Design and synthesis of a MAO-B-selectively activated prodrug based on MPTP: A mitochondria-targeting chemotherapeutic agent for treatment of human malignant gliomas

Article abstract of DOI:10.1002/cmdc.201402562

Malignant gliomas, including glioblastomas, are extremely difficult to treat. The median survival for glioblastoma patients with optimal therapeutic intervention is 15 months. We developed a novel MAO-B-selectively activated prodrug, N,N-bis(2-chloroethyl)-2-(1-methyl-1,2,3,6-tetrahydropyridin-4-yl)propanamide (MP-MUS), for the treatment of gliomas based on 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). The design of neutral MP-MUS involved the use of a seeker molecule capable of binding to mitochondrial MAO-B, which is up-regulated ≥fourfold in glioma cells. Once the binding occurs, MP-MUS is converted into a positively charged moiety, P⁺-MUS, which accumulates inside mitochondria at a theoretical maximal value of 1000:1 gradient. The LD₅₀ of MP-MUS against glioma cells is 75 μM, which is two- to threefold more potent than temozolomide, a primary drug for gliomas. Importantly, MP-MUS was found to be selectively toxic toward glioma cells. In the concentration range of 150-180 μM MP-MUS killed 90-95 % of glioma cells, but stimulated the growth of normal human astrocytes. Moreover, maturation of MP-MUS is highly dependent on MAO-B, and inhibition of MAO-B activity with selegiline protected human glioma cells from apoptosis.

Full text of DOI:10.1002/cmdc.201402562

DOI: 10.1002/cmdc.201402562
Full Papers
Design and Synthesis of a MAO-B-Selectively Activated
Prodrug Based on MPTP: A Mitochondria-Targeting
Chemotherapeutic Agent for Treatment of Human
Malignant Gliomas
Martyn A. Sharpe,[a] Junyan Han,[a] Alexandra M. Baskin,[b] and David S. Baskin*[a]
Malignant gliomas, including glioblastomas, are extremely diffi-
cult to treat. The median survival for glioblastoma patients
with optimal therapeutic intervention is 15 months. We devel-
oped a novel MAO-B-selectively activated prodrug, N,N-bis(2-
chloroethyl)-2-(1-methyl-1,2,3,6-tetrahydropyridin-4-yl)propana-
mide (MP-MUS), for the treatment of gliomas based on 1-
methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). The design
of neutral MP-MUS involved the use of a seeker molecule capa-
ble of binding to mitochondrial MAO-B, which is up-regulated
ꢀfourfold in glioma cells. Once the binding occurs, MP-MUS is
converted into a positively charged moiety, P+-MUS, which ac-
cumulates inside mitochondria at a theoretical maximal value
of 1000:1 gradient. The LD50 of MP-MUS against glioma cells is
75 mm, which is two- to threefold more potent than temozolo-
mide, a primary drug for gliomas. Importantly, MP-MUS was
found to be selectively toxic toward glioma cells. In the con-
centration range of 150–180 mm MP-MUS killed 90–95% of
glioma cells, but stimulated the growth of normal human as-
trocytes. Moreover, maturation of MP-MUS is highly dependent
on MAO-B, and inhibition of MAO-B activity with selegiline pro-
tected human glioma cells from apoptosis.
Introduction
Gliomas, or glial tumors, are the most common primary brain
tumors in humans[1] and include a group of neoplasms with
distinct histological and genetic features. Among them, glio-
blastomas (GBMs) and anaplastic astrocytomas are the most
aggressive primary malignant tumors, with an annual inci-
dence of 5.3 per 100000, or 17000 new diagnoses per year in
the USA.[2] Temozolomide (TMZ) is the primary and most com-
monly used chemotherapeutic drug for the treatment of glio-
mas. TMZ has poor therapeutic indices; it kills healthy tissues
almost as effectively as tumor tissues.[3,4] Its efficacy is limited,
as the median life expectancy for glioma patients after radia-
tion therapy and TMZ treatment is 15 months.[4,5] Therefore, it
is highly desirable to develop new antineoplastic drugs for the
treatment of malignant gliomas. Herein we describe the use of
a prodrug, MP-MUS, based on 1-methyl-4-phenyl-1,2,3,6-tetra-
hydropyridine (MPTP), which is selectively activated by mono-
amine oxidase B (MAO-B). Data are presented to demonstrate
MP-MUS is two- to threefold more potent against glioma cells
than TMZ. Importantly, MP-MUS is selectively activated by
glioma cells and accumulates inside gliomal mitochondria, re-
sulting in the selective destruction of cancer cells without
causing damage to normal cells.[6] In comparison with TMZ,
MP-MUS shows promising features for the treatment of glio-
mas; to our knowledge, it is the first known example of selec-
tive mitochondrial chemotherapy.[7]
Monoamine oxidases (MAOs) are flavoenzymes that are lo-
cated on the outer mitochondrial membranes. Two MAO iso-
forms, MAO-A and MAO-B, are found in human body. The two
enzymes are differentiated by their tissue distribution and sub-
strate specificity (see review[8]). The specific activity of MAO-B
is higher in human brain than in other organs, such as kidney,
liver, heart, and lungs.[9] MAO-B/-A ratios are found to be >14
throughout white matter, and between 5.9 and 9.0 in gray
matter in human brain tissue,[10] with MAO-B primarily found in
non-neuronal cells such as astrocytes and radial glia; it is, how-
ever, present in serotonergic neurons.[11] Brain MAOs modulate
neurotransmitter levels through oxidative deamination of
a number of amines, generating the corresponding aldehydes
and hydrogen peroxide.
[a] Prof. Dr. M. A. Sharpe,+ Dr. J. Han,+ Prof. Dr. D. S. Baskin
Kenneth R. Peak Brain and Pituitary Tumor Center
Department of Neurological Surgery
Neurological Institute, Houston Methodist Hospital
Weill Cornell Medical College of Cornell University
6565 Fannin Street, Houston, TX, 77030 (USA)
E-mail: dbaskin@houstonmethodist.org
By measuring MAO oxidation of phenylethylamine in the ab-
sence or presence of the MAO-B-specific inhibitor l-deprenyl
(selegiline), MAO-B activity was found to be significantly up-
regulated in glial tumors relative to control brain tissue or
non-glial brain tumors.[12] Immuno-histological studies in our
research group have confirmed this observation (unpublished
results). We speculate that glioma-targeting prodrugs could be
[b] A. M. Baskin
St. Johns School, 2401 Claremont Lane, Houston, TX, 77019 (USA)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402562: characterization data for
compounds 1–3, MP-Est, and MP-MUS.
ChemMedChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
developed by using the oxidation capacities of MAO-B. The dif-
ference in substrate pocket shape between the two MAO iso-
forms provides feasibility to develop a prodrug which will be
selectively metabolized by MAO-B over MAO-A, resulting in
fewer side effects.[13]
moiety for MAO-B binding, and MUS is the cytotoxic fragment
capable of alkylating a wide range of biomolecules, such as
DNA. We proposed that MP-MUS should be oxidized by MAO-
B to MD+-MUS, and then further oxidized into the mature
drug, P+-MUS.[24] Lipophilic cations such as MPP+ preferentially
accumulate in the mitochondrial matrix of cells, driven by the
mitochondrial membrane potential; it has been demonstrated
that in energized mitochondria MPP+ displays classical Nerns-
tian behavior,[18] and P+-MUS has been designed to act in the
same manner. The steady-state accumulation of P+-MUS inside
mitochondria can be estimated from the Nernst equation
[Equations (1) and (2)]:[19]
Results and Discussion
Design and synthesis of prodrug MP-MUS
We designed the MAO-B-activated prodrug MP-MUS based on
a known transformation of a nontoxic compound, MPTP, into
the neurotoxin MPP+ (Figure 1a). Mice[14] and primates[15] have
high levels of MAO-B in brain tissue, and are therefore sensitive
to MPTP. Nontoxic MPTP is oxidized by MAO-B or human cyto-
chrome P450 2D6[16] to form MDP+, which undergoes further
oxidation (typically by the mitochondrial quinone pool[17]) to
form the neurotoxin MPP+ (Figure 1a). MPP+ is a substrate for
the dopamine transporter (DAT), which transports MPP+ inside
dopaminergic neurons. As a lipophilic cation, MPP+ diffuses
across the mitochondrial membrane in a Nernstian fashion,[18,19]
and inhibits mitochondrial respiration by interfering with com-
plex I of the electron-transport chain, resulting in neuron
damage, selective loss of dopaminergic neurons, and Parkinso-
nian syndrome.[20] 1-Methyl-1,2,3,6-tetrahydropyridine (MTP)
and some of its derivatives, such as MPTP, are known selective
MAO-B substrates.[21] These MTP substrates could be oxidized
by MAO-B to form pyridinium cations, which accumulate in mi-
tochondria. If covalently conjugated to a cytotoxic DNA-alkylat-
ing moiety, specific MTP substrates can act as anti-glioma pro-
drugs activatable by MAO-B,[22] exhibiting both antitumor activ-
ity through alkylation of mitochondria DNA (mtDNA) and low
toxicity toward normal tissues.
ꢀ
ꢁ
ꢀ
ꢁ
ꢁ
2:3 RT
mF
½Xmþꢁ
½Xmþꢁoiunt
DY ¼
log10
ð1Þ
ð2Þ
ꢀ
½Pþ-MUSꢁin
½Pþ-MUSꢁout
180 mV ¼ 60 log10
in which DY is mitochondrial membrane potential, R is the
gas constant, T is temperature, m is the charge number, and F
is the Faraday constant. In glioma cells, which have a mitochon-
drial membrane potential of ꢂ180–200 mV,[25] the concentra-
tion of [P+-MUS]in will be theoretically ꢂ1000-fold greater than
[P+-MUS]out
.
Nitrogen mustards typically attack DNA and other important
biomolecules via the highly reactive three-membered aziridini-
um ring (Supporting Information (SI), Figure S1).[26] However,
the presence of an amide in MP-MUS allows the formation of
a five-membered dihydrooxazolium ring capable of alkylating
mitochondria DNA and inducing apoptosis. (SI Figure S1). Spe-
cific accumulation of P+-MUS in the mitochondria of MAO-B-
overexpressing glioma cells, but not in those of normal non-
cancerous cells, is a function of the mature drug.
Like MPP+, P+-MUS could be a substrate for DAT. Therefore,
MP-MUS chemotherapy could kill dopaminergic neurons, re-
sulting in Parkinson’s disease. To determine whether
M+-MUS is a good substrate for DAT, we conducted
modeling studies based on the nortriptyline-inhibit-
ed form of the crystal structure of the Drosophila
melanogaster dopamine transporter (PDB ID:
4M48;[27] SI Figure S2). Modeling studies suggest that
P+-MUS is too large to be a DAT substrate. The DAT
transporting channel is like a constraining ring above
the sodium ions which bind the amine of a sub-
strate/inhibitor. S422 plays a crucial role in DAT sub-
strate selectivity, as it can block of the movement of
large amines through the DAT channel. Therefore,
the S422A mutant form, which has a slightly larger
constraining ring, has a far wider substrate specificity
than the wild-type.[28] In Supporting Information Fig-
ures S2A and S2B we show a model of the top and
side views of substrates, dopamine and MPP+, along
We chose nitrogen mustard (MUS),[23] a traditional antineo-
plastic DNA alkylation drug used since 1943, as a prototype
drug for conjugation to MTP. This led to our first functional
MAO-B-activated prodrug, MP-MUS (Figure 1b). MTP is the
Figure 1. Design of MAO-B-activated prodrug MP-MUS based on MPTP: a) MPTP toxicity
mechanism; b) design of the MP-MUS prodrug. MPTP=(1-methyl-4-phenyl-1,2,3,6-tetra-
hydropyridine; MDP+ =1-methyl-4-phenyl-2,3-dihydropyridinium; MPP+ =1-methyl-4-
phenylpyridinium.
with P+-MUS and the inhibitor nortriptyline in the
4M48 DAT crystal structure. Unlike dopamine and
MPP+, the dibenzocycloheptene ‘parasol’ of nortrip-
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
tyline is too wide to go through the DAT annular constraining
ring (SI Figure S2C(I)). The chloroethyl ‘wings’ of P+-MUS make
this compound unlikely to move through this annular ring
either, being blocked by S422 (SI Figure S2C(II)). In addition, P+
-MUS is too large to fit into the substrate binding pocket of
the 4M48 structure, overlapping with the valine residue that
constrains substrate length.
spectively.[24] MP-Est and MP-MUS (2 mm) were incubated with
either MAO-A (14 UmLꢃ1, 94 mgmLꢃ1) or MAO-B (7 UmLꢃ1
,
140 mgmLꢃ1) in 1 mL buffer (pH 7.4) containing 50 mm KPi and
100 mm NaCl at 378C; UV absorbance spectra were taken
every 5 min using a BioTeck Synergy HT spectrophotometer.
