Stabilizing p-Dithiobenzyl Urethane Linkers without Rate-Limiting Self-Immolation for Traceless Drug Release

Article abstract of DOI:10.1002/cmdc.201900248

Exploiting the redox sensitivity of disulfide bonds is a prevalent strategy in targeted prodrug designs. In contrast to aliphatic disulfides, p-thiobenzyl-based disulfides have rarely been used for prodrug designs, given their intrinsic instability caused by the low pK_a of aromatic thiols. Here, we examined the interplay between steric hindrance and the low-pK_a effect on thiol–disulfide exchange reactions and uncovered a new thiol–disulfide exchange process for the self-immolation of p-thiobenzyl-based disulfides. We observed a central leaving group shifting effect in the α,α-dimethyl-substituted p-dithiobenzyl urethane linkers (DMTB linkers), which leads to increased disulfide stability by more than two orders of magnitude, an extent that is significantly greater than that observed with typical aliphatic disulfides. In particular, the DMTB linkers display not only high stability, but also rapid self-immolation kinetics due to the low pK_a of the aromatic thiol, which can be used as a general and robust linkage between targeting reagents and cytotoxic drugs for targeted prodrug designs. The unique and promising stability characteristics of the present DMTB linker will likely inspire the development of novel targeted prodrugs to achieve traceless release of drugs into cells.

Full text of DOI:10.1002/cmdc.201900248

Accepted Article
Title: Stabilizing p-dithiobenzyl Urethane Linkers without Rate-Limiting
Self-Immolation for Traceless Drug Release
Authors: Yiwu zheng, Yang Shen, Xiaoting Meng, Yaqi Wu, Yibing
Zhao, and Chuanliu Wu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.201900248
Link to VoR: http://dx.doi.org/10.1002/cmdc.201900248
A Journal of
www.chemmedchem.org

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
Stabilizing p-dithiobenzyl Urethane Linkers without Rate-Limiting
Self-Immolation for Traceless Drug Release
Yiwu Zheng, Yang Shen, Xiaoting Meng, Yaqi Wu, Yibing Zhao*, and Chuanliu Wu*[a]
[a]
Y. Zheng, Y. Shen, X. Meng, Y. Wu, Prof. Y. Zhao, Prof. C. Wu
The MOE key laboratory of Spectrochemical Analysis and instrumentation, State Key Laboratory of Physical Chemistry of Solid Surfaces, Department of
chemistry, College of Chemistry and Chemical Engineering, Xiamen University, 361005 (P.R. China)
E-mail: chlwu@xmu.edu.cn, ybzhao@xmu.edu.cn
Supporting information for this article is given via a link at the end of the document.
Abstract: Exploiting the redox-sensitivity of disulfide bonds is a
prevalent strategy adopted in targeted prodrug designs. Compared
with aliphatic disulfides, the p-thiobenzyl-based disulfides have
rarely been used for prodrug designs due to their intrinsic instability
caused by the low pKa of aromatic thiols. Here, we examined the
interplay between steric hindrance and the low-pKa effect on the
thiol-disulfide exchange reactions and uncovered a new thiol-
disulfide exchange process for the self-immolation of p-thiobenzyl-
based disulfides. We observed a central-leaving group shifting effect
in the α,α-dimethyl substituted p-dithiobenzyl urethane linkers
(DMTB-linkers), which leads to >2 orders of magnitude increase in
disulfide stability, an extent that is significantly larger than that
observed from the typical aliphatic disulfides. Particularly, the DMTB-
linkers display not only a high stability, but also rapid self-immolation
kinetics due to the low pKa of the aromatic thiol, which can be used
as a general and robust linkage between targeting reagents and
cytotoxic drugs for targeted prodrug designs. The unique and
promising stability characteristics of the present DMTB-linker would
inspire the development of novel targeted prodrugs to achieve
traceless release of drugs into cells.
the cytotoxic activity of the anticancer drug.[5] It has been known
that reduction of cytotoxicity for drugs bearing a free thiol group
may be caused by thiol-disulfide exchange reactions between
the thiol-bearing drug diffusing out from the targeting cells and
protein or small-molecule disulfides in the extracellular space,
which leads to the formation of mixed disulfide species less
capable of crossing cell membranes and killing the surrounding
cells compared to the free drugs, a phenomenon that has been
well-known as the “bystander effect”).[6] Thus, to avoid the
influence of extracellular thiol-disulfide exchange reactions on
the “bystander effect” and to maximize cytotoxic activity of
anticancer drugs, self-immolating disulfide linkers have recently
been utilized for traceless drug delivery.[5, 7]
The most prevalently used self-immolating disulfide linkers
were developed on the basis of a cyclization elimination
mechanism (Figure 1a).[3a, 5] One merit of these disulfide linkers
is that their stability can be easily tuned by introducing steric
hindrance (i.e., methyl groups) adjacent to the disulfide bonds.
