Synthetic and crystallographic studies of a new inhibitor series targeting Bacillus anthracis dihydrofolate reductase

Article abstract of DOI:10.1021/jm800776a

Bacillus anthracis, the causative agent of anthrax, poses a significant biodefense danger. Serious limitations in approved therapeutics and the generation of resistance have produced a compelling need for new therapeutic agents against this organism. Baci

Full text of DOI:10.1021/jm800776a

7532
J. Med. Chem. 2008, 51, 7532–7540
Synthetic and Crystallographic Studies of a New Inhibitor Series Targeting Bacillus anthracis
Dihydrofolate Reductase
Jennifer M. Beierlein,† Kathleen M. Frey,† David B. Bolstad,† Phillip M. Pelphrey,† Tammy M. Joska,† Adrienne E. Smith,‡
Nigel D. Priestley,‡ Dennis L. Wright,*,† and Amy C. Anderson*,†
Department of Pharmaceutical Sciences, UniVersity of Connecticut, 69 N. EagleVille Road, Storrs, Connecticut 06269, and Department of
Chemistry, UniVersity of Montana, Missoula, Montana 59812
ReceiVed June 26, 2008
Bacillus anthracis, the causative agent of anthrax, poses a significant biodefense danger. Serious limitations
in approved therapeutics and the generation of resistance have produced a compelling need for new therapeutic
agents against this organism. Bacillus anthracis is known to be insensitive to the clinically used antifolate,
trimethoprim, because of a lack of potency against the dihydrofolate reductase enzyme. Herein, we describe
a novel lead series of B. anthracis dihydrofolate reductase inhibitors characterized by an extended
trimethoprim-like scaffold. The best lead compound adds only 22 Da to the molecular weight and is 82-fold
more potent than trimethoprim. An X-ray crystal structure of this lead compound bound to B. anthracis
dihydrofolate reductase in the presence of NADPH was determined to 2.25 Å resolution. The structure
reveals several features that can be exploited for further development of this lead series.
Introduction
Bacillus anthracis is the highly pathogenic, Gram-positive
bacteria responsible for the acute and often fatal disease anthrax.
Although known for ages as a general threat to mammals such
as cattle, B. anthracis has more recently attracted attention as
a potential bioterrorism weapon. Humans are susceptible to three
forms of infection: cutaneous, gastrointestinal, and inhalational,
all of which can progress to a systemic infection that ultimately
proves fatal.1 To have the greatest efficacy, the highly virulent
nature of anthrax necessitates that prophylactic treatment with
antibiotics is started prior to the presentation of symptoms.2
Fluoroquinolones, such as ciprofloxacin, are first-line therapy
for anthrax, followed by doxycycline and various third-
generation cephalosporins. However, each of these treatments
has serious limitations: ciprofloxacin and doxycycline are
expensive and are not indicated for use in children less than 8
Figure 1. DHFR inhibitors trimethoprim (TMP), methotrexate (MTX)
and compounds 1 and 2.
years of age, especially as prophylactic measures without
diagnosed exposure. Strains resistant to ciprofloxacin3,4 as well
peutics directed against DHFR have validated this status. DHFR
plays a key role in the folate biosynthesis pathway, responsible
as both ꢀ-lactamase-based and non-ꢀ-lactamase-based penicillin-
resistant forms have emerged, and doxycyclin-resistance has
for the generation of the DNA base, deoxythymidine mono-
been engineered in the laboratory.2,5 The potential of a large-
phosphate, as well as the biosynthesis of purine nucleotides and
scale anthrax attack on a heavily populated area, with either a
the amino acids histidine and methionine. Specifically, DHFR
natural or genetically engineered resistant strain, necessitates
catalyzes the reduction of dihydrofolate, using NADPH, to form
tetrahydrofolate and NADP+.
Since human cells also depend on DHFR for DNA replication,
developing inhibitors that are not only potent but also selective
for the pathogen is critical. Fortunately, active site differences
the stockpiling of a large number of classes of low-cost and
shelf-stable antibiotics. The case for the development of new
classes of antibiotics against B. anthracis is compelling.
Dihydrofolate reductase (DHFRa) has been a widely recog-
nized drug target for at least 50 years; the successful clinical
have allowed the development of very species-specific DHFR
use of anticancer,6,7 antibacterial,8,9 and antiparasitic10 thera-
inhibitors for some bacteria9 and some parasitic protozoa
including Plasmodium11 and Toxoplasma.12 DHFR inhibitors
such as methotrexate (MTX)13 and trimethoprim (TMP) (Figure
1) have been used successfully in the clinic for several years.
MTX is a potent, nonselective inhibitor employed in cancer
chemotherapy, and TMP is a selective inhibitor of E. coli and
S. aureus DHFR.
* To whom correspondence should be addressed. For D.L.W.: phone,
(860) 486-9451; fax, (860) 486-6857; e-mail, dennis.wright@uconn.edu.
For A.C.A.: phone, (860) 486-6145; fax, (860) 486-6857; e-mail,
amy.anderson@uconn.edu.
†
University of Connecticut.
University of Montana.
‡
a Abbreviations: DHFR, dihydrofolate reductase; BaDHFR, dihydrofolate
reductase from Bacillus anthracis; hDHFR, human dihydrofolate reductase;
ChDHFR, dihydrofolate reductase from Cryptosporidium hominis; TMP,
trimethoprim; MTX, methotrexate; NADPH, nicotinamide adenine dinucle-
otide phosphate, reduced form.
Previously, we14 and others5,15 had investigated the potential
of known antifolates to be inhibitors of B. anthracis DHFR
(BaDHFR). BaDHFR has been shown to be naturally resistant
to the widely used antifolate trimethoprim but sensitive to the
10.1021/jm800776a CCC: $40.75
2008 American Chemical Society
Published on Web 11/13/2008

New Bacillus anthracis DHFR Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7533
nonselective inhibitor methotrexate (Figure 1).5 In addition, we
examined a number of previously described antifolates14 and
discovered that extended systems such as 5-deazapteridine
analogues (Figure 1; 1, 2) were good inhibitors of the related
B. cereus DHFR. A recent X-ray crystal structure of BaDHFR
bound to methotrexate has been reported15 and describes
differences with human DHFR as well as a potential mechanism
of trimethoprim resistance.
In order to find a new class of antifolates effective against
BaDHFR and selective for the bacterial enzyme, we screened a
group of propargyl-based DHFR inhibitors that we have
developed for pathogenic species of DHFR.16 Using a homology
model to guide the synthesis of an improved class of compounds,
we synthesized a group of 2′,5′-dimethoxyphenylpyrimidines.
These compounds proved to be superior in both potency and
selectivity and were shown to inhibit bacterial growth. We then
determined a crystal structure of the most potent and selective
compound bound to BaDHFR. The crystal structure rationalizes
the structure-activity relationships and provides guidance for
future compound development.
Table 1. Evaluation of Propargyl Inhibitors in Assays with BaDHFR
and hDHFR
IC50 (µM)
selectivity
compda
R1
R2
BaDHFR
hDHFR
ratio (h/Ba)
TMP
3
4
5
6
7
8
9
71
2.3
3.7
120
1.46
0.4
1.46
14.3
1.16
1.38
5.71
1.22
1.7
H
CH3
H
H
H
CH3
CH3
CH3
H
H
0.63
0.11
0.3
0.67
0.01
0.05
0.4
CH3
OH
OMe
CH3
OH
4.8
21.2
109.2
30.3
14.5
29.1
10
OMe
0.04
a Compounds 5-10 were tested as racemates.
Chemistry, Crystal Structure, and Biological Evaluation
Evaluation and Analysis of Existing Propargyl-Linked
Inhibitors. Since anthrax is resistant to TMP and since MTX
is not selective for the bacterial enzyme and hence toxic, the
development of a new class of compounds is needed to
effectively target the folate pathway in this organism. Interest-
ingly, natural resistance to TMP has been noted in a wide variety
of eukaryotic organisms such as the parasitic protozoa Cryptospo-
ridium and Toxoplasma and the fungi Candida and Pneumocys-
tis.17 In prior work, we examined TMP resistance in Cryptospo-
ridium hominis DHFR (ChDHFR). On the basis of an analysis
of the structure of ChDHFR, we attributed TMP resistance to
an inability of the trimethoxyphenyl ring to adequately occupy
a key hydrophobic pocket in the active site of the enzyme.
