Halogenated trimethoprim derivatives as multidrug-resistant Staphylococcus aureus therapeutics

Article abstract of DOI:10.1016/j.bmc.2018.05.019

Incorporation of halogen atoms to drug molecule has been shown to improve its properties such as enhanced in membrane permeability and increased hydrophobic interactions to its target. To investigate the effect of halogen substitutions on the antibacterial activity of trimethoprim (TMP), we synthesized a series of halogen substituted TMP and tested for their antibacterial activities against global predominant methicillin resistant Staphylococcus aureus (MRSA) strains. Structure-activity relationship analysis suggested a trend in potency that correlated with the ability of the halogen atom to facilitate in hydrophobic interaction to saDHFR. The most potent derivative, iodinated trimethoprim (TMP-I), inhibited pathogenic bacterial growth with MIC as low as 1.25 μg/mL while the clinically used TMP derivative, diaveridine, showed resistance. Similar to TMP, synergistic studies indicated that TMP-I functioned synergistically with sulfamethoxazole. The simplicity in the synthesis from an inexpensive starting material, vanillin, highlighted the potential of TMP-I as antibacterial agent for MRSA infections.

Full text of DOI:10.1016/j.bmc.2018.05.019

Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Halogenated trimethoprim derivatives as multidrug-resistant Staphylococcus
aureus therapeutics
Napon Nilchana, Wanida Phetsanga, Taechin Nowwarata, Soraya Chaturongakulb,c
,
Chutima Jiarpinitnuna,⁎
a
Department of Chemistry and Center for Innovation in Chemistry (PERCH-CIC), Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400, Thailand
Department of Microbiology, Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400, Thailand
b
c
Center for Emerging Bacterial Infections, Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400, Thailand
A R T I C L E I N F O
A B S T R A C T
This manuscript is dedicated to Professor Laura
L. Kiessling in celebration of her career
contributions to organic chemistry.
Incorporation of halogen atoms to drug molecule has been shown to improve its properties such as enhanced in
membrane permeability and increased hydrophobic interactions to its target. To investigate the effect of halogen
substitutions on the antibacterial activity of trimethoprim (TMP), we synthesized a series of halogen substituted
TMP and tested for their antibacterial activities against global predominant methicillin resistant Staphylococcus
aureus (MRSA) strains. Structure-activity relationship analysis suggested a trend in potency that correlated with
the ability of the halogen atom to facilitate in hydrophobic interaction to saDHFR. The most potent derivative,
iodinated trimethoprim (TMP-I), inhibited pathogenic bacterial growth with MIC as low as 1.25 μg/mL while the
clinically used TMP derivative, diaveridine, showed resistance. Similar to TMP, synergistic studies indicated that
TMP-I functioned synergistically with sulfamethoxazole. The simplicity in the synthesis from an inexpensive
starting material, vanillin, highlighted the potential of TMP-I as antibacterial agent for MRSA infections.
Keywords:
Trimethoprim
Methicillin resistant Staphylococcus aureus
Halogenation
1. Introduction
pathway, namely dihydrofolate reductase (DHFR) and dihydropteroate
synthase (DHPS), respectively.6,7 Disruption of folate pathways inhibits
Staphylococcus aureus (S. aureus) is an opportunistic Gram-positive
bacterial pathogen accounted for one of the major public health pro-
blems worldwide.1,2 The key concern of S. aureus infection is due to
rapid development of antibiotic resistance that could complicate the
treatment. Methicillin-resistant S. aureus (MRSA), in particular, ac-
quires resistance to commonly used β-lactam antibiotics such as peni-
cillin. Since its first report in 1961, MRSA infection has now been
considered endemic. The impact of infection is substantial. In the
United States alone, MRSA infection results in an estimate of 19,000
deaths in addition to $3–4 billion antibiotics and hospitalization cost
annually.3 Hence, the developments of effective antibiotics and strate-
gies to combat resistance bacteria are called for.
The combination of antibiotics has been one of the treatment
choices for infections that caused by bacteria with high mutation
rate.4,5 Combination of trimethoprim (TMP)-sulfamethoxazole (SMZ) or
co-trimethoxazole is a successful example of combination antibiotic
therapy, recommended as first-line of treatments for urinary tract in-
fections, soft tissue infections, and MRSA infection, particularly com-
munity-associated MRSA (CA-MRSA). TMP and SMZ are potent and
selective inhibitors of two key enzymes in the folate biosynthetic
productions of key intermediates for DNA nucleoside and protein
syntheses, eventually leading to cell death. Folate pathway has been
widely focused as therapeutic target for antibacterial agent develop-
ment due to the difference in folate sources in pathogens and in
mammalians.8 Bacteria solely depend on the biosynthesis of folate
pathway while mammalian host could uptake folates from nutrients.