The self-immolation rate and efficiency of these disulfide linkers
after the cleavage is usually fast and high enough in the
intracellular environments.[8] However, we realize that the rate of
self-immolation can be slow under acidic conditions such as in
the endocytic or lysosomal compartments (pH 4‒6) and in tumor
microenvironments which have been found to be mildly acidic
(pH ~6), because the thiols of these linkers usually have a pKa of
8‒9. In addition, as thiol-reducing reagents such as the
glutathione (GSH) prefer attacking the less sterically hindered
sulfur atoms, we also realize that the subsequent circulation of
thiol-disulfide exchanges is in principle required to release the
free thiol drugs ready for self-immolation, though depending on
in which side of the disulfide bond the two methyl groups were
introduced (Figure 1a). This disulfide shuffling circulation
constitutes the rate-limiting steps for self-immolation, particularly
in acidic environments where the rate of thiol-disulfide
exchanges is slow.[9] Considering that more and more prodrug
designs exploit the mildly acidic tumor microenvironments and
endocytic compartments as targeting locations for reduction-
induced drug release,[2b, 10] novel self-immolating disulfide linkers
with both good stability and robust self-immolation capability
under these specific environments are definitely a valuable
module for the development of novel cytotoxic prodrugs.
Introduction
Cancer chemotherapies can usually benefit from targeted
delivery of cytotoxic drugs into cancer cells. One of the most
efficient ways to achieve targeted delivery relies on prodrug
designs involving the conjugation of drug molecules to targeting
agents through covalent linkers (e.g., the antibody-drug
conjugates and small-molecule drug conjugates).[1] The linkers
are one of the most critical components in these prodrugs, which
can response to either endogenous stimuli in cells or in tumor
extracellular microenvironments or external stimuli to release the
drugs.[2] Among the diverse linkers used for prodrug designs,
disulfides have attracted extensive interests due to the large
differences between the extra- and intra-cellular redox
environments, which can selectively release the drug molecules
inside the cells.[3] However, to avoid the premature release of
drugs during circulation in the body, disulfide linkers usually
need to be stabilized by regulating the local steric or electronic
environments on the linkers.[4] As very few small-molecule drugs
contain thiol groups, disulfide linkers with an easily tunable
stability that are also easily amenable to drug-targeting agent
conjugations have usually been a virtue for prodrug designs.
Despite the merits of disulfide linkers, the presence of an
appended thiol to the free drug has been found detrimental to
An alternative route for the design of self-immolating disulfide
linkers relies on the 1,6-benzyl elimination mechanism.[11]
However, these disulfide linkers have rarely been used for
prodrug designs, because they are usually ~2‒3 orders of
magnitude less stable than the conventional aliphatic disulfides,
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
though disulfide linkers of this type (i.e., p-thiobenzyl-based
disulfide) were developed over 20 years ago and their capability
of self-immolation under physiological conditions is even more
robust than the aforementioned self-immolating disulfides.[11b, 11c]
The intrinsic instability of the p-thiobenzyl-based disulfides can
be elucidated by Brønsted equation, which indicates that the
lower pKa (~6) of the aromatic thiol can accelerate the cleavage
of the disulfides by serving as a favorable leaving group.[12] In
general, the effect of steric hindrance on disulfide stability can
be less pronounced than the electronic effect. For example, the
presence of 1‒2 methyl groups can usually increase stability of
disulfide bonds by 5- to 20-folds in comparison to unhindered
bonds.[4b, 13] Thus, it is important to examine to what extent the p-
thiobenzyl-based disulfides can be stabilized by manipulation of
steric hindrance and if the steric hindrance can compensate for
the low-pKa induced instability to afford useful linkers for
cytotoxic prodrug designs. Though there have been a few
reports showing that the stability of p-thiobenzyl-based disulfides
can indeed be increased by introducing one or two methyl
groups adjacent to the disulfide bonds,[14] it remains unclear how
the interplay between steric hindrance and low-pKa effect affects
the thiol-disulfide exchange reactions, and how these reactions
further react on the self-immolation of the p-thiobenzyl-based
disulfides.
In this work, we examined the effect of the steric hindrance of
two methyl groups adjacent to the p-thiobenzyl-based disulfides
on their thiol-disulfide exchange reactions with small thiol-
reducing reagents and how these reactions affect the
subsequent 1,6-benzyl elimination of the p-thiobenzyl group. We
found that the low-pKa aromatic thiol prefers serving as the
central group for nucleophilic attacking rather than as leaving
group, because the aliphatic thiol is too sterically hindered to be
attacked (Figure 1b). This central-leaving group shifting effect
results in >2 orders of magnitude increase in stability in GSH
buffers in comparison to the unhindered analogue. After the
disulfide reduction, the GSH-mixed intermediate is subject to a
rapid exchange reaction with another GSH molecule to further
release the free thiol form of the p-thiobenzyl molecule ready for
self-immolation (Figure 1b). Considering that the low pKa of
aromatic thiols can afford the p-thiobenzyl-based disulfides rapid
self-immolation kinetics upon reduction even under mildly acidic
conditions, targeted prodrugs were then designed and
synthesized using a stabilized p-dithiobenzyl urethane, folic acid
(FA) and Doxorubicin (DOX) or monomethyl auristatin E (MMAE)
Figure 1. a) Generic thiol-disulfide exchange reactions for the cleavage of
typical aliphatic disulfides and cyclization-based elimination processes. b)
Thiol-disulfide exchange reactions taking place during the cleavage and self-
immolation of the DMTB linkers.