Subsequently, we were able to develop a highly potent and
efficient lead series effective against both parasitic protozoal
DHFR enzymes.16 This novel lead series is characterized by a
propargyl linker between the two arenes, providing the ideal
spacing and rigidity to produce potent inhibitors. Docking
representative propargyl-based compounds in a homology model
of BaDHFR suggested that we could exploit a similar strategy
to develop inhibitors of BaDHFR.
As we already had a number of inhibitors in hand from prior
work with ChDHFR,16 we were able to gather some preliminary
structure-activity data by evaluating in vitro inhibition of
BaDHFR (Table 1). We also evaluated the activity of these
compounds against human DHFR in order to gauge selectivity.
Although these compounds were not highly potent inhibitors
of BaDHFR, we were encouraged that most of the compounds
were more potent than TMP. From this limited data set, it was
apparent that certain substitutions at the propargyl position
proved deleterious; substitution at the C6 position on the
pyrimidine ring was tolerated. Further exploration of increased
bulk at C6 and the substitution pattern on the phenyl ring was
warranted.
Figure 2. Homology model of BaDHFR bound to compound 4.
mized using tools within Sybyl (Tripos, Inc.). Ramachandran
plots of the model show that 99.4% of the residues fall within
acceptable ranges (the outlier, Ile 93, is far from the active site).
Comparisons of the homology model to the published crystal
structure of BaDHFR show that the two models superimpose
with 1.2 Å rmsd and share the same overall fold. A few residues
at the opening to the active site exhibit rotamer differences.
We examined the interactions of the most potent compounds,
3-5, in the active site of the homology model of BaDHFR.
The pyrimidine rings of compounds 3-5 appear to form the
conserved interactions in the active site. This conserved orienta-
tion includes ionic interactions between the protonated N1 atom
and the 2-amino group of the pyrimidine with the acidic residue
Glu 28 (Figure 2).19-22 The propargyl linker places the
trimethoxyphenyl ring in van der Waals contact with the
hydrophobic pocket containing Asn 47, Ala 50, Ile 51, and Leu
55. Models of 3-5 in the active site led us to discover that a
simple change in the pattern of substitution around the phenyl
ring from 3′,4′,5′-OMe to 2′,5′-OMe may maintain the interac-
tions of the 5′-methoxy group with Leu 55 and Ile 51 while
promoting a 2′ substitution to occupy the upper portion of the
pocket in order to form contacts with Ala 50 and the backbone
of Asn 47. Therefore, we set out to explore this alternative
substitution pattern with a series of compounds that would
concomitantly probe varying steric bulk at the C6 position of
the pyrimidine.
Design, Synthesis, and Evaluation of First Generation
BaDHFR Inhibitors. In an attempt to improve the potency of
this series, we used structural analysis to guide the design of
additional inhibitors. Since an experimental structure of BaDH-
FR was not available at the time, a homology model was created
on the basis of the structure of the E. coli enzyme, which has
39% overall sequence identity and 62% identity in the active
site.14 The model was created with 3D-JIGSAW18 and mini-
For the synthesis of this propargyl-based class of inhibitors
we relied on a key Sonagashira coupling to unite the arene and
pyrimidine fragments, allowing a convergent, modular route.

7534 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Beierlein et al.
Scheme 1a
a (a) Ph3PCH2OMeCl, NaOtBu, THF, 0° C; (b) 10% HCl, THF, reflux; (c) CBr4, PPh3, CH2Cl2, 0° C; (d) Mg, THF; (e) CuI, Pd(PPh3)2Cl2, Et3N, DMF,
60° C.
Table 2. Inhibitory potency of first generation inhibitors
Scheme 2
IC50 (µM)
BaDHFR hDHFR
selectivity
ratio
compd R1
W
X
Y
Z
15
16
17
18
20
22
24
H
OMe
H
H
OMe
OMe
OMe
OMe
1.7
1.3
0.89
5.5
0.942
9.2
3.2
1.3
1.9
1
1.4
0.21
0.06
0.01
0.11
CH3 OMe
Et OMe
n-Pr OMe
Et
Et
Et
H
H
H
H
H
H
1.28
1.18
0.057
0.13
0.36
H
OMe OMe OMe
OMe OMe
H
H
H
H
H
H
3.2
The required propargylarenes are available through homologa-
tion of the corresponding arylaldehydes (Scheme 1).
Table 3. Antibacterial Assay Results
Commercially available 2,5-dimethoxybenzaldehyde 11
was homologated to the corresponding arylacetaldehyde 12
through initial Wittig reaction to give the enol ether followed
by direct hydrolysis to the aldehyde. The resulting crude
material was subjected to a second homologation to give the
vinyl dibromide 13 in good overall yield for the three
operations. Final conversion to the terminal acetylene 14 was
accomplished by a modified Corey-Fuchs reaction using
elemental magnesium. Cross-coupling of 14 to four different
iodinated 2,4-diaminopyrimidines16,23,24 produced the inhibi-
tors 15-18 in moderate to very good yields. The 2′,5′-
dimethoxy compounds were evaluated using an in vitro
enzyme inhibition assay against BaDHFR and human DHFR
(Table 2).
From these assay results, it was apparent that an ethyl group
at C6 was optimal. Therefore, this was maintained and three
other substitution patterns on the aryl ring were explored. A
trimethoxyphenyl derivative was easily prepared by coupling
the previously described 1916 with the ethyliodopyrimidine to
give 20. A 2′3′-dimethoxy analogue 22 was prepared from
commercially available 2,3-dimethoxybenzaldehyde by a route
analogous to that shown in Scheme 1. Finally, a completely
unsubstituted phenyl derivative 24 was synthesized in one step
by coupling with commercially available phenylpropyne 23
(Scheme 2). All analogues were evaluated in enzyme inhibition
assays (Table 2).
minimum inhibitory
concentration (µg/mL)
compd
15
16
17
20
71
37
20
inactivea
a Inactive at 2 mM.
We were pleased to see that these first generation inhibitors
demonstrated moderate ability to kill the target organism. While
growth inhibition is not at the level that would be clinically
useful at this stage, the results do show that compounds in this
series can function as antimicrobial agents. It is evident from
these results that within the 2′,5′-dimethoxy series, as enzyme
inhibition increases, antibacterial growth inhibition increases.
Surprisingly, the similarly potent 3′,4′,5′-trimethoxy derivative
20 failed to inhibit bacterial growth.
X-ray Crystal Structure of BaDHFR/NADPH/
Compound 17. In order to build upon these first generation
compounds, we determined the crystal structure of our best lead
compound, 17, bound to BaDHFR. Crystals were grown in the
presence of the cofactor NADPH as well as compound 17, and
diffraction data were collected to 2.25 Å resolution. The protein
crystallized with two molecules in the asymmetric unit in space
group P42. The structure was solved by molecular replacement
using a previously published structure of BaDHFR bound to
MTX.15 Electron density for the ligand and cofactor was well
resolved (Figure 3), allowing the construction of a model of
the ternary complex. The final model is refined with an Rfree
value of 23.8 and an R-factor of 19.1 with all residues falling
These data show that with the optimal C6-ethyl substituent,
both the 2′5′-dimethoxy and the 3′,4′,5′-trimethoxy patterns are
effective. However, the 2′,5′-dimethoxy pattern appears to garner
a slightly favorable degree of selectivity for the bacterial
enzyme. We selected four compounds, 15-17 and 20, to test
in an antibacterial assay against B. anthracis Sterne (Table 3).

New Bacillus anthracis DHFR Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7535
Figure 3. Electron density for the active site of the ternary complex of BaDHFR including compound 17. Density was contoured at 2σ.