Therefore, blocking sequential steps in the folate synthesis pathway by
SMZ and TMP could lead to death of pathogens and not perturb the
mammalian host. TMP and SMZ have been shown to function sy-
nergistically, resulting in dramatic enhancement in antibacterial ac-
tivity.9 However, the rises of SMZ-resistance, TMP-resistance, and re-
sistance to the combination therapy have called for new antibacterial
agents.10,11
RAB1, a phthalazine analogue of TMP, was reported as a DHFR
inhibitor that exhibited broad spectrum activity with exceptional ef-
fectiveness against S. aureus and MRSA (Fig. 1).12 The reported X-ray
crystal structure of S. aureus DHFR (saDHFR) with trimethoprim and
with RAB1 suggested a unique shallow surface cavity that could be
exploited in drug development.12 The co-crystal complex structure with
DHFR also indicated that the F98Y mutation13, which commonly found
⁎
Corresponding author.
E-mail address: chutima.jia@mahidol.ac.th (C. Jiarpinitnun).
https://doi.org/10.1016/j.bmc.2018.05.019
Received 28 February 2018; Received in revised form 10 May 2018; Accepted 13 May 2018
0968-0896/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Nilchan,N.,Bioorganic&MedicinalChemistry(2018),https://doi.org/10.1016/j.bmc.2018.05.019

N. Nilchan et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
2,4-diamine), commonly known as diaveridine, of which a hydrogen
atom was in place of a halogen, was also prepared for antibacterial
activity comparison. The synthesis of diaveridine was achieved via a
similar methodology using commercially available 3,4-dimethox-
ybenzaldehyde. All compounds were purified and characterized via
nuclear magnetic spectroscopy and mass spectroscopy, as reported in
the experimental section.
NH2
N
O
NH2
OCH3
N
N
N
N
H2N
OCH3
OCH3
H2N
N
OC
H
3
OCH3
TMP
RAB1
Fig. 1. Structure of TMP and TMP derivative RAB1.
2.2. Susceptibility testing
in TMP-resistance strains, has minimal effects on saDHFR binding to
RAB1, highlighting the shallow surface cavity of saDHFR as an excellent
target for MRSA therapeutic development.
All trimethoprim derivatives—TMP-Cl, TMP-Br, TMP-I, TMP, and
diaveridine—were subjected to susceptibility assays against methicillin
susceptible S. aureus (MSSA) control strains (ATCC25923 and
ATCC29213), global predominant CA-MRSA PFGE strain USA300
(SF8300), and highly resistant hospital associated-MRSA (HA-MRSA)
strain COL. Rapid transmission and resistance towards antibiotics of
MRSA USA300 and COL strains lead to global healthcare problem.16,17
Hence, the effective antibiotics against these pathogenic strains are
searched for.
The synthesized compounds were first quantitatively screened for
their ability to inhibit growth of pathogenic bacteria via disk diffusion
assay. We tested all compounds at 5 µg as recommended by Clinical and
Laboratory Standards Institute (CLSI) for TMP.18 The conventional
antibiotic diaveridine exhibited incomplete inhibition zone when tested
with MRSA USA300 and no inhibition zone when tested with COL
strain. The results suggested that CA-MRSA USA300 and HA-MRSA
strain COL are diaveridine-resistant. TMP resulted in complete inhibi-
tion zone in the range of > 16 mm (disk diffusion breakpoint for sus-
ceptibility suggested by CLSI) when tested with MSSA ATCC25923 and
CA-MRSA; however, the results with HA-MRSA strain COL showed in-
complete inhibition zone. Similar to TMP, all halogenated derivatives
could completely inhibit growth of MSSA and CA-MRSA strain USA300
but incompletely inhibit HA-MRSA strain COL. Among halogenated
series, TMP-I showed the highest inhibition zone (> 20 mm), followed
by TMP-Br and TMP-Cl respectively. The trend is consistently observed
in all three strains tested (Table 1).
With regards to chemical structures, the difference between dia-
veridine, TMP, and our synthesized halogenated TMP derivatives are
subtle, only at meta-position of the benzene ring. However, this small
difference in structure leads to significant change in the antibacterial
activity. Replacing –H of diaveridine with electron donating –OMe
moiety or with -halogens group at meta-position could significantly
enhance the antibacterial activity, as evidently shown in > 15 mm in-
hibition zone in TMP and halogenated derivatives and < 9 mm in dia-
veridine. The results suggested that substitution at this particular po-
sition could change the antibacterial activity outcome.
All compounds were subjected to microdilution assay to quantita-
tively evaluate their antibacterial activity against MSSA and MRSA. The
results are shown in Table 1. The minimum inhibitory concentration
(MIC) and minimum bactericidal concentration (MBC) determined from
microdilution assay also suggested similar trend in antibacterial po-
tency correlated to the extents of inhibition zone observed in disk dif-
fusion assay. The substituent at meta-position of the benzene ring was
suggested key for MSSA and MRSA growth inhibitory activity. MSSA
ATCC29213 showed susceptibility to all TMP derivatives while HA-
MRSA COL strain exhibited resistant to diaveridine as its MIC was
clearly above the breakpoint of 5 µg/mL suggested by CLSI.18 The po-
tency of diaveridine was noticeably less potent (approximately 4-folds)
in all strains tested. When substituted hydrogen atom at meta-position
of diaveridine with halogens, the antibacterial potency against MSSA
and MRSA strains were enhanced. All strains showed susceptibility to
our synthesized halogenated derivatives (TMP-Cl, TMP-Br, and TMP-I).