Results and Discussion
To examined the effect of steric hindrance on the stability of p-
thiobenzyl-based disulfides,
a
cysteine (Cys) analog,
penicillamine (Pen, ,-dimethylcysteine), which has two methyl
groups adjacent to the thiol and two functional groups amenable
to further modifications was chosen in our study.[15] In addition,
to facilitate the monitoring of the thiol-disulfide exchanges using
high performance liquid chromatography (HPLC),[16] short
as
a linker, a targeting agent and an anticancer drug,
respectively. We demonstrated that the free cytotoxic drugs can
be efficiently released into the cytosol by exploiting the mildly
acidic and reducing endocytosis pathways mediated by folate
receptor (FR) as a targeting location for the reduction cleavage
and self-immolation. A FRET probe that enables us to visualize
the drug release processes was also developed, which clearly
demonstrates the performance of the p-dithiobenzyl urethane
linker for endocytosis-based prodrug release. This work thus not
only reveals a central-leaving group shifting effect for stabilizing
p-thiobenzyl-based disulfide bonds, but provides useful self-
immolating disulfide linkers for targeted cytotoxic prodrug
designs.
peptides with good water solubility were used as
a
straightforward and flexible platform for presenting the Cys and
Pen thiol. Thus, Cys-containing peptide (P1), Pen-containing
peptide (P2) and 4-mercaptobenzeneacetate (MBA, thiol pKa:
5.95)[17] were used as models to synthesize the mixed disulfides
(i.e., M1: unhindered p-thiobenzyl-based disulfide; M2: α,α-
dimethyl substituted p-dithiobenzyl-based disulfide) (Figure 2a).
HPLC chromatograms shown in Figure 2b reveal the cleavage
process of the 10 μM M1 in 0.2 mM glutathione (GSH) at pH 7.4.
We found that GSH can selectively attack the aliphatic sulfur
atom, which results in the rapid formation of MBA and P1-GSH
mixed disulfide and the negligible formation of P1. This result
indicates that the nucleophilic attack on mixed disulfides
(R'SSR", pKaR'SH > pKaR"SH) would occur favorably with the
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
release of the more acidic thiol (R"SH). By contrast, when 10 μM
M2 was incubated in 0.5 mM GSH at pH 7.4, we observed a
gradual decrease of M2 associated with an increase of P2 and
MBA-GSH mixed disulfide, and MBA-GSH is subsequently
reduced by a reaction with another GSH molecule to give the
MBA (Figure 2c). It is worth mentioning that the formation of
minor amount of P2-GSH mixed disulfide should be resulted
from the further thiol-disulfide exchanges between P2 and
GSSG. To further demonstrate that the aromatic thiol can act as
These results show that the cleavage of the α,α-dimethyl
substituted disulfides by thiol-disulfide exchanges relies on the
nucleophilic attack of the MBA sulfur atom (as a central group),
and the α,α-dimethyl substituted sulfur mainly serves as the
leaving group due to steric hindrance.
We then examined and compared the kinetics of cleavage of
M1 and M2. The half-lives (t ) can be obtained from the non-
½
linear regression of curves shown in Figure 2d using a pseudo-
first-order kinetics model. The M2 disulfide (a half-life of 26.6
min was obtained in 0.5 mM GSH at pH 7.4) is ~127 times more
stable than the M1 disulfide (a half-life of 31.3 s was obtained in
0.2 mM GSH at pH 7.4). Furthermore, it is worth mentioning that
the M2 disulfide is slightly more stable than the Cys-S-S-Pen
disulfides.[15] Additionally, the stability of the α,α-dimethyl
substituted disulfides could be easily adjusted by altering the pKa
value of the aromatic thiol. As shown in Figure 2d, compared to
M2 disulfide, M4 (EIISH, thiol pKa: 4.5; t = 4.7 min) have
½
approximately 5.6-fold faster exchange kinetics under the same
condition.
Based on the above promising results, we then used the 4-
mercaptobenzyl alcohol (MA) as a self-immolative unit to design
α,α-dimethyl substituted p-dithiobenzyl urethane linkers (DMTB-
linkers) for targeted cytotoxic prodrug designs. Folic acid (FA) is
employed as a targeting ligand for the design, because it has a
very high affinity to folate receptor (FR)-overexpressed cells.[10b]
FR is overexpressed in many tumors, which can actively
internalize the bound folates into cells via recycling endosomal
pathways.[18] The recycling endosomal pathways have been
found reducing,[19] which can be potential locations for disulfide
reduction. We designed and synthesized a DMTB-linked folate-
doxorubicin (DOX) conjugates (D1) for targeted delivery of DOX
to the FR-overexpressed cancer cells (Figure 3a). D1 was
synthesized by coupling a thiophilic heterobifunctional cross-
linker to the DOX by the activated carbonate chemistry first,
followed by a substitution of the 2-thiopyridyl group of the linker
with
a
FA-containing penicillamine (see the Supporting
Information for the details). We first examined if the treatment of
the D1 with GSH can result in a release of the active drug.