Table 4. Data Collection and Refinement Statistics for BaDHFR/17/
NADPH
than the coordinate error of the structure (0.34 Å). Electron
density was also observed for four of the six N-terminal
histidines of the His-tag. Finally, there is a bulge in ꢀA at the
N-terminus, which is believed to be a result of the mutation of
Ile 2 to Arg. This mutation was engineered to allow for future
removal of the N-terminal His-tag. At the active site, there are
only minor differences in the positions of residues, other than
RB, which moves into the active site when bound to the more
potent inhibitor MTX (Figure 4C).
parameter
BaDHFR/17/NADPH
space group
P42
unit cell (a, b, c in Å)
a ) 78.42, b ) 78.42,
c ) 67.09
21.75-2.25
97.2 (73.6)
18 002
resolution (Å)
completeness, % (last shell*, %)
unique reflections
redundancy (last shell)
Rsym, % (last shell, %)
I/σ (last shell)
6.5 (6.1)
0.100 (0.299)
11.6 (2.2)
2
Ligand Binding. Compound 17 is bound in the active site
of BaDHFR with the pyrimidine ring demonstrating the
conserved orientation for antifolate inhibitors (Figure 5). The
2-amino group forms an additional hydrogen bond with the
backbone carbonyl of Val 7 and a water molecule. The carbonyl
oxygen of Met 6 forms a hydrogen bond with the 4-amino
group. In addition, there are several van der Waals interactions
involving the pyrimidine ring and Ala 8, Val 32, Met 6, and
Val 7. The ethyl group at the C6 position makes favorable
lipophilic contacts with Leu 21. The acetylene linker forms van
der Waals interactions with Phe 96, Leu 21, and the nicotina-
mide ring of NADPH. The 2′-OMe is pointed up toward a small
hydrophobic pocket with Ala 50 and Leu 21, while the 5′-OMe
is pointed down toward a larger hydrophobic pocket comprising
Ile 51, Leu 55, Leu 29, and Phe 96. An ordered water molecule
in the active site forms hydrogen bonds to Glu 28 and Trp 23
and has been observed in other species of DHFR.21 A second
water molecule forms hydrogen bonds to the 2′-OMe on the
phenyl ring.
no. of monomers in asymmetric unit
Refinement Statistics
R-factor/Rfree
0.191, 0.238
1390, 69, 127
0.014, 1.493
25.1
24.5
29.12
no. of atoms (protein, ligands, solvent)
rms deviation bond lengths (Å) angles (deg)
average B factor (Å2)
average B factor for ligand (Å2)
average B factor for solvent molecules (Å2)
Ramachandran Plot Statistics
residues in most favored regions (%)
89.7
10.3
0.0
residues in additional allowed regions (%)
residues in generously allowed regions (%)
residues in disallowed regions (%)
0.0
into allowed regions of the Ramachandran plot (Table 4). The
model has been deposited in the Protein Data Bank with code
3E0B.
The structure shows the same overall fold seen throughout
several DHFR species, including a canonical eight-stranded
twisted ꢀ sheet with four flanking R helices (Figure 4A). The
structure of the ternary complex is similar to the published
binary structure of BaDHFR bound to MTX,15 with a root-mean-
square deviation of 0.617 Å. However, there are some noticeable
differences between the two structures, as observed in Figure
4B, mainly resulting from the differences in ligand and the
addition of NADPH. The R helix RB, which contains residues
44-51, is positioned farther away from the inhibitor because
of the proximity of the 5′-OMe group in 17. This repositioning
of RB creates additional space between itself and R helix RD,
which contains residues 99-108, allowing more room for the
NADPH cofactor. The ꢀC sheet, which contains residues
60-64, is shifted to allow the residues to make polar contacts
with the adenosine ring of NADPH. Distances between pairs
of representative atoms on RB and ꢀC were 1.2-1.5 Å, greater
The structural analysis rationalizes some of the trends
observed in the preliminary compound evaluation. The C6 ethyl
substitution appears to be optimal because the terminal methyl
group is properly situated to form favorable interactions with
Leu 21. This interaction is not possible with hydrogen or methyl
at C6, while the larger propyl group likely introduces destabiliz-
ing interactions. The acetylenic linker seems ideally suited to
bypass the restricted space introduced by Phe 96 and position
the aryl ring in the hydrophobic pocket. Furthermore, the
deleterious effect of propargylic substitution is rationalized by
destabilizing interactions with either Phe 96 or the nicotinamide
ring of NADPH. Finally, it appears that the 2′-OMe anchors
the phenyl ring by binding the small hydrophobic pocket,
allowing the 5′-OMe to explore the larger pocket below.

7536 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Beierlein et al.
Figure 4. (A) overall structure of BaDHFR bound to NADPH and 17 and a comparison with (B) overall binary complex with MTX (PDB 2QK815
)
and (C) residues and ligands at the active site.
Substituents at the 3′ position are not likely to make any
productive contacts, while groups at the 4′ position may have
an opportunity to interact with Leu 29.
opportunity for increasing selectivity. Additionally, increased
bulk at the 5′ position may result in destabilizing interactions
with a loop at the active site (Pro 61-Glu 62-Lys 63-Asn
64, the PEKN loop) in hDHFR that is absent in BaDHFR.
Finally, substitution at the 4′ position of the aryl ring could
interact with Leu 29, yielding an increase in potency. Current
work is focused on the design and synthesis of second generation
inhibitors of BaDHFR.
In summary, we have identified a novel, flexible lead series
effective as inhibitors of the enzyme BaDHFR as well as the
growth of B. anthracis Sterne. The best compound in this series
makes several key interactions with the active site of BaDHFR,
which results in a greater than 88-fold increase in potency
relative to trimethoprim. Further development of this class will
necessitate both an increase in potency against BaDHFR and
an increase in selectivity over the human form of the enzyme.
Improvements in potency and selectivity, while maintaining
good druglike properties, should lead to a corresponding increase
in antibacterial activity. Analysis of the experimentally deter-
mined structure of BaDHFR bound to compound 17 reveals
several design strategies for superior analogues.
A structure-based sequence alignment and structural com-
parison (Figure 6) show that there are several residue differences
between BaDHFR and hDHFR, providing opportunities to
garner selectivity in future designs. Specifically, optimization
of the substituent at the C6 position of the pyrimidine ring may
lead to increased potency and selectivity. Branching at the arylic
position at C6, such as isopropyl, cyclopropyl, or tert-butyl,
would be expected to project functionality into the larger
hydrophobic pocket below the pyrimidine ring comprising Val
32 and Leu 29 in BaDHFR. The structural comparison with
hDHFR shows that the corresponding residues are both larger
(Phe 31 and Phe 34) in the human enzyme. Therefore, the
potential exists for destabilizing interactions with these branched
substituents at the C6 position, which could lead to an increase
in selectivity. Alternatively, increased bulk at the 5′ position of
the aryl ring could form additional hydrophobic interactions with
the larger pocket comprising Ile 51, Leu 55, Leu 29, and Phe
96 (Figure 7). Again, the phenylalanine residues in hDHFR
restrict the volume of this pocket and as such present an
Experimental Section
Enzyme Cloning. Site-directed mutagenesis was used to change
the earlier reported construct for BcDHFR-pET4114 to BaDHFR
by three point mutations: V77A, I130M, and I138V. Mutations were
verified by ABI Big Dye sequencing. The BaDHFR-pET41
construct was amplified using PCR and inserted into vector pQE2.
BaDHFR-pQE2 clones were verified by sequencing. The resulting
construct contained the BaDHFR gene with an N-terminus histidine
tag and a DAPase stop point (Ile 2 was mutated to Arg) for future
His-tag removal.
Genomic DNA containing the gene for hDHFR was obtained
from ATCC and amplified by PCR. The gene was inserted into the
pET41 vector and verified by sequencing. Expression and purifica-
tion methods described for BcDHFR-pET4114 were followed for
hDHFR.