Interestingly, substitution with iodide led to the most potent S. aureus
growth inhibitor analogue, even more active than the parent antibiotics
TMP. Substitution with smaller halogens such as bromine and chlorine
resulted in less active growth inhibitors. As clearly shown when tested
The unique shallow surface cavity of saDHFR contains the electron-
rich sites from side chains and backbones of amino acids; therefore, we
envisioned that incorporation of halogens into DHFR inhibitor structure
may allow better interactions of the inhibitor with these surrounding
electron-rich via halogen bonds interactions. Halogens could facilitate
in the non-directional hydrophobic interactions that could enhance the
binding contact.14 Here in, we investigated the effect of halogens sub-
stitution on the antibacterial activity of TMP against S. aureus and
MRSA.
2. Results and discussion
2.1. Chemistry
A series of halogenated derivatives of TMP—chlorinated-, bromi-
nated-, and iodinated- (abbreviated as TMP-Cl, TMP-Br, and TMP-I,
respectively)—were synthesized. The syntheses were based on pre-
viously reported method as shown in Scheme 1.15 In brief, an in-
expensive commercially available vanillin was used as a starting ma-
terial to undergo halogenation with various halogen sources. N-
halosuccinimides such as N-chlorosuccinimide (NCS) and N-bromo-
succinimide (NBS) were used for chlorination and bromination, re-
spectively. The desired products 1a and 1b were obtained in moderate
to high yield. For the electrophilic aromatic iodination of vanillin, we
employed potassium iodide (KI) as the iodine source and the common
household bleach solution as an oxidizing agent. 5-iodovanillin (1c)
was obtained in excellent yield. Subsequently, the halogenated vanillin
(1a–1c) was subjected to o-methylation reaction by reacting with me-
thyl iodide (MeI) in the presence of tetrabutylammonium iodide (TBAI).
Compound 2a–2c were obtained in 96–98% yield. Condensation of the
halogenated dimethoxy benzaldehyde adduct (2a–2c) with 3-morpho-
linopropionitrile using sodium methoxide in DMSO afforded alkene
product, which were sequentially cyclized by treating with a guanidine
hydrochloride to yield the corresponding halogenated derivatives of
TMP in moderate yield (47–60%). 5-(3,4-dimethoxybenzyl)pyrimidine-
Scheme 1. Syntheses of halogenated TMP derivatives.
2

N. Nilchan et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Table 1
Antibacterial activities of TMP derivatives determined by disk diffusiona and microdilution susceptibility test.b
Compounds
Diameter of Inhibition zones (mm)
MIC (μg/mL)
MBC (μg/mL)
IC50 (μM)
ATCC25923
SF8300
COL
ATCC29213
SF8300
COL
ATCC29213
SF8300
COL
diaveridine
TMP
TMP-Cl
TMP-Br
TMP-I
8.81
0.56
0.94
0.60
0.93
0.03
6 (+++)
6
5
5
1.25
2.5
1.25
1.25
10
5
5
5
2.5
5
10
2.5
2.5
1.25
1.25
20
10
10
5
> 1000
–
467
321
266
20.19
15.06
17.37
21.01
22.80
18.11
20.39
22.82
0.92
6 (++)
6 (+)
6 (++)
6 (+++)
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
0.61
0.83
0.62
5
a
5 µg of test compounds in DMSO were used for disk diffusion testing. At least three independent experiments were performed and the average diameters of
complete inhibition zones were reported with standard errors. Whatman disk has a diameter of 6 mm; therefore, average inhibition zone of 6 mm indicates inability
to inhibit growth. Incomplete inhibition zones were defined as +, ++, and +++, depending on the thickness of bacterial growth visualized. The incompletion of
+++ was given to a zone with the least thickness of bacterial growth, suggesting near completion of the inhibition zone. Meanwhile, + was giving to that appeared
to be more in thickness, suggesting near resistance.
b
Each compound was 2-fold diluted based on its stock concentration.
against highly resistant HA-MRSA COL strain, TMP-I was the most
potent among all the derivatives with MIC of 2.5 µg/mL; other halo-
genated derivatives and even the clinically used TMP resulted in MIC at
the borderline susceptibility of 5 µg/mL. Noticeably, they still resulted
in better activity than diaveridine.
synergistic results were also within the FICI of TMP with SMZ. Note that
TMP-I exhibited slightly better synergism with lowest FICI of 0.09 when
tested with MSSA control strain. TMP-I could reduce the MIC of SMZ
when used in combination up to 30 folds in the case of MSSA and 16
folds in both CA-MRSA and HA-MRSA.
The synthesized TMP derivatives were subjected to cytotoxicity
experiments with A549 cells. The IC50 values of TMP-Cl, TMP-Br, and
TMP-I were 467 µM, 321 µM, and 266 µM, respectively. Even though
these halogen-containing adducts exhibited higher toxicity than dia-
veridine (IC50 > 1000 µM), their IC50 values were significantly above
their MIC and MBC values, suggesting that these compounds potentially
would not be toxic to treated cells at the concentrations required for
inhibiting growth or killing the pathogen.