We observed large light scattering during the incubation of
MP-MUS at 2 mm with the partly purified MAOs, but especially
with MAO-B, which made UV spectral changes very difficult to
interpret. However, we were able to observe the formation of
a dihydropyridinium peak centered at 380 nm using 600 mm
MP-MUS. No such light-scattering changes were observed
using unreactive MP-Est, which produced UV signatures of
both expected products. The real-time UV absorbance changes
during the oxidation of MP-Est with MAO-A/-B were measured
over a period of 45 min. As shown in Figure 2, both enzymes
The synthesis of MP-MUS is illustrated in Scheme 1. The
Horner–Wadsworth–Emmons reaction of N-methylpiperidone
with triethyl 2-phosphonopropionate produced the a,b-conju-
Scheme 1. Synthesis of MP-MUS. Reagents and conditions: a) NaH, THF, 208C,
overnight; b) LDA, THF, ꢃ728C, 1 h; c) KOH (2n), H2O, 608C, 2 h; d) SOCl2,
708C, 1 h; e) diethanolamine, CH2Cl2, 208C, 1 h; f) SOCl2, benzene, 808C, 5 h.
LDA=lithium diisopropylamide; DMF=N,N’-dimethylformamide.
gated ester 1. Treatment of compound 1 with LDA (THF,
ꢃ728C, 1 h), followed by quenching with NH4Cl, yielded the
critical b,g-unsaturated ester MP-Est, a racemic compound, in
nearly quantitative yield. MP-Est was hydrolyzed (2 equiv 1n
KOH, 2 h, 608C) to form acid 2. Under these harsh hydrolysis
conditions, the double bond of MP-Est was not isomerized, as
Figure 2. Real-time UV absorbance spectra of the reaction between MAO-B
(left, 7 UmLꢃ1) or MAO-A (right, 14 UmLꢃ1) and MP-Est (2.0 mm) over the
course of 45 min.
produced a significant amount of the first and final oxidation
products MD+-Est (dihydropyridinium, 365 nm) and P+-Est
(pyridium, 292 nm), respectively (two structures shown in SI
Figure S3). The substantially larger absorbance peak exhibited
during the MAO-B oxidation than with MAO-A, despite a two-
fold higher MAO-A concentration, indicated that oxidation of
MP-Est by MAO-B is much faster than by MAO-A.
1
indicated by the H NMR spectra of 2 and MP-Est (SI). Carboxyl-
ic acid 2 was converted into acid chloride with SOCl2, then
joined diethanolamine (1 equiv) to afford crude amide 3.[20] Fi-
nally, the racemic prodrug MP-MUS was synthesized by con-
verting the dihydroxy groups in compound 3 into dichlorine
with SOCl2.[20] Direct coupling of compound 2 to diethanola-
mine via other common coupling reagents, such as DCC or
BOP, could not be satisfactorily accomplished probably as
a result of the low activity of the carboxylic acid.
Kinetics of MAO-A/-B on substrates tyramine, MP-Est, and
MP-MUS
The kinetics of MP-MUS oxidation catalyzed by MAO-A and -B
were measured by monitoring the rate of hydrogen peroxide
generation with the Amplex Red/horse radish peroxidase (HRP)
system.[30] H2O2 is one of the key products generated during
MAO-catalyzed oxidative deamination of a number of primary
amines in the brain.[31] The rates of H2O2 generation during the
oxidation of substrates tyramine (Tyr), MP-Est, and MP-MUS by
either MAO-A or MAO-B are shown in Figure 3. Tyr is a known
nonselective substrate;[32] therefore, it is expected that the
H2O2 generation rate for Tyr metabolism by MAO-A is similar to
that by MAO-B. This is exactly what we measured, as the Vmax
of MAO-A is the same as that of MAO-B, indicating both en-
UV absorbance spectra changes in the reaction of MP-Est
with MAO-A/-B
Partially purified recombinant human MAO-A and MAO-B
(Sigma–Aldrich; M7316 (150 Umgꢃ1), and M7441 (52 Umgꢃ1))
were used to examine the enzyme specificity of MP-MUS and
its precursors. Using high chloride buffer to decrease the reac-
tivity of MP-MUS,[29] we attempted to monitor the oxidation of
MP-Est and MP-MUS by both MAOs, observing the formation
of the oxidation intermediate 1-methyl-2,3-dihydropyridinium
and the final oxidation product 1-methyl-pyridinium, which
have UV absorbance spectra peaking at ~365 and 292 nm, re-
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Table 1. Parameters for the kinetics of oxidation of tyramine (Tyr), benzyl-
amine (BA), 4-fluorobenzylamine (4-FBA), MP-Est, and MP-MUS by MAO-
A/MAO-B.
Substrate
MAO-A[a]
MAO-B[a]
B/A[b]
KM
Vmax
Vmax/KM
KM
Vmax
Vmax/KM
Tyr
BA
4-FBA
MP-Est
MP-MUS
0.11
–
–
–
0.3
1.33
<0.01
<0.01
<0.01
0.05
12.1
–
–
–
0.17
0.35
0.12
0.12
1.86
0.27
1.39
2.3
2.4
1.58
1.39
3.97
19.2
20
0.85
5.15
0.33
–
–
–
30.3
[a] Units: KM [mm], Vmax [nmUꢃ1 sꢃ1], Vmax/KM [10ꢃ6 Uꢃ1 sꢃ1]. [b] Ratio of Vmax
KM for MAO-B/MAO-A.