Recombinant Protein Expression and Purification. BaDHFR
recombinant protein was expressed in M15 cells upon induction
with 1 mM IPTG at mid-log phase. Protein expression was extended
for an additional 6 h at 37 °C after induction. Cells were harvested
by centrifugation. Pellets were lysed with 1× Bugbuster and DNase
for 20 min at room temperature and then centrifuged at high speed
to collect the supernatant. The supernatant was loaded onto a nickel
affinity column and washed with 20 mM Tris, 1 mM DTT, 200
mM KCl (pH 8.0). Bound protein was eluted across a linear gradient
with 20 mM Tris, pH 8.0, 1 mM DTT, 50 mM KCl, and 250 mM
imidazole. The fractions with pure protein were concentrated to
∼1 mL and loaded onto a size exclusion column (S200) for
desalting. Protein was eluted into a final buffer of 20 mM Tris, 50

New Bacillus anthracis DHFR Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7537
quality crystals were cryoprotected in 15% ethylene glycol and
flash-cooled with liquid nitrogen. Data were collected at Brookhaven
National Synchotron Light Source on beamline X29A. All data sets
were collected at 100 K.
Structure Determination. The structure of the BaDHFR-ligand
complex was solved by molecular replacement. The program Phaser
and a published model of BaDHFR (PDB code 2QK815) were used
to determine initial phase information. The program Coot was used
to visualize the electron density map and build the model. The
model was refined with the program Refmac5. The refined model
satisfies the conditions of the Ramachandran plot (Table 2).
1
Synthesis: General. The H and 13C spectra were recorded on
Bruker instruments at 500 and 125 MHz or 300 and 75 MHz,
respectively. Melting points were recorded on Mel-Temp 3.0
apparatus and are uncorrected. High-resolution mass spectrometry
was provided by the Washington University Mass Spectrometry
Laboratory or the Notre Dame Mass Spectrometry Laboratory. IR
data were obtained a Shimadzu 8400-s FTIR spectrometer. Anhy-
drous dichloromethane, ether, and tetrahydrofuran were used
directly from Baker cycletainers. Anhydrous dimethylformamide
was purchased from Acros and degassed by purging with argon.
Anhydrous triethylamine was purchased from Aldrich and degassed
by purging with argon. TLC analyses were performed on Whatman
Partisil K6F silica gel 60 plates and visualized at 254 nm and/or
by staining with potassium permanganate. All reagents were used
directly from commercial sources unless otherwise stated. All
glassware was oven-dried and allowed to cool under an argon
atmosphere. Compounds 3-10 were previously synthesized ac-
cording to literature procedures.16 2,4-Diamino-5-iodopyrimidine,16
2,4-diamino-5-iodo-6-methylpyrimidine,16 2,4-diamino-6-ethyl-5-
iodopyrimidine,23 and 2,4-diamino-6-n-propylpyrimidine24 were
synthesized according to literature procedures.
1,1-Dibromo-3-(2,5-dimethoxyphenyl)propene (13). To a 0 °C
suspension of methoxymethyltriphenylphosphonium chloride (10.8
g, 31.6 mmol) in dry THF (90 mL) under an argon atmosphere is
added NaOtBu (3.90 g, 40.6 mmol) in one portion. The red-orange
suspension is stirred for a further 10 min at 0 °C, and then solid
2,5-dimethoxybenzaldehyde 11 (3.0 g, 18.1 mmol) was added in
small portions. After 10 min, the reaction was quenched with water
(50 mL) and diluted with ether (50 mL). The organic phase was
separated and the aqueous phase extracted with additional ether (2
× 50 mL). The combined organics were washed with brine (50
mL), dried over sodium sulfate, and concentrated to afford the crude
product that was filtered through a column of silica (SiO2 75 g,
5% EtOAc/hexanes) to afford the crude enol ether that was
immediately hydrolyzed in the subsequent step.
To a solution of crude enol ether in THF (110 mL) was added
10% aqueous HCl (10 mL). The solution was heated to refllux and
monitored by TLC. The reaction was proceeding sluggishly after
1 h, so an additional 0.5 mL of concentrated HCl was added. Once
the starting material had been consumed (∼1 h), the reaction
mixture was diluted with saturated NaHCO3 (100 mL). THF was
removed at the rotovap, and the aqueous mixture was extracted
with ether (3 × 50 mL). The combined organics were washed with
brine (2 × 50 mL), dried over sodium sulfate, and concentrated to
afford the crude aldehyde 12 (3.02 g, 93%, two steps) that was
taken immediately to the next step. TLC Rf ) 0.31 (15% EtOAc/
hexanes); 1H NMR (300 MHz, CDCl3) δ 9.67 (t, J ) 2.1 Hz, 1H),
6.83 (m, 2H), 6.73 (m, 1H), 3.78 (s, 3H), 3.77 (s, 3H), 3.62 (d, J
) 2.1 Hz, 2H).
Figure 5. Detailed interactions of the ligand with the active site
residues. (A) Stereoview of the active site with NADPH in magenta
and compound 17 in blue. Water molecules are shown as “x”, and
orange dashed lines indicate hydrogen bonds. (B) Two-dimensional
depiction of protein-ligand interactions. Hydrogen bonds are noted
with dashed lines and distance measurements.
mM KCl, 5 mM DTT, and 0.5 mM EDTA. Fractions were analyzed
by SDS-PAGE. Protein was concentrated to ∼5 mg/mL and stored
at -20 °C until crystal tray setup.
Enzyme Assays. Enzyme activity assays were performed at 25
°C by monitoring the rate of enzyme-dependent NADPH oxidation
at an absorbance of 340 nm over several minutes.14 Reactions were
performed in a buffer containing 20 mM TES, pH 7.0, 50 mM
KCl, 10 mM 2-mercaptoethanol, 0.5 mM EDTA, and 1 mg/mL
BSA. All enzyme assays were performed with a single, limiting
concentration of enzyme and saturating concentrations of NADPH
and dihydrofolate. IC50 values were calculated as the average of
three independent experiments.
Antibacterial Assays. Minimum inhibitory concentrations (MIC)
against B. anthracis Sterne were determined using a broth microdi-
lution approach based upon CLSI (formerly NCCLS) standards and
the use of the colorimetric reporter Alamar Blue. The MIC value
is the lowest concentration of test compound that inhibits growth
such that less than 1% reduction of the blue resazurin (νmax ) 570
nm) component of the Alamar Blue to the pink resorufin (νmax
600 nm) is observed.
)
To a 0 °C solution of CBr4 (8.36 g, 25.2 mmol) in dry CH2Cl2
(100 mL) was added PPh3 (13.20 g, 50.3 mmol) in a single portion.
The resulting dark-yellow solution was stirred a further 5 min, and
then crude aldehyde 12 dissolved in CH2Cl2 (10 mL) was added
dropwise. The solution was stirred for 30 min and then poured into
ice cold ether (450 mL), producing a white precipitate and yellow
oil. The mixture was filtered through a column of silica gel (100
g) equilibrated with hexanes, and rinsed with hexanes (100 mL)
and 15% EtOAc/hexanes (300 mL). The filtrate was concentrated
and the residue purified by flash chromatography (SiO2 100 g, 10%
EtOAc/hexanes) to afford dibromoalkene 13 as a clear viscous oil
Crystallization. Protein at 5 mg/mL concentration was incubated
with 2 mM NADPH and 1 mM compound 17 for 1 h at 4 °C.
After incubation, the protein-ligand mix was concentrated to 15
mg/mL using a microcon (Amicon). All crystallization trials were
conducted at 25 °C. Initial hits were grown by hanging-drop vapor
diffusion in 25% (w/v) PEG 10,000, 0.1 M MES, pH 6.5, at an
equal ratio of protein to crystallization solution. Microseeding was
used to obtain isolated crystals in 10% (w/v) PEG 10,000 and 0.1
M MES, pH 6.5, at a protein concentration of 10 mg/mL. Good

7538 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Beierlein et al.