Indeed, D1 was completely converted to DOX in the presence of
10 mM GSH within 5 h at pH 7.4 (Figure 3b). In addition, we
evaluate the kinetics of cleavage of the D1 in the presence of 10
mM and 0.2 mM GSH at pH 7.4 phosphate buffer solutions,
which mimics the redox environments in the cytosol and in the
blood circulation, respectively. The reconversion kinetics of the
DOX prodrug D1 was rapid in 10 mM GSH with a half-life of 1.3
h, but the prodrug D1 are extremely stable in 0.2 mM GSH, the
percentage of cleavage is only 8% after 6 h incubation (Figure
3c), which meets the demand that FA-drug conjugates can
usually be eliminated from the blood circulation with a half-life of
around half of an hour.[20] Of note, it is likely that the hydrophobic
stacking interaction between the ligand and the drug as well as
the negatively charged microenvironment surrounding the
disulfide bond should contribute to the remarkable increase in
the stability of the DMTB-linker, which is reflected by the slow
kinetics of cleavage compared to the kinetics shown in Figure
2b). This result indicates that the DMTB-linker shows not only a
rapid responsiveness to highly reducing conditions, but an
ultrahigh stability in the weakly reducing extracellular
Figure 2. a) Chemical structure of disulfide linkers; note that flanking residues
in peptides were omitted (P1: Ac-GGRCGW-NH2; P2: Ac-GGRPenGW-NH2).
b) Representative chromatograms for thiol-disulfide exchanges of M1 (10 μM)
with GSH (0.2 mM) at pH 7.4 (phosphate buffer, 100 mM); absorbance was
recorded at 280 nm. c) Representative chromatograms for thiol-disulfide
exchanges of M2 (10 μM) with GSH (0.5 mM) at pH 7.4 (phosphate buffer, 100
mM); absorbance was recorded at 280 nm; the chromatograph peaks were
characterized by mass spectrometry (Figure S1). d) Representative plots of
disappearance of disulfide linkers as a function of time.
the central atom in thiol-disulfide exchange reactions, aromatic
disulfide (M3) was then synthesized, and the rate of cleavage of
M3 is very rapid in 0.2 mM GSH at pH 7.4 with a half-life of 37.6
s (Figure 2d), suggesting that GSH can attack the M3 sulfur
atom. In stark contrast to the aromatic disulfide, Pen-S-S-Pen
disulfide is extremely stable even in 10 mM GSH buffers.[4a]
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
environments such as in the blood circulation. We next
examined if the prodrug D1 would take advantage of the
reducing power in the recycling endosomal for drug release. The
prodrug D1 was incubated with FR-positive KB cells, D1 shows
a dose-dependent inhibition of proliferation in KB cells after 48 h
of incubation, as determined by Cell Counting Kit-8 (CCK-8).
Free DOX was also tested in an analogous manner. The
corresponding IC50 values were estimated to be 0.87 µM and
3.25 µM for the free DOX and the D1 (Table 1), respectively.
indicates that the reduction of the DMTB-linker and the rapid
self-immolation can take place in the mildly acidic and reducing
endosomal compartment.
microtubule inhibitor agent MMAE[21]). The prodrug D2 was
synthesized in a similar way to synthesizing D1. The release
kinetics of D2 was also studied, and we obtained similar results
to that from D1 (Figure 5a and Figure S2). However, D2 (IC50
=
4.54 nM) is around 1000-fold more cytotoxic that D1 (Table 1
and Figure 4), indicating the potent therapeutic effect of D2.
We then compared the DMTB-linker with the well-established
self-immolating linkers based on cyclization. A prodrug C1 was
synthesized for the comparison (Figure 5a), which has two
methyl groups adjacent to the sulfur atom on the ligand side.
Interestingly, the cytotoxic activity of D2 is slightly higher than
C1 (IC50 = 8.90 nM), though the D2 disulfide has 1.5-fold slower
exchange kinetics in comparison to the C1 disulfide (Figure S6)
(Table 1 and Figure 4). A potential explanation for this is the
complex thiol–disulfide exchange processes shown in Figure 1a.
The disulfide shuffling circulation (Figure 1a) should be the rate-
limiting steps for self-immolation, which slows down the release
rate of free drugs. Indeed, we observed the formation of GSH-
mixed MMAE (1a) when C1 was incubated in a GSH/GSSG-
mixed buffer, whereas no formation of GSH-mixed MMAE (2a)
was observed from D2 under the same condition (Figure 5b).
Because the rate of disulfide cleavage for 2a is ~15-fold faster
than 1a according to the prediction of Brønsted equation,[22] and
2a can undergo a rapid thiol–disulfide exchange reaction with
GSH to directly yield 2b. In contrast, GSH can attack either of
the two sulfur atoms in 1a, which results in two distinguished
thiol-disulfide exchange pathways. Thus, the complete cleavage
of 1a disulfide with concomitant release of 1b involves multistep
thiol–disulfide exchange processes. Indeed, the formation of the
unstable intermediates 2a and 2b was not observed (Figure 5b),
suggesting that the kinetics of self-immolation are very fast.
Overall, the promising results presented in this section indicate
that the DMTB-linker provides a useful tool for developing
targeted cytotoxic prodrugs.
To expand our understanding of the role that the DMTB-linked
self-immolation plays in drug release and cytotoxicity, the
nonself-immolative prodrug C2 with an amide conjugation was
designed and synthesized (Figure 5a). The cytotoxicity of C2 to
KB cells (IC50 = 18.6 nM; Table 1) was slightly lower than that of
D2. In addition, MMAE-SH was also evaluated for comparison,
and its IC50 was estimated to be 11.4 nM. The free MMAE (0.30
nM) was 38-fold more potent than MMAE-SH, indicating that the
sulfhydryl modification on MMAE has a great influence on the
cytotoxicity. A possible explanation is that when the disulfide is
cleaved in cells to release a thiol-bearing MMAE, upon efflux out
of the cells, thiol-bearing MMAE can react with disulfide-
containing compounds in the extracellular space, a process that
can significantly reduce the bystander effect of cytotoxic
drugs.[6b] Indeed, the thiol-bearing MMAE (MMAE-SH) is also
less cytotoxicity than the S-methylated MMAE (MMAE-S-CH3)
(Table 1). In addition, we are not able to rule out the possibility
that the thiol modification on the MMAE might affect its
intercalation into the target tubulin proteins. To confirm the
importance of the disulfide bond reduction, the non-cleavable
maleimide control C3 was synthesized and evaluated; it only has
very weak cytotoxicity to KB cells (Table 1 and Figure 4).
