Methotrexate prodrugs sensitive to reactive oxygen species for the improved treatment of rheumatoid arthritis

Article abstract of DOI:10.1016/j.ejmech.2018.07.045

Methotrexate (MTX) is the standard of care in the treatment of rheumatoid arthritis (RA), a common autoimmune disease that is characterized by chronic inflammation in the synovial membrane of joints. Unfortunately, MTX suffers from high discontinuation ra

Full text of DOI:10.1016/j.ejmech.2018.07.045

European Journal of Medicinal Chemistry 156 (2018) 738e746
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Methotrexate prodrugs sensitive to reactive oxygen species for the
improved treatment of rheumatoid arthritis
a
a
a
b
Nikolaj S. Andersen , Jorge Peir oꢀ Cadahía , Viola Previtali , Jon Bondebjerg ,
c
d
e
a, *
Christian A. Hansen , Anders E. Hansen , Thomas L. Andresen , Mads H. Clausen
a
Center for Nanomedicine & Theranostics, Department of Chemistry, Technical University of Denmark, Kemitorvet 207, DK, 2800, Kongens Lyngby,
Denmark
b
MC2 Therapeutics, Agern Alle 24-26, 2970, Hørsholm, Denmark
Capdelta Group Aps, C/O Kavsbjerglund 30, DK, 2740, Skovlunde, Denmark
Department of Clinical Physiology, Nuclear Medicine & PET and Cluster for Molecular Imaging, Rigshospitalet and University of Copenhagen, DK, 2100,
c
d
Copenhagen, Denmark
e
Center for Nanomedicine & Theranostics, Department of Micro- and Nanotechnology, Technical University of Denmark, Ørsteds Plads, Building 345, DK,
2
800, Kongens Lyngby, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 May 2018
Received in revised form
Methotrexate (MTX) is the standard of care in the treatment of rheumatoid arthritis (RA), a common
autoimmune disease that is characterized by chronic inflammation in the synovial membrane of joints.
Unfortunately, MTX suffers from high discontinuation rates due to a large variability in efficacy and, in
particular, adverse effects. As inflammation is associated with elevated levels of reactive oxygen species
8
July 2018
Accepted 17 July 2018
Available online 20 July 2018
(
ROS) like H
matory tissue by use of a H
sensitive, thiazolidinone-based MTX prodrugs were synthesized and evaluated for this purpose.
-thiazolidinone (MTX- -TZ) exhibited the most promising properties e good to high chemical and
metabolic stability, excellent aqueous solubility, while being activated when subjected to patho-
physiological concentrations of H O . In vivo, MTX- -TZ exhibited comparable efficacy to MTX in a
2
O
2
, we propose to improve treatment through site-selective delivery of MTX to inflam-
2 2
O
sensitive MTX prodrug. To establish proof proof-of-concept, two novel
2 2
H O
MTX-
Chemical compounds studied in this article:
Methotrexate (PubChem CID: 126941)
g
g
Keywords:
Rheumatoid arthritis
Inflammation
Methotrexate
Prodrug
g
2
2
murine collagen type II-induced arthritis (CIA) model while treated mice showed indications of reduced
toxicity as their body weight decreased less towards the end of the study, compared to the MTX-treated
group.
©
2018 Elsevier Masson SAS. All rights reserved.
Thiazolidinone
Reactive oxygen species
1
. Introduction
characterized by chronic synovial inflammation and is associated
with progressive disability, systemic complications, early death,
and socioeconomic costs [1]. The pathogenesis of RA is still
incompletely understood but involves a complex interplay of ge-
netic and environmental factors. It is estimated that 0.5e1% of the
adult populations in developed countries suffer from RA, with a
prevalence three times greater for women than men [2].
1.1. Background
Rheumatoid arthritis (RA) is a common autoimmune disease
Currently, no cure for RA exists and treatment strategies consist
of life-long palliative care, primarily using disease-modifying
antirheumatic drugs (DMARDs) to relieve inflammatory symp-
toms and retard joint destruction. Since the 1980s, methotrexate
Abbreviations: ADMET, absorption, distribution, metabolism, excretion, toxicity;
CFA, Complete Freund's Adjuvant; CIA, collagen-induced arthritis; CL'int, apparent
intrinsic clearance; DBA, dilute brown non-agouti; dec., decompose; DMA, dime-
thylacetamide; DMARD, disease-modifying antirheumatic drug; HD, high-dose; i.p.,
intraperitoneal; LD, low-dose; MTX, methotrexate; PBS, phosphate buffered saline;
RA, rheumatoid arthritis; ROS, reactive oxygen species; SEM, standard error of the
mean; TNF, tumor necrosis factor; TZ, 1,3-thiazolidin-2-one.
(MTX, 1) has been the standard of care in the treatment of RA
(Fig. 1). MTX was originally developed as a folic acid antagonist for
high-dose cancer therapy (HD-MTX: 1000e5000 mg/week). How-
ever, low-dose treatment (LD-MTX: 5e25 mg/week), administered
either orally, subcutaneously, or intravenously, has shown potent
*
Corresponding author. Department of Chemistry, Center for Nanomedicine &
Theranostics, Technical University of Denmark, Kemitorvet 207, DK, 2800, Kongens
Lyngby, Denmark.
E-mail address: mhc@kemi.dtu.dk (M.H. Clausen).
https://doi.org/10.1016/j.ejmech.2018.07.045
0223-5234/© 2018 Elsevier Masson SAS. All rights reserved.

N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
739
and in vivo evaluation in a murine collagen-induced arthritis (CIA)
model.
1.2. Strategy
The thiazolidinone group can be introduced through coupling to
a carboxylic acid. As MTX contains two carboxylic acids ( and ),
we planned at first, to synthesize three MTX prodrugs: MTX- -TZ
2), MTX- -TZ (3), and MTX-TZ (4) to explore the potentially
different behavior of the compounds (Fig. 2).
a
g
a
(
g
2
Fig. 1. Methotrexate (1) is the standard of care for treatment of RA.
anti-inflammatory effects in patients suffering from RA [3e5].
Unfortunately, in spite of LD-MTX being the standard of care in
managing RA, treatment is still unsatisfactory for many patients.
Low-dose treatment is associated with several prominent adverse
effects, in particular gastrointestinal toxicities but also hepatotox-
icity, lethargy, fatigue, nodulosis, hepatic and pulmonary fibrosis,
renal insufficiency, anemia, and neutropenia [4]. Furthermore, LD-
MTX suffers from high interindividual variability, leading to un-
predictable treatment outcomes and, in many cases, poor patient
response or lack of efficacy [4,6]. Consequently, nearly half of pa-
tients discontinue treatment within 3 years of therapy [7].
Currently, the best alternative is combinational therapy using MTX
2
2
2
. Results and discussion
.1. MTX-TZ (4)
2
.1.1. Synthesis
2
As a preliminary test of the hypothesis, MTX-TZ (4) was initially
synthesized to investigate its properties for use as a prodrug.