Figure 6. Structural alignment of BaDHFR and hDHFR: (left) structure-based sequence alignment, where residues in the active site are shown in
red; (right) superposition of BaDHFR (green) and hDHFR (purple) with active site residue substitutions labeled.
method. The vial was sealed under argon and heated at 50 °C for
1 h. After cooling, the orange solution was diluted with EtOAc
(60 mL) and washed with a water/saturated NaHCO3 solution (1:
2, 30 mL × 3) and brine (2 × 30 mL). The organic phase was
dried over sodium sulfate and concentrated to afford the crude
product that was purified by flash chromatography (SiO2 15 g,
EtOAc) to provide the coupled pyrimidine 15 as a pale solid (0.144
g, 92%). An analytical sample was generated by triturating under
DCM. TLC Rf ) 0.61 (9:1, CHCl3/MeOH); mp, decomposed above
180 °C; 1H NMR (500 MHz, DMSO-d6) δ 7.88 (vbs, 1H), 7.05 (d,
J ) 3.1 Hz, 1H), 6.91 (d, J ) 8.9 Hz, 1H), 6.79 (dd, J ) 8.9, 3.1
Hz, 1H), 6.41 (vbs, 2H), 6.22 (s, 2H), 3.76 (s, 3H), 3.73 (s, 2H),
3.70 (s, 3H); 13C NMR (125 MHz, DMSO-d6) δ 163.8, 162.0,
158.5, 153.1, 150.6, 126.0, 115.3, 111.7, 111.5, 92.6, 90.5, 76.4,
55.8, 55.3, 20.2; HRFAB [MLi+] 291.1442 (calculated for
C15H16N4O2Lim 291.1433); HPLC (a) tR ) 4.71 min, 100%, (b) tR
Figure 7. Surface depiction of binding pocket surrounding compound
17. The surface is colored with a lipophilicity gradient. Compound 17
and NADPH are shown as gray sticks with atom colors. Figure was
generated by Sybyl 8.0.
) 8.27 min, 98.8%. Anal. (C15H16N4O2) C, H, N.
2,4-Diamino-5-[3-(2,5-dimethoxyphenyl)prop-1-ynyl]-6-
methylpyrimidine (16). To an oven-dried 8 mL screw-cap vial
was added 2,4-diamino-5-iodo-6-methylpyrimidine (0.162 g, 0.648
mmol), CuI (18 mg, 0.094 mmol, ∼15%), and Pd(PPh3)2Cl2 (32
mg, 0.046 mmol, ∼7%). Degassed (argon purge) anhydrous DMF
(1.5 mL) was added followed by alkyne 14 (0.285 g, 1.62 mmol)
as a solution in DMF (1 mL). Degassed (argon purge) anhydrous
triethylamine was added (2.5 mL), and the mixture was degassed
once using the freeze-pump-thaw method. The vial was sealed
under argon and heated at 50 °C for 2 h. After cooling, the orange
solution was diluted with EtOAc (50 mL) and washed twice with
a water/saturated NaHCO3 solution (1:2, 30 mL) and then brine
(30 mL). The organic phase was dried over sodium sulfate and
concentrated to afford the crude product that was purified by flash
chromatography (SiO2 14 g, EtOAc) to provide coupled pyrimidine
16 (0.145 g, 75%) as a white powder. Rf ) 0.61 (9:1, CHCl3/
MeOH); mp, decomposed above 180 °C; 1H NMR (500 MHz,
DMSO-d6) δ 7.07 (d, J ) 3.1 Hz, 1H), 6.92 (d, J ) 8.9 Hz, 1H),
6.79 (dd, J ) 8.9, 3.1 Hz, 1H), 6.28 (vbs, 2H), 6.13 (s, 2H),
3.77-3.76 (m, 5H), 3.70 (s, 3H), 2.20 (s, 3H); 13C NMR (125 MHz,
DMSO-d6) δ 166.8, 164.3, 160.9, 153.1, 150.6, 126.2, 115.0, 111.8,
111.5, 95.4, 88.7, 76.5, 55.8, 55.3, 22.4, 20.5; HRFAB [MLi+]
305.1591 (calculated for C16H18N4O2Li, 305.1590); HPLC (a) tR
) 5.33 min, 98.6%, (b) tR ) 8.92 min, 99.1%.
(4.63 g, 76% from 11, three steps). TLC Rf ) 0.61 (15% EtOAc/
hexanes); 1H NMR (300 MHz, CDCl3) δ 6.81-6.71 (m, 3H), 6.57
(t, J ) 7.3 Hz, 1H), 3.80 (s, 3H), 3.77 (s, 3H), 3.39 (d, J ) 7.3 Hz,
2H); 13C NMR (75 MHz, CDCl3) δ 153.6, 151.5, 136.7, 127.0,
116.3, 111.9, 111.3, 89.5, 55.9, 55.7, 34.1; IR (neat, KBr, cm-1
)
2949, 2831, 1591, 1504, 1227, 1045, 787; HRMS (FAB, M+) m/z
333.9188 (calculated for C11H12Br2O2, 333.9204).
3-(2,5-Dimethoxyphenyl)propyne (14). To the dibromoalkene
13 (0.313 g, 0.93 mmol) in an 8 mL screw-cap vial was added
magnesium (0.045 g, 1.88 mmol) and dry THF (1 mL). The vial
was flushed with argon and sealed tightly with a rubber septum.
The mixture was heated in a 75 °C oil bath for 15 min when a
check by TLC showed consumption of the starting material. The
mixture was cooled and the residue purified by flash chromatog-
raphy (SiO2 17 g, 10% EtOAc/hexanes) to afford acetylene 14 as
a clear viscous oil (0.160 g, 97%): TLC Rf ) 0.39 (5% EtOAc/
hexanes); 1H NMR (300 MHz, CDCl3) δ 7.15 (m, 1H), 6.80-6.72
(m, 2H), 3.800 (s, 3H), 3.797 (s, 3H), 3.58 (d, J ) 2.7 Hz, 2H),
2.20 (t, J ) 2.7 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ 153.6,
150.9, 125.6, 115.3, 111.9, 110.9, 81.7, 70.5, 55.8, 55.6, 19.3; IR
(neat, KBr, cm-1) 3290, 2999, 2833, 1593, 1499, 1220, 1049;
HRFAB [MLi+] 183.1000 (calculated for C11H12O2Li, 183.0997).
2,4-Diamino-5-[3-(2,5-dimethoxyphenyl)prop-1-ynyl]pyrimi-
dine (15). To an oven-dried 8 mL screw-cap vial was added 2,4-
diamino-5-iodopyrimidine (0.145 g, 0.614 mmol), CuI (18 mg,
0.094 mmol, ∼15%), and Pd(PPh3)2Cl2 (30 mg, 0.043 mmol, ∼7%).