/
30-fold faster by MAO-B than MAO-A. The much more efficient
oxidation of MP-MUS by MAO-B than MAO-A suggests that
MP-MUS preferentially gains access to the larger substrate
cavity of MAO-B,[34] resulting in a MAO-B-selective substrate.[35]
Cytotoxicity of MP-MUS on human glioma cells and normal
human astrocytes
Finally, we investigated the cytotoxic effect of MP-MUS on
human primary glioma cells and normal human astrocytes
(NHAs), the latter of which are the most abundant normal glial
cells in human brains. MP-MUS showed distinct toxic effects
toward glioma cells and NHAs. The toxic effects of dose escala-
tion of MP-MUS on glioma cells with 48 h treatment are shown
in Figure 4a. The number of viable cells decreased with in-
creasing MP-MUS concentration. If [MP-MUS] is >150 mm,
<13% glioma cells remain alive. The LD50 of MP-MUS against
glioma cells is ~75 mm. In comparison with halting cell prolifer-
ation, MP-MUS caused the opposite pattern of dead cell frac-
tions (blue trace in Figure 4a). TMZ is the most commonly
used chemotherapeutic drug for the clinical treatment of
glioma, and has an LD50 value of ~150–200 mm.[36] MP-MUS is
about two- to threefold more potent against glioma cells than
TMZ. The effects of dose escalation of MP-MUS on NHAs are
shown in Figure 4b. MP-MUS showed not only an absence of
cytotoxicity toward normal human astrocytes up to 180 mm,
but stimulated cell growth, possibly due to the generation of
mitogen H2O2 in the oxidation process.[37] At an MP-MUS con-
centration of 210 mm, some 86% NHAs survived, whereas the
same treatment caused 95% gliomal cell death. The distinct
cytotoxicity effects of MP-MUS on human primary glioma cells
and not NHAs were further validated by fluorescence micros-
copy. As shown in Figure 4c (I–III), treatment of GBMs with
MP-MUS (90 mm) induced apoptosis, as indicated by shrunken
cytoplasm, condensed nuclei, and disrupted mitochondria. At
210 mm MP-MUS, there were no indications of morphological
changes related to apoptosis in NHAs, whereas the gliomal
cells are clearly in a pathological state (Figure 4c (VI–IV)).
Figure 3. Rate of H2O2 production by MAO-A/-B activities with substrates
a) Tyr and MP-Est and b) MP-MUS, at the indicated concentration ranges.
[Concentration series are different between panels a) and b).] Data are the
meanꢄSD of n=3 experiments.
zymes can oxidize Tyr efficiently. The KM for MAO-A is slightly
smaller than for MAO-B; therefore, MAO-A oxidizes Tyr faster
than MAO-B. In addition, Figure 3a shows MP-Est to have been
rapidly oxidized by MAO-B, but very slowly by MAO-A. Finally,
the data presented in Figure 3b demonstrate that MP-MUS is
a remarkably good substrate for MAO-B, but a very poor MAO-
A substrate. Pre-incubating MAO-B with the inhibitor selegiline
caused a block in MAO-B-catalyzed H2O2 generation by MP-
MUS and other substrates.
The data shown in Figure 3 and those of other kinetic stud-
ies were fitted to the Michaelis–Menten equation, and the pa-
rameters, such as KM, Vmax, and Vmax/KM, were calculated and are
presented in Table 1. Vmax/KM is a measure of how efficiently
MAOs convert a substrate into the corresponding oxidization
products. As for the substrate Tyr, Vmax/KM values for MAO-A
and MAO-B are 12.1 and 3.97ꢁ10ꢃ6 Uꢃ1 sꢃ1 respectively, indicat-
ing both enzymes oxidized Tyr efficiently, but MAO-A did this
threefold faster than MAO-B. The MAO-B-specific substrates BA
and 4-FBA were only oxidized by MAO-B, and no activity was
observed with MAO-A.[33] MAO-B oxidized MP-MUS with Vmax
/
To determine whether the differential toxic effects of MP-
MUS on glioma cells and NHAs are caused by the difference in
intracellular MAO-B levels, we first measured the concentration
of MAO-B in both cells via measurement of fluorescence inten-
sity of Alexa Fluor 488-labeled anti-MAO-B antibody. The data
KM at 5.15ꢁ10ꢃ6 Uꢃ1 sꢃ1, a slightly higher rate than for Tyr. It is
clear that MP-MUS is poorly oxidized by MAO-A, with Vmax/KM
at 1.7ꢁ10ꢃ7 Uꢃ1 sꢃ1. The Vmax/KM ratio of MAO-B/-A is 30.3ꢁ
10ꢃ6 Uꢃ1 sꢃ1, indicating that MP-MUS is metabolized at a rate
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
tion to generate the mature
drug pyridinium P+-MUS, which
alkylates mtDNA and induces
cell death. The enzyme kinetics
measured by monitoring the
rate of H2O2 generation showed
that MAO-B oxidized MP-MUS
to form the toxic mature drug
30 times as efficient as MAO-A.
This indicates that MP-MUS can
be selectively converted into
toxic P+-MUS by MAO-B. We
demonstrated that MP-MUS is
capable of selectively killing pri-
mary human glioma cells with
an LD50 value of 75 mm at 48 h,
while causing little damage to
NHA. The relatively high doses
of MP-MUS are needed to kill
primary human glioma possibly
because of the non-enzymatic
hydrolysis of the mustard
groups, with the chlorines
being replaced by hydroxy
groups, generating nontoxic al-
cohols. Nevertheless, our MP-
MUS showed greater potency
against brain glioma cells than
TMZ, which has an LD50 value in
the range of 150–200 mm.[36] In
addition, the specific MAO-B in-
Figure 4. Studies of MAO-B-dependent cytotoxicity of MP-MUS. The cytotoxicity effects of MP-MUS (0–210 mm,
48 h) on a) glioma cells (GBMs) and b) normal human astrocytes (NHAs). Data are the meanꢄSEM of n=8 experi-
ments. c) Fluorescence images of human glioma cells (I, II, III) and NHAs (VI, V, IV) after treatment with MP-MUS at
the indicated concentrations. Red and blue are signals from MitoTracker Red FM and nucleus marker Hoechst
33342, respectively. d) Fluorescence signals of MAO-B antibody in I) glioma cells and II) NHAs; III) 5ꢁ green signals
in (II). e) Effects of the MAO-B inhibitor selegiline (10 mm) on the toxicity of MP-MUS (210 mm) against glioma cells
for 48 h; data are the meanꢄSEM of n=8 experiments.