3. Conclusion
In conclusion, in this study we incorporated various halogens into
trimethoprim structures to exploit the potential halogen bond interac-
tions with saDHFR. The susceptibility of MSSA and global epidemic
MRSA USA300 strain SF8300 and COL strains to halogenated TMP
derivatives showed high potency in growth inhibitory activity, parti-
cularly TMP-I that inhibited growth of all strains with MIC in the range
of 1.25–2.5 μg/mL. In comparison to the clinically used antibiotics
diaveridine, incorporation of halogen could significantly enhance the
ability to inhibit growth of MSSA and MRSA. Similar to TMP, all ha-
logenated TMP could function synergistically with sulfamethoxazole.
Interestingly, we observed a trend in potency that correlates to the
ability of halogens to participate halogen bonding with Ser49 side chain
of the targeted saDHFR binding site. TMP-I was more potent than TMP-
Br and TMP-Cl. The binding activity of these halogenated TMP deri-
vatives to saDHFR and the potential role of halogen bond interactions
to the protein should be further studied and reported in due course.
To better understand the trend in potency observed, we performed
molecular docking of the halogen-substituted derivatives with saDHFR
using Schrӧdinger Maestro suites (Schrӧdinger suites 2017-4). The
binding was compared to the binding of TMP and diaveridine, known
inhibitors of DHFR. The structures of all compounds were constructed
from TMP structure and their geometries were subsequently optimized
using LigPrep in Schrӧdinger program suite. The structure of S. aureus
DHFR was obtained from Protein Data Bank (PDB: 2W9H). The binding
modes of halogenated TMP derivatives in the binding site of saDHFR
were evaluated using Glide dock in standard precision (SP) with core
constrains based on co-crystal structure of TMP and saDHFR.12 The
docking analyses showed that the diaminopyrimidine (DAP) ring of all
derivatives retain key hydrogen bonding interactions with Leu5, Asp27
and Phe92 residues. Interestingly, the methoxy group at meta-position
of TMP forms a contact interaction with Ser49 side chain. In case of
diaveridine, this particular interaction with Ser49 was absent. Halogen-
substituted derivative exhibited halogen bonding interaction with
Ser49. TMP-I showed strongest halogen bond with 2.28 Å in length,
followed by TMP-Cl, and TMP-Br, respectively (see Supporting in-
formation). Interestingly, the trend in halogen bonding strength cor-
relates with the antibacterial potency observed. Hence, the potential
interaction of the substituent at the meta-position of TMP with Ser49
may affect the antibacterial potency against MSSA and MRSA.
In the treatment of S. aureus and MRSA infections, TMP is com-
monly used in combination with sulfonamide antibiotics such as sul-
famethoxazole (SMZ). Such combination exploits the exceptional sy-
nergistic effect of the two compounds.6,7 We, therefore, investigated
whether our synthesized halogenated derivatives possess similar sy-
nergistic effects with SMZ. Synergistic studies were performed using
checkerboard method against reference strain ATCC29213, CA-MRSA
strain USA300, and HA-MRSA COL strain. The results were shown in
the Supplementary materials. Based on the fractional inhibitory con-
centration index (FICI), the combination of halogenated TMP and SMZ
exhibited synergistic effect with FICI of less than 0.5 (calculated FICI
index is in the range of 0.09–0.18) in all strains tested. These excellent
4. Experimental
4.1. Chemistry
All materials, unless otherwise noted, were obtained from com-
mercial suppliers and used as provided. Reactions were conducted
under an inert atmosphere of nitrogen. The progress of reactions was
monitored by thin layer chromatography (TLC). 1H NMR spectra were
obtained using either a Bruker DPX-300 (300 MHz), or a Bruker Avance
300, or a Bruker Avance-500 spectrometer. 13C NMR spectra were ob-
tained using a Bruker Avance-300 spectrometer (75 MHz) spectrometer
or a Bruker Avance-500 spectrometer (125 MHz). Chemical shifts were
reported relative to residual solvent signals in parts per million (δ)
(CDCl3: 1H: δ 7.27, 13C: δ 77.23; DMSO‑d6: 1H: δ 2.50, 13C: δ 39.51). 1H
NMR data were assumed to be first order with apparent doublets and
triplets reported as d and t, respectively. Multiplets were reported as m,
and resonances that appear broad were designated as br. High-resolu-
tion mass spectra (HRMS) were obtained on a Bruker micro TOF
spectrometer in the ESI and/or APCI modes.
4.1.1. Synthesis of halogenated vanillin 1a–1c
4.1.1.1. Synthesis
(1a). To a solution of vanillin (2.00 g, 13.1 mmol) in glacial acetic
of
3-chloro-4-hydroxy-5-methoxybenzaldehyde
3

N. Nilchan et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
acid (20 mL), N-chlorosuccinimide (NCS) (1.76 g, 13.1 mmol) was
added. The reaction was stirred at room temperature for 16 h. The
reaction was extracted with CH2Cl2. An organic layer was washed with
H2O followed by saturated NaCl solution. The resulting organic layer
was dried over anhydrous Na2SO4 and concentrated to dryness to afford
an off-white solid. The solid was washed with CHCl3. Upon filtration, an
off-white precipitated crystal product (1a) was obtained (1.14 g, 47%).