Starting from commercially available MTX, MTX-TZ was obtained
in one step using DCC and DMAP for the coupling (Scheme 1).
2
2.1.2. Preliminary evaluation of prodrug stability
As a first measure of prodrug potential, the aqueous stability of
with newer biological therapeutics such as the anti-TNF-a agents
the synthesized prodrug candidate was examined in PBS (pH 7.4) at
etanercept or infliximab. However, such treatments are extremely
costly and are associated with additional side effects [3,4]. Thus,
there is still an urgent need for improving the efficacy and safety of
RA therapies.
Prodrugs are inactive forms of pharmaceuticals that undergo
chemical or enzymatic conversion in vivo to release the active agent
ꢁ
2
0 C using RP-UPLC-MS analysis. Interestingly, the
a
-thiazolidi-
(4) was highly labile and rapidly hydro-
-TZ (3) within 30 min (Figure S1, SI). On the
other hand and to our delight, MTX- -TZ (3) was significantly more
stable and no hydrolysis of this compound was observed even after
4 h. Based on the short half-life of the -thiazolidinone moiety,
MTX- -TZ (2) was deemed to likely be a poor prodrug candidate
and we decided to pursue MTX- -TZ as the most promising of the
three thiazolidinone-based MTX prodrugs shown in Fig. 2.
none moiety of MTX-TZ
lyzed to release MTX-
2
g
g
2
a
[
8e10]. Prodrug design is typically implemented to improve un-
desirable absorption, distribution, metabolism, excretion, toxicity
ADMET) properties, typically poor solubility or absorption, but it
a
g
(
may also be exploited to increase tissue selectivity. While several
prodrugs, drug conjugates, and drug delivery systems for MTX have
been reported in the literature over the years, so far none have been
2.2. MTX-g-TZ (3)
approved for clinical use [11].
ꢀ
2.2.1. Synthesis
MTX- -TZ (3) was synthesized from the commercially available
pteridine alcohol 5 (Scheme 2). The commercially available hy-
drochloride salt of 5 was first neutralized with aqueous NaOH and
Reactive oxygen species (ROS), e.g. H
2
O
2
, O
2
, HO _, and HOCl,
g
serve important physiological roles that include signaling func-
tions, host defense, and oxidative biosynthesis [12,13]. However,
elevated levels of ROS can lead to harmful oxidative stress, which is
a central component of the pathogenesis of chronic inflammation
then converted to the corresponding
PPh Br followed by nucleophilic substitution with 4-(methyl-
amino)benzoic acid to form 6. Coupling of H-Glu (OMe)-OtBu with
using PyBOP and Et N with subsequent hydrolysis of the
methyl ester using Ba(OH) , afforded MTX- -OtBu (7). The thiazo-
lidinone moiety was then introduced to the position through a
DCC/DMAP-mediated coupling with the unprotected -carboxylic
a-bromo species in situ using
[
14e17], including autoimmune disease such as RA [18,19]. H
the most stable of the ROS and under pathological conditions,
extracellular concentrations up to 1 mM of H have been
measured [12,20e23]. This increased concentration of H can
2 2
O is
3
2
6
3
g-
2 2
O
a
2 2
O
2
g
potentially serve as a stimulus for site-selective drug delivery of
ROS sensitive prodrugs. Development of such prodrugs has
emerged as a novel approach to potentially increase target selec-
g
tivity of drugs. Examples of H
phenylboronic acids and esters [24e29], phenylsulfonate esters
30,31], N-(2,5-dihydroxyphenyl)acetamides [32], -boryl ethers,
2 2
O sensitive promoieties include
[
a
carbonates, and acetals [33], and 1,3-thiazolidin-2-one has also
been proposed as a promising ROS labile promoiety [34].
Recently, our group published promising preliminary results for
the treatment of RA using a boronic acid based MTX prodrug in vivo
[
35]. Based on these results and the promising ROS sensitive
properties of the 1,3-thiazolidin-2-one moiety (henceforward
simply thiazolidinone or TZ), we here propose a new strategy for
the treatment of RA using a thiazolidinone-based MTX prodrug for
site-selective delivery of MTX. Our aim is to localize and accumulate
MTX in inflammatory tissue in order to improve the safety profile
and potentially the efficacy of the drug. We herein report the
synthesis, in vitro pharmacokinetic and physiochemical studies,
Fig. 2. The three targeted thiazolidinone (TZ) based MTX prodrug candidates: Meth-
otrexate- -thiazolidinone (MTX- -TZ, 2), methotrexate- -thiazolidinone (MTX- -TZ,
), and methotrexate- -dithiazolidinone (MTX-TZ , 4).
a
a
g
g
3
a,
g
2

740
N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
2
Scheme 1. Synthesis of MTX-TZ . Reagents and conditions: (a) Thiazolidin-2-one, DCC, DMAP, DMF, 16 h (18%).
ꢁ
Scheme 2. Synthesis of MTX-
g
-TZ (3). Reagents and conditions: (a) NaOH/H
2
O, 80 C, 5 min (b) PPh
3
2
Br , DMA, 20 h, then 4-(methylamino)benzoic acid, DIPEA, 72 h (90%). (c) PyBOP,
Et N, DMF, 16 h. (d) Ba(OH) , H
3
2
2
O/EtOH, 16 h (79%). (e) thiazolidin-2-one, DCC, DMAP, DMF, 16 h (65%). (f) HCO
2
H, 20 h (74%).
ꢁ
acid to obtain compound 8. Finally,
a-tert-butyl ester deprotection
DMSO/PBS at 37 C and was followed by RP-UPLC-MS (Fig. 3, left).
was achieved in neat formic acid to give the desired MTX-
g-TZ (3)
Diclofenac was used as internal standard and its stability in a so-
in reasonable yield (74%) and excellent purity (>95%) after pre-
2 2
lution of H O was confirmed (Figure S2, SI). Importantly, MTX was
parative HPLC.
confirmed to be the product of prodrug activation (Figure S3, SI, top
left and top right) and to be stable under different concentrations of
H
5
2
O
2
(10 and 0 mM, Figure S3, SI, bottom). The prodrug half-life at
mM was determined to be 1.7 ± 0 h. As expected, the rate of
and at
2, the half-life was reduced to 16.5 ± 0 h. Fig. 3 (right)
shows the appearance of MTX as a consequence of the activation of
MTX- -TZ (3) within the first hour of the experiment. MTX reached
2
2
.3. Physiochemical and pharmacokinetic evaluation
activation was dependent on the concentration of H
.5 mM H
2 2
O
.3.1. Prodrug activation under oxidative conditions and PBS
0
2
O
stability
Activation of MTX-
different concentrations of H
g
-TZ (3) was examined in vitro under
(10e5 e 1e0.5e0.25 mM) in 20%
g
2 2
O
Fig. 3. Activation of prodrug MTX-
g
-TZ (3) (left) and appearance of MTX (right) in the presence of varying concentrations of H
2
O
2
. The experiments were carried out at a compound
¼ 306 nm) using diclofenac as
ꢁ
concentration of 100
m
M in a mixture of 20% DMSO in PBS at 37 C. Experiments were performed in triplicates (n ¼ 3) and analyzed by RP-UPLC-MS (
l
internal standard. Error bars are calculated as the standard deviation. Hidden error bars are smaller than symbols.