Degassed (argon purge) anhydrous DMF (1.5 mL) was added
followed by alkyne 14 (0.271 g, 1.54 mmol) as a solution in DMF
(1 mL). Degassed (argon purge) anhydrous triethylamine was added
(2.5mL),andthemixturewasdegassedonceusingthefreeze-pump-thaw
2,4-Diamino-5-[3-(2,5-dimethoxyphenyl)prop-1-ynyl]-6-
ethylpyrimidine (17). To an oven-dried 15 mL sealed tube was
added 2,4-diamino-6-ethyl-5-iodopyrimidine (0.132 g, 0.50 mmol),
Pd(PPh3)2Cl2 (24.6 mg, 0.035 mmol, 7% Pd), CuI (6.6 mg, 0.035
mmol, 7%), and acetylene 14 (0.176 g, 1.00 mmol). Degassed
(argon purge) anhydrous DMF and triethylamine (2.5 mL each)
were added, and the tube was sealed and the mixture degassed by
one cycle of freeze-pump-thaw. The mixture was stirred at 50
°C for 15 h and then added to a separatory funnel containing EtOAc

New Bacillus anthracis DHFR Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7539
(20 mL). The organic layer was washed twice with a water/saturated
NaHCO3 solution (1:2, 8 mL) and then brine (8 mL). The organic
layer was then dried over MgSO4 and concentrated under reduced
pressure. The residue was purified by flash chromatography (SiO2,
40 g, 1% MeOH in CHCl3) to afford coupled product 17 as a white
powder (0.110 g, 70%). An analytical sample was obtained by
crystallization from MeCN. TLC Rf ) 0.61 (9:1, CHCl3/MeOH);
The organic phase was dried over sodium sulfate and concentrated
to afford the crude product that was purified by flash chromatog-
raphy (SiO2 13 g, EtOAc) to afford coupled pyrimidine 20 as a tan
powder (0.086 g, 54%). TLC Rf ) 0.22 (EtOAc); mp 176 - 178
°C; 1H NMR (500 MHz, DMSO-d6) δ 6.71 (s, 2H), 6.30 (vbs, 2H),
6.15 (bs, 2H), 3.85 (s, 2H), 3.77 (s, 6H), 3.64 (s, 3H), 2.56 (q, J )
7.5 Hz, 2H), 1.12 (t, J ) 7.5 Hz, 3H); 13C NMR (125 MHz, DMSO-
d6) δ 171.5, 164.4, 161.2, 152.8, 136.0, 133.0, 105.0, 95.7, 87.9,
76.2, 60.0, 55.7, 28.8, 25.7, 12.5; HRMS (FAB, MH+) m/z
343.1784 (calculated for C18H23 N4O3, 343.1770); HPLC (a) tR )
4.21 min, 99.3%, (b) tR ) 8.27 min, 98.7%. Anal. (C18H22N4O3)
C, H, N.
3-(2,3-Dimethoxyphenyl)propyne (21). To a 0 °C suspension
of methoxymethyltriphenylphosphonium chloride (1.80 g, 5.25
mmol) in dry THF (15 mL) under an argon atmosphere is added
NaOtBu (0.65 g, 6.76 mmol) in one portion. The red-orange
suspension was stirred for a further 3 min at 0 °C, and then 2,3-
dimethoxybenzaldehyde (0.50 g, 3.0 mmol) was added directly in
small portions. After 5 min, the reaction was quenched with water
(15 mL) and allowed to stir for 16 h (can be worked up
immediately). The mixture was diluted with ether (15 mL), and
the organic phase was separated. The aqueous phase was extracted
with additional ether (2 × 10 mL) and the combined organics were
washed with brine (15 mL), dried over sodium sulfate, and
concentrated to afford the crude product that was filtered through
a column of silica (SiO2 19 g, 5% EtOAc/hexanes) to afford the
crude enol that was immediately hydrolyzed in the subsequent step.
TLC Rf ) 0.36 (15% EtOAc/hexanes).
1
mp, decomposed above 180 °C; H NMR (500 MHz, DMSO-d6)
δ 7.06 (d, J ) 2.9 Hz, 1H), 6.91 (d, J ) 8.8 Hz, 1H), 6.79 (dd, J
) 8.8, 2.9 Hz, 1H), 6.18 (s, 2H), 3.77-3.76 (m, 5H), 3.70 (s, 3H),
2.54 (q, J ) 7.6 Hz, 2H), 1.12 (t, J ) 7.6 Hz, 3H); 13C NMR (125
MHz, DMSO-d6) δ 171.5, 164.5, 161.2, 153.1, 150.6, 126.3, 115.0,
111.8, 111.4, 95.2, 87.9, 76.2, 55.8, 55.3, 28.9, 20.6, 12.6; HRFAB
[MLi+] 319.1736 (calculated for C17H20N4O2Li, 319.1746); HPLC
(a) tR ) 6.13 min, 96.0%, (b) tR ) 8.85 min, 97.0%.
2,4-Diamino-5-iodo-6-n-propylpyrimidine. To a flame-dried
200 mL flask was added 2,4-diamino-6-n-propylpyrimidine (3.0 g,
0.016 mol) in MeOH (56 mL) followed by dropwise addition of
1.0 M ICl solution in CH2Cl2 (56 mL, 0.056 mol). The solution
was stirred at 25 °C for 18 h and then the solvent removed under
reduced pressure. The resulting viscous oil was stirred in Et2O (150
mL) for 2 h. The resulting solid was filtered off and washed with
Et2O (3 × 10 mL) to afford the HCl salt as a yellow solid. The
crude salt was suspended in 1.0 N NaOH (150 mL) and stirred at
25 °C for 2 h. The solid was filtered, washed with cold water (2 ×
10 mL) followed by cold Et2O (3 × 10 mL), and dried under
vacuum to afford the product as a brown powder (1.86 g, 41.5%).
An analytical sample was prepared by recrystallization from MeCN
to give title compound as colorless crystals. Rf ) 0.63 (9:1, CHCl3/
MeOH); mp, 187.0-188.5 °C; 1H NMR (500 MHz, DMSO-d6) δ
6.32 (bs, 2H), 6.03 (bs, 2H), 2.53 (t, J ) 7.5 Hz, 2H), 1.58 (sextet,
J ) 7.5 Hz, 2H), 0.93 (t, J ) 7.5 Hz, 3H); 13C NMR (125 MHz,
DMSO-d6) δ 170.1, 163.6, 163.0, 64.5, 42.7, 21.6, 14.4; HRMS
(FAB, MH+) m/z 279.0106 (calculated for C7H12IN4, 279.0107).
2,4-Diamino-5-[3-(2,5-dimethoxyphenyl)prop-1-ynyl]-6-n-
propylpyrimidine (18). To an oven-dried 8 mL vial was added
2,4-diamino-5-iodo-6-n-propylpyrimidine (0.1202 g, 0.43 mmol),
Pd(PPh3)2Cl2 (34 mg, 0.043 mmol, 10% Pd), CuI (18 mg, 0.095
mmol, 22%). Degassed (argon purge) anhydrous DMF (0.75 mL)
and triethylamine (1.25 mL) were added, followed by acetylene
14 (0.156 g, 0.86 mmol) in DMF (0.50 mL). The vial was sealed
and the mixture degassed by one cycle of freeze-pump-thaw. The
mixture was stirred at 60 °C for 17 h. The reaction mixture was
then added to a separatory funnel containing EtOAc (20 mL). The
organic layer was washed twice with a water/saturated NaHCO3
solution (1:2, 20 mL) and then brine (20 mL). The organic layer
was then dried over MgSO4 and concentrated under reduced
pressure. The residue was preloaded onto silica gel and purified
by flash chromatography (SiO2, 15 g), eluting with straight EtOAc
to afford the coupled pyrimidine 18 as a pale-yellow powder (0.0889
g, 63%). An analytical sample was obtained by crystallization from
MeCN. TLC Rf ) 0.63 (9:1 CHCl3/MeOH); mp, decomposed above
To a solution of crude enol ether in THF (13 mL) was added
10% aqueous HCl (3 mL). The solution was heated to reflux and
monitored by TLC. Once the starting material had been consumed
(∼1.5 h), the mixture was cooled and diluted with saturated
NaHCO3 and ether (15 mL each). The organic phase was separated
and the aqueous phase extracted with additional ether (2 × 15 mL).
The combined organics were washed with saturated NaHCO3 (20
mL) and brine (20 mL), dried over sodium sulfate, and concentrated
to afford the crude product that was used directly in the next step.
TLC Rf ) 0.30 (15% EtOAc/hexanes).
To a 0 °C solution of CBr4 (2.61 g, 7.87 mmol) in dry CH2Cl2
(45 mL) was added PPh3 (4.11 g, 15.7 mmol) in a single portion.
The resulting dark-yellow solution was stirred a further 5 min, and
then the crude aldehyde dissolved in CH2Cl2 (3.5 mL) was added
dropwise. The resulting solution was stirred for 30 min and then
poured into ice cold ether (250 mL), producing a white precipitate.