1H NMR (300 MHz, DMSO‑d6) δ: 3.89 (s, 3H), 6.45 (s, 1H), 7.38 (d,
J = 1.77 Hz, 1H), 7.57 (d, J = 1.79 Hz, 1H), 9.76 (s, 1H); 13C NMR
(75 MHz, DMSO‑d6) δ: 56.32, 109.16, 120.05, 125.64, 128.18, 148.79,
(s, 1H); 13C NMR (75 MHz, CDCl3) δ: 56.21, 60.92, 109.44, 125.74,
128.84, 132.39, 150.76, 154.31, 189.96; ESI-MS calc for C9H9ClNaO3
[M+Na+ +: 223.0132, found 223.0131.
]
4.1.2.2. 3-Bromo-4,5-dimethoxybenzaldehyde (2b). White solid product
was obtained in 98% yield. 1H NMR (300 MHz, CDCl3) δ: 3.86 (s, 3H),
3.88 (s, 3H), 7.31 (d, J = 1.76 Hz, 1H), 7.58 (d, J = 1.78 Hz, 1H), 9.77
(s, 1H); 13C NMR (75 MHz, CDCl3) δ: 56.24, 60.80, 110.17, 117.92,
128.73, 133.06, 151.83, 154.185, 189.80; ESI-MS calc for C9H9BrNaO3
[M+Na+ +: 266.9627, found 266.9626.
]
148.94, 190.49; ESI-MS calc for C8H7ClNaO3 [M+Na+
found 208.9975.
]
+: 208.9976,
4.1.2.3. 3-Iodo-4,5-dimethoxybenzaldehyde (2c). White solid product
was obtained in 98% yield. 1H NMR (300 MHz, CDCl3) δ: 3.91 (s,
6H), 7.83 (d, J = 1.39 Hz, 1H), 7.83 (d, J = 1.53 Hz, 1H), 9.81 (s, 1H);
13C NMR (75 MHz, CDCl3) δ: 56.09, 60.63, 92.09, 111.04, 133.91,
134.67, 152.98, 154.15, 189.68; ESI-MS calc for C9H9INaO3 [M
4.1.1.2. Synthesis
of
3-bromo-4-hydroxy-5-methoxybenzaldehyde
(1b). To a solution of vanillin (2.00 g, 13.1 mmol) in glacial acetic
acid (20 mL), N-bromosuccinimide (NBS) (2.57 g, 14.6 mmol) was
added. The reaction was stirred at room temperature for 3 h. The
reaction was extracted with CH2Cl2. An organic layer was washed with
H2O followed by saturated NaCl solution. The resulting organic layer
was dried over anhydrous Na2SO4 and concentrated to under vacuum to
give an orange to brown solid. The solid was washed with CHCl3. An
orange crystal of product (1b) was filtered and washed with cold CHCl3
(2.29 g, 75%). 1H NMR (300 MHz, DMSO‑d6) δ: 4.00 (s, 3H), 6.59 (s,
1H), 7.38 (d, J = 1.70 Hz, 1H), 7.65 (d, J = 1.71 Hz, 1H), 9.80 (s, 1H);
13C NMR (75 MHz, DMSO‑d6) δ: 189.66, 148.87, 147.65, 130.06,
129.99, 108.16, 108.00, 56.59; ESI-MS calc for C8H7BrNaO3 [M
+Na+ +: 314.9489, found 314.9487.
]
4.1.3. General procedure for cyclization of 2,4-diamino pyrimidine ring
(TMP-Cl, TMP-Br, and TMP-I)
The condensation reactions of halogenated dimethoxy benzalde-
hyde adducts (2a, 2b, or 2c) and 3-morpholinopropio nitrile (3) were
conducted based on the previously reported procedure.15 Compound 3
(11.49 mmol), prepared based on previously reported procedure.20 was
dissolved in DMSO (3.0 mL) and heated to 65 °C. Into a preheated so-
lution, a freshly prepared sodium methoxide (4.59 mmol) was slowly
added and stirred for 45 min at 80 °C. Then, a solution of 2 (9.19 mmol)
in DMSO (5.0 mL) was added into the heated solution. The reaction was
stirred at 80 °C for 16 h. The reaction was quenched with diluted HCl
(1:2 of 1 M HCl in H2O) and extracted with CH2Cl2. An organic layer
was subsequently washed with water and saturated NaCl solution. The
washed organic layer was dried over Na2SO4. The solvent was removed
under reduced pressure, affording dark brown viscous liquid. The crude
product was partially purified by column chromatography to give a
yellow to brown viscous liquid mixture. The mixture was carried on to
the next step.