N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
741
5
0% with 10 mM H
2
O
2
and at patho-physiological concentrations of
experienced a significantly more stable body weight compared to
the MTX group, indicating a safer toxicity profile.
2 2
H O
(0.5 mM), MTX formed at around 7%. To evaluate the rate of
non-H
2
O
2
-mediated hydrolysis, the aqueous stability of MTX-
g
-TZ
Compared to recently published results for the in vivo effects of a
boronic acid based MTX prodrug [35], MTX-g-TZ (3) shows very
ꢁ
(3) under the same conditions (20% DMSO/PBS at 37 C) was also
examined. Here, MTX-
g
-TZ (3) exhibited excellent stability with a
comparable behavior in the murine CIA model, both with regards to
half-life of 115.5 ± 0 h indicating that release of MTX is primarily
efficacy and mean body weight of the treated mice.
driven by hydrogen peroxide activation of the prodrug.
To further evaluate the drug-like properties of MTX-
g-TZ (3), the
3. Conclusion
prodrug candidate was subjected to a series of pharmacokinetic
and physiochemical assays (Table 1).
Of the two thiazolidinone-based MTX prodrugs that were syn-
thesized for targeted treatment of inflammatory tissue in RA, MTX-
2.3.2. Plasma stability
g-TZ (3) had the most promising properties. While the
a-thiazoli-
Plasma stability was examined in reconstituted freeze dried
dinone of MTX-TZ showed poor stability in aqueous media, MTX-
2
human plasma diluted 1:1 with PBS to ensure a pH of 7.4 [36]. At
g
-TZ (3) was selectively activated at patho-physiological concen-
ꢁ
3
7 C, the half-life of MTX-
g-TZ was determined as 7 ± 1 h indi-
trations of H O . This prodrug candidate displayed excellent
2
2
cating good plasma stability [37] (Table 1 and Figure S4, SI).
chemical stability in PBS, good stability in human plasma, excellent
metabolic stability, and excellent aqueous solubility. In vivo, MTX-
2.3.3. Metabolic stability
g-TZ (3) exhibited comparable efficacy to the standard of care, MTX,
Metabolic stability was evaluated by determining the apparent
in a murine CIA model. Mice being treated with MTX- -TZ (3)
g
intrinsic clearances (CL'int) in human pooled liver microsomes. The
apparent intrinsic clearance was calculated to be 1.4 ± 0.4 mL/min/
kg indicating excellent metabolic stability [38] (Table 1 and
Figure S5, SI).
experienced a significantly more stable body weight compared to
the MTX group, indicating reduced toxicity. These preliminary re-
sults show promise in improving the treatment of RA and future
work will focus on additional in vivo studies to establish optimal
dosing and to further investigate the toxicity profile.
2.3.4. Thermodynamic solubility
The thermodynamic solubility of MTX-
PBS at 20 C and determined as 400 ± 50
g
-TZ was measured in
4. Materials and methods
ꢁ
m
g/mL (Table 1 and
Figure S6, SI). This is considered high solubility according to
guidelines for oral administration in humans and animal dosing
formulations [39].
4.1. Chemistry
Thiazolidin-2-one was obtained from Fluorochem Ltd. All other
starting materials, reagents, and solvents were purchased from
Sigma-Aldrich and used without further purification. All reactions
2.4. In vivo anti-arthritic efficacy and preliminary toxicity
2
were run under a N atmosphere and were monitored by thin layer
Based on the promising in vitro physiochemical and pharma-
cokinetic data, MTX-g-TZ (3) was tested for anti-inflammatory ac-
chromatography (TLC) and/or reversed-phase ultra-performance
liquid chromatography mass spectrometry (RP-UPLC-MS).
tivity in vivo using a murine collagen-induced arthritis (CIA) model.
This model was chosen as it has been extensively studied using
many agents including MTX [40], and is known to have the best
record of predictability across species as compared to other in vivo
arthritis models [41,42]. A dosing regime equimolar to 7 mg/kg
MTX with daily intraperitoneal (i.p.) administration was selected.
To minimize anti-inflammatory effects of DMSO [43,44], the con-
centration of DMSO was kept to 2% (v/v).
Therapeutic intervention was initiated at the onset of disease, 27
days after induction of chronic and destructive arthritis using type-
II collagen (CII) in Complete Freund's Adjuvant (CFA), and treatment
was prolonged for 14 days. Disease evaluation was based on a
macroscopic scoring system and performed three times a week (see
Analytical TLC was conducted on Merck aluminum sheets
covered with silica (C60). The plates were either visualized under
UV-light or stained by dipping in a developing agent followed by
4 2 3
heating. KMnO [3 g in water (300 mL) along with K CO (20 g) and
5% aqueous NaOH (5 mL)] were used as developing agent. Flash
®
column chromatography was performed using Merck Geduran Si
1
13
60 (40e63 mm) silica gel. For the recording of H NMR and C NMR
a Bruker Ascend equipped with a Prodigy cryoprobe (operating at
400 MHz for proton and 101 MHz for carbon) was used. The
chemical shifts (
coupling constants (J) in Hz. For spectra recorded in DMSO‑d
signal positions were measured relative to the signal for DMSO (
d) are reported in parts per million (ppm) and the
6
,
d
1
13
2.50 ppm for H NMR and
recorded in CDCl , signal positions were measured relative to the
signal for CHCl
d 39.43 ppm for C NMR). For spectra
Fig. 4). As evident from the figure, MTX-g-TZ (3) showed compa-
3
1
13
rable efficacy to MTX while both compounds effectively reduced
the severity of arthritis compared to the vehicle.
As a first estimate of drug tolerability, the general health of the
mice was evaluated as a function of body weight (Fig. 5). While the
MTX group suffered from a reduction in body weight towards the
3
(d
7.26 ppm for H NMR and
d
77.0 ppm for
C
NMR). IR analysis was performed on a Bruker Alpha FT-IR spec-
trometer. Analytical RP-UPLC-MS (ESI) analysis was performed on a
S2 Waters AQUITY RP-UPLC system equipped with a diode array
detector using an Thermo Accucore C18 column (d 2.6
2.1 ꢂ 50 mm; column temp: 50 C; flow: 1.0 mL/min). Eluents A
m
m,
ꢁ
end of the study, mice being treated with MTX-
g
-TZ (3)
Table 1
Pharmacokinetic and physiochemical properties of prodrug candidate MTX-
otherwise stated.
g-TZ (3). Values are an average of at least three measurements ± standard deviation (SD), unless
Solubility (
m
g/mL)a
Human plasma
stab. t1/2 (h)
PBS stab.
Metabolic stability,
CL'int (mL/min/kg)b
5 mM H
1/2 (h)b
2
O
2
activation
0.5 mM H
t
1/2 (h)b
2 2
O activation
b
b
t
1/2 (h)
t
400 ± 50
7 ± 1.0
115.5 ± 0.01
1.4 ± 0.4
1.7 ± 0.01
16.5 ± 0.01
a
ꢁ
ꢁ
pH 7.4, 20 C.
b
pH 7.4, 37 C.

742
N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
Fig. 4. Suppression of CIA development in mice after treatment with MTX (7 mg/kg), MTX-
g
-TZ (8.3 mg/kg), and vehicle (n ¼ 8 per group). DBA/1J mice were given the indicated
amounts of compound daily, starting at onset of disease (day 27), and disease progression was evaluated three times per week. A macroscopic scoring system of the four limbs 2
ranging from 0 to 15 (1 point per swollen toe, 1 point per swollen foot knuckle, and 5 points for swollen ankle) was used for a maximal score of 60 per mouse. One animal in the
vehicle group was sacrificed pre-termination due to high arthritis severity score. Data represents mean values of arthritic score ± SEM. * represents a p-value <0.05 and ** rep-
resents a p-value <0.01 when comparing MTX and vehicle groups while y represents a p-value <0.05 between MTX-
g-TZ (3) and vehicle. Statistics were calculated using a one-tailed
nonparametric MannꢀWhitney.
Fig. 5. The general health of mice was evaluated as the average body weight during collagen-induced arthritis as an indication of drug tolerability. Measurements were performed
three times per week (n ¼ 8 per group). One animal in the vehicle group was sacrificed pre-termination due to high arthritis severity score. Data represents mean values of body
weight ± SEM. * represents a p-value <0.05 when comparing MTX and MTX-
one-tailed nonparametric MannꢀWhitney.
g
-TZ (3) while y represents a p-value <0.05 between MTX and vehicle. Statistics were calculated using a
ꢁ
(
0.1% HCO
linear gradient (5% B to 100% B) in 2.4 min and then held for
.1 min at 100% B (total run time: 2.6 min). The LC system was
2
H in H
2
O) and B (0.1% HCO
2
H in MeCN) were used in a
temperature for all recordings was approximately 20 C. Melting
points were obtained using a Stuart SMP30 melting point apparatus
and are uncorrected. Preparative RP-HPLC was carried out on a
Waters Alliance reversed-phase HPLC system consisting of a Waters
2545 Binary Gradient Module equipped with either an xBridge BEH
0
coupled to a SQD mass spectrometer. RP-UPLC-UV method valida-
tion is reported in the Supporting Information. Analytical LC-HRMS
(
ESI) analysis was performed on an Agilent 1100 RP-LC system
C18 OBD Prep Column (130 Å, 5
m
m, 30 ꢂ 150 mm) or an xBridge
equipped with a diode array detector using a Phenomenex Luna
C18 column (d 3
0
MeCN) were used in a linear gradient (20% B to 100% B) in a total
run time of 15 min. The LC system was coupled to a Micromass LCT
orthogonal time-of-flight mass spectrometer equipped with a Lock
Mass probe operating in positive electrospray mode. Optical rota-
tion was carried out using a Perkin-Elmer polarimeter 341. The
Peptide BEH C18 OBD Prep Column (130 Å, 5
m
m, 19 mm ꢂ 100 mm)
ꢁ
ꢁ
m
m, 2.1 ꢂ 50 mm; column temp: 40 C; flow:
both operating at 20 C and a flow rate of 20 mL/min, a Waters
Photodiode Array Detector (detecting at 210e600 nm), a Waters UV
Fraction Manager, and a Waters 2767 Sample Manager. Eluents A
.4 mL/min). Eluents A (0.1% HCO
2
H in H
2
O) and B (0.1% HCO
2
H in
2 2 2
(0.1% HCO H in H O) and B (0.1% HCO H in MeCN) were used in the
following gradient: 5% B to 70% B in 10 min, hold for 3.5 min, then
70% B to 100% B in 1.5 min, and hold 3 min (total run time: 20 min).
The purity of the compounds was assessed by RP-UPLC-UV and

N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
743
NMR, and purities ꢃ95% were considered acceptable for evaluation
purposes both in in vitro and in vivo assays.
2H), 7.21 (br. s, 2H), 6.83 (d, J ¼ 9.1 Hz, 2H), 4.83 (s, 2H), 3.24 (s, 3H),
13
(Litt. [45]). C NMR (101 MHz, DMSO‑d
6
) d 167.4, 162.7, 159.5, 151.8,
150.7, 148.9, 148.2, 131.0, 121.9, 117.7, 111.2, 54.8, 39.2. IR (neat) cm-
4
4
.2. Synthesis
1: 3341, 3189, 2956, 2823, 1660, 1599, 1294, 1188 (Litt. [45]).
.2.1. (S)-4-(((2,4-Diaminopteridin-6-yl)methyl)(methyl)amino)-
4.2.3. (S)-5-(Tert-butoxy)-4-(4-(((2,4-diaminopteridin-6-yl)
N-(1,5-dioxo-1,5-bis(2-oxothiazolidin-3-yl)pentan-2-yl)benzamide
methyl)(methyl)amino)-benzamido)-5-oxopentanoic acid (7)
(4)
To a solution of pteroic acid 6 (3.15 g, 8.71 mmol, 90% purity) in
anhydrous DMF (200 mL) were added PyBOP (6.55 g, 12.6 mmol)
and triethylamine (6.00 mL, 45.2 mmol) and the turbid reaction
To a solution of MTX (209 mg, 0.460 mmol) in anhydrous DMF
5 mL) were added DCC (209 mg, 1.01 mmol) and DMAP (124 mg,
.01 mmol) and the turbid reaction mixture was stirred for
(
1
ꢁ
mixture was stirred 30 min at 20 C. H-Glu (OMe)-OtBu hydro-
ꢁ
2
0 min at 20 C. Thiazolidin-2-one (95.1 mg, 0.922 mmol) was
chloride (2.58 g, 10.2 mmol) was added and the mixture was stirred
for another 16 h. The reaction mixture was filtered over celite,
concentrated in vacuo to 75 mL, and then poured slowly into a
vigorously stirred solution of ice-cooled diethyl ether (2 L). The
suspension was left on ice for 3 h and precipitate was collected by
filtration and washed with cold diethyl ether. The filter cake was
added and the turbid mixture was stirred overnight. The precipitate
was removed by filtration and the filtrate was purified by prepar-
ative HPLC to give the desired product 4 as a yellow solid (53.1 mg,
ꢁ
20
ꢁ
1
1
8%). mp: > 151 C (dec.); ½
a
ꢄ
¼ þ1.8 (c 0.73, DMSO); H NMR
D
(
400 MHz, DMSO‑d
6
)
d
8.57 (s, 1H), 8.25 (d, J ¼ 7.3 Hz, 1H), 7.72 (d,
J ¼ 8.9 Hz, 2H), 7.65 (br. s,1H), 7.43 (br. s,1H), 6.82 (d, J ¼ 9.0 Hz, 2H),
suspended in
Ba(OH)
ꢅ8H
2
a mixture of EtOH:H O 1:1 (150 mL) and
6
4
2
.61 (br. s, 2H), 5.28 (q, J ¼ 7.0 Hz, 1H), 4.78 (s, 2H), 4.14e3.98 (m,
H), 3.40 (t, J ¼ 7.3 Hz, 2H), 3.33 (td, J ¼ 7.2, 2.5 Hz, 2H), 3.21 (s, 3H),
.98 (dt, J ¼ 17.6, 7.5 Hz, 1H), 2.81 (dt, J ¼ 17.6, 7.6 Hz, 1H), 1.98 (q,
2
2
O (6.11 g, 13.4 mmol) was added. The suspension was
ꢁ
stirred for 24 h at 20 C and precipitate was then removed by
filtration. To the filtrate was added sat. aq. Na
2
SO
4
(25 mL), stored at
1
3
J ¼ 7.2 Hz, 2H); C NMR (101 MHz, DMSO‑d
6
)
d
173.4, 173.2, 172.8,
ꢁ
4
4
C for 5 h, and filtered over celite to remove precipitated BaSO .