presented in Figure 4d suggest that NHAs expressed ꢂ5-fold
less MAO-B than glioma cells, which is in good agreement
with reported data.[12] To further study the MAO-B-dependent
toxicity of MP-MUS, we exposed human primary glioma cells
to selegiline prior to treatment with MP-MUS at 210 mm. With-
out selegiline, MP-MUS was found to be highly toxic, and only
5% of cells survived a 48 h incubation (middle column in Fig-
ure 4e), whereas 83% of glioma cells were viable with selegi-
line pre-treatment. Incubation of cells with the MAO-A inhibi-
tor clorgyline had little effect on the toxicity of MP-MUS in
a parallel experiment (data not shown). Our results demon-
strate that cells lacking high MAO-B activity, such as a patient’s
noncancerous cells/tissues, have significantly lower sensitivity
to MP-MUS toxicity.
hibitor selegiline was shown to protect glioma cells from MP-
MUS, indicating the action of our prodrug toxicity is highly
contingent on the activity of MAO-B. Importantly, we demon-
strated that MP-MUS is selectively toxic toward glioma cells,
but not toxic to NHAs, which have lower MAO-B expression
levels.
Treatment of gliomas with our new drug MP-MUS in animal
studies has demonstrated promising results, and will be report-
ed soon. We envision that MP-MUS, and similar bifunctional
mitochondrial-targeting MAO-B-activated prodrugs, will be
a promising clinical strategy for the treatment of human glio-
mas.
Experimental Section
Reagents and materials: Tested compounds including tyramine
(Tyr) and selegiline were purchased from Sigma–Aldrich with purity
>95% and were used directly without further purification. MP-Est
was synthesized with >95% purity, as determined by a high-pres-
sure liquid chromatography (HPLC) system (Phenomenex Luna C-
18(2), 250 mmꢁ10 mm, 5 mm) mobile phase: A=H2O, B=MeOH;
flow rate: 4.0 mLminꢃ1; gradient: 10!82% B in 9.0 min. MAO-A
and MAO-B were obtained from Sigma (M7316, batch SLBJ7414V,
M7441, batch SLBJ6906V). MP-MUS was synthesized with purity
>95% as determined by HPLC (Phenomenex Luna C-18(2),
250 mmꢁ4.6 mm, 5 mm) mobile phase: A=H2O (0.5% TFA), B=
CH3CN (0.5% TFA); flow rate: 1.0 mLminꢃ1; gradient: 10!50% B in
Conclusions
In summary, we designed, synthesized, and tested the first
MAO-B-selectively activated prodrug MP-MUS based on MPTP,
which exerts its anticancer action inside the mitochondria of
brain cancer cells. To our knowledge, this is the first example
of selective mitochondrial chemotherapy. The maturation of
the prodrug MP-MUS is like the mechanism by which nontoxic
MPTP is converted into neurotoxin MPP+. The maturation of
MP-MUS is established first by the oxidation by MAO-B to form
1-methyl-2,3-dihydropyridinium MD+-MUS, and further oxida-
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
10.0 min. MP-MUS is a fragile compound and decomposes slowly
in solvents; therefore, it must be stored under nitrogen at ꢃ808C.
MitoTracker Red FM (MT, M22425) and Hoechst 33342 for DNA
were purchased from Life Technologies (Invitrogen/Molecular
Probes, Eugene, OR, USA).
The mixture was stirred at 608C for 2 h. The solution was carefully
neutralized to pH 6 (HCl, 1m), and the solvent was removed under
reduced pressure. The crude compound was further purified by
chromatography on regular silica gel eluting with EtOH/NH4OH
(5:1 v/v) to yield the desired product as a gelatinous white solid
(23 mg, 100% yield). Rf =0.23 (EtOH/NH4OH 5:1 v/v); 1H NMR
(CD3OD, 500 MHz): d=5.55 (brs, 1H), 3.70 (brs, 2H), 3.35–3.33 (m,
J=6.0 Hz, 1H), 3.03 (q, J=7.0 Hz, 2H), 2.89 (s, 3H), 2.54–2.51 (m,
1H), 2.45–2.42 (m, 1H), 1.24 ppm (d, J=7.0 Hz, 3H); 13C NMR
(CDCl3, 125 MHz): d=179.5, 137.8, 114.1, 51.8, 50.4, 48.5, 41.5, 24.0,
Synthesis of compounds 1, 2, 3, MP-Est, and MP-MUS. All re-
agents and solvents were purchased from Sigma–Aldrich. All syn-
theses were performed under an atmosphere of nitrogen unless in-
dicated otherwise. 1H and 13C NMR spectra were obtained with
a Bruker 500/600 MHz and 125 MHz magnetic resonance spectrom-
eter, respectively. Solvent peaks of CDCl3 at 7.26 and 77.1 ppm
15.0 ppm; LRMS calcd for C9H16NO2 [M+H+]+ 170.2, found 170.1.
+
1
were used as internal standards for H and 13C NMR, respectively.
N,N-Bis(2-hydroxyethyl)-2-(1-methyl-1,2,3,6-tetrahydropyridin-4-
Data for 1H NMR spectra are reported as follows: chemical shift,
multiplicity (s=singlet, d=doublet, t=triplet, q=quartet, b=
broad, m=multiplet), coupling constants, and number of protons.
Low-resolution mass spectra (LRMS) were recorded on a quadru-
pole spectrometer using electrospray ionization (ESI). Analytical
thin-layer chromatography was conducted on glass sheets coated
with silica gel 60 F254. Flash chromatography was performed on
regular grade silica gel (60 ꢂ, Fisher Scientific).
yl)propanamide (3): The mixture of compound
2 (30 mg,
0.178 mmol) and SOCl2 (0.30 mL) were stirred at 708C for 1 h. The
solvent was removed under reduced pressure. Then the residue
was dissolved in anhydrous CH2Cl2, and diethanolamine (17 mL,
0.178 mmol) was added, followed by anhydrous pyridine (0.1 mL).
The reaction mixture was stirred at room temperature for 1 h. The
solvent was removed under reduced pressure. The crude com-
pound was used directly for the next step without further purifica-
tion; LRMS calcd for C13H25N2O3 [M+H+]+ 257.2, found 257.1.