Figure 3. a) Schematic representation of the drug release process and
chemical structure of FA-Pen and MA; b) HPLC chromatograms show the
release of DOX from D1 (10 μM) in the presence of 10 mM GSH at pH 7.4
after a reaction of 5h. HPLC chromatograms were recorded at 254 nm. c).
Kinetic traces for DOX release from D1 in the presence of 10 mM and 0.2 mM
GSH at pH = 7.4 phosphate buffer. Data are presented as mean ± SD (n = 3).
In the light of the stability characteristics of D1 in GSH buffers,
we next turned our attention to the utility of the DMTB-linker for
conjugating an anti-cancer drug with a higher cytotoxicity (a
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
Figure 4. Dose response curves for cell viability treated with D2, C1 and C3,
respectively. All compounds were incubated with the KB cells for 48 h. Data
are presented as mean ± SD (n = 3).
We then examined if the recycling endosomal pathways are
indeed the locations where the disulfide linkers were cleaved for
the release of drugs by using a DMTB-FRET probe (Figure 6).
Kinetics measurements based on fluorescence were carried out
in the presence of 10 mM and 0.2 mM GSH at pH 7.4 phosphate
buffers. The incubation of DMTB-FRET probe in 10 mM GSH
buffer led to a rapid increase of fluorescence intensity with a
half-life of 0.92 h-1. At 0.2 mM GSH concentration, only a
negligible fluorescence recovery was observed after 6 h
incubation (Figure S9). In addition, the boron dipyrromethene
(BODIPY) fluorophore is very suitable for mimicking the behavior
of the anticancer drugs (MMAE and DOX) in living cells due to
their similar hydrophobic properties. Thus, DMTB-FRET probe
can serve as a fluorescent prodrug mimic for the fluorescent
monitoring of the targeted cellular uptake and the subsequent
drug release process. Interestingly, intense intracellular
fluorescence was observed after the incubation of cells with the
DMTB-FRET probe (Figure 6b), though a portion of green
fluorescence merged well with the red fluorescence
(LysoTracker Red DND-99) from the lysosomes. However, the
coincubation of N-(3-maleimidopropionyl)biocytin (MPB, a potent
blocker for cell-surface thiols and thiols in the endocytosis
pathways[23]) with the DMTB-FRET probe during the cell culture
can result in a complete suppression of the DMTB-linker
reduction (Figure 6c). These results indicate that the reduction of
the DMTB-linker takes place on the cell surface, or during the
early stages of uptake. This result suggests that the reduction of
DMTB-linker stems only from thiol-disulfide exchange reactions,
thus excluding other possible ways of disulfide reduction, though
the detailed mechanism of disulfide reduction in reducing
compartments is still unknown. To distinguish the relative
contribution of the cell surface and the early stages of
endocytosis to DMTB-linker reduction, KB cells were incubated
with DMTB-FRET probe at 4 °C (to inhibit endocytosis); and no
Figure 5. a) Abbreviated chemical structure of FA-Drug conjugates. b) HPLC
chromatograms show the reduction of C1 and D2 (30 μM) by GSH/GSSG (at -
220 mV redox potential) in phosphate buffer pH 7.4 supplemented with 10%
DMSO as co-solvent after a reaction of 3h; the peaks were characterized by
mass spectrometry (Figure S7-S8). The HPLC chromatograms were recorded
at 210 nm.
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
Figure 6. a) Design of DMTB-FRET probe based on intramolecular energy transfer mechanism. b) Confocal fluorescence images of KB cells incubated with the
DMTB-FRET probe at 37°C , evolution of fluorescence was observed with time. LysoTracker Red DND-99 (DND-99) was used to track cell lysosomes. c) Confocal
fluorescence images of KB cells treated with DMTB-FRET probe over 4h in the presence of MPB and at 4°C . Scale bar: 10 μm.
fluorescence inside cells was observed (Figure 6c), suggesting
that the cell surface was not involved in the fluorescence
recovery. These results suggest that the mildly acidic reducing
endosomal milieu is strongly involved in the reduction of the
DMTB-FRET probe (Figure 7a).
in our DMTB-linkers, which leads to >2 orders of magnitude
increase in disulfide stability, an extent that is significantly larger
than that observed from the typical aliphatic disulfides.
Particularly, the DMTB-linkers display not only a high stability,
but also rapid self-immolation kinetics due to the low pKa of the
aromatic thiol, which can be a useful tool for developing targeted
cytotoxic prodrugs. Indeed, the DMTB-linked cytotoxic prodrugs
exhibit a high potency in killing cancer cells. A DMTB-FRET
probe was also designed and synthesized to enable us to
monitor the process of DMTB-linked prodrug release. Rapid
release of cytotoxic drugs in recycling endosomal pathways
mainly contribute to the high potency of the prodrugs. Recently,
non-internalizing targeted prodrugs which take advantage of
reduction of disulfide-linked prodrugs on the tumor cell-surface
microenvironments have been found to be of high efficiency in
anticancer therapy.[24] However, considering that the reduction of
conventional aliphatic disulfides in the tumor microenvironments
can usually lead to unavoidable formation biothiols-linked drugs
with weaker bystander killing capability,[25] our DMTB-linked
prodrug design provides an alternative strategy to solve this
problem. Achieving high stability in blood circulation without
comprising fast drug release is of fundamental importance to
targeted prodrug design. We believe that novel targeted prodrug
designs would greatly benefit from the stability characteristics of
the present DMTB-linker.