The mixture was filtered through a column of silica gel (37 g)
equilibrated with hexanes and rinsed with 15% EtOAc/hexanes until
product elution ceased. The filtrate was concentrated and the residue
purified by flash chromatography (SiO2 23 g, 5-10% EtOAc/
hexanes) to afford 1,1-dibromo-3-(2,3-dimethoxyphenyl)propene as
a clear, viscous oil (0.752 g, 74%, three steps). TLC Rf ) 0.46
1
(15% EtOAc/hexanes); H NMR (500 MHz, CDCl3) δ 7.01 (at, J
) 8.0 Hz, 1H), 6.83 (dd, J ) 8.2, 1.3 Hz, 1H), 6.79 (dd, J ) 7.7,
1.3 Hz, 1H), 6.55 (t, J ) 7.2 Hz, 1H), 3.87 (s, 3H), 3.84 (s, 3H),
3.45 (d, J ) 7.2 Hz, 2H); 13C NMR (125 MHz, CDCl3) δ 152.8,
146.9, 137.0, 131.4, 124.1, 121.6, 111.0, 89.5, 60.6, 55.7, 33.7; IR
(neat, KBr, cm-1) 2997, 2935, 2833, 1585, 1481, 1275, 1080, 789;
HRMS (FAB, M+) m/z 333.9189 (calculated for C11H12Br2O2,
333.9204).
1
158.5 °C; H NMR (500 MHz, CDCl3) δ 7.12 (m, 1 H), 6.82 (d,
J ) 5.0 Hz, 1H), 6.80 (m, 1H), 5.16 (bs, 2H), 4.76 (bs, 2H), 3.85
(s, 5H), 3.80 (s, 3H), 2.68 (t, J ) 5.9 Hz, 2H), 1.73 (sextet, J )
5.9 Hz, 2H), 0.99 (t, J ) 5.9 Hz, 3H); 13C NMR (125 mHz, CDCl3)
δ 172.1, 164.4, 160.5, 153.7, 151.1, 126.5, 115.4, 112.0, 111.1,
96.2, 91.4, 75.4, 55.9, 55.7, 38.3, 21.88, 21.0, 14.1; HRMS (FAB,
MH+) m/z 327.1826 (calculated for C18H23N4O2, 327.1821); HPLC
(a) tR ) 7.19 min, 98.2%, (b) tR ) 9.93 min, 99.0%.
To the dibromoalkene (0.530 g, 1.58 mmol) in an 8 mL screw-
cap vial was added magnesium (0.078 g, 3.25 mmol) and dry THF
(1.5 mL). The vial was flushed with argon and sealed tightly with
a rubber septum. The mixture was heated in a 75 °C oil bath for
40 min when a check by TLC showed consumption of the starting
material. The mixture was cooled and the residue purified by flash
chromatography (SiO2 22 g, 10% EtOAc/hexanes) to afford
acetylene 21 as a clear viscous oil (0.252 g, 91%). TLC Rf ) 0.39
(5% EtOAc/hexanes); 1H NMR (500 MHz, CDCl3) δ 7.09 (m, 1H),
7.04 (t, J ) 7.9 Hz, 1H), 6.84 (dd, J ) 8.0, 1.6 Hz, 1H), 3.86 (s,
3H), 3.85 (s, 3H), 3.62 (d, J ) 2.7 Hz, 2H), 2.14 (t, J ) 2.7 Hz,
1H); 13C NMR (125 MHz, CDCl3) δ 152.6, 146.5, 130.2, 124.0,
2,4-Diamino-5-[3-(3,4,5-trimethoxyphenyl)prop-1-ynyl]-6-
ethylpyrimidine (20). To an oven-dried 8 mL screw-cap vial
was added alkyne 19 (0.190 g, 0.921 mmol), 2,4-diamino-6-ethyl-
5-iodopyrimidine (0.122 g, 0.462 mmol), CuI (0.015 g, 0.079 mmol,
∼17%), and Pd(PPh3)2Cl2 (23 mg, 0.033 mmol, ∼7%). Degassed
(argon purge) anhydrous DMF and degassed anhydrous triethy-
lamine were added (1.25 mL each), and the mixture was degassed
once using the freeze-pump-thaw method. The vial was sealed
under argon and heated at 60 °C for 4 h. After cooling, the orange
solution was diluted with EtOAc (20 mL) and washed twice with
a water/saturated NaHCO3 solution (1:2, 20 mL) and brine (20 mL).
121.0. 111.2, 82.2, 69.9, 60.5, 55.8, 19.2; IR (neat, KBr, cm-1
)

7540 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Beierlein et al.
3290, 2937, 1587, 1483, 1273, 1074, 748; HRMS (FAB, M+) m/z
176.0845 (calculated for C11H12O2, 176.0837).
(2) Coker, P.; Smith, K.; Hugh-Jones, M. Antimicrobial susceptibilities
of diverse Bacillus anthracis isolates. Antimicrob. Agents Chemother.
2002, 46, 3843–3845.
(3) Davidson, R.; Cavalcanti, R.; Brunton, J.; Bast, D.; DeAzavedo, J.;
Kibsey, P.; Fleming, C.; Low, D. Resistance to levofloxacin and failure
of treatment of Pneumococcal pneumonia. N. Engl. J. Med. 2002, 346,
747–750.
(4) Neuhauser, M.; Weinstein, R.; Rydman, R.; Danziger, L.; Karam, G.;
Quinn, J. Antibiotic resistance among gram-negative Bacilli in US
intensive care units. JAMA, J. Am. Med. Assoc. 2003, 289, 885–888.
(5) Barrow, E.; Bourne, P.; Barrow, W. Functional cloning of Bacillus
anthracis dihydrofolate reductase and confirmation of natural resistance
to trimethoprim. Antimicrob. Agents Chemother. 2004, 48, 4643–4649.
(6) Sirotnak, F.; Burchall, J.; Ensminger, W.; Montgomery, J. Folate
Antagonists as Therapeutic Agents; Academic Press: New York, 1984.
(7) Fleming, G.; Schilsky, R. Antifolates: the next generation. Semin.
Oncol. 1992, 19, 707–719.
(8) Salter, A. Trimethoprim-sulfamethoxazole: an assessment of more than
12 years of use. ReV. Infect. Dis. 1982, 4, 196–236.
(9) Roth, B.; Rauckman, B.; Ferone, R.; Baccanari, D.; Champness, J.;
Hyde, R. 2,4-Diamino-5-benzylpyrimidines as antibacterial agents. 7.
Analysis of the effect of 3,5-dialkyl substituent size and shape on
binding to four different dihydrofolate reductase enzymes. J. Med.
Chem. 1987, 30, 348–356.
(10) Plowe, C.; Kublin, J.; Dzinjalamala, F. K.; Kamwendo, D.; Mukadam,
R.; Chimpeni, P.; Molyneux, M.; Taylor, T.; Terrie, E. Sustained
efficacy of sulfadoxine-pyrimethamine for uncomplicated falciparum
malaria in Malawi after 10 years as first-line treatment: five-year
prospective study. Br. Med. J. 2004, 328, 545–8.
(11) Armstrong, C.; Smith, C. Cyclization and N-dealkylation of chloro-
guanide by rabbit and rat hepatic microsomes. Toxicol. Appl. Phar-
macol. 1974, 29, 90.
(12) Rosowsky, A.; Forsch, R.; Queener, S. Further studies on 2,4-diamino-
5-(2′,5′-disubstituted benzyl)pyrimidines as potent and selective inhibi-
tors of dihydrofolate reductases from three major opportunistic
pathogens of AIDS. J. Med. Chem. 2003, 46, 1726–1736.
(13) Bertino, J. Karnofsky memorial lecture: ode to methotrexate. J. Clin.
Oncol. 1993, 11, 5–14.
(14) Joska, T.; Anderson, A. Structure-activity relationships of Bacillus
cereus and Bacillus anthracis dihydrofolate reductase: toward the
identification of new potent drug leads. Antimicrob. Agents Chemother.
2006, 50, 3435–3443.