+
Ethyl 2-(1-methylpiperidin-4-ylidene)propanoate (1): NaH (3.83 g,
0.096 mol) was added slowly to a solution of ethyl 2-diethoxyphos-
phorylpropanoate (28 mL, 0.13 mol) in anhydrous THF (87 mL) in
a 250-mL flask kept in an ice bath. The mixture was stirred at 08C
for 15 min, and then 1-methylpiperidin-4-one (10 mL, 0.087 mol)
was added dropwise to the solution. The solution was slowly
warmed to room temperature, and stirring continued overnight.
The reaction was quenched by the addition of H2O (50 mL), and
the product was extracted with EtOAc (3ꢁ35 mL). The combined
organic solvent was washed with H2O (1ꢁ20 mL) and brine (1ꢁ
20 mL), dried over anhydrous Na2SO4, and finally removed under
reduced pressure. The crude residue was purified by regular silica
gel chromatography eluting with MeOH/CH2Cl2 (1:10) to yield the
desired product as light-yellow oil (10.5 g) in 61% yield. Rf =0.23
(MeOH/CH2Cl2 =1:10); 1H NMR (CDCl3, 500 MHz): d=4.17 (q, J=
7.0 Hz, 2H), 2.64–2.62 (m, 2H), 2.45–2.40 (m, 4H), 2.38–2.36 (m,
2H), 2.26 (s, 3H), 1.85 (s, 3H), 1.28 ppm (t, J=7.0 Hz, 3H); 13C NMR
(CDCl3, 150 MHz): d=170.0, 143.5, 121.3, 60.3, 56.5, 56.1, 45.8, 31.3,
N,N-Bis(2-chloroethyl)-2-(1-methyl-1,2,3,6-tetrahydropyridin-4-
yl)propanamide (MP-MUS): Compound 3 (22 mg, 0.086 mmol)
and SOCl2 (0.1 mL) were stirred at 708C for 3 h. The solvent was re-
moved under reduced pressure. Then the residue was purified by
flash chromatography on basic alumina eluting with CH2Cl2/MeOH
(30:1 v/v) to yield the desired product as a light-yellow solid
(16 mg, 65% yield). Rf =0.2 (CH2Cl2/MeOH 30:1 v/v); 1H NMR
(CDCl3, 600 MHz): d=5.56 (brs, 1H), 3.86 (q, J=7.2 Hz, 1H), 3.80 (t,
J=6.0 Hz, 4H), 3.70–3.67 (m, 3H), 3.54 (brs, 2H), 3.43–3.36 (m, 1H),
3.16 (t, J=6.0 Hz, 4H), 2.84 (s, 3H), 1.30 ppm (d, J=7.2 Hz, 3H);
LRMS calcd for C13H23N2Cl2O+ [M+H+]+ 293.1, 295.1, found 293.0,
295.0.
4-(1-(Bis(2-chloroethyl)amino)-1-oxopropan-2-yl)-1-methyl-
1,2,3,6-tetrahydropyridinium chloride (MP-MUS·HCl): The neutral
form MP-MUS that was prepared by following the protocol above
was dissolved in anhydrous EtOH (2m HCl, 3 equiv). The removal
of solvent under reduced pressure gave the desired product as
a white solid.
30.4, 22.4, 15.1, 14.3 ppm; LRMS calcd for C11H20NO2 [M+H+]+
+
198.2, found 198.2.
Primary human GBM: A glioblastoma (GBM) tumor was taken at
the time of excision and given the laboratory ID of BT150. It was
chopped with a scalpel and then homogenized with a 5-mL pip-
ette. The cells were grown in Dulbecco’s modified Eagle’s medium
(DMEM) with fetal bovine serum (FBS, 20%), 1ꢁ GlutaMax-I,
sodium pyruvate (1 mm), penicillin (100 UmLꢃ1), and streptomycin
(100 mgmLꢃ1). BT150 cells are spontaneously immortal and were
frozen at fourth passage and used between seventh and ninth pas-
sages. Glioma cells were grown to achieve confluency in either
Costar 96-well plates (Corning, NYC, NY, USA) or 16-well Lab-Tek
slide chambers (Nalge Nunc International, Rochester, NY, USA).
Low-passage human glioma primary culture, BT150, was used in all
the in vitro data presented, although we have used a number of
other low-passage primary glioblastoma cultures to confirm re-
sults.
Ethyl
2-(1-methyl-1,2,3,6-tetrahydropyridin-4-yl)propanoate
(MP-Est): Compound 1 (119 mg, 0.604 mmol) was dissolved in an-
hydrous THF (1.0 mL) and the solution was merged in a ꢃ728C
bath (dry ice in EtOH). A solution of LDA (0.91 mL, 2m) was then
added dropwise to the former solution. The final solution was
stirred at ꢃ728C for 1 h, and then the reaction was quenched with
saturated NH4Cl. The product was extracted from H2O with EtOAc
(3ꢁ20 mL). The combined organic phase was washed with deion-
ized H2O, brine, and then dried over anhydrous Na2SO4. The salt
was removed by filtration, and the solvent was removed under re-
duced pressure. The residue was purified by chromatography on
silica gel eluting with EtOAc (3% Et3N) to yield the desired product
1
as a yellow oil (117 mg, 99% yield). H NMR (CDCl3, 500 MHz): d=
5.53 (brs, 1H), 4.11 (q, J=7.0 Hz, 2H), 3.08 (q, J=7.0 Hz, 1H), 3.03
(brs, 2H), 2.65–2.60 (m, 2H), 2.38 (s 3H), 2.25 (brd, J=16.5 Hz, 1H),
2.16 (brd, J=16.5 Hz, 1H), 1.25 (d, J=7.0 Hz, 3H), 1.23 ppm (t, J=
7.0 Hz, 3H); 13C NMR (CDCl3, 125 MHz): d=174.2, 135.0, 120.3, 60.6,
53.6, 51.4, 46.1, 44.7, 26.4, 15.1, 14.3 ppm; LRMS calcd for
Normal human astrocytes (NHAs): NHAs were obtained from
Lonza (Walkersville, MD, USA) and subjected to their recommenda-
tions for growth. NHAs were grown to confluency in astrocyte cell
basal medium supplemented with 3% FBS, 1% glutamine, 0.1% in-
sulin, 0.1% rhEGF, 0.1% GA-1000, and 0.25% ascorbic acid (from
CloneticsTM AGMTM BulletKitTM (CC-3186)) in 96-well plates or in 16-
well Lab-Tek slide chambers in a total volume of 250 mL.