1
72.0, 167.0, 163.4, 163.2, 155.6, 151.5, 149.6, 146.4, 129.5, 121.2 (2C),
The filtrate was acidified to pH 3 using 0.5 M HCl and the desired
product 7 was collected as an orange solid after washing H O, ether,
1
11.5, 55.3, 52.8, 47.7, 47.4, 40.6, 33.9, 25.7, 25.6, 25.3; IR (neat)
2
ꢀ
1
cm : 3344, 3325, 3066, 3034, 2953, 2924, 1702, 1644, 1236, 1153;
HRMS (ESI) calcd. for [C26
6
1
and pentane and drying in vacuo (4.48 g, 79%, 90% purity). H NMR
400 MHz, DMSO‑d 12.17 (br. s, 1H), 9.24 (br. s, 1H), 9.04 (br. s,
H), 8.72 (s, 1H), 8.56 (br. s, 1H), 8.21 (d, J ¼ 7.6 Hz, 1H), 7.74 (d,
29
H N
10
O
5
S
2
] [MþH]þ 625.1758, found
(
6
) d
25.1779.
1
J ¼ 9.0 Hz, 2H), 7.35 (br. s, 1H), 6.82 (d, J ¼ 9.0 Hz, 2H), 4.87 (s, 2H),
4.2.2. 4-(N-((2,4-Diaminopteridin-6-yl)methyl)-N-methylamino)
4
2
.29 (ddd, J ¼ 9.7, 7.5, 5.2 Hz,1H), 3.25 (s, 3H), 2.32 (t, J ¼ 7.5 Hz, 2H),
.06e1.96 (m, 1H), 1.90 (ddt, J ¼ 14.1, 9.3, 7.2 Hz, 1H), 1.39 (s, 9H),
benzoic acid (6)
13
6
(Litt. [46]). C NMR (101 MHz, DMSO‑d ) d 173.9, 171.5, 166.4, 162.8
(2C), 155.9, 151.2, 150.7, 148.8, 129.0, 122.3, 121.4, 111.2, 80.5, 54.9,
5
2
2.5, 40.2, 30.4, 27.7, 26.0, (Litt. [46]). IR (neat) cm-1: 3342, 3118,
976, 2931, 1716, 1639, 1601, 1506, 1364, 1152, 832.
4.2.4. tert-Butyl (S)-2-(4-(((2,4-diaminopteridin-6-yl)
methyl)(methyl)amino)-benzamido)-5-oxo-5-(2-oxothiazolidin-3-
yl)pentanoate (8)
Synthesis was performed following a published procedure [45].
Pteridine alcohol 5 hydrochloride (3.34 g, 14.6 mmol) was dissolved
ꢁ
in hot water (120 mL) and the solution was left to cool to 20 C. The
solution was neutralized with 1 M NaOH (ca. 12 mL) and the pre-
cipitate was collected by filtration, washed with water, and dried in
vacuo over P
2 5
O . The orange-beige solid was suspended in anhy-
drous DMA (20 mL) and triphenylphosphine dibromide (13.9 g,
3
2.9 mmol) was then added. The turbid reaction mixture was stir-
ꢁ
red for 20 h at 20 C and 4-(methylamino)benzoic acid (2.45 g,
1
6.2 mmol) and DIPEA (5.34 mL, 30.7 mmol) were added. The
ꢁ
turbid mixture was stirred for another 3 days at 20 C and then
To a solution of MTX-a-OtBu (7) (211 mg, 0.372 mmol, 90% pu-
poured into 0.33 M NaOH (190 mL). The precipitate was filtered off
and the filtrate was acidified to approximately pH 4.5 with 10%
AcOH in water (ca. 20 mL). The precipitate was collected by filtra-
tion, washed with water, triturated with hot methanol (23 mL), and
rity) in anhydrous DMF (5 mL) were added DCC (170 mg,
0.827 mmol) and DMAP (202 mg, 1.65 mmol) and the turbid reac-
tion mixture was stirred for 30 min at 20 C. Then, thiazolidin-2-
one (85.3 mg, 0.827 mmol) was added and the mixture was stir-
red another 16 h. Water (1 mL) was added, the precipitate was
removed by filtration, and the filtrate was concentrated and puri-
ꢁ
dried in vacuo over P
beige solid (4.25 g, 90%). H NMR (400 MHz, DMSO‑d
s, 1H), 8.66 (s, 1H), 8.42 (br. s, 1H), 8.19 (br. s, 1H), 7.73 (d, J ¼ 9.0 Hz,
2 5
O to give the title compound as an orange-
1
6
) d 12.2 (br.
fied by column chromatography (water/acetonitrile 1:19, R
f
¼ 0.28)

744
N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
to give the desired product 8 as a yellow solid (160 mg, 65%, 90%
compound was performed using RP-UPLC-UV (
l
¼ 306 nm). A
ꢁ
20
ꢁ
1
purity). mp: > 142 C (dec.); ½
a
ꢄ
¼ -0.7 (c 0.71, DMSO); H NMR
control experiment (no H addition but PBS, ‘PBS stability’) was
2 2
O
D
(
400 MHz, DMSO‑d
6
)
d
8.58 (s, 1H), 8.18 (d, J ¼ 7.5 Hz, 1H), 8.13 (s,
performed in parallel. Every prodrug activation assay was carried
out in triplicates.