Finally, we examined the stability of DMTB in the presence of
fetal bovine serum (FBS). As can be seen in Figure 7b, the
recovery of fluorescence is extremely slow in 50% FBS buffer.
When 50% FBS was supplemented with 10 mM GSH, as
expected, a very rapid dequenching was observed. In addition,
when the probes were co-incubated in 50% serum with N-
ethylmaleimide (NEM, a thiol capping reagent), no change of
fluorescence intensity was observed over time, indicating that
the cleaved of DMTB was resulted mainly from thiol-disulfide
exchange reactions. These results demonstrate the excellent
stability of the DMTB linker in serum.
Conclusions
In summary, we examined the effect of two methyl groups
adjacent to the p-thiobenzyl-based disulfides on their thiol-
disulfide exchange reactions and elucidated the mechanism of
the p-dithiobenzyl-based disulfides reduction and self-
immolation. A central-leaving group shifting effect was observed
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
(λex = 488 nm; λem = 511 nm) was measured using a fluorescence
spectrometer. To block thiol-disulfide exchanges in serum, pre-treat FBS
with 0.2 mM N-ethylmaleimide (NEM) for 2 h. Then, 3 μL of a 50 μM
stock solution of DMTB-FRET probe was added and fluorescence was
measured with time.
Cell culture: KB cells were cultured in DMEM medium (high glucose)
containing 10 % FBS and 1 % penicillin/streptomycin (penicillin: 10,000 U
·mL–1, streptomycin: 10,000 U·mL–1) under an atmosphere of 5% CO2 at
37 °C. Cells were passaged at about 80 % cell confluency using a
0.25 % trypsin solution.
Cell viability assay: Cytotoxicity of prodrugs and MMAE-derived drugs
was evaluated by Cell Counting Kit-8 (CCK-8). KB cells were seeded in a
96-well plate at an initial cell density of 8000 cells per well and grown
overnight at 37 °C, 5 % CO2. On the second day, the supernate was
removed and 100 μL prodrugs in DMEM (without FBS) was added to
each well at several different concentrations, the cells were incubated at
37 °C, 5 % CO2 for 48 h. Then the cell viability was measured using
CCK-8 according to the manufacture’s protocol. The absorbance was
then measured at 450 nm using an ELIASA reader (PerkinElmer Enspire).
The obtained absorbance was blank-corrected (blank: DMEM+CCK-8, no
cells) and the cell viability in percent was calculated according to the
following equation:
Cell viability = (OD450, sample / OD450, control) × 100%
where OD450, sample represents the optical density of the cells treated with
prodrugs and OD450, control is the cells treated with DMEM (without FBS).
Experiments were conducted in triplicate with n = 3. Error bars represent
standard deviation of the mean.
Confocal cell imaging: KB cells were plated into glass bottom cell
culture dishes at an initial cell density of 100,000 cells per well. After 24 h
at 37 °C, 5 % CO2, the cells were grown to about 80 % confluence. The
medium was then removed and the cells were washed twice with PBS.
Cells were then incubated with 1 mL serum-free complete medium
containing 200 nM DMTB-FRET (stock solution: 50 μM in DMSO) at
37 °C, 5 % CO2 for desired lengths of time. For cell lysosome staining,
cells were incubated with 200 nM LysoTracker Red DND-99 for 2 h. After
that, cells were imaged in a Leica TCS SP5X Confocal Microscope
System at different detection channels. (BODIPY channel: λex = 488 nm,
λem = 510-560 nm; LysoTracker Red DND-99 channel: λex = 543 nm, λem
= 600-670 nm).
Figure 7. a) Schematic representation of the reduction of DMTB-FRET probe
through the recycling endosomal route. b) Thiol-disulfide exchange kinetics of
DMTB-FRET probe (0.1 μM) in 50% FBS at pH 7.4 (black line), 50% FBS at
pH 7.4 pretreated with 0.2mM NEM for 2 h (red line), and 50% FBS
supplemented with 10 mM GSH (blue line). Data are presented as mean ± SD
(n = 3).
Confocal cell imaging with MPB coincubation: Cells were seeded as
described above. Upon medium removal, the cells were washed twice
with PBS, coincubated with 1 mL complete medium containing 500 μM
MPB (stock solution: 100 mM in DMSO) and 200 nM FRET-DMTB for 4 h,
After that, cells were imaged in a Leica TCS SP5X Confocal Microscope
System.
Experimental Section
Kinetics of thiol–disulfide exchanges and drug release: All solutions
used in kinetic studies were deoxygenated by bubbling with nitrogen in a
round-bottomed flask sealed with a rubber stopper. To further exclude
oxygen, experiments were carried out in an anaerobic incubator. In a
typical experiment, 10 μM peptide disulfide or prodrugs were incubated
with 0.2 or 0.5 mM reducing agent (GSH) in pH 7.4 phosphate buffers. At
predefined times, disulfide exchanges were quenched by addition of an
equal volume of 10% meta-phosphoric acid. The samples were then
analyzed by UPLC or HPLC.