C11H20NO2 [M+H+]+ 198.2, found 198.1.
+
2-(1-Methyl-1,2,3,6-tetrahydropyridin-4-yl)propanoic acid (2):
MP-Est (27 mg, 0.137 mmol) was dissolved in KOH (0.137 mL, 2m).
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Measurements of H2O2 generation with Amplex Red: For Tyr and
MP-Est: substrates, Amplex Red (150 mm), and HRP (3 UmLꢃ1) were
incubated with MAOs (1 UmLꢃ1) in pH 7.4 buffer (50 mm KPi,
100 mm NaCl) at 378C for 15 min in 96-well format (75 mL volume
per well). The formation of fluorescent resorufin (lex 500–560 nm,
lem 565–625 nm) was measured in a BioTeck Synergy HT spectro-
photometer. The maximal signal generated from Amplex Red
(150 nmolmLꢃ1) was established by the addition of H2O2 (1 mm) to
each well at the end of the assay period.
ing effects of brain tumors and central nervous system disease.
We thank the MD Anderson Cancer Center for use of their NMR
facility.
Keywords: DNA damage · gliomas · MAO-B · MPTP · prodrugs
[1] A. Omuro, L. M. DeAngelis, JAMA J. Am. Med. Assoc. 2013, 310, 1842–
1850.
[2] T. A. Dolecek, J. M. Propp, N. E. Stroup, C. Kruchko, J. Neuro-Oncol. 2012,
14, v1–49.
[3] J. P. Mannas, D. D. Lightner, S. R. Defrates, T. Pittman, J. L. Villano, J. Clin.
Neurosci. 2014, 21, 121–123.
[4] Y. P. Ramirez, J. L. Weatherbee, R. T. Wheelhouse, A. H. Ross, Pharmaceut-
icals 2013, 6, 1475–1506.
For MP-MUS: We found that MP-MUS was a potent inhibitor of
HRP, and therefore was eliminated from the 15 min incubation
period of dye, MAO, and MP-MUS. HRP (3UmLꢃ1) was added to the
mix at the end of 15 min, just prior to fluorescence measurement.
Fluorescence microscopy: Images were captured using a Nikon
Eclipse TE2000-E at 4ꢁ or 20ꢁ magnification using a CoolSnap ES
digital camera system (Roper Scientific) containing a CCD-1300-Y/
HS 1392ꢁ1040 imaging array that is cooled by Peltier. Images
were recorded and analyzed using Nikon NIS-Elements software
(Elements 3.22.11). All images were saved as JPEG2000 files using
Nikon NIS-Elements. The emission of goat anti-rabbit IgG antibody
labeled with Alexa Fluor 488 was collected through the FITC
(lex 450–490 nm, lem 500–550 nm). The fluorescence of MTR was
obtained through the Texas Red filter set (lex 510–550 nm, lem 573–
648 nm), whereas Hoechst 33342 was taken via the DAPI filter set
(lex 325–375 nm, lem 435–485 nm).
[5] M. Preusser, S. de Ribaupierre, A. Wohrer, S. C. Erridge, M. Hegi, M.
Weller, R. Stupp, Ann. Neurol. 2011, 70, 9–21.
[6] Y. Singh, M. Palombo, P. J. Sinko, Curr. Med. Chem. 2008, 15, 1802–1826.
[7] L. Galluzzi, N. Larochette, N. Zamzami, G. Kroemer, Oncogene 2006, 25,
4812–4830.
[8] J. C. Shih, K. Chen, M. J. Ridd, Annu. Rev. Neurosci. 1999, 22, 197–217.
[9] a) N. M. Freedman, E. Mishani, Y. Krausz, J. Weininger, H. Lester, E. Blau-
grund, D. Ehrlich, R. Chisin, J. Nucl. Med. 2005, 46, 1618–1624; b) J.
Saura, E. Nadal, B. van den Berg, M. Vila, J. A. Bombi, N. Mahy, Life Sci.
1996, 59, 1341–1349.
[10] J. Tong, J. H. Meyer, Y. Furukawa, I. Boileau, L. J. Chang, A. A. Wilson, S.
Houle, S. J. Kish, J. Cereb. Blood Flow Metab. 2013, 33, 863–871.
[11] a) K. N. Westlund, R. M. Denney, R. M. Rose, C. W. Abell, J. Neurosci.
1988, 25, 439–456; b) K. N. Westlund, R. M. Denney, L. M. Kochersperg-
er, R. M. Rose, C. W. Abell, Science 1985, 230, 181–183; c) P. Levitt, J. E.
Pintar, X. O. Breakefield, Proc. Natl. Acad. Sci. USA 1982, 79, 6385–6389.
[12] A. M. Gabilondo, C. Hostalot, J. M. Garibi, J. J. Meana, L. F. Callado, Neu-
rochem. Int. 2008, 52, 230–234.
[13] J. P. M. Finberg, Pharmacol. Ther. 2014, 143, 133–152.
[14] G. E. Meredith, D. J. Rademacher, J. Parkinsons Dis. 2011, 1, 19–33.
[15] G. Porras, Q. Li, E. Bezard, Cold Spring Harbor Perspect. Med. 2012, 2,
a009308.
[16] P. Bajpai, M. C. Sangar, S. Singh, W. Tang, S. Bansal, G. Chowdhury, Q.
Cheng, J.-K. Fang, M. V. Martin, F. P. Guengerich, N. G. Avadhani, J. Biol.
Chem. 2013, 288, 4436–4451.
[17] H. Shi, N. Noguchi, Y. Xu, E. Niki, FEBS Lett. 1999, 461, 196–200.
[18] G. P. Davey, K. F. Tipton, M. P. Murphy, Biochem. J. 1992, 288, 439–443.
[19] D. G. Nicholls, S. J. Ferguson, Bioenergetics 2, Academic Press Ltd.,
London, 1992, P. 59.
[20] Y. Watanabe, T. Himeda, T. Araki, Med. Sci. Monit. 2005, 11, RA17–RA23.
[21] H. Rollema, E. A. Johnson, R. G. Booth, P. Caldera, P. Lampen, S. K.
Youngster, A. J. Trevor, N. Naiman, N. Castagnoli, J. Med. Chem. 1990,
33, 2221–2230.
