Angewandte
Communications
Chemie
including a single-electron transfer, radical opening of the
cyclopropyl ring, and formation of a covalent bond with
FAD.[5] During the inactivation of LSD1, the imine inter-
mediate is hydrolyzed, which leads to the extrusion of the
nitrogen atom of PCPA in the form of ammonia. Based on the
mechanism of the PCPA-induced inhibition of LSD1, we
propose that PCPA-drug conjugates (PDCs) should be able to
target cancer cells, in which LSD1 is highly expressed
(Figure 1b). As shown in Figure 1b, PDCs should be recog-
nized by LSD1 and deactivate it in a similar manner to PCPA
itself, that is, through a single-electron transfer mechanism.
Subsequently, the drug should be released together with the
linker moiety of the PDCs through hydrolysis of the imine
intermediate, and an ensuing intramolecular cyclization
should eventually separate the linker from the drug. Thus,
PDCs could serve as prodrugs that selectively release an
anticancer drug upon binding to LSD1. This method would
induce significantly lower levels of side effects, as such
molecules are inactive against normal cells, where the
expression of LSD1 is lower. Furthermore, PDCs can expect
the combined effect of LSD1 inhibition and the release of an
anticancer drug to further increase their anticancer activity.
As a proof-of-concept study, we designed PCPA-tamox-
ifen conjugates 1a and 1b (Figure 2) as PDC prototypes.
These consist of PCPA, a linker, and 4-hydroxytamoxifen
(4OHT),[6] that is, an anti-estrogen agent for breast cancer
treatment (Figure 2a). According to the mechanism shown in
Figure 2b, we expected 1a and 1b to exhibit anticancer
Scheme S1 in the Supporting Information shows the
synthetic route to 1a and 1b, which were obtained as
diastereomers, as one isomer of 4OHT easily isomerizes
under physiological conditions.[7]
Initially, we examined the recognition of 1a and 1b by
LSD1 and their ability to inhibit the catalytic activity of the
enzyme. As expected, the inhibitory activity toward LSD1
was much higher for 1a and 1b relative to PCPA (IC50 values:
PCPA, 24.8 mm; 1a, 0.339 mm; 1b, 0.155 mm; Supporting
Information, Table S1). In addition, 1a and 1b exhibited
weak activity toward other FAD-dependent monoamine
oxidases such as MAOA and MAOB (Supporting Informa-
tion, Table S1). These results suggest that 1a and 1b are
recognized selectively by LSD1, thereby effectively achieving
its inhibition. Moreover, a kinetic analysis suggested that this
inhibition is irreversible (Figure 2a).[5] The kinetic analysis,
performed using two different substrate concentrations,
exhibited nonlinear progress curves, which ultimately reach
a plateau, thus indicating that 1a and 1b inhibit LSD1 in
a time-dependent manner (Supporting Information, Figur-
es S5 and S6). Furthermore, significantly higher kinact/Ki
values were observed for 1a and 1b relative to those of
PCPA, which demonstrates the highly selective recognition
and inhibition of LSD1 by 1a and 1b (Supporting Informa-
tion, Table S2). Subsequently, we carried out a MALDI-TOF-
MS analysis to examine the potential inhibition of LSD1 by
1a and 1b through the formation of PCPA-FAD adducts
(Figure 2a). Peaks at m/z = 918 and 900, corresponding to the
PCPA-FAD adduct and its dehydrated form, respectively
(Figure S7b,d), were observed in the presence of LSD1 but
not in its absence (Supporting Information, Figure S7a,c).
The results obtained from the kinetic and MALDI-TOF-MS
analyses are thus consistent with the irreversible inhibition of
LSD1 by 1a and 1b through the formation of PCPA-FAD
adducts as shown in Figure 2a.
To confirm that the release of 4OHT is triggered through
the inhibition of LSD1 by 1a and 1b, we also carried out an
ESI-MS analysis. If 1a and 1b engage with FAD on the active
site of LSD1, 4OHT should be released upon formation of the
PCPA-FAD adduct (Figure 2a). As expected, the release of
4OHT was detected by ESI-MS in the presence of LSD1
(Supporting Information, Figure S8b,d) and not observed in
its absence (Supporting Information, Figure S8a,c). More-
over, the release was found to be time-dependent (Fig-
ure 3a,b and the Supporting Information, Figure S8) and
significantly suppressed by known LSD1 inhibitors such as
PCPA and NCD38[4b] (Figure 3c and the Supporting Infor-
mation, Figure S9), which provides further support for our
hypothesis that the release of 4OHT depends on the enzyme
activity of LSD1.
Because of the simultaneous inhibition of LSD1 and ERa
(Figure 2b), 1a and 1b were expected to inhibit the growth of
breast cancer cells. To evaluate the in-cell activity of 1a and
1b, we selected ERa-positive breast cancer MCF7 cells, in
which LSD1 is overexpressed (Supporting Information, Fig-
ure S2a).
The level of the LSD1 substrate H3K4me2 present in
MCF7 cells upon treatment with 1a and 1b was examined by
western blotting analysis. As shown in Figure 4a, a dose-
Figure 2. a) Design of PCPA-tamoxifen conjugates 1a and 1b.
b) Scheme for the expected mechanism of action for 1a and 1b in
LSD1/ERa-positive breast cancer cells.
activity, and theoretical simulations suggested that the
recognition of 1a and 1b by LSD1 should be effective, as
isomers of the conjugates are likely to fit well within the active
pocket of LSD1 (Supporting Information, Figure S4). Sub-
sequently, 4OHT should be released in LSD1-expressing
breast cancer cells to antagonize the estrogen receptor
a (ERa). Conjugates 1a and 1b should furthermore induce
a synergistic anticancer effect by the simultaneous inhibition
of LSD1 and ERa, as the interaction of LSD1 and ERa in
breast cancer cells promotes cell growth.[3g,h]
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!