1
H), 7.77 (br. s, 1H), 7.71 (d, J ¼ 9.0 Hz, 2H), 7.56 (br. s, 1H), 6.82 (d,
J ¼ 9.0 Hz, 2H), 6.71 (br. s, 2H), 4.79 (s, 2H), 4.27 (ddd, J ¼ 9.5, 7.5,
5
.6 Hz, 1H), 4.04 (t, J ¼ 7.3 Hz, 2H), 3.32 (td, J ¼ 7.2, 1.7 Hz, 3H), 3.21
4.3.2. Human plasma stability
(
(
s, 3H), 2.88 (td, J ¼ 7.4, 2.4 Hz, 2H), 2.16e2.01 (m, 1H), 2.00e1.88
Freeze-dried human plasma (2 mL, Sigma-Aldrich) was recon-
13
m, 1H), 1.39 (s, 9H); C NMR (101 MHz, DMSO‑d
71.9, 166.8, 163.5, 163.2, 154.9, 151.4, 149.6, 146.8, 129.4, 121.9,
21.5, 111.5, 80.9, 55.3, 52.9, 47.4, 40.5, 33.2, 28.1, 25.8, 25.3; IR
6
) d 173.2, 172.1,
stituted from Milli-Q water (2 mL), diluted with PBS (2 mL), and
1
1
ꢁ
incubated for 10 min at 37 C and 300 rpm. A 1 mg/mL stock solu-
tion of 3 in DMSO was prepared. 475
mL of pre-incubated human
ꢀ
1
(
neat) cm : 3325, 3183, 3116, 2974, 2924, 1635, 1604, 1506, 1446,
plasma solution was added to 25
samples were incubated at 37 C at 300 rpm along with a blank
mL prodrug stock solution and the
1
362, 1151; HRMS (ESI) calcd. for [C27
H
34
N
9
O
5
S] [MþH]þ 596.2398,
ꢁ
found 596.2397.
containing only human plasma (500
mL). A zero-time reference was
obtained by quenching an aliquot (50
m
L) in ice-cold MeCN (50 L)
m
4
.2.5. (S)-2-(4-(((2,4-Diaminopteridin-6-yl)methyl)(methyl)
amino)benzamido)-5-oxo-5-(2-oxothiazolidin-3-yl)pentanoic acid
3)
immediately after addition of plasma to the stock solution. The
quenched sample was vortexed for 30 s and the resulting precipi-
tate was removed by centrifugation (14,000 rpm, 5 min). The su-
pernatant was analyzed directly by RP-UPLC-MS. Measurements
were done at 30 min, 1 h, 2 h, 4 h, and 8 h following the same
procedure.
(
4.3.3. Microsomal stability
In quadruplicates, 10 mg/mL human liver microsome solution
ꢁ
(25
mL) was added to PBS (425
m
L) and incubated at 37 C and
1
000 rpm for 10 min. A 10 mM stock solution of compound 3 in
MTX-
a
-OtBu-
g
-TZ (8) (844 mg, 90% purity, 1.29 mmol) was
DMSO and a 30 mM stock solution of b-NADPH tetrasodium salt
ꢁ
dissolved in pure formic acid (60 mL) and stirred for 24 h at 20 C.
The reaction mixture was concentrated in vacuo and the crude was
dissolved in DMF (6 mL) and purified by preparative HPLC to give
the desired product 3 as a light orange solid (507 mg, 74%). mp.:
hydrate (27.2 mg/mL) in PBS were prepared. In quadruplicates,
prodrug stock solution (1.5 L) was added to the diluted and pre-
incubated microsome solution. Then, to three of the solutions
NADPH stock solution (25 L) was added while PBS (25 L) was
added to the remaining solution (negative control). Aliquots (50 L)
were removed after 0 min, 5 min, 10 min, 20 min, 30 min, and
0 min and added to ice-cold MeCN, vortexed for 10 s, and stored
m
m
m
ꢁ
20
D
ꢁ
1
>
184 C (dec.); ½
a
ꢄ
¼ -8.0 (c 0.48, DMSO); H NMR (400 MHz,
m
DMSO‑d
6
)
d
12.57 (br. s, 1H), 8.57 (s, 1H), 8.18 (d, J ¼ 7.7 Hz, 1H), 7.71
(
d, J ¼ 8.9 Hz, 2H), 7.65 (br. s, 1H), 7.44 (br. s, 1H), 6.82 (d, J ¼ 9.0 Hz,
4
2
4
2
5
1
5
H), 6.61 (br. s, 2H), 4.78 (s, 2H), 4.34 (ddd, J ¼ 9.9, 7.6, 4.9 Hz, 1H),
.13e3.96 (m, 2H), 3.32 (dt, J ¼ 7.3, 1.9 Hz, 2H), 3.20 (s, 3H),
.97e2.79 (m, 2H), 2.19e2.05 (m, 1H), 1.95 (dddd, J ¼ 13.8, 9.7, 8.0,
on ice for 40 min. Precipitated protein was removed by centrifu-
gation (14,000 rpm, 5 min) and the supernatants was analyzed by
UPLC-MS. The in vitro intrinsic clearance (CL'int in mL/min/kg pro-
tein) was calculated according to the following equation [38]:
.8 Hz, 1H); 13C NMR (101 MHz, DMSO‑d
) d 174.3, 173.2, 172.2,
6
66.8,163.6,163.3, 155.6,151.4, 149.7, 146.5, 129.4,121.9, 121.6, 111.5,
5.3, 52.1, 47.5, 40.7, 33.4, 25.9, 25.3; IR (neat) cm : 3325, 3137,
117, 2923, 2853, 1690, 1631, 1604, 1502, 1443, 1362, 1149; HRMS
ESI) calcd. for [C23
.3. In vitro assays
Phosphate buffered saline (PBS) was prepared by dissolving 1
ꢀ1
0
:693
ml incubation 45 mg microsomes
ꢂ
mg microsomes gm liver
0
CL
¼
ꢂ
3
(
int
in vitro T1=2
S] [MþH]þ 540.1772, found 540.1779.
26
H N
9
O
5
ꢂ 20 gm liver
kg b:w:
4
PBS tablet in Milli-Q water (200 mL) to give a solution of 0.01 M
4
.3.4. Thermodynamic solubility in PBS
A 1 mg/mL stock solution of 3 in DMF was prepared. A calibra-
tion curve was constructed at 25, 50, 75, 100, and 125 g/mL con-
phosphate buffer, 0.0027 M KCl, and 0.137 M NaCl with pH 7.4 at
ꢁ
2
5 C. PBS tablets, human liver microsomal fractions, freeze-dried
m
human plasma, and NADPH tetrasodium salt were purchased
from Sigma-Aldrich. Assays were run in Eppendorf tubes (1.5 mL)
and shaken using an Eppendorf Thermomixer C (1.5 mL). Analysis
of the assays was performed by RP-UPLC-MS (see above,
centrations of this stock solution and were analyzed by RP-UPLC-
MS (Figure S6, SI). Compound 3 (2 mg) was suspended in PBS
(1 mL) and sonicated for 10 min. The samples were incubated at
ꢁ
2
0 C and 1000 rpm for 15 h. The samples were centrifuged
l
¼ 306 nm). All experiments were performed in triplicates unless
(14,000 rpm, 1 min) and the supernatant was diluted in DMF (1:5)
stated otherwise.
and analyzed by RP-UPLC-MS.