[22] Y.-X. Wang, N. Castagnoli, Jr., Bioorg. Med. Chem. 1997, 5, 1321–1325.
[23] V. T. DeVita, Jr., E. Chu, Cancer Res. 2008, 68, 8643–8653.
[24] R. R. Fritz, C. W. Abell, N. T. Patel, W. Gessner, A. Brossi, FEBS Lett. 1985,
186, 224–228.
[25] a) X. Yan, Y. Zhou, S. Liu, Theranostics 2012, 2, 988–998; b) S. Salvioli, A.
Ardizzoni, C. Franceschi, A. Cossarizza, FEBS Lett. 1997, 411, 77–82; c) A.
Cossarizza, M. Baccaranicontri, G. Kalashnikova, C. Franceschi, Biochem.
Biophys. Res. Commun. 1993, 197, 40–45.
Cell viability and cell death assay: MP-MUS·HCl was added to
a DMSO solution to make a stock solution (7.5 mm). GBM cells or
NHAs were treated with MP-MUS (0, 30, 60, 90, 120, 150, 180,
210 mm), and all wells received the same 8-mL DMSO vehicle
volume (total volume of 250 mL). After treatment, cells were grown
in the presence of all effectors (378C, 5% CO2). At t=47 h, cells
were incubated with Hoechst 33342 (10 mm) in the presence and
absence of MitoTracker Red FM (1 mm, M22425) for 1 h and then
were fixed with ice-cold 4% paraformaldehyde (PFA). We conduct-
ed cell counts in center field at 20ꢁ magnification, so that there
was an average of 150 cells per image and between one and eight
dead or dying cells visualized in the same field. Dead/dying cells
were identified as having condensed nuclei with signal intensities
over threefold that of the median cell nuclei. Dividing cells identi-
fied by paired and brightly labeled nuclei were counted as two
living cells.
Measurement of MAO-B levels in glioma cells and NHAs: Glioma
cells were permeabilized with 0.1% Triton X-100 in PBS and treated
with Dako protein blocking solution (X0909, Dako North America,
Carpinteria, CA, USA). Rabbit anti-MAO-B (clone EPR7103, Cat
#ab125010, Abcam, Cambridge, MA, USA) was used to identify
MAO-B and was diluted 1:500 using Dako antibody diluent (S3022).
Goat anti-rabbit IgG antibody labeled with Alexa Fluor 488
(A11034) was used to visualize for primary rabbit antibody to
MAO-B.
[26] A. Polavarapu, J. A. Stillabower, S. G. W. Stubblefield, W. M. Taylor, M.-H.
Baik, J. Org. Chem. 2012, 77, 5914–5921.
[27] A. Penmatsa, K. H. Wang, E. Gouaux, Nature 2013, 503, 85–90.
[28] T. Beuming, J. Kniazeff, M. L. Bergmann, L. Shi, L. Gracia, K. Raniszewska,
A. H. Newman, J. A. Javitch, H. Weinstein, U. Gether, C. J. Loland, Nat.
Neurosci. 2008, 11, 780–789.
[29] E. Watson, P. Dea, K. K. Chan, J. Pharm. Sci. 1985, 74, 1283–1292.
[30] M. J. Zhou, Z. J. Diwu, N. PanchukVoloshina, R. P. Haugland, Anal. Bio-
chem. 1997, 253, 162–168.
[31] M. B. H. Youdim, E. Heldman, H. B. Pollard, P. Fleming, E. McHugh, Neu-
roscience 1986, 19, 1311–1318.
[32] L. B. Pearce, J. A. Roth, Arch. Biochem. Biophys. 1983, 224, 464–472.
[33] C. H. Williams, Biochem Pharmacol 1974, 23, 615–628; A. K. Tan, R. R.
Ramsay, Biochemistry 1993, 32, 2137–2143.
Acknowledgements
This work was supported by the Donna and Kenneth R. Peak
Foundation, the Kenneth R. Peak Brain and Pituitary Center at
Houston Methodist Hospital, the Taub Foundation, the Blanche
Green Estate Fund of the Pauline Sterne Wolff Memorial Founda-
tion, the Verelan Foundation, the Houston Methodist Hospital
Foundation, the American Brain Tumor Association, and many
patients and families who have been impacted by the devastat-
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[34] C. Binda, M. Li, F. Hubalek, N. Restelli, D. E. Edmondson, A. Mattevi, Proc.
Natl. Acad. Sci. USA 2003, 100, 9750–9755.
[36] M. Taspinar, S. Ilgaz, M. Ozdemir, T. Ozkan, D. Oztuna, H. Canpinar, J. A.
Rey, A. Sunguroglu, J. S. Castresana, H. C. Ugur, Tumor Biol. 2013, 34,
1935–1947.
[37] A. Jekabsone, P. K. Mander, A. Tickler, M. Sharpe, G. C. Brown, J. Neuroin-
flammation 2006, 3, 24.
[35] a) C. Binda, F. Hubalek, M. Li, Y. Herzig, J. Sterling, D. E. Edmondson, A.
Mattevi, J. Med. Chem. 2004, 47, 1767–1774; b) G. Engberg, T. Elebring,
H. Nissbrandt, J. Pharmacol. Exp. Ther. 1991, 259, 841–847; c) J. R. Miller,
D. E. Edmondson, Biochemistry 1999, 38, 13670–13683; d) L. De Colibus,
M. Li, C. Binda, A. Lustig, D. E. Edmondson, A. Mattevi, Proc. Natl. Acad.
Sci. USA 2005, 102, 12684–12689.
Received: December 26, 2014
Published online on && &&, 0000
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Specificity is key: We developed
a MAO-B-activated prodrug, MP-MUS,
for the treatment of brain gliomas. MP-
MUS is nontoxic, and can be selectively
oxidized by MAO-B, which is overex-
pressed in glioma cells, to form toxic
P+-MUS. P+-MUS translocates inside mi-
tochondria passively and alkylates frag-
ile mitochondrial DNA, leading to specif-
ic apoptosis of glioma cells, but not
normal cells.
M. A. Sharpe, J. Han, A. M. Baskin,
D. S. Baskin*
&& – &&
Design and Synthesis of a MAO-B-
Selectively Activated Prodrug Based
on MPTP: A Mitochondria-Targeting
Chemotherapeutic Agent for
Treatment of Human Malignant
Gliomas
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
9
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