Acknowledgements
We would like to acknowledge the financial support from the
National Natural Science Foundation of China (21822404 and
21675132), The fundamental Research Funds for Central
Universities (20720180034), the Program for Changjiang
Scholars and Innovative Research Team in University (13036),
and the Foundation for Innovative Research Groups of the
National Natural Science Foundation of China (21521004).
Kinetics of DMTB-FRET cleavage in redox buffers: Stock solution of
DMTB-FRET probe (50 μM) was prepared in DMSO. The concentration
of the probe was calculated spectrophotometrically (ε500 nm = 86652 cm-
1·M -1). In a typical experiment, FRET-DMTB probe (final concentration:
100 nM) was incubated with GSH solution (0.2 mM or 10 mM in 100 mM
degassed phosphate buffer; pH 7.4). At predefined intervals, the
fluorescence of the solution (λex = 488 nm; λem = 511 nm) was measured
using a fluorescence spectrometer.
Keywords: Disulfide • Traceless drug release • Self-immolating
linkers • Penicillamine • p-Dithiobenzyl urethane
References:
Kinetics of DMTB-FRET cleavage in Serum: 150 μL of fetal bovine
serum (FBS) and 147 μL phosphate buffer (100mM, pH 7.4) were added
into a fluorescence cuvette. Then, 3 μL of a 50 μM stock solution of
DMTB-FRET probe was added. At predefined intervals, the fluorescence
[1] G. Casi, D. Neri, J. Med. Chem. 2015, 58, 8751-8761.
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
[2] a) L. R. Staben, S. G. Koenig, S. M. Lehar, R. Vandlen, D. Zhang, J.
Chuh, S. F. Yu, C. Ng, J. Guo, Y. Liu, A. Fourie-O'Donohue, M. Go, X.
Linghu, N. L. Segraves, T. Wang, J. Chen, B. Wei, G. D. Phillips, K. Xu,
K. R. Kozak, S. Mariathasan, J. A. Flygare, T. H. Pillow, Nat. Chem.
2016, 8, 1112-1119; b) G. J. Bernardes, G. Casi, S. Trussel, I. Hartmann,
K. Schwager, J. Scheuermann, D. Neri, Angew. Chem. Int. Ed. 2012, 51,
941-944; c) Y. Zheng, J. Ren, Y. Wu, X. Meng, Y. Zhao, C. Wu,
Bioconjugate Chem. 2017, 28, 2620-2626; d) R. M. Versteegen, R.
Rossin, W. ten Hoeve, H. M. Janssen, M. S. Robillard, Angew. Chem. Int.
Ed. 2013, 52, 14112-14116.
[3] a) M. H. Lee, J. L. Sessler, J. S. Kim, Acc. Chem. Res. 2015, 48, 2935-
2946; b) L. Brulisauer, M. A. Gauthier, J. C. Leroux, J. Control. Release
2014, 195, 147-154.
[4] a) C. Wu, S. Wang, L. Brulisauer, J. C. Leroux, M. A. Gauthier,
Biomacromolecules 2013, 14, 2383-2388; b) B. A. Kellogg, L. Garrett, Y.
Kovtun, K. C. Lai, B. Leece, M. Miller, G. Payne, R. Steeves, K. R.
Whiteman, W. Widdison, H. Xie, R. Singh, R. V. Chari, J. M. Lambert, R.
J. Lutz, Bioconjugate Chem. 2011, 22, 717-727.
[5] D. Zhang, T. H. Pillow, Y. Ma, J. d. Cruz-Chuh, K. R. Kozak, J. D.
Sadowsky, G. D. Lewis Phillips, J. Guo, M. Darwish, P. Fan, J. Chen, C.
He, T. Wang, H. Yao, Z. Xu, J. Chen, J. Wai, Z. Pei, C. E. C. A. Hop, S.
C. Khojasteh, P. S. Dragovich, ACS Med. Chem. Lett. 2016, 7, 988-993.
[6] a) W. C. Widdison, J. F. Ponte, J. A. Coccia, L. Lanieri, Y. Setiady, L.
Dong, A. Skaletskaya, E. E. Hong, R. Wu, Q. Qiu, R. Singh, P. Salomon,
N. Fishkin, L. Harris, E. K. Maloney, Y. Kovtun, K. Veale, S. D. Wilhelm,
C. A. Audette, J. A. Costoplus, R. V. Chari, Bioconjugate Chem. 2015,
26, 2261-2278; b) X. Sun, W. Widdison, M. Mayo, S. Wilhelm, B. Leece,
R. Chari, R. Singh, H. Erickson, Bioconjugate Chem. 2011, 22, 728-735.
[7] a) J. D. Sadowsky, T. H. Pillow, J. Chen, F. Fan, C. He, Y. Wang, G. Yan,
H. Yao, Z. Xu, S. Martin, D. Zhang, P. Chu, J. Dela Cruz-Chuh, A.
O'Donohue, G. Li, G. Del Rosario, J. He, L. Liu, C. Ng, D. Su, G. D.
Lewis Phillips, K. R. Kozak, S. F. Yu, K. Xu, D. Leipold, J. Wai,
Bioconjugate Chem. 2017, 28, 2086-2098; b) S. Maiti, N. Park, J. H. Han,
H. M. Jeon, J. H. Lee, S. Bhuniya, C. Kang, J. S. Kim, J. Am. Chem. Soc.
2013, 135, 4567-4572; c) Z. Pei, C. Chen, J. Chen, J. D. Cruz-Chuh, R.