(15) Bennett, B.; Xu, H.; Simmerman, R.; Lee, R.; Dealwis, C. Crystal
structure of the anthrax drug target Bacillus anthracis dihydrofolate
reductase. J. Med. Chem. 2007, 50, 4374–4381.
(16) Pelphrey, P.; Popov, V.; Joska, T.; Beierlein, J.; Bolstad, E.; Fillingham,
Y.; Wright, D.; Anderson, A. Highly efficient ligands for DHFR from
Cryptosporidium hominis and Toxoplasma gondii inspired by structural
analysis. J. Med. Chem. 2007, 50, 940–950.
(17) Chan, D.; Anderson, A. Towards species-specific antifolates. Curr.
Med. Chem. 2006, 13, 377–398.
(18) Bates, P.; Kelley, L.; MacCallum, R.; Sternberg, M. Enhancement of
protein modeling by human intervention in applying the automatic
programs 3D-JIGSAW and 3D-PSSM. Proteins 2001, 45 (S5), 39–
46.
(19) Anderson, A. Two crystal structures of dihydrofolate reductase-
thymidylate synthase from Cryptosporidium hominis reveal protein:
ligand interactions including a structural basis for observed antifolate
resistance. Acta Crystallogr. 2005, F61, 258–262.
(20) Bolin, J.; Filman, D.; Matthews, D.; Hamlin, R.; Kraut, J. Crystal
structures of Escherichia coli and Lactobacillus casei dihydrofolate
reductase refined at 1.7 Å resolution. I. General features and binding
of methotrexate. J. Biol. Chem. 1982, 257, 13650–13662.
(21) Cody, V.; Schwalbe, C. Structural characteristics of antifolate dihy-
drofolate reductase enzyme interactions. Crystallogr. ReV. 2006, 12,
301–333.
2,4-Diamino-5-[3-(2,3-dimethoxyphenyl)prop-1-ynyl]-6-
ethylpyrimidine (22). To an oven-dried 8 mL screw-cap vial
was added alkyne 21 (0.145 g, 0.823 mmol), 2,4-diamino-6-ethyl-
5-iodopyrimidine (0.108 g, 0.409 mmol), CuI (0.010 g, 0.052 mmol,
∼13%), and Pd(PPh3)2Cl2 (20 mg, 0.028 mmol, ∼7%). Degassed
(argon purge) anhydrous DMF (1.0 mL) and triethylamine were
added (1.0 mL each), and the mixture was degassed once using
the freeze-pump-thaw method. The vial was sealed under argon
with a rubber septum and heated at 60 °C for 3 h. After cooling,
the orange mixture was diluted with EtOAc (60 mL) and washed
with a water/saturated NaHCO3 solution (1:2, 20 mL × 3) and brine
(20 mL). The organic phase was dried over sodium sulfate and
concentrated to afford the crude product that was purified by flash
chromatography (SiO2 14 g, EtOAc) to afford the still pyrimidine
22 (0.094 g, 73%) as an amber powder. An analytical sample was
generated by triturating under ether: TLC Rf ) 0.22 (EtOAc); mp
1
158-160 °C; H NMR (500 MHz, CDCl3) δ 7.08 (dd, J ) 7.8,
1.7 Hz, 1H), 7.05 (t, J ) 7.8 Hz, 1H), 6.85 (dd, J ) 7.8, 1.7 Hz,
1H), 5.16 (bs, 2H), 4.76 (bs, 2H), 3.89 (s, 2H), 3.88 (s, 6H), 2.69
(q, J ) 7.6 Hz, 2H), 1.22 (t, J ) 7.6 Hz, 3H); 13C NMR (125
MHz, CDCl3) δ 173.3, 164.5, 160.7, 152.7, 146.7, 131.0, 124.1,
121.0, 111.3, 96.6, 90.8, 74.9, 60.5, 55.8, 29.6, 20.8, 12.6; HRMS
(FAB, MH+) m/z 313.1645 (calculated for C17H21N4O2, 313.1665);
HPLC (a) tR ) 5.34 min, 97.7%, (b) tR ) 9.42 min, 97.2%.
2,4-Diamino-5-(3-phenylprop-1-ynyl)-6-ethylpyrimidine (24).
To an oven-dried 8 mL screw-cap vial was added 2,4-diamino-6-
ethyl-5-iodopyrimidine (0.132 g, 0.50mmol), CuI (7.0 mg, 0.035
mmol, 7%), and Pd(PPh3)2Cl2 (25.0 mg, 0.035 mmol, 7% Pd).
Degassed (argon purge) anhydrous DMF (2.0 mL) was added
followed by 3-phenyl-1-propyne 23 (0.087, 0.75 mmol) as a solution
in DMF (0.5 mL). Degassed (argon purge) anhydrous triethylamine
was added (2.5 mL), and the mixture was degassed once using the
freeze-pump-thaw method. The vial was sealed under argon and
heated at 50 °C for 8 h, then cooled to room temperature and stirred
for another 18 h. After cooling, the orange solution was diluted
with EtOAc (20 mL) and washed twice with a water/saturated
NaHCO3 solution (1:2, 7.5 mL) and brine 7.5 mL). The organic
phase was dried over MgSO4 and concentrated to afford the crude
product that was purified by flash chromatography (SiO2 40 g, 2%
MeOH/CHCl3) to afford coupled pyrimidine 24 as a pale solid (78.6
mg, 62%). An analytical sample was obtained by crystallization
from MeCN. TLC Rf ) 0.56 (9:1 CHCl3/MeOH); mp, decomposed
1
above 118.5 °C; H NMR (500 MHz, CDCl3) δ 7.42 (d, J ) 8.3
Hz, 2H), 7.37 (t, J ) 8.3, 2H Hz), 7.29 (t, J ) 8.3 Hz,1H), 5.22
(bs, 2H), 4.96 (bs, 2H), 3.92 (s, 2H), 2.72 (q, J ) 6.0 Hz, 2H),
1.24 (t, J ) 6.0 Hz, 3H); 13C NMR (125 MHz, CDCl3) δ 173.5,
164.5, 160.8, 136.9, 128.7, 127.8, 126.8, 96.4, 90.5, 75.4, 29.7,
26.2, 12.6; HRMS (FAB, MH+) m/z 253.1461 (calculated for
C15H17N4, 253.1453); HPLC (a) tR ) 5.32 min, 95.9%, (b) tR
9.18 min, 95.9%.
)
Acknowledgment. The authors gratefully acknowledge
Justin Fair (University of Connecticut) for his assistance in
performing the HPLC analyses and funding from NIAID Grant
AI073375.
(22) Yuvaniyama, J.; Chitnumsub, P.; Kamchonwongpaisan, S.; Vanich-
tanankul, J.; Sirawaraporn, W.; Taylor, P.; Walkinshaw, M.; Yuthavong,
Y. Insights into antifolate resistance from malarial DHFR-TS struc-
tures. Nat. Struct. Biol. 2003, 10, 357–365.
(23) Richardson, M. L.; Stevens, M. F. G. Structural studies on bioactive
compounds. Part 37. Suzuki coupling of diaminopyrimidines: a new
synthesis of the antimalarial drug pyrimethamine. J. Chem. Res., Synop.
2002, 10, 482.
(24) Roth, B.; Aig, E.; Lane, K.; Rauckman, B. S. 2,4-Diamino-5-
benzylpyrimidines as antibacterial agents. 4. 6-Substituted trimethoprim
derivatives from phenolic Mannich intermediates. Application to the
synthesis of trimethoprim and 3,5-dialkylbenzyl analogs. J. Med.
Chem. 1980, 23 (5), 535–541.
Supporting Information Available: Details of HPLC purity
determinations for compounds 15-18, 20, 22, and 24 and instru-
mentation; tabulated data; copies of chromatograms; 1H NMR and
13C NMR spectra of compounds 13-18, 20-22, and 24 and
appropriate intermediates; and elemental analysis data for com-
pounds 15 and 20. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Mock, M.; Fouet, A. Anthrax. Annu. ReV. Microbiol. 2001, 55, 647–
671.
JM800776A