that the former should play a
more important role.
The most decisive difference between Cu(DTC)2 and other
copper(II) complexes was observed from their reactions with two
thiols—EtSH and dithiothreitol—mimicking intracellular thiols
such as glutathione and cysteinyl thiols of proteins. Similar to
TpMe,MeZnSEt, both thiols reduced three simple copper(II)
complexes—Cu(OAc)2, CuCl2, and CuBr2—virtually completely
within 2 minutes under the same conditions (Figure S20), which
is consistent with the role of intracellular thiols in quenching
aberrant oxidants. In stark contrast, < 5% reduction of Cu(DTC)2
by both thiols was observed even after 3 hours (Figure 5b, S20),
indicating that thiols are much less reactive than zinc(II) thiolates
in their reactions with Cu(DTC)2. Such extraordinary redox
specificity is unprecedented, clearly showcasing the unique ability
of DTC in altering the common reactivity of copper(II) via tuning
its oxidation power to discriminate between thiols and zinc(II)
thiolates. Addition of base was found to promote the oxidation of
EtSH by Cu(DTC)2 (ca. 20% conversion, Figure 5b, S21),
suggesting that the formation of proton acid due to the oxidation
of EtSH should account for the reactivity difference between
TpMe,MeZnSEt and thiols. This interesting discovery suggests that
acidic cancerous environments might enhance the resistance of
thiol oxidation by Cu(DTC)2. A further experiment showed that
Cu(DTC)2 is similarly inert in the presence of NADH (Figure S22,
a close analogue of nicotinamide adenine dinucleotide phosphate
(NADPH)). At this point, whether intracellular zinc(II)-binding
histidine/cysteine sites are susceptible to similar oxidations by
Figure 3. a, Cu K-edge XANES spectra of Cu-containing products and
reference samples; b, Fourier-transform (FT) plots of the Cu K-edge EXAFS (FT
range: 2.9-10.6 Å-1).
The clean model reactions with simple CuX2 clearly point to a
two-step mechanism (steps 1 and 2, Figure 4). TpMe,MeZnSR is
first oxidized by 1 equivalent of CuX2 to afford a 0.5:1:1 ratio of
RSSR, TpMe,MeZnX, and CuIX. Subsequent metathesis between
CuIX and the second equivalent of TpMe,MeZnSR leads to CuISR
and another equivalent of TpMe,MeZnX. As observed previously in
the literature, copper(II) thiolate can dimerize and undergo further
redox tautomerizations to form copper(I) and disulfides.[17a],[17b]
A
similar process might also be involved in step 1. The yield relative
to TpMe,MeZnSR (50%) is constant for the disulfides under all
circumstances suggests that the second step is much faster than
the first step; otherwise, higher yields are expected for the
disulfides and TpMe,MeZnSR would formally act more similar to a
typical 1-electron reductant. In fact, the 50% yield remained
unchanged even upon increasing the starting copper(II)/zinc(II)
ratio from 0.5:1 to 1:1 (Figure S11), further supporting the notion
that the first step is rate-determining and the abstraction of thiolate
from the zinc(II) center by copper(I) is much faster than the redox
step.
Figure 4. a proposed reaction mechanism between TpMe,MeZnSR and CuX2.
Competitive oxidations of both TpMe,MeZnSR complexes were
performed to further probe the mechanism (Figure S13, Table S8).
In addition to EtSSEt and PhSSPh, the cross coupling product
EtSSPh was also formed. Since PhSSPh doesn’t react with
TpMe,MeZnSEt (Figure S14), the formation of EtSSPh should arise
from cross coupling reaction of relevant intermediates in step 1
(Figure 4). As observed previously in the literature, copper(II)
thiolate can dimerize and undergo further redox tautomerizations
to form copper(I) and disulfides[17]. A similar process might also
be involved in step 1 (Figure S15), which is consistent with
statistical analyses of disulfide product distributions in the
competition reactions upon a dimeric mechanism (Table S8, S9).
We next studied the impact of DTC ligation onto the reactivity
between copper(II) and the zinc(II) model complexes. A 2:1
mixture of TpMe,MeZnSR and Cu(DTC)2 shows a redox reactivity
1
similar to those with simple CuX2, albeit not as clean. Broad H
NMR signals typical of paramagnetic species were observed
This article is protected by copyright. All rights reserved.
Xu, Luyan