Delarosa, Y. Deng, A. Fourie-O'Donohue, I. Figueroa, J. Guo, W. Jin, S.
C. Khojasteh, K. R. Kozak, B. Latifi, J. Lee, G. Li, E. Lin, L. Liu, J. Lu, S.
Martin, C. Ng, T. Nguyen, R. Ohri, G. Lewis Phillips, T. H. Pillow, R. K.
Rowntree, N. J. Stagg, D. Stokoe, S. Ulufatu, V. A. Verma, J. Wai, J.
Wang, K. Xu, Z. Xu, H. Yao, S. F. Yu, D. Zhang, P. S. Dragovich, Mol.
Pharm. 2018, 15, 3979-3996.
[8] T. H. Pillow, J. D. Sadowsky, D. Zhang, S. F. Yu, G. Del Rosario, K. Xu,
J. He, S. Bhakta, R. Ohri, K. R. Kozak, E. Ha, J. R. Junutula, J. A.
Flygare, Chem. Sci. 2017, 8, 366-370.
[9] C. D. Austin, X. Wen, L. Gazzard, C. Nelson, R. H. Scheller, S. J. Scales,
Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 17987-17992.
[10] a) N. Joubert, C. Denevault-Sabourin, F. Bryden, M. C. Viaud-Massuard,
Eur. J. Med. Chem. 2017, 142, 393-415; b) I. R. Vlahov, C. P. Leamon,
Bioconjugate Chem. 2012, 23, 1357-1369.
[11] a) L. H. Qian, J. Q. Fu, P. Y. Yuan, S. B. Du, W. Huang, L. Li, S. Q. Yao,
Angew. Chem. Int. Ed. 2018, 57, 1532-1536; b) T. Sun, A. Morger, B.
Castagner, J. C. Leroux, Chem. Commun. 2015, 51, 5721-5724; c) P. D.
Senter, W. E. Pearce, R. S. Greenfield, J. Org. Chem 1990, 55, 2975-
2978.
[12] R. Freter, E. R. Pohl, J. M. Wilson, D. J. Hupe, J. Org. Chem. 1979, 44,
1771-1774.
[13] E. K. Lei, S. O. Kelley, J. Am. Chem. Soc. 2017, 139, 9455-9458.
[14] L. Greenfield, W. Bloch, M. Moreland, W. B. Lawrence Greenfield, and
Margaret Moreland, Bioconjugate Chem. 1990, 1, 400-410.
[15] Y. Zheng, L. Zhai, Y. Zhao, C. Wu, J. Am. Chem. Soc. 2015, 137, 15094-
15097.
[16] C. Wu, C. Belenda, J. C. Leroux, M. A. Gauthier, Chem. Eur. J. 2011, 17,
10064-10070.
[17] J. D. Gough, W. J. Lees, J. Biotechnol. 2005, 115, 279-290.
[18] S. Sabharanjak, S. Mayor, Adv. Drug. Deliv. Rev. 2004, 56, 1099-1109.
[19] J. Yang, H. Chen, I. R. Vlahov, J. X. Cheng, P. S. Low, Proc. Natl. Acad.
Sci. U. S. A. 2006, 103, 13872-13877.
[20] J. Li, E. A. Sausville, P. J. Klein, D. Morgenstern, C. P. Leamon, R. A.
Messmann, P. LoRusso, J. Clin. Pharmacol. 2009, 49, 1467-1476.
[21] A. Maderna, C. A. Leverett, Mol. Pharm. 2015, 12, 1798-1812.
[22] J. D. Gough, R. H. Williams, A. F. Donofrio, W. J. Lees, J. Am. Chem.
Soc. 2002, 124, 3885-3892.
[23] L. Brulisauer, G. Valentino, S. Morinaga, K. Cam, J. Thostrup Bukrinski,
M. A. Gauthier, J. C. Leroux, Angew. Chem. Int. Ed. 2014, 53, 8392-
8396.
[24] G. Casi, D. Neri, Mol. Pharm. 2015, 12, 1880-1884.
[25] E. Perrino, M. Steiner, N. Krall, G. J. Bernardes, F. Pretto, G. Casi, D.
Neri, Cancer Res. 2014, 74, 2569-2578.
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900248
ChemMedChem
FULL PAPER
Entry for the Table of Contents
Insert graphic for Table of Contents here. ((Please ensure your graphic is in one of following formats))
Insert text for Table of Contents here.
We examined the effect of two methyl groups adjacent to the p-thiobenzyl-based disulfides on their thiol-disulfide exchange reactions
and elucidated the mechanism of the p-dithiobenzyl-based disulfides reduction and self-immolation. We observed a central-leaving
group shifting effect in these p-thiobenzyl-based disulfide linkers, which leads to >2 orders of magnitude increase in disulfide stability.
This work thus not only reveals a central-leaving group shifting effect for stabilizing p-thiobenzyl-based disulfide bonds, but provides
useful self-immolating disulfide linkers for targeted cytotoxic prodrug designs.
((The Table of Contents text should give readers a short preview of the main theme of the research and results included in the paper
to attract their attention into reading the paper in full. The Table of Contents text should be different from the abstract and should
be no more than 450 characters including spaces.))
9
This article is protected by copyright. All rights reserved.