4
.3.1. Prodrug activation under different oxidative conditions and
PBS stability
To aq. PBS buffer (750
drug (100 L, 1 mM in DMSO) followed by addition of internal
standard solution (diclofenac, 100 L, 1 mM in DMSO). The assay
was initiated by addition of a solution of H2O2 in PBS (50 L, 10 -
e1e0.5e0.25 mM) followed by vortex mixing. The resulting
4.4. Collagen type-II-induced arthritis
m
L, pH 7.4) was added a solution of pro-
m
Animals: DBA/1J mice (male, 8e9 weeks) were obtained from
Janvier, France. The mice were maintained in the animal house at
Redoxis, Medicon Village, Lund, Sweden, where they were accli-
matized for approximately one week before initiation of the
experiment. All animal experiments were approved by the local
animal ethic committee Malm o€ /Lund, Sweden, under the license
m
m
5
ꢁ
mixture was incubated at 37 C in a Eppendorf Thermomixer C
1.5 mL, 1000 rpm) and samples were taken after 5, 15, 30, 60,
0 min and 2, 4, 24 h. Analysis of the percentage of remaining
(
9
N165-15.

N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
745
Induction of disease: collagen-induced arthritis (CIA) was
induced by intradermal immunization with 100 g of chicken type-
II collagen (CII, Chondrex) in Complete Freund's Adjuvant (CFA,
Difco) on day ꢀ1 via subcutaneous injection approximately 0.5 cm
from the root of the tail. On day 21 a boost injection was admin-
Redox Signal. 8 (2008) 243e270.
[
13] C. Wittmann, P. Chockley, S.K. Singh, L. Pase, G.J. Lieschke, C. Grabher,
Hydrogen peroxide in inflammation: messenger, guide, and assassin, Adv.
Hematol (2012) 1e6.
m
[14] M. Gutowski, S. Kowalczyk, A study of free radical chemistry: their role and
pathophysiological significance, Acta Biochim. Pol. 60 (2013) 1e16.
[
15] M. Valko, D. Leibfritz, J. Moncol, M.T. Cronin, M. Mazur, J. Telser, Free radicals
and antioxidants in normal physiological functions and human disease, Int. J.
Biochem. Cell Biol. 39 (2007) 44e84.
16] M. Nagata, Inflammatory cells and oxygen radicals, Curr. Drug Targets -
Inflamm. Allergy 4 (2005) 503e504.
istered in the same way with 50
immunization injection, onset of disease started to be observed
day 26).
Anti-arthritic effect of test compounds and health evaluation:
mg CII. One week after the second
[
(
[
17] M. Mittal, M.R. Siddiqui, K. Tran, S.P. Reddy, A.B. Malik, Reactive oxygen
species in inflammation and tissue injury, Antioxidants Redox Signal. 20
mice were randomly divided in 3 groups (n ¼ 8 per group): group I
(2014) 1126e1167.
(
(
vehicle), group II (MTX, Sigma Aldrich, 7.0 mg/kg, i.p.), group III
MTX- -TZ, 8.3 mg/kg, i.p.). Vehicle or test compound (2% DMSO in
L) were dosed daily
[
18] C.A. Hitchon, H.S. El-Gabalawy, Oxidation in rheumatoid arthritis, Arthritis
g
Res. Ther. 6 (2004) 265.
PBS, Life Technologies, injection volume 370
m
[19] C.J. Wruck, A. Fragoulis, A. Gurzynski, L.-O. Brandenburg, Y.W. Kan, K. Chan,
J. Hassenpflug, S. Freitag-Wolf, D. Varoga, S. Lippross, T. Pufe, Role of oxidative
stress in rheumatoid arthritis: insights from the Nrf2-knockout mice, Ann.
Rheum. Dis. 70 (2011) 844e850.
[20] R. Weinstain, E.N. Savariar, C.N. Felsen, R.Y. Tsien, In vivo targeting of
hydrogen peroxide by activatable cell-penetrating peptides, J. Am. Chem. Soc.
intraperitoneally for 14 days, starting at onset of disease (day 27).
Disease was evaluated three times per week in a blinded fashion,
starting at day 18 until the end of the experiment (day 40). A
macroscopic scoring system of the four limbs ranging from 0 to 15
136 (2014) 1e18.
(
1 point per swollen toe, 1 point per swollen foot knuckle, and 5
[
21] R. Kumar, J. Han, H.J. Lim, W.X. Ren, J.Y. Lim, J.H. Kim, J.S. Kim, Mitochondrial
induced and self-monitored intrinsic apoptosis by antitumor theranostic
prodrug: in vivo imaging and precise cancer treatment, J. Am. Chem. Soc. 136
points for swollen ankle) was used, meaning a maximal score of 60
per mice. For ethical reasons and restrictions, mice with score
exceeding 45 were removed from the experiment. The general
health of mice was evaluated three times per week after disease
induction. As an indicator of general health, animal body weight
was used.
(2014) 17836e17843.
[
22] A. Boveris, S. Alvarez, J. Bustamante, L. Valdez, Measurement of superoxide
radical and hydrogen peroxide production in isolated cells and subcellular
organelles, Meth. Enzymol. 349 (2002) 280e287.
23] S. Mueller, Sensitive and nonenzymatic measurement of hydrogen peroxide in
biological systems, Free Radic. Biol. Med. 29 (2000) 410e415.
24] X. Peng, V. Gandhi, ROS-activated anticancer prodrugs: a new strategy for
tumorspecific damage, Ther. Deliv. 3 (2012) 823e833.
25] H. Hagen, P. Marzenell, E. Jentzsch, F. Wenz, M.R. Veldwijk, A. Mokhir, Ami-
noferrocene-based prodrugs activated by reactive oxygen species, J. Med.
Chem. 55 (2012) 924e934.
[
[
[
Acknowledgements
We are grateful for financial support from the Independent
Research Fund Denmark (grant no. 7017-00026), the Novo Nordisk
Foundation (Novo Scholarship to NSA) and the Technical University
of Denmark (PoC funding and a PhD scholarship for JPC). NSA, JPC,
CAH, JB and MHC are inventors on patents filed that are relevant to
these data. CAH, JB and MHC are co-founders of ROS Therapeutics
that licenses the technology. All other authors declare that there is
no conflict of interests in respect to the publication.
[
26] Y. Kuang, K. Balakrishnan, V. Gandhi, X. Peng, Hydrogen peroxide inducible
DNA cross-linking agents targeted, J. Am. Chem. Soc. 133 (2011)
19278e19281.
[27] W. Chen, K. Balakrishnan, Y. Kuang, Y. Han, M. Fu, V. Gandhi, X. Peng, Reactive
oxygen species (ROS) inducible DNA cross-linking agents and their effect on
cancer cells and normal lymphocytes, J. Med. Chem. 57 (2014) 4498e4510.
[
28] S. Cao, Y. Wang, X. Peng, ROS-inducible DNA cross-linking agent as a new
anticancer prodrug building block, Chem. Eur J. 18 (2012) 3850e3854.
[29] E.J. Kim, S. Bhuniya, H. Lee, H.M. Kim, C. Cheong, S. Maiti, K.S. Hong, J.S. Kim,
An activatable prodrug for the treatment of metastatic tumors, J. Am. Chem.
Soc. 136 (2014) 13888e13894.