The cyclization reactions with guanidine were carried out based on
previously reported procedure.15 In brief, into a solution mixture of
guanidine hydrochloride (23.59 mmol) and NaOEt (23.59 mmol) in dry
EtOH (6.0 mL), a solution of halogenated dimethoxybenzyl morpholi-
noacrylonitrile derivative (4.72 mmol) in the mixed solvent of dry EtOH
(8.0 mL) and dry DMSO (3.0 mL) was added. The reaction mixture was
stirred under refluxing temperature of 80 °C for 16 h. Then, EtOH was
removed by rotary evaporation under reduced pressure. The remaining
solution was quenched with diluted HCl (1:3 of 1 M HCl in H2O). The
quenched reaction mixture was extracted with CH2Cl2. The combined
organic layers were washed with H2O and saturated NaCl solution,
respectively. The resulting solution was dried over Na2SO4, and con-
centrated to dryness to obtain crude product as dark brown viscous
liquid. The crude product was purified by column chromatography to
give partially pure product, which was then further purified by re-
crystallization in MeOH. The crystal product was washed with 1:1
CH2Cl2/Hexane to yield the desired product.
+Na+ +: 252.9471, found 252.9473.
]
4.1.1.3. Synthesis of 3-iodo-4-hydroxy-5-methoxybenzaldehyde (1c). To a
solution of vanillin (0.50 g, 3.29 mmol) in EtOH (70 mL), KI (0.61 g,
3.65 mmol) was added and stirred to mix. The mixture was cooled
down to 0 °C with an ice bath. Household bleach solution (CLOROX®)
(12.5 mL, equivalent to sodium hypochlorite 8.23 mmol) was added
dropwise into the solution mixture. After addition was complete, the
reaction was stirred at room temperature for 1 h as EtOH was removed
in vacuo. The remaining aqueous solution was neutralized with diluted
HCl (1:3 ratio of 1 M HCl to H2O). Subsequently, the reaction mixture
was extracted with CH2Cl2. An organic layer was washed with H2O
followed by saturated NaCl solution, then dried over Na2SO4. The
solution was concentrated to dryness to yield crude product as
yellowish solid powder. The powder was washed with CHCl3. A white
to off-white powder product (1c) was obtained upon filtration and
washing with cold chloroform (0.74 g, 80%). 1H NMR (300 MHz,
DMSO‑d6) δ: 3.98 (s, 3H), 6.69 (s, 1H), 7.39 (s, 1H), 7.83 (s, 1H),
9.78 (s, 1H); 13C NMR (75 MHz, DMSO‑d6) δ: 56.19, 84.12, 110.10,
130.00, 134.76, 147.30, 152.20, 190.23; ESI-MS calc for C8H7INaO3 [M
+Na+ +: 300.9332, found 300.9399.
]
4.1.2. General procedure for methylation of 1a–1c
The methylation reaction of halogenated vanillin derivative (1a–1c)
were generally conducted based on the previously reported proce-
dure.19 In brief, into a solution of halogenated vanillin (1a, 1b, or 1c)
(10.17 mmol) in CH2Cl2 (135.0 mL), an aqueous solution of 1.6 M
NaOH (75.0 mL) and tetrabutylammonium iodide (TBAI) (15.25 mmol)
were added. The reaction was stirred until a clear solution was ob-
served. Methyl iodide (120.5 mmol) was added to a solution mixture
and stirred for 16 h at room temperature. The reaction was quenched
with 6 M HCl and extracted with CH2Cl2. The organic layer was sub-
sequently washed with H2O followed by saturated NaCl solution. The
solution was subsequently dried over Na2SO4 and concentrated to
dryness, affording crude product as yellow solid. The crude product was
purified by column chromatography to give white solid product 2a–2c.
4.1.3.1. 5-(3-Chloro-4,5-dimethoxybenzyl)pyrimidine-2,4-diamine (TMP-
Cl). A light brown crystal product was obtained in 47% yield. 1H NMR
(300 MHz, DMSO‑d6) δ: 3.53 (s, 2H), 3.66 (s, 3H), 3.77 (s, 3H), 5.72 (s,
2H), 6.12 (2, 2H), 6.94 (d, J = 1.90 Hz, 1H), 6.96 (d, J = 1.88 Hz, 1H),
7.55 (s, 1H); 13C NMR (75 MHz, DMSO‑d6) δ: 31.95, 55.95, 59.96,
105.17, 112.91, 116.26, 123.40, 138.17, 143.74, 153.07, 156.00,
+
162.12, 162.35; ESI-MS calc for
295.0956, found 295.0954.