Appendix A. Supplementary data
[30] K.B. Daniel, J.L. Major Jourden, K.E. Negoescu, S.M. Cohen, Activation of sul-
fonate ester based matrix metalloproteinase proinhibitors by hydrogen
peroxide, J. Biol. Inorg. Chem. 16 (2011) 313e323.
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.ejmech.2018.07.045.
[31] J.L. Major Jourden, S.M. Cohen, Hydrogen peroxide activated matrix metal-
loproteinase inhibitors: a prodrug approach, Angew. Chem. Int. Ed. 49 (2010)
6795e6797.
[
32] A.K. Vadukoot, S.F. Abdulsalam, M. Wunderlich, E.D. Pullen, J. Landero-Fig-
ueroa, J.C. Mulloy, E.J. Merino, Design of a hydrogen peroxide-activatable
agent that specifically targets cancer cells, Bioorg. Med. Chem. 22 (2014)
6885e6892.
References
[1] I.B. McInnes, G. Schett, The pathogenesis of rheumatoid arthritis, N. Engl. J.
Med. 365 (2011) 2205e2219.
[33] R.D. Hanna, Y. Naro, A. Deiters, P.E. Floreancig, Alcohol, aldehyde, and ketone
[2] J.S. Smolen, D. Aletaha, I.B. McInnes, Rheumatoid arthritis, Lancet 388 (2016)
liberation and intracellular cargo release through peroxide-mediated a-boryl
ether fragmentation, J. Am. Chem. Soc. 138 (2016) 13353e13360.
2023e2038.
[
[
3] D.M. Lee, M.E. Weinblatt, Rheumatoid arthritis, Lancet 358 (2001) 903e911.
4] M.E. Weinblatt, Methotrexate in rheumatoid arthritis: a quarter century of
development, Trans. Am. Clin. Climatol. Assoc. 124 (2013) 16e25.
[34] C. Perez, J.-P. Monserrat, Y. Chen, S.M. Cohen, Exploring hydrogen peroxide
responsive thiazolidinone-based prodrugs, Chem. Commun. (J. Chem. Soc.
Sect. D) 51 (2015) 7116e7119.
[35] J. Peir
oꢀ Cadahía, J. Bondebjerg, C.A. Hansen, V. Previtali, A.E. Hansen,
[
[
5] B.N. Cronstein, Low-dose Methotrexate : a mainstay in the treatment of
rheumatoid arthritis, Rev. Pharmacol 57 (2005) 163e172.
6] J. Grim, J. Chl aꢀ dek, J. Martínkov aꢀ , Pharmacokinetics and pharmacodynamics of
T.L. Andresen, M.H. Clausen, Synthesis and evaluation of hydrogen peroxide
sensitive prodrugs of methotrexate and aminopterin for the treatment of
rheumatoid arthritis, J. Med. Chem. 61 (2018) 3503e3515.
methotrexate in non-neoplastic diseases, Clin. Pharmacokinet. 42 (2003)
1
39e151.
[36] L. Di, E.H. Kerns, Y. Hong, H. Chen, Development and application of high
throughput plasma stability assay for drug discovery, Int. J. Pharm. (Amst.)
297 (2005) 110e119.
[37] E.H. Kerns, L. Di, Plasma stability, in: Drug-like Prop. Concepts, Struct. Des.
Methods from ADME to Toxic. Optim, first ed., Elsevier Inc., London, 2008,
pp. 169e177.
[38] R.S. Obach, Prediction of human clearance of twenty-nine drugs from hepatic
microsomal intrinsic clearance data: an examination of in vitro half-life
approach and nonspecific binding to microsomes, Drug Metab. Dispos. 27
(1999) 1350e1359.
[
[
[
7] B. Bannwarth, F. P eꢀ hourcq, T. Schaeverbeke, J. Dehais, Clinical pharmacoki-
netics of low-dose pulse methotrexate in rheumatoid arthritis, Clin. Phar-
macokinet. 30 (1996) 194e210.
8] J. Rautio, H. Kumpulainen, T. Heimbach, R. Oliyai, D. Oh, T. J €a rvinen,
J. Savolainen, Prodrugs: design and clinical applications, Nat. Rev. Drug Discov.
7
9] J.B. Zawilska, J. Wojcieszak, A.B. Olejniczak, Prodrugs: a challenge for the drug
(2008) 255e270.
development, Pharmacol. Rep. 65 (2013) 1e14.
[
10] J. Rautio, N.A. Meanwell, L. Di, M.J. Hageman, The expanding role of prodrugs
and development, Nat. Rev. Drug Discov. (2018), https://doi.org/10.1038/
nrd.2018.46.
11] S.S. Abolmaali, A.M. Tamaddon, R. Dinarvand, A review of therapeutic chal-
lenges and achievements of methotrexate delivery systems for treatment of
cancer and rheumatoid arthritis, Canc. Chemother. Pharmacol. 71 (2013)
[39] E.H. Kerns, L. Di, Solubility, in: Drug-like Prop. Concepts, Struct. Des. Methods
from ADME to Toxic. Optim, first ed., Elsevier Inc., London, 2008, pp. 56e85.
[40] M.F. Neurath, K. Hildner, C. Becker, J.F. Schlaak, K. Barbulescu, T. Germann,
E. Schmitt, P. Schirmacher, S. Haralambous, M. Pasparakis, K.H. Meyer Zum
[
Büschenfelde, G. Kollias, E.
modulates cytokine production by
collagen-induced arthritis (CIA): a mechanism for methotrexate-mediated
Marker-Hermann, Methotrexate specifically
€
1
115e1130.
T cells and macrophages in murine
[
12] J.R. Stone, S. Yang, Hydrogen peroxide: a signaling messenger, Antioxidants

746
N.S. Andersen et al. / European Journal of Medicinal Chemistry 156 (2018) 738e746
immunosuppression, Clin. Exp. Immunol. 115 (1999) 42e55.
41] S. Sardar, Å. Andersson, Old and new therapeutics for Rheumatoid Arthritis:
in vivo models and drug development, Immunopharmacol. Immunotoxicol.
2016) 2e13.
[44] M. Colucci, F. Maione, M.C. Bonito, A. Piscopo, A. Di Giannuario, S. Pieretti,
[
New insights of dimethyl sulphoxide effects (DMSO) on experimental in vivo
models of nociception and inflammation, Pharmacol. Res. 57 (2008) 419e425.
[45] S.M. Robillard, H. Cruell, Patent: US 2010/0227798 A1, 2010.
[46] A. Warnecke, I. Fichtner, G. Saß, F. Kratz, Synthesis, cleavage profile, and
antitumor efficacy of an albumin-binding prodrug of methotrexate that is
cleaved by plasmin and cathepsin B, Arch. Pharm. (Weinheim) 340 (2007)
389e395.
(
[42] D.E. Trentham, Collagen arthritis as a relevant model for rheumatoid arthritis,
Arthritis Rheum. (1982) 911e916.
[
43] T. Keleva, I. Cavar, F. Culo, Biological actions of drug solvent, Period. Biol. 113
(
2011) 311e320.