C
13H15ClN4NaO2 [M+H+
]
:
4.1.2.1. 3-Chloro-4,5-dimethoxybenzaldehyde (2a). White solid product
was obtained in 96% yield. 1H NMR (300 MHz, CDCl3) δ: 3.86 (s, 1H),
3.88 (s, 1H),7.27 (d, J = 1.69 Hz, 1H), 7.42 (d, J = 1.79 Hz, 1H), 9.76
4.1.3.2. 5-(3-Chloro-4,5-dimethoxybenzyl)pyrimidine-2,4-diamine (TMP-
Br). A dark brown crystal product was obtained in 49% yield. 1H NMR
(300 MHz, DMSO‑d6) δ: 3.53 (s, 2H), 3.68 (s, 3H), 3.77 (s, 3H), 5.71 (s,
4

N. Nilchan et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
2H), 6.11 (2, 2H), 6.81 (d, J = 1.90 Hz, 1H), 6.93 (d, J = 1.86 Hz, 1H),
7.55 (s, 1H); 13C NMR (75 MHz, DMSO‑d6) δ: 32.07, 55.97, 60.09,
DMSO were used to prepare series of concentration of each compound
by diluting in sterilized Muller-Hinton broth (MHB). An 180 µL aliquot
of each compounds at different dilutions was transferred to sterilized
96-well plates. To each aliquot of tested compounds, 20 µL of bacterial
inoculum was added to reach final inoculums density of approximately
105 CFU/mL. The assay plates were incubated at 37 °C for 20 h. After
incubation, the optical density of each well was measured using Wallac
Victor V1420 Multilable HTS Counter (PerkinElmer, USA). The MIC is
determined by the lowest concentration showing no visible growth,
quantitatively indicated by the basal optical density. The MBC were
determined based on the results obtained from microdilution testings.
The solution of well that appeared as clear solution, or quantitatively
indicted by basal optical density, was spread on TSA agar plate and
subsequently incubated at 37 °C for 18–20 h. The lowest concentration
that shows no colony of bacteria is recorded as MBC. The experiments
were repeated at least in triplicate.
105.13, 112.24, 120.65, 126.42, 137.54, 142.72, 153.20, 156.06,
+
162.12, 162.38; ESI-MS calc for
339.0451, found 339.0453.
C
13H15BrN4NaO2 [M+H+
]
:
4.1.3.3. 5-(3-Iodo-4,5-dimethoxybenzyl)pyrimidine-2,4-diamine
(TMP-
I). A yellow crystal product was obtained in 60% yield. 1H NMR
(300 MHz, DMSO‑d6) δ: 3.51 (s, 2H), 3.63 (s, 3H), 3.75 (s, 3H), 5.70 (s,
2H), 6.10 (2, 2H), 6.96 (d, J = 1.81 Hz, 1H), 7.12 (d, J = 1.83 Hz, 1H),
7.54 (s, 1H); 13C NMR (75 MHz, DMSO‑d6) δ: 31.71, 55.82, 59.76,
92.38, 105.26, 113.81, 129.09, 138.91, 146.34, 151.98, 156.02,
+
162.12, 162.35; ESI-MS calc for
387.0312, found 387.0301.
C
13H15IN4NaO2 [M+H+
]
:
4.2. Molecular docking
The molecular docking experiment was performed on Schrӧdinger
Maestro suites using the ‘standard protocol’. The saDHFR crystal
structure from the Protein Data Bank with co-crystal TMP (PDB
ID:2W9H) was refined with Protein Preparation Wizard. The Receptor
grid for Glide dock was generated from the refined protein structure
and the co-crystal TMP. Ligand structures were constructed from TMP
structure and prepared with LigPrep in Schrӧdinger suites. Glide dock
SP was performed with core constrain based on co-crystal TMP. The
binding poses from the docking were then subsequently analyzed.
4.3.4. Cytotoxicity assay
To test cytotoxicity of the synthetic derivatives, we followed pre-
viously reported protocol.21 A549 cells were harvested. Cells (ap-
proximately 104 cells) in 45 µL of DMEM culture medium was added to
sterilized 96-well plate. The stock of TMP derivatives tested were pre-
pared by diluting with DMSO. To obtain series of compound tested, the
stock solution were subsequently 2-fold diluted with DMEM culture
medium. 45 µL of tested compounds in DMEM medium was placed to
each well that already contained A549. The assay plate was cultured at
37 °C and 5% CO2 incubator for 24–48 h. Subsequently, 10 µL of Pre-
stoblue® was added into each well and incubated at 37 °C and 5% CO2
for 1 h. Then, the fluorescence intensity of each well was determined
using Wallac Victor V1420 Multilable HTS Counter (PerkinElmer, USA).
The excitation wavelength was 535 nm and the emission wavelength
was 610 nm. Numbers of viable cells were indirectly determined by
measuring fluorescent intensity as the dye would transform to fluor-
escent adduct in viable cells.
4.3. Microbiology
Overnight bacterial cultures were prepared in Tryptic soy broth
(TSB). The susceptibility tests were performed, using Muller-Hinton
broth (MHB) supplemented with Bacto Agar. All media were purchased
from BD (Sparks, MD). Antibiotics were from Sigma-Aldrich. The tri-
methoprim-sulfamethoxazole (SXT) disk was purchased from Oxoid
(Hampshire, UK).
4.3.5. Synergy testing by broth microdilution method (Checkerboard
Titration)
4.3.1. Bacterial strains and inoculums preparation
A single colony of each strain was picked and transferred into TSB
(5 mL) in a sterilized 15-mL test tube, which was then capped and in-
cubated at 37 °C for 16–18 h. The suspension was subsequently diluted
in sterilized phosphate buffer saline (PBS) to yield the appropriate in-
oculum density of approximately 108 CFU/mL for disk diffusion sus-
ceptibility test and 106 CFU/mL for microdilution assay. The inoculum
was used within 15 min after adjusting to its appropriate density.
Checkerboard titration method was performed based on previously
reported protocols.22 In brief, two-fold serial dilutions of our synthetic
trimethoprim derivatives and sulfamethoxazole were prepared by di-
luting in MHB. The range of dilution was chosen to cover the higher end
and the lower end of each MIC. 90 µL of each serial dilution of first
compound of the combination was dispensed into the successive row,
while 90 µL of each serial dilution of second compound was dispensed
into the successive column. 20 µL of test strain suspension, of which the
inoculum density was adjusted to 106 CFU/mL, was added. The final
inoculum density of each well was approximately 105 CFU/mL. The
plates were incubated at 37 °C for 18 h. After incubation, the optical
density of each well was measured. The fractional inhibitory con-
centration (FIC) and fractional inhibitory concentration index (FICI)
calculations were described in the Supplementary material.
4.3.2. Disk diffusion
The disk diffusion was performed based on previously reported
procedure.18 Muller-Hinton Agar was sterilized and poured into Petri
dishes to acquire 4 mm in thickness. After solidification, bacterial sus-
pension with inoculum density of approximately 108 CFU/mL were
spread on the agar surface using sterile swap. The inoculated plate was
let stand at room temperature for 5–10 min before placing a Whatman
paper disk (6 mm). Subsequently, a solution of test compound (typically
5 µg, based on suggested breakpoint of TMP, otherwise noted) diluted
in DMSO was impregnated onto each paper disk. After 18–20 h of in-
cubation at 37 °C in ambient air incubator, the diameters of inhibition
zones were measured in millimeters using sliding callipers. The ex-
periments were performed at least in triplicate.
Acknowledgments
This work was supported in part by the Thailand Research Fund
(RSA6180037 grant to C.J.) and the Office of the Higher Education
Commission, Mahidol University under the National Research
Universities Initiative and the Center of Excellence for Innovation in
Chemistry (PERCH-CIC). We would like to thank Professor William R.
Roush for allowing access to Schrӧdinger Maestro suites and James
Alburger for assisting in the docking experiment.
The zone diameter was measured from the back of the plate. The
inhibition zones were classified as complete and incomplete inhibition.
The complete zone margin is identified as the area where no obvious
visible growth is observed by the naked eyes.
A. Supplementary data
4.3.3. Microdilution assay
The microdilution susceptibility protocol reported by CLSI was used
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.bmc.2018.05.019.
as a guideline.18 The stock solutions of test compounds at 100 mM in
5

N. Nilchan et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
References
Agents Chemother. 2010;54:3825–3833.
13. Dale GE, Broger D, D'Arcy A, et al. J Mol Biol. 1997;266:23–30.
14. Scholfield MR, Zanden CMV, Carter M, Ho PS. Protein Sci. 2013;22(2):139–152.
15. (a) Sirichaiwat C, et al. J Med Chem. 2004;47:345–354;
1. Otto M. Annu Rev Microbiol. 2010;64:143–162.
2. Saini V, Riekerink RG, McClure JT, Barkema HW. J Clin Microbiol.
2011;49:1568–1577.
(b) Nammalwar B, et al. Eur J Med Chem. 2012;54:387–396.
16. Tenover FC, Goering RV. J Antimicrob Chemother. 2009;64:441.
17. Otto M. Cell Microbiol. 2012;14:1513.
18. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. CLSI supplement
M100 27th ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2017.
19. Bang HB, Han SY, Choi DH, et al. Synth Commun. 2009;39(3):506–515.
20. Holcomb WF, Hamilton CS. J Am Chem Soc. 1942;64(6):1309–1311.
21. Boncler M, Różalski M, Krajewska U, Podsędek A, Watala C. J Pharmacol Toxicol
Methods. 2014;69:9–16.
3. Fischbach MA, Walsh CT. Science. 2009;325:1089–1093.
4. Melander RJ, Melander C. ACS Infect Dis. 2017;3:559–563.
5. Tamma PD, Cosgrove SE, Maragakis LL. Clin Microbiol Rev. 2012;25(3):450–470.
6. Masters PA, O'Bryan TA, Zurlo J, Miller DQ, Joshi N. Arch Intern Med.
2003;163(4):402–410.
7. Yun MK, Wu Y, Li Z, et al. Science. 2012;335:1110–1114.
8. Murima P, McKinney JD, Pethe K. Chem Biol. 2014;21:1423–1432.
9. Zander J, Besier S, Ackermann H, Wichelhaus TA. Antimicrob Agents Chemother.
2010;54:1226–1231.
22. Schwalbe R, Steele-Moore L, Goodwin AC. CRC Press: Boca Raton, FL, 2007; pp
275–298.
10. Huovinen P. Clin Infect Dis. 2001;32:1608–1614.
11. Charpentier E, Courvalin P. Antimicrob Agents Chemother. 1997;41:1134–1136.
12. Bourne CR, Barrow EW, Bunce RA, Bourne PC, Berlin KD, Barrow WW. Antimicrob
6